diff options
Diffstat (limited to 'thirdparty')
63 files changed, 7954 insertions, 1756 deletions
diff --git a/thirdparty/README.md b/thirdparty/README.md index 4ed17168e8..56b4a07d9b 100644 --- a/thirdparty/README.md +++ b/thirdparty/README.md @@ -107,7 +107,7 @@ Files extracted from upstream source: ## libpng - Upstream: http://libpng.org/pub/png/libpng.html -- Version: 1.6.31rc01 +- Version: 1.6.32 - License: libpng/zlib Files extracted from upstream source: @@ -129,6 +129,19 @@ Files extracted from upstream source: TODO. +## libtheora + +- Upstream: https://www.theora.org +- Version: 1.1.1 +- License: BSD-3-Clause + +Files extracted from upstream source: + +- all .c, .h in lib/ +- all .h files in include/theora/ as theora/ +- COPYING and LICENSE + + ## libvorbis - Upstream: https://www.xiph.org/vorbis @@ -180,7 +193,8 @@ Files extracted from the upstream source: - contrib/minizip/{crypt.h,ioapi.{c,h},zip.{c,h},unzip.{c,h}} Important: Some files have Godot-made changes for use in core/io. -TODO: Properly sync with version 1.2.4 and document changes. +They are marked with `/* GODOT start */` and `/* GODOT end */` +comments and a patch is provided in the minizip/ folder. ## misc @@ -246,18 +260,30 @@ Collection of single-file libraries used in Godot components. * License: zlib - `stb_truetype.h` * Upstream: https://github.com/nothings/stb - * Version: 1.11 + * Version: 1.17 * License: Public Domain (Unlicense) or MIT - `stb_vorbis.c` * Upstream: https://github.com/nothings/stb - * Version: 1.09 + * Version: 1.11 * License: Public Domain (Unlicense) or MIT +## nanosvg + +- Upstream: https://github.com/memononen/nanosvg +- Version: 9a74da4 (git) +- License: zlib + +Files extracted from the upstream source: + +- All .h files in `src/` +- LICENSE.txt + + ## openssl - Upstream: https://www.openssl.org -- Version: 1.0.2h +- Version: 1.0.2l - License: OpenSSL license / BSD-like Files extracted from the upstream source: @@ -330,19 +356,6 @@ Files extracted from upstream source: - all .cpp, .h and .inl files -## theora - -- Upstream: https://www.theora.org -- Version: 1.1.1 -- License: BSD-3-Clause - -Files extracted from upstream source: - -- all .c, .h in lib/ -- all .h files in include/theora/ as theora/ -- COPYING and LICENSE - - ## tinyexr - Upstream: https://github.com/syoyo/tinyexr @@ -368,13 +381,14 @@ Files extracted from upstream source: - all .c and .h files + ## zstd - Upstream: https://github.com/facebook/zstd -- Version: 1.3.0 +- Version: 1.3.1 - License: BSD-3-Clause Files extracted from upstream source: -- all .c and .h under lib/ -- README.md, LICENSE, PATENTS +- lib/{common/,compress/,decompress/,zstd.h} +- README.md, LICENSE diff --git a/thirdparty/enet/enet/godot.h b/thirdparty/enet/enet/godot.h index 75645153dd..937c4fa60e 100644 --- a/thirdparty/enet/enet/godot.h +++ b/thirdparty/enet/enet/godot.h @@ -3,7 +3,7 @@ /*************************************************************************/ /* This file is part of: */ /* GODOT ENGINE */ -/* http://www.godotengine.org */ +/* https://godotengine.org */ /*************************************************************************/ /* Copyright (c) 2007-2017 Juan Linietsky, Ariel Manzur. */ /* Copyright (c) 2014-2017 Godot Engine contributors (cf. AUTHORS.md) */ diff --git a/thirdparty/enet/godot.cpp b/thirdparty/enet/godot.cpp index f050b7b916..2fc264345b 100644 --- a/thirdparty/enet/godot.cpp +++ b/thirdparty/enet/godot.cpp @@ -3,7 +3,7 @@ /*************************************************************************/ /* This file is part of: */ /* GODOT ENGINE */ -/* http://www.godotengine.org */ +/* https://godotengine.org */ /*************************************************************************/ /* Copyright (c) 2007-2017 Juan Linietsky, Ariel Manzur. */ /* Copyright (c) 2014-2017 Godot Engine contributors (cf. AUTHORS.md) */ diff --git a/thirdparty/libpng/arm/filter_neon.S b/thirdparty/libpng/arm/filter_neon.S index 09267c6030..000764cd21 100644 --- a/thirdparty/libpng/arm/filter_neon.S +++ b/thirdparty/libpng/arm/filter_neon.S @@ -3,7 +3,7 @@ * * Copyright (c) 2014,2017 Glenn Randers-Pehrson * Written by Mans Rullgard, 2011. - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.31 [July 27, 2017] * * This code is released under the libpng license. * For conditions of distribution and use, see the disclaimer diff --git a/thirdparty/libpng/png.c b/thirdparty/libpng/png.c index 87fda0b306..2352df13cb 100644 --- a/thirdparty/libpng/png.c +++ b/thirdparty/libpng/png.c @@ -1,7 +1,7 @@ /* png.c - location for general purpose libpng functions * - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.32 [August 24, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) @@ -14,7 +14,7 @@ #include "pngpriv.h" /* Generate a compiler error if there is an old png.h in the search path. */ -typedef png_libpng_version_1_6_31rc01 Your_png_h_is_not_version_1_6_31rc01; +typedef png_libpng_version_1_6_32 Your_png_h_is_not_version_1_6_32; #ifdef __GNUC__ /* The version tests may need to be added to, but the problem warning has @@ -619,8 +619,18 @@ png_free_data(png_const_structrp png_ptr, png_inforp info_ptr, png_uint_32 mask, /* Free any eXIf entry */ if (((mask & PNG_FREE_EXIF) & info_ptr->free_me) != 0) { - png_free(png_ptr, info_ptr->exif); - info_ptr->exif = NULL; +# ifdef PNG_READ_eXIf_SUPPORTED + if (info_ptr->eXIf_buf) + { + png_free(png_ptr, info_ptr->eXIf_buf); + info_ptr->eXIf_buf = NULL; + } +# endif + if (info_ptr->exif) + { + png_free(png_ptr, info_ptr->exif); + info_ptr->exif = NULL; + } info_ptr->valid &= ~PNG_INFO_eXIf; } #endif @@ -806,14 +816,14 @@ png_get_copyright(png_const_structrp png_ptr) #else # ifdef __STDC__ return PNG_STRING_NEWLINE \ - "libpng version 1.6.31rc01 - July 19, 2017" PNG_STRING_NEWLINE \ + "libpng version 1.6.32 - August 24, 2017" PNG_STRING_NEWLINE \ "Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson" \ PNG_STRING_NEWLINE \ "Copyright (c) 1996-1997 Andreas Dilger" PNG_STRING_NEWLINE \ "Copyright (c) 1995-1996 Guy Eric Schalnat, Group 42, Inc." \ PNG_STRING_NEWLINE; # else - return "libpng version 1.6.31rc01 - July 19, 2017\ + return "libpng version 1.6.32 - August 24, 2017\ Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson\ Copyright (c) 1996-1997 Andreas Dilger\ Copyright (c) 1995-1996 Guy Eric Schalnat, Group 42, Inc."; diff --git a/thirdparty/libpng/png.h b/thirdparty/libpng/png.h index 80ecc82c3a..51ac8abe74 100644 --- a/thirdparty/libpng/png.h +++ b/thirdparty/libpng/png.h @@ -1,7 +1,7 @@ /* png.h - header file for PNG reference library * - * libpng version 1.6.31rc01, July 19, 2017 + * libpng version 1.6.32, August 24, 2017 * * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) @@ -12,7 +12,7 @@ * Authors and maintainers: * libpng versions 0.71, May 1995, through 0.88, January 1996: Guy Schalnat * libpng versions 0.89, June 1996, through 0.96, May 1997: Andreas Dilger - * libpng versions 0.97, January 1998, through 1.6.31rc01, July 19, 2017: + * libpng versions 0.97, January 1998, through 1.6.32, August 24, 2017: * Glenn Randers-Pehrson. * See also "Contributing Authors", below. */ @@ -25,7 +25,7 @@ * * This code is released under the libpng license. * - * libpng versions 1.0.7, July 1, 2000 through 1.6.31rc01, July 19, 2017 are + * libpng versions 1.0.7, July 1, 2000 through 1.6.32, August 24, 2017 are * Copyright (c) 2000-2002, 2004, 2006-2017 Glenn Randers-Pehrson, are * derived from libpng-1.0.6, and are distributed according to the same * disclaimer and license as libpng-1.0.6 with the following individuals @@ -213,7 +213,7 @@ * ... * 1.5.28 15 10527 15.so.15.28[.0] * ... - * 1.6.31 16 10631 16.so.16.31[.0] + * 1.6.32 16 10632 16.so.16.32[.0] * * Henceforth the source version will match the shared-library major * and minor numbers; the shared-library major version number will be @@ -241,13 +241,13 @@ * Y2K compliance in libpng: * ========================= * - * July 19, 2017 + * August 24, 2017 * * Since the PNG Development group is an ad-hoc body, we can't make * an official declaration. * * This is your unofficial assurance that libpng from version 0.71 and - * upward through 1.6.31rc01 are Y2K compliant. It is my belief that + * upward through 1.6.32 are Y2K compliant. It is my belief that * earlier versions were also Y2K compliant. * * Libpng only has two year fields. One is a 2-byte unsigned integer @@ -309,8 +309,8 @@ */ /* Version information for png.h - this should match the version in png.c */ -#define PNG_LIBPNG_VER_STRING "1.6.31rc01" -#define PNG_HEADER_VERSION_STRING " libpng version 1.6.31rc01 - July 19, 2017\n" +#define PNG_LIBPNG_VER_STRING "1.6.32" +#define PNG_HEADER_VERSION_STRING " libpng version 1.6.32 - August 24, 2017\n" #define PNG_LIBPNG_VER_SONUM 16 #define PNG_LIBPNG_VER_DLLNUM 16 @@ -318,13 +318,13 @@ /* These should match the first 3 components of PNG_LIBPNG_VER_STRING: */ #define PNG_LIBPNG_VER_MAJOR 1 #define PNG_LIBPNG_VER_MINOR 6 -#define PNG_LIBPNG_VER_RELEASE 31 +#define PNG_LIBPNG_VER_RELEASE 32 /* This should match the numeric part of the final component of * PNG_LIBPNG_VER_STRING, omitting any leading zero: */ -#define PNG_LIBPNG_VER_BUILD 01 +#define PNG_LIBPNG_VER_BUILD 0 /* Release Status */ #define PNG_LIBPNG_BUILD_ALPHA 1 @@ -341,7 +341,7 @@ #define PNG_LIBPNG_BUILD_SPECIAL 32 /* Cannot be OR'ed with PNG_LIBPNG_BUILD_PRIVATE */ -#define PNG_LIBPNG_BUILD_BASE_TYPE PNG_LIBPNG_BUILD_RC +#define PNG_LIBPNG_BUILD_BASE_TYPE PNG_LIBPNG_BUILD_STABLE /* Careful here. At one time, Guy wanted to use 082, but that would be octal. * We must not include leading zeros. @@ -349,7 +349,7 @@ * version 1.0.0 was mis-numbered 100 instead of 10000). From * version 1.0.1 it's xxyyzz, where x=major, y=minor, z=release */ -#define PNG_LIBPNG_VER 10631 /* 1.6.31 */ +#define PNG_LIBPNG_VER 10632 /* 1.6.32 */ /* Library configuration: these options cannot be changed after * the library has been built. @@ -459,7 +459,7 @@ extern "C" { /* This triggers a compiler error in png.c, if png.c and png.h * do not agree upon the version number. */ -typedef char* png_libpng_version_1_6_31rc01; +typedef char* png_libpng_version_1_6_32; /* Basic control structions. Read libpng-manual.txt or libpng.3 for more info. * @@ -2014,6 +2014,11 @@ PNG_EXPORT(246, png_uint_32, png_get_eXIf, (png_const_structrp png_ptr, png_inforp info_ptr, png_bytep *exif)); PNG_EXPORT(247, void, png_set_eXIf, (png_const_structrp png_ptr, png_inforp info_ptr, const png_bytep exif)); + +PNG_EXPORT(248, png_uint_32, png_get_eXIf_1, (png_const_structrp png_ptr, + png_const_inforp info_ptr, png_uint_32 *num_exif, png_bytep *exif)); +PNG_EXPORT(249, void, png_set_eXIf_1, (png_const_structrp png_ptr, + png_inforp info_ptr, const png_uint_32 num_exif, const png_bytep exif)); #endif #ifdef PNG_gAMA_SUPPORTED @@ -3259,7 +3264,7 @@ PNG_EXPORT(244, int, png_set_option, (png_structrp png_ptr, int option, * one to use is one more than this.) */ #ifdef PNG_EXPORT_LAST_ORDINAL - PNG_EXPORT_LAST_ORDINAL(247); + PNG_EXPORT_LAST_ORDINAL(249); #endif #ifdef __cplusplus diff --git a/thirdparty/libpng/pngconf.h b/thirdparty/libpng/pngconf.h index 9308d6a60a..c0f15547be 100644 --- a/thirdparty/libpng/pngconf.h +++ b/thirdparty/libpng/pngconf.h @@ -1,7 +1,7 @@ /* pngconf.h - machine configurable file for libpng * - * libpng version 1.6.31rc01, July 19, 2017 + * libpng version 1.6.32, August 24, 2017 * * Copyright (c) 1998-2002,2004,2006-2016 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) diff --git a/thirdparty/libpng/pngerror.c b/thirdparty/libpng/pngerror.c index 37f26c6558..ad48bfb986 100644 --- a/thirdparty/libpng/pngerror.c +++ b/thirdparty/libpng/pngerror.c @@ -1,7 +1,7 @@ /* pngerror.c - stub functions for i/o and memory allocation * - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.31 [July 27, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) diff --git a/thirdparty/libpng/pngget.c b/thirdparty/libpng/pngget.c index ace9e6351f..26e9fb1c35 100644 --- a/thirdparty/libpng/pngget.c +++ b/thirdparty/libpng/pngget.c @@ -1,8 +1,8 @@ /* pngget.c - retrieval of values from info struct * - * Last changed in libpng 1.6.26 [October 20, 2016] - * Copyright (c) 1998-2002,2004,2006-2016 Glenn Randers-Pehrson + * Last changed in libpng 1.6.32 [August 24, 2017] + * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) * @@ -778,11 +778,22 @@ png_uint_32 PNGAPI png_get_eXIf(png_const_structrp png_ptr, png_inforp info_ptr, png_bytep *exif) { + png_warning(png_ptr, "png_get_eXIf does not work; use png_get_eXIf_1"); + PNG_UNUSED(info_ptr) + PNG_UNUSED(exif) + return 0; +} + +png_uint_32 PNGAPI +png_get_eXIf_1(png_const_structrp png_ptr, png_const_inforp info_ptr, + png_uint_32 *num_exif, png_bytep *exif) +{ png_debug1(1, "in %s retrieval function", "eXIf"); if (png_ptr != NULL && info_ptr != NULL && (info_ptr->valid & PNG_INFO_eXIf) != 0 && exif != NULL) { + *num_exif = info_ptr->num_exif; *exif = info_ptr->exif; return (PNG_INFO_eXIf); } diff --git a/thirdparty/libpng/pnginfo.h b/thirdparty/libpng/pnginfo.h index 6e6d46a62b..d5f6149dbd 100644 --- a/thirdparty/libpng/pnginfo.h +++ b/thirdparty/libpng/pnginfo.h @@ -186,8 +186,11 @@ defined(PNG_READ_BACKGROUND_SUPPORTED) #endif #ifdef PNG_eXIf_SUPPORTED - int num_exif; + int num_exif; /* Added at libpng-1.6.31 */ png_bytep exif; +# ifdef PNG_READ_eXIf_SUPPORTED + png_bytep eXIf_buf; /* Added at libpng-1.6.32 */ +# endif #endif #ifdef PNG_hIST_SUPPORTED diff --git a/thirdparty/libpng/pnglibconf.h b/thirdparty/libpng/pnglibconf.h index 8738ebdfc2..9e45f73129 100644 --- a/thirdparty/libpng/pnglibconf.h +++ b/thirdparty/libpng/pnglibconf.h @@ -1,8 +1,8 @@ -/* libpng 1.6.31rc01 STANDARD API DEFINITION */ +/* libpng 1.6.32 STANDARD API DEFINITION */ /* pnglibconf.h - library build configuration */ -/* Libpng version 1.6.31rc01 - July 19, 2017 */ +/* Libpng version 1.6.32 - August 24, 2017 */ /* Copyright (c) 1998-2017 Glenn Randers-Pehrson */ diff --git a/thirdparty/libpng/pngpread.c b/thirdparty/libpng/pngpread.c index 650ba1e232..fbe361dc34 100644 --- a/thirdparty/libpng/pngpread.c +++ b/thirdparty/libpng/pngpread.c @@ -1,8 +1,8 @@ /* pngpread.c - read a png file in push mode * - * Last changed in libpng 1.6.24 [August 4, 2016] - * Copyright (c) 1998-2002,2004,2006-2016 Glenn Randers-Pehrson + * Last changed in libpng 1.6.32 [August 24, 2017] + * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) * @@ -189,6 +189,7 @@ png_push_read_chunk(png_structrp png_ptr, png_inforp info_ptr) png_crc_read(png_ptr, chunk_tag, 4); png_ptr->chunk_name = PNG_CHUNK_FROM_STRING(chunk_tag); png_check_chunk_name(png_ptr, png_ptr->chunk_name); + png_check_chunk_length(png_ptr, png_ptr->push_length); png_ptr->mode |= PNG_HAVE_CHUNK_HEADER; } diff --git a/thirdparty/libpng/pngpriv.h b/thirdparty/libpng/pngpriv.h index 7c273bd5e3..1f2e90f2b3 100644 --- a/thirdparty/libpng/pngpriv.h +++ b/thirdparty/libpng/pngpriv.h @@ -1,7 +1,7 @@ /* pngpriv.h - private declarations for use inside libpng * - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.32 [August 24, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) @@ -1143,6 +1143,11 @@ PNG_INTERNAL_FUNCTION(void,png_write_sRGB,(png_structrp png_ptr, int intent),PNG_EMPTY); #endif +#ifdef PNG_WRITE_eXIf_SUPPORTED +PNG_INTERNAL_FUNCTION(void,png_write_eXIf,(png_structrp png_ptr, + png_bytep exif, int num_exif),PNG_EMPTY); +#endif + #ifdef PNG_WRITE_iCCP_SUPPORTED PNG_INTERNAL_FUNCTION(void,png_write_iCCP,(png_structrp png_ptr, png_const_charp name, png_const_bytep profile), PNG_EMPTY); @@ -1522,8 +1527,11 @@ PNG_INTERNAL_FUNCTION(void,png_handle_zTXt,(png_structrp png_ptr, png_inforp info_ptr, png_uint_32 length),PNG_EMPTY); #endif -PNG_INTERNAL_FUNCTION(void,png_check_chunk_name,(png_structrp png_ptr, - png_uint_32 chunk_name),PNG_EMPTY); +PNG_INTERNAL_FUNCTION(void,png_check_chunk_name,(png_const_structrp png_ptr, + const png_uint_32 chunk_name),PNG_EMPTY); + +PNG_INTERNAL_FUNCTION(void,png_check_chunk_length,(png_const_structrp png_ptr, + const png_uint_32 chunk_length),PNG_EMPTY); PNG_INTERNAL_FUNCTION(void,png_handle_unknown,(png_structrp png_ptr, png_inforp info_ptr, png_uint_32 length, int keep),PNG_EMPTY); diff --git a/thirdparty/libpng/pngread.c b/thirdparty/libpng/pngread.c index b44d4dfbb5..e34ddd99a0 100644 --- a/thirdparty/libpng/pngread.c +++ b/thirdparty/libpng/pngread.c @@ -1,7 +1,7 @@ /* pngread.c - read a PNG file * - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.32 [August 24, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) @@ -175,6 +175,11 @@ png_read_info(png_structrp png_ptr, png_inforp info_ptr) png_handle_cHRM(png_ptr, info_ptr, length); #endif +#ifdef PNG_READ_eXIf_SUPPORTED + else if (chunk_name == png_eXIf) + png_handle_eXIf(png_ptr, info_ptr, length); +#endif + #ifdef PNG_READ_gAMA_SUPPORTED else if (chunk_name == png_gAMA) png_handle_gAMA(png_ptr, info_ptr, length); @@ -534,6 +539,7 @@ png_read_row(png_structrp png_ptr, png_bytep row, png_bytep dsp_row) png_error(png_ptr, "Invalid attempt to read row data"); /* Fill the row with IDAT data: */ + png_ptr->row_buf[0]=255; /* to force error if no data was found */ png_read_IDAT_data(png_ptr, png_ptr->row_buf, row_info.rowbytes + 1); if (png_ptr->row_buf[0] > PNG_FILTER_VALUE_NONE) @@ -842,6 +848,11 @@ png_read_end(png_structrp png_ptr, png_inforp info_ptr) png_handle_cHRM(png_ptr, info_ptr, length); #endif +#ifdef PNG_READ_eXIf_SUPPORTED + else if (chunk_name == png_eXIf) + png_handle_eXIf(png_ptr, info_ptr, length); +#endif + #ifdef PNG_READ_gAMA_SUPPORTED else if (chunk_name == png_gAMA) png_handle_gAMA(png_ptr, info_ptr, length); diff --git a/thirdparty/libpng/pngrtran.c b/thirdparty/libpng/pngrtran.c index 9285983848..9a30ddf22b 100644 --- a/thirdparty/libpng/pngrtran.c +++ b/thirdparty/libpng/pngrtran.c @@ -1,7 +1,7 @@ /* pngrtran.c - transforms the data in a row for PNG readers * - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.31 [July 27, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) diff --git a/thirdparty/libpng/pngrutil.c b/thirdparty/libpng/pngrutil.c index 9cde64845c..a4fa71457b 100644 --- a/thirdparty/libpng/pngrutil.c +++ b/thirdparty/libpng/pngrutil.c @@ -1,7 +1,7 @@ /* pngrutil.c - utilities to read a PNG file * - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.32 [August 24, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) @@ -181,6 +181,9 @@ png_read_chunk_header(png_structrp png_ptr) /* Check to see if chunk name is valid. */ png_check_chunk_name(png_ptr, png_ptr->chunk_name); + /* Check for too-large chunk length */ + png_check_chunk_length(png_ptr, length); + #ifdef PNG_IO_STATE_SUPPORTED png_ptr->io_state = PNG_IO_READING | PNG_IO_CHUNK_DATA; #endif @@ -1377,11 +1380,13 @@ png_handle_iCCP(png_structrp png_ptr, png_inforp info_ptr, png_uint_32 length) * chunk is just ignored, so does not invalidate the color space. An * alternative is to set the 'invalid' flags at the start of this routine * and only clear them in they were not set before and all the tests pass. - * The minimum 'deflate' stream is assumed to be just the 2 byte header and - * 4 byte checksum. The keyword must be at least one character and there is - * a terminator (0) byte and the compression method. */ - if (length < 9) + + /* The keyword must be at least one character and there is a + * terminator (0) byte and the compression method byte, and the + * 'zlib' datastream is at least 11 bytes. + */ + if (length < 14) { png_crc_finish(png_ptr, length); png_chunk_benign_error(png_ptr, "too short"); @@ -1413,6 +1418,16 @@ png_handle_iCCP(png_structrp png_ptr, png_inforp info_ptr, png_uint_32 length) png_crc_read(png_ptr, (png_bytep)keyword, read_length); length -= read_length; + /* The minimum 'zlib' stream is assumed to be just the 2 byte header, + * 5 bytes minimum 'deflate' stream, and the 4 byte checksum. + */ + if (length < 11) + { + png_crc_finish(png_ptr, length); + png_chunk_benign_error(png_ptr, "too short"); + return; + } + keyword_length = 0; while (keyword_length < 80 && keyword_length < read_length && keyword[keyword_length] != 0) @@ -1431,7 +1446,7 @@ png_handle_iCCP(png_structrp png_ptr, png_inforp info_ptr, png_uint_32 length) if (png_inflate_claim(png_ptr, png_iCCP) == Z_OK) { - Byte profile_header[132]; + Byte profile_header[132]={0}; Byte local_buffer[PNG_INFLATE_BUF_SIZE]; png_alloc_size_t size = (sizeof profile_header); @@ -2014,36 +2029,61 @@ void /* PRIVATE */ png_handle_eXIf(png_structrp png_ptr, png_inforp info_ptr, png_uint_32 length) { unsigned int i; - png_bytep eXIf_buf; png_debug(1, "in png_handle_eXIf"); if ((png_ptr->mode & PNG_HAVE_IHDR) == 0) png_chunk_error(png_ptr, "missing IHDR"); - else if (info_ptr != NULL && (info_ptr->valid & PNG_INFO_eXIf) != 0) + if (length < 2) + { + png_crc_finish(png_ptr, length); + png_chunk_benign_error(png_ptr, "too short"); + return; + } + + else if (info_ptr == NULL || (info_ptr->valid & PNG_INFO_eXIf) != 0) { png_crc_finish(png_ptr, length); png_chunk_benign_error(png_ptr, "duplicate"); return; } - eXIf_buf = png_voidcast(png_bytep, + info_ptr->free_me |= PNG_FREE_EXIF; + + info_ptr->eXIf_buf = png_voidcast(png_bytep, png_malloc_warn(png_ptr, length)); + if (info_ptr->eXIf_buf == NULL) + { + png_crc_finish(png_ptr, length); + png_chunk_benign_error(png_ptr, "out of memory"); + return; + } + for (i = 0; i < length; i++) { png_byte buf[1]; png_crc_read(png_ptr, buf, 1); - eXIf_buf[i] = buf[0]; + info_ptr->eXIf_buf[i] = buf[0]; + if (i == 1 && buf[0] != 'M' && buf[0] != 'I' + && info_ptr->eXIf_buf[0] != buf[0]) + { + png_crc_finish(png_ptr, length); + png_chunk_benign_error(png_ptr, "incorrect byte-order specifier"); + png_free(png_ptr, info_ptr->eXIf_buf); + info_ptr->eXIf_buf = NULL; + return; + } } if (png_crc_finish(png_ptr, 0) != 0) return; - info_ptr->num_exif = length; + png_set_eXIf_1(png_ptr, info_ptr, length, info_ptr->eXIf_buf); - png_set_eXIf(png_ptr, info_ptr, eXIf_buf); + png_free(png_ptr, info_ptr->eXIf_buf); + info_ptr->eXIf_buf = NULL; } #endif @@ -2624,23 +2664,28 @@ png_handle_zTXt(png_structrp png_ptr, png_inforp info_ptr, png_uint_32 length) { png_text text; - /* It worked; png_ptr->read_buffer now looks like a tEXt chunk except - * for the extra compression type byte and the fact that it isn't - * necessarily '\0' terminated. - */ - buffer = png_ptr->read_buffer; - buffer[uncompressed_length+(keyword_length+2)] = 0; - - text.compression = PNG_TEXT_COMPRESSION_zTXt; - text.key = (png_charp)buffer; - text.text = (png_charp)(buffer + keyword_length+2); - text.text_length = uncompressed_length; - text.itxt_length = 0; - text.lang = NULL; - text.lang_key = NULL; - - if (png_set_text_2(png_ptr, info_ptr, &text, 1) != 0) - errmsg = "insufficient memory"; + if (png_ptr->read_buffer == NULL) + errmsg="Read failure in png_handle_zTXt"; + else + { + /* It worked; png_ptr->read_buffer now looks like a tEXt chunk + * except for the extra compression type byte and the fact that + * it isn't necessarily '\0' terminated. + */ + buffer = png_ptr->read_buffer; + buffer[uncompressed_length+(keyword_length+2)] = 0; + + text.compression = PNG_TEXT_COMPRESSION_zTXt; + text.key = (png_charp)buffer; + text.text = (png_charp)(buffer + keyword_length+2); + text.text_length = uncompressed_length; + text.itxt_length = 0; + text.lang = NULL; + text.lang_key = NULL; + + if (png_set_text_2(png_ptr, info_ptr, &text, 1) != 0) + errmsg = "insufficient memory"; + } } else @@ -3076,20 +3121,58 @@ png_handle_unknown(png_structrp png_ptr, png_inforp info_ptr, */ void /* PRIVATE */ -png_check_chunk_name(png_structrp png_ptr, png_uint_32 chunk_name) +png_check_chunk_name(png_const_structrp png_ptr, const png_uint_32 chunk_name) { int i; + png_uint_32 cn=chunk_name; png_debug(1, "in png_check_chunk_name"); for (i=1; i<=4; ++i) { - int c = chunk_name & 0xff; + int c = cn & 0xff; if (c < 65 || c > 122 || (c > 90 && c < 97)) png_chunk_error(png_ptr, "invalid chunk type"); - chunk_name >>= 8; + cn >>= 8; + } +} + +void /* PRIVATE */ +png_check_chunk_length(png_const_structrp png_ptr, const png_uint_32 length) +{ + png_alloc_size_t limit = PNG_UINT_31_MAX; + + if (png_ptr->chunk_name != png_IDAT) + { +# ifdef PNG_SET_USER_LIMITS_SUPPORTED + if (png_ptr->user_chunk_malloc_max > 0 && + png_ptr->user_chunk_malloc_max < limit) + limit = png_ptr->user_chunk_malloc_max; +# elif PNG_USER_CHUNK_MALLOC_MAX > 0 + if (PNG_USER_CHUNK_MALLOC_MAX < limit) + limit = PNG_USER_CHUNK_MALLOC_MAX; +# endif + } + else + { + size_t row_factor = + (png_ptr->width * png_ptr->channels * (png_ptr->bit_depth > 8? 2: 1) + + 1 + (png_ptr->interlaced? 6: 0)); + if (png_ptr->height > PNG_UINT_32_MAX/row_factor) + limit=PNG_UINT_31_MAX; + else + limit = png_ptr->height * row_factor; + limit += 6 + 5*(limit/32566+1); /* zlib+deflate overhead */ + limit=limit < PNG_UINT_31_MAX? limit : PNG_UINT_31_MAX; + } + + if (length > limit) + { + png_debug2(0," length = %lu, limit = %lu", + (unsigned long)length,(unsigned long)limit); + png_chunk_error(png_ptr, "chunk data is too large"); } } diff --git a/thirdparty/libpng/pngset.c b/thirdparty/libpng/pngset.c index 998473557f..6f3a1ee11e 100644 --- a/thirdparty/libpng/pngset.c +++ b/thirdparty/libpng/pngset.c @@ -1,7 +1,7 @@ /* pngset.c - storage of image information into info struct * - * Last changed in libpng 1.6.30 [June 28, 2017] + * Last changed in libpng 1.6.32 [August 24, 2017] * Copyright (c) 1998-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) @@ -139,6 +139,15 @@ void PNGAPI png_set_eXIf(png_const_structrp png_ptr, png_inforp info_ptr, const png_bytep eXIf_buf) { + png_warning(png_ptr, "png_set_eXIf does not work; use png_set_eXIf_1"); + PNG_UNUSED(info_ptr) + PNG_UNUSED(eXIf_buf) +} + +void PNGAPI +png_set_eXIf_1(png_const_structrp png_ptr, png_inforp info_ptr, + const png_uint_32 num_exif, const png_bytep eXIf_buf) +{ int i; png_debug1(1, "in %s storage function", "eXIf"); @@ -146,7 +155,13 @@ png_set_eXIf(png_const_structrp png_ptr, png_inforp info_ptr, if (png_ptr == NULL || info_ptr == NULL) return; - png_free_data(png_ptr, info_ptr, PNG_FREE_EXIF, 0); + if (info_ptr->exif) + { + png_free(png_ptr, info_ptr->exif); + info_ptr->exif = NULL; + } + + info_ptr->num_exif = num_exif; info_ptr->exif = png_voidcast(png_bytep, png_malloc_warn(png_ptr, info_ptr->num_exif)); @@ -154,13 +169,12 @@ png_set_eXIf(png_const_structrp png_ptr, png_inforp info_ptr, if (info_ptr->exif == NULL) { png_warning(png_ptr, "Insufficient memory for eXIf chunk data"); - return; } info_ptr->free_me |= PNG_FREE_EXIF; - for (i = 0; i < info_ptr->num_exif; i++) + for (i = 0; i < (int) info_ptr->num_exif; i++) info_ptr->exif[i] = eXIf_buf[i]; info_ptr->valid |= PNG_INFO_eXIf; @@ -1388,6 +1402,7 @@ png_set_keep_unknown_chunks(png_structrp png_ptr, int keep, static PNG_CONST png_byte chunks_to_ignore[] = { 98, 75, 71, 68, '\0', /* bKGD */ 99, 72, 82, 77, '\0', /* cHRM */ + 101, 88, 73, 102, '\0', /* eXIf */ 103, 65, 77, 65, '\0', /* gAMA */ 104, 73, 83, 84, '\0', /* hIST */ 105, 67, 67, 80, '\0', /* iCCP */ diff --git a/thirdparty/libpng/pngstruct.h b/thirdparty/libpng/pngstruct.h index 44b79dea74..d83f971253 100644 --- a/thirdparty/libpng/pngstruct.h +++ b/thirdparty/libpng/pngstruct.h @@ -1,7 +1,7 @@ /* pngstruct.h - header file for PNG reference library * - * Last changed in libpng 1.6.28 [January 5, 2017] + * Last changed in libpng 1.6.32 [August 24, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) @@ -479,8 +479,5 @@ struct png_struct_def png_colorspace colorspace; #endif #endif - -/* New member added in libpng-1.6.31 */ - int num_exif; }; #endif /* PNGSTRUCT_H */ diff --git a/thirdparty/libpng/pngwrite.c b/thirdparty/libpng/pngwrite.c index 8c2952e794..a7662acb71 100644 --- a/thirdparty/libpng/pngwrite.c +++ b/thirdparty/libpng/pngwrite.c @@ -1,7 +1,7 @@ /* pngwrite.c - general routines to write a PNG file * - * Last changed in libpng 1.6.31 [(PENDING RELEASE)] + * Last changed in libpng 1.6.32 [August 24, 2017] * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) @@ -237,6 +237,11 @@ png_write_info(png_structrp png_ptr, png_const_inforp info_ptr) png_write_bKGD(png_ptr, &(info_ptr->background), info_ptr->color_type); #endif +#ifdef PNG_WRITE_eXIf_SUPPORTED + if ((info_ptr->valid & PNG_INFO_eXIf) != 0) + png_write_eXIf(png_ptr, info_ptr->exif, info_ptr->num_exif); +#endif + #ifdef PNG_WRITE_hIST_SUPPORTED if ((info_ptr->valid & PNG_INFO_hIST) != 0) png_write_hIST(png_ptr, info_ptr->hist, info_ptr->num_palette); @@ -432,6 +437,12 @@ png_write_end(png_structrp png_ptr, png_inforp info_ptr) } } #endif + +#ifdef PNG_WRITE_eXIf_SUPPORTED + if ((info_ptr->valid & PNG_INFO_eXIf) != 0) + png_write_eXIf(png_ptr, info_ptr->exif, info_ptr->num_exif); +#endif + #ifdef PNG_WRITE_UNKNOWN_CHUNKS_SUPPORTED write_unknown_chunks(png_ptr, info_ptr, PNG_AFTER_IDAT); #endif diff --git a/thirdparty/libpng/pngwutil.c b/thirdparty/libpng/pngwutil.c index 348bb524dd..0d4fb1336c 100644 --- a/thirdparty/libpng/pngwutil.c +++ b/thirdparty/libpng/pngwutil.c @@ -1,8 +1,8 @@ /* pngwutil.c - utilities to write a PNG file * - * Last changed in libpng 1.6.26 [October 20, 2016] - * Copyright (c) 1998-2002,2004,2006-2016 Glenn Randers-Pehrson + * Last changed in libpng 1.6.32 [August 24, 2017] + * Copyright (c) 1998-2002,2004,2006-2017 Glenn Randers-Pehrson * (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger) * (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.) * @@ -1479,24 +1479,15 @@ void /* PRIVATE */ png_write_eXIf(png_structrp png_ptr, png_bytep exif, int num_exif) { int i; - png_byte buf[3]; + png_byte buf[1]; png_debug(1, "in png_write_eXIf"); - if (num_exif > (int)png_ptr->num_exif) - { - png_debug2(3, "num_exif = %d, png_ptr->num_exif = %d", num_exif, - png_ptr->num_exif); - - png_warning(png_ptr, "Invalid number of exif bytes specified"); - return; - } - png_write_chunk_header(png_ptr, png_eXIf, (png_uint_32)(num_exif)); for (i = 0; i < num_exif; i++) { - buf[i] = exif[i]; + buf[0] = exif[i]; png_write_chunk_data(png_ptr, buf, (png_size_t)1); } diff --git a/thirdparty/minizip/LICENSE-InfoZip.txt b/thirdparty/minizip/LICENSE-InfoZip.txt deleted file mode 100644 index bcfe47e978..0000000000 --- a/thirdparty/minizip/LICENSE-InfoZip.txt +++ /dev/null @@ -1,60 +0,0 @@ -This is version 2007-Mar-4 of the Info-ZIP license. -The definitive version of this document should be available at -ftp://ftp.info-zip.org/pub/infozip/license.html indefinitely and -a copy at http://www.info-zip.org/pub/infozip/license.html. - - -Copyright (c) 1990-2007 Info-ZIP. All rights reserved. - -For the purposes of this copyright and license, "Info-ZIP" is defined as -the following set of individuals: - - Mark Adler, John Bush, Karl Davis, Harald Denker, Jean-Michel Dubois, - Jean-loup Gailly, Hunter Goatley, Ed Gordon, Ian Gorman, Chris Herborth, - Dirk Haase, Greg Hartwig, Robert Heath, Jonathan Hudson, Paul Kienitz, - David Kirschbaum, Johnny Lee, Onno van der Linden, Igor Mandrichenko, - Steve P. Miller, Sergio Monesi, Keith Owens, George Petrov, Greg Roelofs, - Kai Uwe Rommel, Steve Salisbury, Dave Smith, Steven M. Schweda, - Christian Spieler, Cosmin Truta, Antoine Verheijen, Paul von Behren, - Rich Wales, Mike White. - -This software is provided "as is," without warranty of any kind, express -or implied. In no event shall Info-ZIP or its contributors be held liable -for any direct, indirect, incidental, special or consequential damages -arising out of the use of or inability to use this software. - -Permission is granted to anyone to use this software for any purpose, -including commercial applications, and to alter it and redistribute it -freely, subject to the above disclaimer and the following restrictions: - - 1. Redistributions of source code (in whole or in part) must retain - the above copyright notice, definition, disclaimer, and this list - of conditions. - - 2. Redistributions in binary form (compiled executables and libraries) - must reproduce the above copyright notice, definition, disclaimer, - and this list of conditions in documentation and/or other materials - provided with the distribution. The sole exception to this condition - is redistribution of a standard UnZipSFX binary (including SFXWiz) as - part of a self-extracting archive; that is permitted without inclusion - of this license, as long as the normal SFX banner has not been removed - from the binary or disabled. - - 3. Altered versions--including, but not limited to, ports to new operating - systems, existing ports with new graphical interfaces, versions with - modified or added functionality, and dynamic, shared, or static library - versions not from Info-ZIP--must be plainly marked as such and must not - be misrepresented as being the original source or, if binaries, - compiled from the original source. Such altered versions also must not - be misrepresented as being Info-ZIP releases--including, but not - limited to, labeling of the altered versions with the names "Info-ZIP" - (or any variation thereof, including, but not limited to, different - capitalizations), "Pocket UnZip," "WiZ" or "MacZip" without the - explicit permission of Info-ZIP. Such altered versions are further - prohibited from misrepresentative use of the Zip-Bugs or Info-ZIP - e-mail addresses or the Info-ZIP URL(s), such as to imply Info-ZIP - will provide support for the altered versions. - - 4. Info-ZIP retains the right to use the names "Info-ZIP," "Zip," "UnZip," - "UnZipSFX," "WiZ," "Pocket UnZip," "Pocket Zip," and "MacZip" for its - own source and binary releases. diff --git a/thirdparty/minizip/LICENSE-MiniZip.txt b/thirdparty/minizip/LICENSE-MiniZip.txt deleted file mode 100644 index 0e8950f86f..0000000000 --- a/thirdparty/minizip/LICENSE-MiniZip.txt +++ /dev/null @@ -1,32 +0,0 @@ -Credits - - Gilles Vollant - Original MiniZip author - Even Rouault - ZIP64 unzip Support - Daniel Borca - BZip Compression method support in unzip - Mathias Svensson - ZIP64 zip support - Mathias Svensson - BZip Compression method support in zip - - This version has been modified for Godot Engine - - -License ----------------------------------------------------------------------------- - Condition of use and distribution are the same than zlib : - - This software is provided 'as-is', without any express or implied - warranty. In no event will the authors be held liable for any damages - arising from the use of this software. - - Permission is granted to anyone to use this software for any purpose, - including commercial applications, and to alter it and redistribute it - freely, subject to the following restrictions: - - 1. The origin of this software must not be misrepresented; you must not - claim that you wrote the original software. If you use this software - in a product, an acknowledgment in the product documentation would be - appreciated but is not required. - 2. Altered source versions must be plainly marked as such, and must not be - misrepresented as being the original software. - 3. This notice may not be removed or altered from any source distribution. - ----------------------------------------------------------------------------- diff --git a/thirdparty/minizip/MiniZip_info.txt b/thirdparty/minizip/MiniZip_info.txt new file mode 100644 index 0000000000..57d7152420 --- /dev/null +++ b/thirdparty/minizip/MiniZip_info.txt @@ -0,0 +1,74 @@ +MiniZip - Copyright (c) 1998-2010 - by Gilles Vollant - version 1.1 64 bits from Mathias Svensson + +Introduction +--------------------- +MiniZip 1.1 is built from MiniZip 1.0 by Gilles Vollant ( http://www.winimage.com/zLibDll/minizip.html ) + +When adding ZIP64 support into minizip it would result into risk of breaking compatibility with minizip 1.0. +All possible work was done for compatibility. + + +Background +--------------------- +When adding ZIP64 support Mathias Svensson found that Even Rouault have added ZIP64 +support for unzip.c into minizip for a open source project called gdal ( http://www.gdal.org/ ) + +That was used as a starting point. And after that ZIP64 support was added to zip.c +some refactoring and code cleanup was also done. + + +Changed from MiniZip 1.0 to MiniZip 1.1 +--------------------------------------- +* Added ZIP64 support for unzip ( by Even Rouault ) +* Added ZIP64 support for zip ( by Mathias Svensson ) +* Reverted some changed that Even Rouault did. +* Bunch of patches received from Gulles Vollant that he received for MiniZip from various users. +* Added unzip patch for BZIP Compression method (patch create by Daniel Borca) +* Added BZIP Compress method for zip +* Did some refactoring and code cleanup + + +Credits + + Gilles Vollant - Original MiniZip author + Even Rouault - ZIP64 unzip Support + Daniel Borca - BZip Compression method support in unzip + Mathias Svensson - ZIP64 zip support + Mathias Svensson - BZip Compression method support in zip + + Resources + + ZipLayout http://result42.com/projects/ZipFileLayout + Command line tool for Windows that shows the layout and information of the headers in a zip archive. + Used when debugging and validating the creation of zip files using MiniZip64 + + + ZIP App Note http://www.pkware.com/documents/casestudies/APPNOTE.TXT + Zip File specification + + +Notes. + * To be able to use BZip compression method in zip64.c or unzip64.c the BZIP2 lib is needed and HAVE_BZIP2 need to be defined. + +License +---------------------------------------------------------- + Condition of use and distribution are the same than zlib : + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + +---------------------------------------------------------- + diff --git a/thirdparty/minizip/godot-zlib-1.2.4-minizip-seek.patch b/thirdparty/minizip/godot-zlib-1.2.4-minizip-seek.patch new file mode 100644 index 0000000000..8e66416a43 --- /dev/null +++ b/thirdparty/minizip/godot-zlib-1.2.4-minizip-seek.patch @@ -0,0 +1,295 @@ +diff --git a/thirdparty/minizip/ioapi.c b/thirdparty/minizip/ioapi.c +index 49958f61f..0afbdc06a 100644 +--- a/thirdparty/minizip/ioapi.c ++++ b/thirdparty/minizip/ioapi.c +@@ -68,8 +68,15 @@ void fill_zlib_filefunc64_32_def_from_filefunc32(zlib_filefunc64_32_def* p_filef + p_filefunc64_32->zfile_func64.opaque = p_filefunc32->opaque; + p_filefunc64_32->zseek32_file = p_filefunc32->zseek_file; + p_filefunc64_32->ztell32_file = p_filefunc32->ztell_file; ++ /* GODOT start */ ++ p_filefunc64_32->zfile_func64.alloc_mem = p_filefunc32->alloc_mem; ++ p_filefunc64_32->zfile_func64.free_mem = p_filefunc32->free_mem; ++ /* GODOT end */ + } + ++/* GODOT start */ ++/* ++// GODOT end + + + static voidpf ZCALLBACK fopen_file_func OF((voidpf opaque, const char* filename, int mode)); +@@ -233,3 +240,6 @@ void fill_fopen64_filefunc (zlib_filefunc64_def* pzlib_filefunc_def) + pzlib_filefunc_def->zerror_file = ferror_file_func; + pzlib_filefunc_def->opaque = NULL; + } ++// GODOT start ++*/ ++/* GODOT end */ +diff --git a/thirdparty/minizip/ioapi.h b/thirdparty/minizip/ioapi.h +index 8309c4cf8..f25ab6464 100644 +--- a/thirdparty/minizip/ioapi.h ++++ b/thirdparty/minizip/ioapi.h +@@ -145,6 +145,10 @@ typedef struct zlib_filefunc_def_s + close_file_func zclose_file; + testerror_file_func zerror_file; + voidpf opaque; ++ /* GODOT start */ ++ alloc_func alloc_mem; ++ free_func free_mem; ++ /* GODOT end */ + } zlib_filefunc_def; + + typedef ZPOS64_T (ZCALLBACK *tell64_file_func) OF((voidpf opaque, voidpf stream)); +@@ -161,6 +165,10 @@ typedef struct zlib_filefunc64_def_s + close_file_func zclose_file; + testerror_file_func zerror_file; + voidpf opaque; ++ /* GODOT start */ ++ alloc_func alloc_mem; ++ free_func free_mem; ++ /* GODOT end */ + } zlib_filefunc64_def; + + void fill_fopen64_filefunc OF((zlib_filefunc64_def* pzlib_filefunc_def)); +diff --git a/thirdparty/minizip/unzip.c b/thirdparty/minizip/unzip.c +index 7617f41f1..32e27bd65 100644 +--- a/thirdparty/minizip/unzip.c ++++ b/thirdparty/minizip/unzip.c +@@ -157,6 +157,9 @@ typedef struct + uLong compression_method; /* compression method (0==store) */ + ZPOS64_T byte_before_the_zipfile;/* byte before the zipfile, (>0 for sfx)*/ + int raw; ++ /* GODOT start */ ++ int extra_size; ++ /* GODOT end */ + } file_in_zip64_read_info_s; + + +@@ -606,9 +609,10 @@ local unzFile unzOpenInternal (const void *path, + us.z_filefunc.zseek32_file = NULL; + us.z_filefunc.ztell32_file = NULL; + if (pzlib_filefunc64_32_def==NULL) +- fill_fopen64_filefunc(&us.z_filefunc.zfile_func64); +- else +- us.z_filefunc = *pzlib_filefunc64_32_def; ++ /* GODOT start */ ++ return NULL; // standard i/o not supported ++ us.z_filefunc = *pzlib_filefunc64_32_def; ++ /* GODOT end */ + us.is64bitOpenFunction = is64bitOpenFunction; + + +@@ -800,6 +804,18 @@ extern unzFile ZEXPORT unzOpen64 (const void *path) + return unzOpenInternal(path, NULL, 1); + } + ++/* GODOT start */ ++extern void* unzGetOpaque(unzFile file) { ++ ++ unz64_s* s; ++ if (file==NULL) ++ return NULL; ++ s=(unz64_s*)file; ++ ++ return s->z_filefunc.zfile_func64.opaque; ++}; ++/* GODOT end */ ++ + /* + Close a ZipFile opened with unzipOpen. + If there is files inside the .Zip opened with unzipOpenCurrentFile (see later), +@@ -1018,10 +1034,20 @@ local int unz64local_GetCurrentFileInfoInternal (unzFile file, + + if (lSeek!=0) + { +- if (ZSEEK64(s->z_filefunc, s->filestream,lSeek,ZLIB_FILEFUNC_SEEK_CUR)==0) +- lSeek=0; +- else +- err=UNZ_ERRNO; ++ /* GODOT start */ ++ if (lSeek<0) { ++ // WORKAROUND for backwards seeking ++ z_off_t pos = ZTELL64(s->z_filefunc, s->filestream); ++ if (ZSEEK64(s->z_filefunc, s->filestream,pos+lSeek,ZLIB_FILEFUNC_SEEK_SET)==0) ++ lSeek=0; ++ else ++ err=UNZ_ERRNO; ++ } else { ++ if (ZSEEK64(s->z_filefunc, s->filestream,lSeek,ZLIB_FILEFUNC_SEEK_CUR)==0) ++ lSeek=0; ++ else ++ err=UNZ_ERRNO; ++ } + } + + while(acc < file_info.size_file_extra) +@@ -1575,8 +1601,10 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, int* method, + } + else if ((s->cur_file_info.compression_method==Z_DEFLATED) && (!raw)) + { +- pfile_in_zip_read_info->stream.zalloc = (alloc_func)0; +- pfile_in_zip_read_info->stream.zfree = (free_func)0; ++ /* GODOT start */ ++ pfile_in_zip_read_info->stream.zalloc = s->z_filefunc.zfile_func64.alloc_mem; ++ pfile_in_zip_read_info->stream.zfree = s->z_filefunc.zfile_func64.free_mem; ++ /* GODOT end */ + pfile_in_zip_read_info->stream.opaque = (voidpf)0; + pfile_in_zip_read_info->stream.next_in = 0; + pfile_in_zip_read_info->stream.avail_in = 0; +@@ -1608,6 +1636,9 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, int* method, + iSizeVar; + + pfile_in_zip_read_info->stream.avail_in = (uInt)0; ++ /* GODOT start */ ++ pfile_in_zip_read_info->extra_size = iSizeVar; ++ /* GODOT end */ + + s->pfile_in_zip_read = pfile_in_zip_read_info; + s->encrypted = 0; +@@ -1638,6 +1669,85 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, int* method, + return UNZ_OK; + } + ++/* GODOT start */ ++extern int ZEXPORT unzSeekCurrentFile(unzFile file, int pos) { ++ ++ unz64_s* s; ++ file_in_zip64_read_info_s* pfile_in_zip_read_info; ++ if (file==NULL) ++ return UNZ_PARAMERROR; ++ s=(unz64_s*)file; ++ pfile_in_zip_read_info=s->pfile_in_zip_read; ++ ++ if (pfile_in_zip_read_info==NULL) ++ return UNZ_PARAMERROR; ++ ++ if (pfile_in_zip_read_info->compression_method==Z_BZIP2ED) { // don't know how to support bzip ++ return UNZ_INTERNALERROR; ++ }; ++ ++ if ((pfile_in_zip_read_info->compression_method==0) || (pfile_in_zip_read_info->raw)) { ++ ++ pfile_in_zip_read_info->rest_read_compressed = ++ s->cur_file_info.compressed_size - pos; ++ pfile_in_zip_read_info->rest_read_uncompressed = ++ s->cur_file_info.uncompressed_size - pos; ++ ++ pfile_in_zip_read_info->pos_in_zipfile = ++ s->cur_file_info_internal.offset_curfile + SIZEZIPLOCALHEADER + ++ pfile_in_zip_read_info->extra_size + pos; ++ ++ pfile_in_zip_read_info->stream.avail_in = (uInt)0; ++ pfile_in_zip_read_info->stream.total_out = pos; ++ ++ return ZSEEK64(pfile_in_zip_read_info->z_filefunc, ++ pfile_in_zip_read_info->filestream, ++ pfile_in_zip_read_info->byte_before_the_zipfile + pfile_in_zip_read_info->pos_in_zipfile, ++ ZLIB_FILEFUNC_SEEK_SET); ++ ++ } else { // gzip ++ ++ if (pos < pfile_in_zip_read_info->stream.total_out) { // negative seek, rewind ++ ++ pfile_in_zip_read_info->rest_read_compressed = ++ s->cur_file_info.compressed_size ; ++ pfile_in_zip_read_info->rest_read_uncompressed = ++ s->cur_file_info.uncompressed_size ; ++ ++ pfile_in_zip_read_info->pos_in_zipfile = ++ s->cur_file_info_internal.offset_curfile + SIZEZIPLOCALHEADER + ++ pfile_in_zip_read_info->extra_size; ++ ++ (void)inflateReset(&pfile_in_zip_read_info->stream); ++ ++ pfile_in_zip_read_info->stream.avail_in = (uInt)0; ++ pfile_in_zip_read_info->stream.total_out = 0; ++ pfile_in_zip_read_info->stream.next_in = 0; ++ }; ++ ++ // not sure where to read, so read on the stack ++ { ++ char buf[512]; ++ int to_read = pos - pfile_in_zip_read_info->stream.total_out; ++ while (to_read) { ++ ++ int len = to_read > sizeof(buf)?sizeof(buf):to_read; ++ int read = unzReadCurrentFile(file, buf, len); ++ if (read < 0) { ++ return read; ++ }; ++ to_read -= read; ++ if (read == UNZ_EOF) { ++ return pos; ++ }; ++ }; ++ }; ++ }; ++ ++ return pos; ++}; ++/* GODOT end */ ++ + extern int ZEXPORT unzOpenCurrentFile (unzFile file) + { + return unzOpenCurrentFile3(file, NULL, NULL, 0, NULL); +diff --git a/thirdparty/minizip/unzip.h b/thirdparty/minizip/unzip.h +index 3183968b7..54e65ad8a 100644 +--- a/thirdparty/minizip/unzip.h ++++ b/thirdparty/minizip/unzip.h +@@ -202,6 +202,10 @@ extern int ZEXPORT unzClose OF((unzFile file)); + these files MUST be closed with unzipCloseCurrentFile before call unzipClose. + return UNZ_OK if there is no problem. */ + ++/* GODOT start */ ++extern void* unzGetOpaque(unzFile file); ++/* GODOT end */ ++ + extern int ZEXPORT unzGetGlobalInfo OF((unzFile file, + unz_global_info *pglobal_info)); + +@@ -390,6 +394,13 @@ extern int ZEXPORT unzReadCurrentFile OF((unzFile file, + (UNZ_ERRNO for IO error, or zLib error for uncompress error) + */ + ++/* GODOT start */ ++extern int ZEXPORT unzSeekCurrentFile(unzFile file, int pos); ++/* ++ Seek to position in uncompressed data ++*/ ++/* GODOT end */ ++ + extern z_off_t ZEXPORT unztell OF((unzFile file)); + + extern ZPOS64_T ZEXPORT unztell64 OF((unzFile file)); +diff --git a/thirdparty/minizip/zip.c b/thirdparty/minizip/zip.c +index 3c34fc8bd..d7093e745 100644 +--- a/thirdparty/minizip/zip.c ++++ b/thirdparty/minizip/zip.c +@@ -854,9 +854,11 @@ extern zipFile ZEXPORT zipOpen3 (const void *pathname, int append, zipcharpc* gl + + ziinit.z_filefunc.zseek32_file = NULL; + ziinit.z_filefunc.ztell32_file = NULL; +- if (pzlib_filefunc64_32_def==NULL) +- fill_fopen64_filefunc(&ziinit.z_filefunc.zfile_func64); +- else ++ /* GODOT start */ ++ if (pzlib_filefunc64_32_def==NULL) { ++ //fill_fopen64_filefunc(&ziinit.z_filefunc.zfile_func64); ++ } else ++ /* GODOT end */ + ziinit.z_filefunc = *pzlib_filefunc64_32_def; + + ziinit.filestream = ZOPEN64(ziinit.z_filefunc, +@@ -1210,8 +1212,10 @@ extern int ZEXPORT zipOpenNewFileInZip4_64 (zipFile file, const char* filename, + { + if(zi->ci.method == Z_DEFLATED) + { +- zi->ci.stream.zalloc = (alloc_func)0; +- zi->ci.stream.zfree = (free_func)0; ++ /* GODOT start */ ++ zi->ci.stream.zalloc = zi->z_filefunc.zfile_func64.alloc_mem; ++ zi->ci.stream.zfree = zi->z_filefunc.zfile_func64.free_mem; ++ /* GODOT end */ + zi->ci.stream.opaque = (voidpf)0; + + if (windowBits>0) diff --git a/thirdparty/minizip/godot-zlib-1.2.4-minizip-unbreak-gentoo.patch b/thirdparty/minizip/godot-zlib-1.2.4-minizip-unbreak-gentoo.patch new file mode 100644 index 0000000000..9292e32ac6 --- /dev/null +++ b/thirdparty/minizip/godot-zlib-1.2.4-minizip-unbreak-gentoo.patch @@ -0,0 +1,27 @@ +diff --git a/thirdparty/minizip/ioapi.h b/thirdparty/minizip/ioapi.h +index f25ab6464..6043d34ce 100644 +--- a/thirdparty/minizip/ioapi.h ++++ b/thirdparty/minizip/ioapi.h +@@ -44,6 +44,22 @@ + #include <stdlib.h> + #include "zlib.h" + ++/* GODOT start */ ++/* Mighty Gentoo saves the day by breaking the API of their zlib.h, ++ * removing this definition of OF(args) for no practical reason ++ * worth breaking compatibility with all projects that embed minizip ++ * while trying not to diverge too much from upstream zlib. ++ * Cf. https://github.com/godotengine/godot/issues/10539 ++ * ++ * "By and large, this is good open source behaviour, and fits with ++ * the gentoo _don't fuck with upstream's releases_ philosophy" ++ * -- Gentoo philosopher ++ */ ++#ifndef OF /* function prototypes */ ++ #define OF(args) args ++#endif ++/* GODOT end */ ++ + #if defined(USE_FILE32API) + #define fopen64 fopen + #define ftello64 ftell diff --git a/thirdparty/minizip/ioapi.c b/thirdparty/minizip/ioapi.c index d6063a5fe6..2b42df4abd 100644 --- a/thirdparty/minizip/ioapi.c +++ b/thirdparty/minizip/ioapi.c @@ -6,7 +6,7 @@ Modifications for Zip64 support Copyright (C) 2009-2010 Mathias Svensson ( http://result42.com ) - For more info read LICENSE-MiniZip.txt + For more info read MiniZip_info.txt */ @@ -68,11 +68,15 @@ void fill_zlib_filefunc64_32_def_from_filefunc32(zlib_filefunc64_32_def* p_filef p_filefunc64_32->zfile_func64.opaque = p_filefunc32->opaque; p_filefunc64_32->zseek32_file = p_filefunc32->zseek_file; p_filefunc64_32->ztell32_file = p_filefunc32->ztell_file; + /* GODOT start */ p_filefunc64_32->zfile_func64.alloc_mem = p_filefunc32->alloc_mem; p_filefunc64_32->zfile_func64.free_mem = p_filefunc32->free_mem; + /* GODOT end */ } +/* GODOT start */ /* +// GODOT end static voidpf ZCALLBACK fopen_file_func OF((voidpf opaque, const char* filename, int mode)); @@ -119,6 +123,7 @@ static voidpf ZCALLBACK fopen64_file_func (voidpf opaque, const void* filename, return file; } + static uLong ZCALLBACK fread_file_func (voidpf opaque, voidpf stream, void* buf, uLong size) { uLong ret; @@ -140,6 +145,7 @@ static long ZCALLBACK ftell_file_func (voidpf opaque, voidpf stream) return ret; } + static ZPOS64_T ZCALLBACK ftell64_file_func (voidpf opaque, voidpf stream) { ZPOS64_T ret; @@ -234,4 +240,6 @@ void fill_fopen64_filefunc (zlib_filefunc64_def* pzlib_filefunc_def) pzlib_filefunc_def->zerror_file = ferror_file_func; pzlib_filefunc_def->opaque = NULL; } +// GODOT start */ +/* GODOT end */ diff --git a/thirdparty/minizip/ioapi.h b/thirdparty/minizip/ioapi.h index cb6cb7e766..6043d34cea 100644 --- a/thirdparty/minizip/ioapi.h +++ b/thirdparty/minizip/ioapi.h @@ -1,11 +1,12 @@ -/* this file is part of the MiniZip project - ( http://www.winimage.com/zLibDll/minizip.html ) +/* ioapi.h -- IO base function header for compress/uncompress .zip + part of the MiniZip project - ( http://www.winimage.com/zLibDll/minizip.html ) Copyright (C) 1998-2010 Gilles Vollant (minizip) ( http://www.winimage.com/zLibDll/minizip.html ) Modifications for Zip64 support Copyright (C) 2009-2010 Mathias Svensson ( http://result42.com ) - For more info read LICENSE-MiniZip.txt + For more info read MiniZip_info.txt Changes @@ -40,8 +41,25 @@ #endif #include <stdio.h> +#include <stdlib.h> #include "zlib.h" +/* GODOT start */ +/* Mighty Gentoo saves the day by breaking the API of their zlib.h, + * removing this definition of OF(args) for no practical reason + * worth breaking compatibility with all projects that embed minizip + * while trying not to diverge too much from upstream zlib. + * Cf. https://github.com/godotengine/godot/issues/10539 + * + * "By and large, this is good open source behaviour, and fits with + * the gentoo _don't fuck with upstream's releases_ philosophy" + * -- Gentoo philosopher + */ +#ifndef OF /* function prototypes */ + #define OF(args) args +#endif +/* GODOT end */ + #if defined(USE_FILE32API) #define fopen64 fopen #define ftello64 ftell @@ -122,14 +140,14 @@ extern "C" { -typedef voidpf (ZCALLBACK *open_file_func) (voidpf opaque, const char* filename, int mode); -typedef uLong (ZCALLBACK *read_file_func) (voidpf opaque, voidpf stream, void* buf, uLong size); -typedef uLong (ZCALLBACK *write_file_func) (voidpf opaque, voidpf stream, const void* buf, uLong size); -typedef int (ZCALLBACK *close_file_func) (voidpf opaque, voidpf stream); -typedef int (ZCALLBACK *testerror_file_func) (voidpf opaque, voidpf stream); +typedef voidpf (ZCALLBACK *open_file_func) OF((voidpf opaque, const char* filename, int mode)); +typedef uLong (ZCALLBACK *read_file_func) OF((voidpf opaque, voidpf stream, void* buf, uLong size)); +typedef uLong (ZCALLBACK *write_file_func) OF((voidpf opaque, voidpf stream, const void* buf, uLong size)); +typedef int (ZCALLBACK *close_file_func) OF((voidpf opaque, voidpf stream)); +typedef int (ZCALLBACK *testerror_file_func) OF((voidpf opaque, voidpf stream)); -typedef long (ZCALLBACK *tell_file_func) (voidpf opaque, voidpf stream); -typedef long (ZCALLBACK *seek_file_func) (voidpf opaque, voidpf stream, uLong offset, int origin); +typedef long (ZCALLBACK *tell_file_func) OF((voidpf opaque, voidpf stream)); +typedef long (ZCALLBACK *seek_file_func) OF((voidpf opaque, voidpf stream, uLong offset, int origin)); /* here is the "old" 32 bits structure structure */ @@ -143,13 +161,15 @@ typedef struct zlib_filefunc_def_s close_file_func zclose_file; testerror_file_func zerror_file; voidpf opaque; - alloc_func alloc_mem; - free_func free_mem; + /* GODOT start */ + alloc_func alloc_mem; + free_func free_mem; + /* GODOT end */ } zlib_filefunc_def; -typedef ZPOS64_T (ZCALLBACK *tell64_file_func) (voidpf opaque, voidpf stream); -typedef long (ZCALLBACK *seek64_file_func) (voidpf opaque, voidpf stream, ZPOS64_T offset, int origin); -typedef voidpf (ZCALLBACK *open64_file_func) (voidpf opaque, const void* filename, int mode); +typedef ZPOS64_T (ZCALLBACK *tell64_file_func) OF((voidpf opaque, voidpf stream)); +typedef long (ZCALLBACK *seek64_file_func) OF((voidpf opaque, voidpf stream, ZPOS64_T offset, int origin)); +typedef voidpf (ZCALLBACK *open64_file_func) OF((voidpf opaque, const void* filename, int mode)); typedef struct zlib_filefunc64_def_s { @@ -161,13 +181,14 @@ typedef struct zlib_filefunc64_def_s close_file_func zclose_file; testerror_file_func zerror_file; voidpf opaque; - alloc_func alloc_mem; - free_func free_mem; - + /* GODOT start */ + alloc_func alloc_mem; + free_func free_mem; + /* GODOT end */ } zlib_filefunc64_def; -void fill_fopen64_filefunc (zlib_filefunc64_def* pzlib_filefunc_def); -void fill_fopen_filefunc (zlib_filefunc_def* pzlib_filefunc_def); +void fill_fopen64_filefunc OF((zlib_filefunc64_def* pzlib_filefunc_def)); +void fill_fopen_filefunc OF((zlib_filefunc_def* pzlib_filefunc_def)); /* now internal definition, only for zip.c and unzip.h */ typedef struct zlib_filefunc64_32_def_s @@ -186,9 +207,9 @@ typedef struct zlib_filefunc64_32_def_s #define ZCLOSE64(filefunc,filestream) ((*((filefunc).zfile_func64.zclose_file)) ((filefunc).zfile_func64.opaque,filestream)) #define ZERROR64(filefunc,filestream) ((*((filefunc).zfile_func64.zerror_file)) ((filefunc).zfile_func64.opaque,filestream)) -voidpf call_zopen64 (const zlib_filefunc64_32_def* pfilefunc,const void*filename,int mode); -long call_zseek64 (const zlib_filefunc64_32_def* pfilefunc,voidpf filestream, ZPOS64_T offset, int origin); -ZPOS64_T call_ztell64 (const zlib_filefunc64_32_def* pfilefunc,voidpf filestream); +voidpf call_zopen64 OF((const zlib_filefunc64_32_def* pfilefunc,const void*filename,int mode)); +long call_zseek64 OF((const zlib_filefunc64_32_def* pfilefunc,voidpf filestream, ZPOS64_T offset, int origin)); +ZPOS64_T call_ztell64 OF((const zlib_filefunc64_32_def* pfilefunc,voidpf filestream)); void fill_zlib_filefunc64_32_def_from_filefunc32(zlib_filefunc64_32_def* p_filefunc64_32,const zlib_filefunc_def* p_filefunc32); diff --git a/thirdparty/minizip/unzip.c b/thirdparty/minizip/unzip.c index 7aa0a86d13..32e27bd657 100644 --- a/thirdparty/minizip/unzip.c +++ b/thirdparty/minizip/unzip.c @@ -10,7 +10,7 @@ Modifications for Zip64 support on both zip and unzip Copyright (C) 2009-2010 Mathias Svensson ( http://result42.com ) - For more info read LICENSE-MiniZip.txt + For more info read MiniZip_info.txt ------------------------------------------------------------------------------------ @@ -157,7 +157,9 @@ typedef struct uLong compression_method; /* compression method (0==store) */ ZPOS64_T byte_before_the_zipfile;/* byte before the zipfile, (>0 for sfx)*/ int raw; - int extra_size; + /* GODOT start */ + int extra_size; + /* GODOT end */ } file_in_zip64_read_info_s; @@ -205,10 +207,10 @@ typedef struct */ -local int unz64local_getByte ( +local int unz64local_getByte OF(( const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, - int *pi); + int *pi)); local int unz64local_getByte(const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, int *pi) { @@ -232,10 +234,10 @@ local int unz64local_getByte(const zlib_filefunc64_32_def* pzlib_filefunc_def, v /* =========================================================================== Reads a long in LSB order from the given gz_stream. Sets */ -local int unz64local_getShort ( +local int unz64local_getShort OF(( const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, - uLong *pX); + uLong *pX)); local int unz64local_getShort (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, @@ -259,10 +261,10 @@ local int unz64local_getShort (const zlib_filefunc64_32_def* pzlib_filefunc_def, return err; } -local int unz64local_getLong ( +local int unz64local_getLong OF(( const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, - uLong *pX); + uLong *pX)); local int unz64local_getLong (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, @@ -294,10 +296,10 @@ local int unz64local_getLong (const zlib_filefunc64_32_def* pzlib_filefunc_def, return err; } -local int unz64local_getLong64 ( +local int unz64local_getLong64 OF(( const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, - ZPOS64_T *pX); + ZPOS64_T *pX)); local int unz64local_getLong64 (const zlib_filefunc64_32_def* pzlib_filefunc_def, @@ -410,7 +412,7 @@ extern int ZEXPORT unzStringFileNameCompare (const char* fileName1, Locate the Central directory of a zipfile (at the end, just before the global comment) */ -local ZPOS64_T unz64local_SearchCentralDir (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream); +local ZPOS64_T unz64local_SearchCentralDir OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream)); local ZPOS64_T unz64local_SearchCentralDir(const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream) { unsigned char* buf; @@ -472,9 +474,9 @@ local ZPOS64_T unz64local_SearchCentralDir(const zlib_filefunc64_32_def* pzlib_f Locate the Central directory 64 of a zipfile (at the end, just before the global comment) */ -local ZPOS64_T unz64local_SearchCentralDir64 ( +local ZPOS64_T unz64local_SearchCentralDir64 OF(( const zlib_filefunc64_32_def* pzlib_filefunc_def, - voidpf filestream); + voidpf filestream)); local ZPOS64_T unz64local_SearchCentralDir64(const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream) @@ -606,9 +608,11 @@ local unzFile unzOpenInternal (const void *path, us.z_filefunc.zseek32_file = NULL; us.z_filefunc.ztell32_file = NULL; - if (pzlib_filefunc64_32_def==NULL) - return NULL; // standard i/o not supported - us.z_filefunc = *pzlib_filefunc64_32_def; + if (pzlib_filefunc64_32_def==NULL) + /* GODOT start */ + return NULL; // standard i/o not supported + us.z_filefunc = *pzlib_filefunc64_32_def; + /* GODOT end */ us.is64bitOpenFunction = is64bitOpenFunction; @@ -617,10 +621,8 @@ local unzFile unzOpenInternal (const void *path, path, ZLIB_FILEFUNC_MODE_READ | ZLIB_FILEFUNC_MODE_EXISTING); - if (us.filestream==NULL) { - printf("no stream\n"); + if (us.filestream==NULL) return NULL; - }; central_pos = unz64local_SearchCentralDir64(&us.z_filefunc,us.filestream); if (central_pos) @@ -743,7 +745,6 @@ local unzFile unzOpenInternal (const void *path, if (err!=UNZ_OK) { - printf("err is %i, %x\n", err, err); ZCLOSE64(us.z_filefunc, us.filestream); return NULL; } @@ -803,15 +804,17 @@ extern unzFile ZEXPORT unzOpen64 (const void *path) return unzOpenInternal(path, NULL, 1); } +/* GODOT start */ extern void* unzGetOpaque(unzFile file) { - unz64_s* s; - if (file==NULL) - return NULL; - s=(unz64_s*)file; + unz64_s* s; + if (file==NULL) + return NULL; + s=(unz64_s*)file; - return s->z_filefunc.zfile_func64.opaque; + return s->z_filefunc.zfile_func64.opaque; }; +/* GODOT end */ /* Close a ZipFile opened with unzipOpen. @@ -878,7 +881,7 @@ local void unz64local_DosDateToTmuDate (ZPOS64_T ulDosDate, tm_unz* ptm) /* Get Info about the current file in the zipfile, with internal only info */ -local int unz64local_GetCurrentFileInfoInternal (unzFile file, +local int unz64local_GetCurrentFileInfoInternal OF((unzFile file, unz_file_info64 *pfile_info, unz_file_info64_internal *pfile_info_internal, @@ -887,7 +890,7 @@ local int unz64local_GetCurrentFileInfoInternal (unzFile file, void *extraField, uLong extraFieldBufferSize, char *szComment, - uLong commentBufferSize); + uLong commentBufferSize)); local int unz64local_GetCurrentFileInfoInternal (unzFile file, unz_file_info64 *pfile_info, @@ -1031,19 +1034,20 @@ local int unz64local_GetCurrentFileInfoInternal (unzFile file, if (lSeek!=0) { - if (lSeek<0) { - // WORKAROUND for backwards seeking - z_off_t pos = ZTELL64(s->z_filefunc, s->filestream); - if (ZSEEK64(s->z_filefunc, s->filestream,pos+lSeek,ZLIB_FILEFUNC_SEEK_SET)==0) - lSeek=0; - else - err=UNZ_ERRNO; - } else { - if (ZSEEK64(s->z_filefunc, s->filestream,lSeek,ZLIB_FILEFUNC_SEEK_CUR)==0) - lSeek=0; - else - err=UNZ_ERRNO; - } + /* GODOT start */ + if (lSeek<0) { + // WORKAROUND for backwards seeking + z_off_t pos = ZTELL64(s->z_filefunc, s->filestream); + if (ZSEEK64(s->z_filefunc, s->filestream,pos+lSeek,ZLIB_FILEFUNC_SEEK_SET)==0) + lSeek=0; + else + err=UNZ_ERRNO; + } else { + if (ZSEEK64(s->z_filefunc, s->filestream,lSeek,ZLIB_FILEFUNC_SEEK_CUR)==0) + lSeek=0; + else + err=UNZ_ERRNO; + } } while(acc < file_info.size_file_extra) @@ -1597,8 +1601,10 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, int* method, } else if ((s->cur_file_info.compression_method==Z_DEFLATED) && (!raw)) { + /* GODOT start */ pfile_in_zip_read_info->stream.zalloc = s->z_filefunc.zfile_func64.alloc_mem; pfile_in_zip_read_info->stream.zfree = s->z_filefunc.zfile_func64.free_mem; + /* GODOT end */ pfile_in_zip_read_info->stream.opaque = (voidpf)0; pfile_in_zip_read_info->stream.next_in = 0; pfile_in_zip_read_info->stream.avail_in = 0; @@ -1608,7 +1614,6 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, int* method, pfile_in_zip_read_info->stream_initialised=Z_DEFLATED; else { - printf("NO OPEN ZLIB %i\n",err); TRYFREE(pfile_in_zip_read_info); return err; } @@ -1631,7 +1636,9 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, int* method, iSizeVar; pfile_in_zip_read_info->stream.avail_in = (uInt)0; - pfile_in_zip_read_info->extra_size = iSizeVar; + /* GODOT start */ + pfile_in_zip_read_info->extra_size = iSizeVar; + /* GODOT end */ s->pfile_in_zip_read = pfile_in_zip_read_info; s->encrypted = 0; @@ -1662,82 +1669,84 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, int* method, return UNZ_OK; } +/* GODOT start */ extern int ZEXPORT unzSeekCurrentFile(unzFile file, int pos) { - unz64_s* s; - file_in_zip64_read_info_s* pfile_in_zip_read_info; - if (file==NULL) - return UNZ_PARAMERROR; - s=(unz64_s*)file; - pfile_in_zip_read_info=s->pfile_in_zip_read; - - if (pfile_in_zip_read_info==NULL) - return UNZ_PARAMERROR; - - if (pfile_in_zip_read_info->compression_method==Z_BZIP2ED) { // don't know how to support bzip - return UNZ_INTERNALERROR; - }; - - if ((pfile_in_zip_read_info->compression_method==0) || (pfile_in_zip_read_info->raw)) { - - pfile_in_zip_read_info->rest_read_compressed = - s->cur_file_info.compressed_size - pos; - pfile_in_zip_read_info->rest_read_uncompressed = - s->cur_file_info.uncompressed_size - pos; - - pfile_in_zip_read_info->pos_in_zipfile = - s->cur_file_info_internal.offset_curfile + SIZEZIPLOCALHEADER + - pfile_in_zip_read_info->extra_size + pos; - - pfile_in_zip_read_info->stream.avail_in = (uInt)0; - pfile_in_zip_read_info->stream.total_out = pos; - - return ZSEEK64(pfile_in_zip_read_info->z_filefunc, - pfile_in_zip_read_info->filestream, - pfile_in_zip_read_info->byte_before_the_zipfile + pfile_in_zip_read_info->pos_in_zipfile, - ZLIB_FILEFUNC_SEEK_SET); - - } else { // gzip - - if (pos < pfile_in_zip_read_info->stream.total_out) { // negative seek, rewind - - pfile_in_zip_read_info->rest_read_compressed = - s->cur_file_info.compressed_size ; - pfile_in_zip_read_info->rest_read_uncompressed = - s->cur_file_info.uncompressed_size ; - - pfile_in_zip_read_info->pos_in_zipfile = - s->cur_file_info_internal.offset_curfile + SIZEZIPLOCALHEADER + - pfile_in_zip_read_info->extra_size; - - (void)inflateReset(&pfile_in_zip_read_info->stream); - - pfile_in_zip_read_info->stream.avail_in = (uInt)0; - pfile_in_zip_read_info->stream.total_out = 0; - pfile_in_zip_read_info->stream.next_in = 0; - }; - - // not sure where to read, so read on the stack - { - char buf[512]; - int to_read = pos - pfile_in_zip_read_info->stream.total_out; - while (to_read) { - - int len = to_read > sizeof(buf)?sizeof(buf):to_read; - int read = unzReadCurrentFile(file, buf, len); - if (read < 0) { - return read; - }; - to_read -= read; - if (read == UNZ_EOF) { - return pos; - }; - }; - }; - }; - - return pos; + unz64_s* s; + file_in_zip64_read_info_s* pfile_in_zip_read_info; + if (file==NULL) + return UNZ_PARAMERROR; + s=(unz64_s*)file; + pfile_in_zip_read_info=s->pfile_in_zip_read; + + if (pfile_in_zip_read_info==NULL) + return UNZ_PARAMERROR; + + if (pfile_in_zip_read_info->compression_method==Z_BZIP2ED) { // don't know how to support bzip + return UNZ_INTERNALERROR; + }; + + if ((pfile_in_zip_read_info->compression_method==0) || (pfile_in_zip_read_info->raw)) { + + pfile_in_zip_read_info->rest_read_compressed = + s->cur_file_info.compressed_size - pos; + pfile_in_zip_read_info->rest_read_uncompressed = + s->cur_file_info.uncompressed_size - pos; + + pfile_in_zip_read_info->pos_in_zipfile = + s->cur_file_info_internal.offset_curfile + SIZEZIPLOCALHEADER + + pfile_in_zip_read_info->extra_size + pos; + + pfile_in_zip_read_info->stream.avail_in = (uInt)0; + pfile_in_zip_read_info->stream.total_out = pos; + + return ZSEEK64(pfile_in_zip_read_info->z_filefunc, + pfile_in_zip_read_info->filestream, + pfile_in_zip_read_info->byte_before_the_zipfile + pfile_in_zip_read_info->pos_in_zipfile, + ZLIB_FILEFUNC_SEEK_SET); + + } else { // gzip + + if (pos < pfile_in_zip_read_info->stream.total_out) { // negative seek, rewind + + pfile_in_zip_read_info->rest_read_compressed = + s->cur_file_info.compressed_size ; + pfile_in_zip_read_info->rest_read_uncompressed = + s->cur_file_info.uncompressed_size ; + + pfile_in_zip_read_info->pos_in_zipfile = + s->cur_file_info_internal.offset_curfile + SIZEZIPLOCALHEADER + + pfile_in_zip_read_info->extra_size; + + (void)inflateReset(&pfile_in_zip_read_info->stream); + + pfile_in_zip_read_info->stream.avail_in = (uInt)0; + pfile_in_zip_read_info->stream.total_out = 0; + pfile_in_zip_read_info->stream.next_in = 0; + }; + + // not sure where to read, so read on the stack + { + char buf[512]; + int to_read = pos - pfile_in_zip_read_info->stream.total_out; + while (to_read) { + + int len = to_read > sizeof(buf)?sizeof(buf):to_read; + int read = unzReadCurrentFile(file, buf, len); + if (read < 0) { + return read; + }; + to_read -= read; + if (read == UNZ_EOF) { + return pos; + }; + }; + }; + }; + + return pos; }; +/* GODOT end */ extern int ZEXPORT unzOpenCurrentFile (unzFile file) { @@ -1797,7 +1806,7 @@ extern int ZEXPORT unzReadCurrentFile (unzFile file, voidp buf, unsigned len) return UNZ_PARAMERROR; - if (pfile_in_zip_read_info->read_buffer==NULL) + if ((pfile_in_zip_read_info->read_buffer == NULL)) return UNZ_END_OF_LIST_OF_FILE; if (len==0) return 0; diff --git a/thirdparty/minizip/unzip.h b/thirdparty/minizip/unzip.h index f67c3b2fa8..54e65ad8ab 100644 --- a/thirdparty/minizip/unzip.h +++ b/thirdparty/minizip/unzip.h @@ -10,7 +10,7 @@ Modifications for Zip64 support on both zip and unzip Copyright (C) 2009-2010 Mathias Svensson ( http://result42.com ) - For more info read LICENSE-MiniZip.txt + For more info read MiniZip_info.txt --------------------------------------------------------------------------------- @@ -150,9 +150,9 @@ typedef struct unz_file_info_s tm_unz tmu_date; } unz_file_info; -extern int ZEXPORT unzStringFileNameCompare (const char* fileName1, +extern int ZEXPORT unzStringFileNameCompare OF ((const char* fileName1, const char* fileName2, - int iCaseSensitivity); + int iCaseSensitivity)); /* Compare two filename (fileName1,fileName2). If iCaseSenisivity = 1, comparision is case sensitivity (like strcmp) @@ -163,8 +163,8 @@ extern int ZEXPORT unzStringFileNameCompare (const char* fileName1, */ -extern unzFile ZEXPORT unzOpen (const char *path); -extern unzFile ZEXPORT unzOpen64 (const void *path); +extern unzFile ZEXPORT unzOpen OF((const char *path)); +extern unzFile ZEXPORT unzOpen64 OF((const void *path)); /* Open a Zip file. path contain the full pathname (by example, on a Windows XP computer "c:\\zlib\\zlib113.zip" or on an Unix computer @@ -181,44 +181,45 @@ extern unzFile ZEXPORT unzOpen64 (const void *path); */ -extern unzFile ZEXPORT unzOpen2 (const char *path, - zlib_filefunc_def* pzlib_filefunc_def); +extern unzFile ZEXPORT unzOpen2 OF((const char *path, + zlib_filefunc_def* pzlib_filefunc_def)); /* Open a Zip file, like unzOpen, but provide a set of file low level API for read/write the zip file (see ioapi.h) */ -extern unzFile ZEXPORT unzOpen2_64 (const void *path, - zlib_filefunc64_def* pzlib_filefunc_def); +extern unzFile ZEXPORT unzOpen2_64 OF((const void *path, + zlib_filefunc64_def* pzlib_filefunc_def)); /* Open a Zip file, like unz64Open, but provide a set of file low level API for read/write the zip file (see ioapi.h) */ -extern int ZEXPORT unzClose (unzFile file); +extern int ZEXPORT unzClose OF((unzFile file)); /* Close a ZipFile opened with unzipOpen. If there is files inside the .Zip opened with unzOpenCurrentFile (see later), these files MUST be closed with unzipCloseCurrentFile before call unzipClose. return UNZ_OK if there is no problem. */ +/* GODOT start */ extern void* unzGetOpaque(unzFile file); +/* GODOT end */ +extern int ZEXPORT unzGetGlobalInfo OF((unzFile file, + unz_global_info *pglobal_info)); -extern int ZEXPORT unzGetGlobalInfo (unzFile file, - unz_global_info *pglobal_info); - -extern int ZEXPORT unzGetGlobalInfo64 (unzFile file, - unz_global_info64 *pglobal_info); +extern int ZEXPORT unzGetGlobalInfo64 OF((unzFile file, + unz_global_info64 *pglobal_info)); /* Write info about the ZipFile in the *pglobal_info structure. No preparation of the structure is needed return UNZ_OK if there is no problem. */ -extern int ZEXPORT unzGetGlobalComment (unzFile file, +extern int ZEXPORT unzGetGlobalComment OF((unzFile file, char *szComment, - uLong uSizeBuf); + uLong uSizeBuf)); /* Get the global comment string of the ZipFile, in the szComment buffer. uSizeBuf is the size of the szComment buffer. @@ -229,22 +230,22 @@ extern int ZEXPORT unzGetGlobalComment (unzFile file, /***************************************************************************/ /* Unzip package allow you browse the directory of the zipfile */ -extern int ZEXPORT unzGoToFirstFile (unzFile file); +extern int ZEXPORT unzGoToFirstFile OF((unzFile file)); /* Set the current file of the zipfile to the first file. return UNZ_OK if there is no problem */ -extern int ZEXPORT unzGoToNextFile (unzFile file); +extern int ZEXPORT unzGoToNextFile OF((unzFile file)); /* Set the current file of the zipfile to the next file. return UNZ_OK if there is no problem return UNZ_END_OF_LIST_OF_FILE if the actual file was the latest. */ -extern int ZEXPORT unzLocateFile (unzFile file, +extern int ZEXPORT unzLocateFile OF((unzFile file, const char *szFileName, - int iCaseSensitivity); + int iCaseSensitivity)); /* Try locate the file szFileName in the zipfile. For the iCaseSensitivity signification, see unzStringFileNameCompare @@ -288,23 +289,23 @@ extern int ZEXPORT unzGoToFilePos64( /* ****************************************** */ -extern int ZEXPORT unzGetCurrentFileInfo64 (unzFile file, +extern int ZEXPORT unzGetCurrentFileInfo64 OF((unzFile file, unz_file_info64 *pfile_info, char *szFileName, uLong fileNameBufferSize, void *extraField, uLong extraFieldBufferSize, char *szComment, - uLong commentBufferSize); + uLong commentBufferSize)); -extern int ZEXPORT unzGetCurrentFileInfo (unzFile file, +extern int ZEXPORT unzGetCurrentFileInfo OF((unzFile file, unz_file_info *pfile_info, char *szFileName, uLong fileNameBufferSize, void *extraField, uLong extraFieldBufferSize, char *szComment, - uLong commentBufferSize); + uLong commentBufferSize)); /* Get Info about the current file if pfile_info!=NULL, the *pfile_info structure will contain somes info about @@ -321,7 +322,7 @@ extern int ZEXPORT unzGetCurrentFileInfo (unzFile file, /** Addition for GDAL : START */ -extern ZPOS64_T ZEXPORT unzGetCurrentFileZStreamPos64 (unzFile file); +extern ZPOS64_T ZEXPORT unzGetCurrentFileZStreamPos64 OF((unzFile file)); /** Addition for GDAL : END */ @@ -331,24 +332,24 @@ extern ZPOS64_T ZEXPORT unzGetCurrentFileZStreamPos64 (unzFile file); from it, and close it (you can close it before reading all the file) */ -extern int ZEXPORT unzOpenCurrentFile (unzFile file); +extern int ZEXPORT unzOpenCurrentFile OF((unzFile file)); /* Open for reading data the current file in the zipfile. If there is no error, the return value is UNZ_OK. */ -extern int ZEXPORT unzOpenCurrentFilePassword (unzFile file, - const char* password); +extern int ZEXPORT unzOpenCurrentFilePassword OF((unzFile file, + const char* password)); /* Open for reading data the current file in the zipfile. password is a crypting password If there is no error, the return value is UNZ_OK. */ -extern int ZEXPORT unzOpenCurrentFile2 (unzFile file, +extern int ZEXPORT unzOpenCurrentFile2 OF((unzFile file, int* method, int* level, - int raw); + int raw)); /* Same than unzOpenCurrentFile, but open for read raw the file (not uncompress) if raw==1 @@ -358,11 +359,11 @@ extern int ZEXPORT unzOpenCurrentFile2 (unzFile file, but you CANNOT set method parameter as NULL */ -extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, +extern int ZEXPORT unzOpenCurrentFile3 OF((unzFile file, int* method, int* level, int raw, - const char* password); + const char* password)); /* Same than unzOpenCurrentFile, but open for read raw the file (not uncompress) if raw==1 @@ -373,15 +374,15 @@ extern int ZEXPORT unzOpenCurrentFile3 (unzFile file, */ -extern int ZEXPORT unzCloseCurrentFile (unzFile file); +extern int ZEXPORT unzCloseCurrentFile OF((unzFile file)); /* Close the file in zip opened with unzOpenCurrentFile Return UNZ_CRCERROR if all the file was read but the CRC is not good */ -extern int ZEXPORT unzReadCurrentFile (unzFile file, +extern int ZEXPORT unzReadCurrentFile OF((unzFile file, voidp buf, - unsigned len); + unsigned len)); /* Read bytes from the current file (opened by unzOpenCurrentFile) buf contain buffer where data must be copied @@ -393,26 +394,28 @@ extern int ZEXPORT unzReadCurrentFile (unzFile file, (UNZ_ERRNO for IO error, or zLib error for uncompress error) */ +/* GODOT start */ extern int ZEXPORT unzSeekCurrentFile(unzFile file, int pos); /* Seek to position in uncompressed data */ +/* GODOT end */ -extern z_off_t ZEXPORT unztell (unzFile file); +extern z_off_t ZEXPORT unztell OF((unzFile file)); -extern ZPOS64_T ZEXPORT unztell64 (unzFile file); +extern ZPOS64_T ZEXPORT unztell64 OF((unzFile file)); /* Give the current position in uncompressed data */ -extern int ZEXPORT unzeof (unzFile file); +extern int ZEXPORT unzeof OF((unzFile file)); /* return 1 if the end of file was reached, 0 elsewhere */ -extern int ZEXPORT unzGetLocalExtrafield (unzFile file, +extern int ZEXPORT unzGetLocalExtrafield OF((unzFile file, voidp buf, - unsigned len); + unsigned len)); /* Read extra field from the current file (opened by unzOpenCurrentFile) This is the local-header version of the extra field (sometimes, there is diff --git a/thirdparty/minizip/zip.c b/thirdparty/minizip/zip.c index 27a3d3cdc1..d7093e7457 100644 --- a/thirdparty/minizip/zip.c +++ b/thirdparty/minizip/zip.c @@ -7,7 +7,7 @@ Modifications for Zip64 support Copyright (C) 2009-2010 Mathias Svensson ( http://result42.com ) - For more info read LICENSE-MiniZip.txt + For more info read MiniZip_info.txt Changes Oct-2009 - Mathias Svensson - Remove old C style function prototypes @@ -283,7 +283,7 @@ local int add_data_in_datablock(linkedlist_data* ll, const void* buf, uLong len) nbByte == 1, 2 ,4 or 8 (byte, short or long, ZPOS64_T) */ -local int zip64local_putValue (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, ZPOS64_T x, int nbByte); +local int zip64local_putValue OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, ZPOS64_T x, int nbByte)); local int zip64local_putValue (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, ZPOS64_T x, int nbByte) { unsigned char buf[8]; @@ -307,7 +307,7 @@ local int zip64local_putValue (const zlib_filefunc64_32_def* pzlib_filefunc_def, return ZIP_OK; } -local void zip64local_putValue_inmemory (void* dest, ZPOS64_T x, int nbByte); +local void zip64local_putValue_inmemory OF((void* dest, ZPOS64_T x, int nbByte)); local void zip64local_putValue_inmemory (void* dest, ZPOS64_T x, int nbByte) { unsigned char* buf=(unsigned char*)dest; @@ -344,7 +344,7 @@ local uLong zip64local_TmzDateToDosDate(const tm_zip* ptm) /****************************************************************************/ -local int zip64local_getByte (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, int *pi); +local int zip64local_getByte OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, int *pi)); local int zip64local_getByte(const zlib_filefunc64_32_def* pzlib_filefunc_def,voidpf filestream,int* pi) { @@ -368,7 +368,7 @@ local int zip64local_getByte(const zlib_filefunc64_32_def* pzlib_filefunc_def,vo /* =========================================================================== Reads a long in LSB order from the given gz_stream. Sets */ -local int zip64local_getShort (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, uLong *pX); +local int zip64local_getShort OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, uLong *pX)); local int zip64local_getShort (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, uLong* pX) { @@ -390,7 +390,7 @@ local int zip64local_getShort (const zlib_filefunc64_32_def* pzlib_filefunc_def, return err; } -local int zip64local_getLong (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, uLong *pX); +local int zip64local_getLong OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, uLong *pX)); local int zip64local_getLong (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, uLong* pX) { @@ -420,7 +420,7 @@ local int zip64local_getLong (const zlib_filefunc64_32_def* pzlib_filefunc_def, return err; } -local int zip64local_getLong64 (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, ZPOS64_T *pX); +local int zip64local_getLong64 OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, ZPOS64_T *pX)); local int zip64local_getLong64 (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream, ZPOS64_T *pX) @@ -475,7 +475,7 @@ local int zip64local_getLong64 (const zlib_filefunc64_32_def* pzlib_filefunc_def Locate the Central directory of a zipfile (at the end, just before the global comment) */ -local ZPOS64_T zip64local_SearchCentralDir (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream); +local ZPOS64_T zip64local_SearchCentralDir OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream)); local ZPOS64_T zip64local_SearchCentralDir(const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream) { @@ -537,7 +537,7 @@ local ZPOS64_T zip64local_SearchCentralDir(const zlib_filefunc64_32_def* pzlib_f Locate the End of Zip64 Central directory locator and from there find the CD of a zipfile (at the end, just before the global comment) */ -local ZPOS64_T zip64local_SearchCentralDir64 (const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream); +local ZPOS64_T zip64local_SearchCentralDir64 OF((const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream)); local ZPOS64_T zip64local_SearchCentralDir64(const zlib_filefunc64_32_def* pzlib_filefunc_def, voidpf filestream) { @@ -854,9 +854,11 @@ extern zipFile ZEXPORT zipOpen3 (const void *pathname, int append, zipcharpc* gl ziinit.z_filefunc.zseek32_file = NULL; ziinit.z_filefunc.ztell32_file = NULL; + /* GODOT start */ if (pzlib_filefunc64_32_def==NULL) { - //fill_fopen64_filefunc(&ziinit.z_filefunc.zfile_func64); + //fill_fopen64_filefunc(&ziinit.z_filefunc.zfile_func64); } else + /* GODOT end */ ziinit.z_filefunc = *pzlib_filefunc64_32_def; ziinit.filestream = ZOPEN64(ziinit.z_filefunc, @@ -1114,9 +1116,9 @@ extern int ZEXPORT zipOpenNewFileInZip4_64 (zipFile file, const char* filename, zi->ci.flag = flagBase; if ((level==8) || (level==9)) zi->ci.flag |= 2; - if (level==2) + if ((level==2)) zi->ci.flag |= 4; - if (level==1) + if ((level==1)) zi->ci.flag |= 6; if (password != NULL) zi->ci.flag |= 1; @@ -1210,9 +1212,10 @@ extern int ZEXPORT zipOpenNewFileInZip4_64 (zipFile file, const char* filename, { if(zi->ci.method == Z_DEFLATED) { - zi->ci.stream.zalloc = zi->z_filefunc.zfile_func64.alloc_mem; - zi->ci.stream.zfree = zi->z_filefunc.zfile_func64.free_mem; - + /* GODOT start */ + zi->ci.stream.zalloc = zi->z_filefunc.zfile_func64.alloc_mem; + zi->ci.stream.zfree = zi->z_filefunc.zfile_func64.free_mem; + /* GODOT end */ zi->ci.stream.opaque = (voidpf)0; if (windowBits>0) diff --git a/thirdparty/minizip/zip.h b/thirdparty/minizip/zip.h index 37478b34c0..8aaebb6234 100644 --- a/thirdparty/minizip/zip.h +++ b/thirdparty/minizip/zip.h @@ -1,4 +1,5 @@ -/* Version 1.1, February 14h, 2010 +/* zip.h -- IO on .zip files using zlib + Version 1.1, February 14h, 2010 part of the MiniZip project - ( http://www.winimage.com/zLibDll/minizip.html ) Copyright (C) 1998-2010 Gilles Vollant (minizip) ( http://www.winimage.com/zLibDll/minizip.html ) @@ -6,7 +7,7 @@ Modifications for Zip64 support Copyright (C) 2009-2010 Mathias Svensson ( http://result42.com ) - For more info read LICENSE-MiniZip.txt + For more info read MiniZip_info.txt --------------------------------------------------------------------------- @@ -39,8 +40,6 @@ #ifndef _zip12_H #define _zip12_H -#include <stdlib.h> - #ifdef __cplusplus extern "C" { #endif @@ -114,8 +113,8 @@ typedef const char* zipcharpc; #define APPEND_STATUS_CREATEAFTER (1) #define APPEND_STATUS_ADDINZIP (2) -extern zipFile ZEXPORT zipOpen (const char *pathname, int append); -extern zipFile ZEXPORT zipOpen64 (const void *pathname, int append); +extern zipFile ZEXPORT zipOpen OF((const char *pathname, int append)); +extern zipFile ZEXPORT zipOpen64 OF((const void *pathname, int append)); /* Create a zipfile. pathname contain on Windows XP a filename like "c:\\zlib\\zlib113.zip" or on @@ -135,17 +134,17 @@ extern zipFile ZEXPORT zipOpen64 (const void *pathname, int append); Of couse, you can use RAW reading and writing to copy the file you did not want delte */ -extern zipFile ZEXPORT zipOpen2 (const char *pathname, +extern zipFile ZEXPORT zipOpen2 OF((const char *pathname, int append, zipcharpc* globalcomment, - zlib_filefunc_def* pzlib_filefunc_def); + zlib_filefunc_def* pzlib_filefunc_def)); -extern zipFile ZEXPORT zipOpen2_64 (const void *pathname, +extern zipFile ZEXPORT zipOpen2_64 OF((const void *pathname, int append, zipcharpc* globalcomment, - zlib_filefunc64_def* pzlib_filefunc_def); + zlib_filefunc64_def* pzlib_filefunc_def)); -extern int ZEXPORT zipOpenNewFileInZip (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -154,9 +153,9 @@ extern int ZEXPORT zipOpenNewFileInZip (zipFile file, uInt size_extrafield_global, const char* comment, int method, - int level); + int level)); -extern int ZEXPORT zipOpenNewFileInZip64 (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip64 OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -166,7 +165,7 @@ extern int ZEXPORT zipOpenNewFileInZip64 (zipFile file, const char* comment, int method, int level, - int zip64); + int zip64)); /* Open a file in the ZIP for writing. @@ -185,7 +184,7 @@ extern int ZEXPORT zipOpenNewFileInZip64 (zipFile file, */ -extern int ZEXPORT zipOpenNewFileInZip2 (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip2 OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -195,10 +194,10 @@ extern int ZEXPORT zipOpenNewFileInZip2 (zipFile file, const char* comment, int method, int level, - int raw); + int raw)); -extern int ZEXPORT zipOpenNewFileInZip2_64 (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip2_64 OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -209,12 +208,12 @@ extern int ZEXPORT zipOpenNewFileInZip2_64 (zipFile file, int method, int level, int raw, - int zip64); + int zip64)); /* Same than zipOpenNewFileInZip, except if raw=1, we write raw file */ -extern int ZEXPORT zipOpenNewFileInZip3 (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip3 OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -229,9 +228,9 @@ extern int ZEXPORT zipOpenNewFileInZip3 (zipFile file, int memLevel, int strategy, const char* password, - uLong crcForCrypting); + uLong crcForCrypting)); -extern int ZEXPORT zipOpenNewFileInZip3_64 (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip3_64 OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -248,7 +247,7 @@ extern int ZEXPORT zipOpenNewFileInZip3_64 (zipFile file, const char* password, uLong crcForCrypting, int zip64 - ); + )); /* Same than zipOpenNewFileInZip2, except @@ -257,7 +256,7 @@ extern int ZEXPORT zipOpenNewFileInZip3_64 (zipFile file, crcForCrypting : crc of file to compress (needed for crypting) */ -extern int ZEXPORT zipOpenNewFileInZip4 (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip4 OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -275,10 +274,10 @@ extern int ZEXPORT zipOpenNewFileInZip4 (zipFile file, uLong crcForCrypting, uLong versionMadeBy, uLong flagBase - ); + )); -extern int ZEXPORT zipOpenNewFileInZip4_64 (zipFile file, +extern int ZEXPORT zipOpenNewFileInZip4_64 OF((zipFile file, const char* filename, const zip_fileinfo* zipfi, const void* extrafield_local, @@ -297,7 +296,7 @@ extern int ZEXPORT zipOpenNewFileInZip4_64 (zipFile file, uLong versionMadeBy, uLong flagBase, int zip64 - ); + )); /* Same than zipOpenNewFileInZip4, except versionMadeBy : value for Version made by field @@ -305,25 +304,25 @@ extern int ZEXPORT zipOpenNewFileInZip4_64 (zipFile file, */ -extern int ZEXPORT zipWriteInFileInZip (zipFile file, +extern int ZEXPORT zipWriteInFileInZip OF((zipFile file, const void* buf, - unsigned len); + unsigned len)); /* Write data in the zipfile */ -extern int ZEXPORT zipCloseFileInZip (zipFile file); +extern int ZEXPORT zipCloseFileInZip OF((zipFile file)); /* Close the current file in the zipfile */ -extern int ZEXPORT zipCloseFileInZipRaw (zipFile file, +extern int ZEXPORT zipCloseFileInZipRaw OF((zipFile file, uLong uncompressed_size, - uLong crc32); + uLong crc32)); -extern int ZEXPORT zipCloseFileInZipRaw64 (zipFile file, +extern int ZEXPORT zipCloseFileInZipRaw64 OF((zipFile file, ZPOS64_T uncompressed_size, - uLong crc32); + uLong crc32)); /* Close the current file in the zipfile, for file opened with @@ -331,14 +330,14 @@ extern int ZEXPORT zipCloseFileInZipRaw64 (zipFile file, uncompressed_size and crc32 are value for the uncompressed size */ -extern int ZEXPORT zipClose (zipFile file, - const char* global_comment); +extern int ZEXPORT zipClose OF((zipFile file, + const char* global_comment)); /* Close the zipfile */ -extern int ZEXPORT zipRemoveExtraInfoBlock (char* pData, int* dataLen, short sHeader); +extern int ZEXPORT zipRemoveExtraInfoBlock OF((char* pData, int* dataLen, short sHeader)); /* zipRemoveExtraInfoBlock - Added by Mathias Svensson diff --git a/thirdparty/misc/stb_truetype.h b/thirdparty/misc/stb_truetype.h index 016972785a..cec2425471 100644 --- a/thirdparty/misc/stb_truetype.h +++ b/thirdparty/misc/stb_truetype.h @@ -1,11 +1,12 @@ -// stb_truetype.h - v1.11 - public domain -// authored from 2009-2015 by Sean Barrett / RAD Game Tools +// stb_truetype.h - v1.17 - public domain +// authored from 2009-2016 by Sean Barrett / RAD Game Tools // // This library processes TrueType files: // parse files // extract glyph metrics // extract glyph shapes // render glyphs to one-channel bitmaps with antialiasing (box filter) +// render glyphs to one-channel SDF bitmaps (signed-distance field/function) // // Todo: // non-MS cmaps @@ -20,16 +21,19 @@ // // Mikko Mononen: compound shape support, more cmap formats // Tor Andersson: kerning, subpixel rendering +// Dougall Johnson: OpenType / Type 2 font handling // // Misc other: // Ryan Gordon // Simon Glass +// github:IntellectualKitty +// Imanol Celaya // // Bug/warning reports/fixes: -// "Zer" on mollyrocket (with fix) +// "Zer" on mollyrocket // Cass Everitt // stoiko (Haemimont Games) -// Brian Hook +// Brian Hook // Walter van Niftrik // David Gow // David Given @@ -48,9 +52,17 @@ // Higor Euripedes // Thomas Fields // Derek Vinyard +// Cort Stratton +// github:oyvindjam // // VERSION HISTORY // +// 1.17 (2017-07-23) make more arguments const; doc fix +// 1.16 (2017-07-12) SDF support +// 1.15 (2017-03-03) make more arguments const +// 1.14 (2017-01-16) num-fonts-in-TTC function +// 1.13 (2017-01-02) support OpenType fonts, certain Apple fonts +// 1.12 (2016-10-25) suppress warnings about casting away const with -Wcast-qual // 1.11 (2016-04-02) fix unused-variable warning // 1.10 (2016-04-02) user-defined fabs(); rare memory leak; remove duplicate typedef // 1.09 (2016-01-16) warning fix; avoid crash on outofmem; use allocation userdata properly @@ -60,20 +72,12 @@ // fix stbtt_GetFontOFfsetForIndex (never worked for non-0 input?); // fixed an assert() bug in the new rasterizer // replace assert() with STBTT_assert() in new rasterizer -// 1.06 (2015-07-14) performance improvements (~35% faster on x86 and x64 on test machine) -// also more precise AA rasterizer, except if shapes overlap -// remove need for STBTT_sort -// 1.05 (2015-04-15) fix misplaced definitions for STBTT_STATIC -// 1.04 (2015-04-15) typo in example -// 1.03 (2015-04-12) STBTT_STATIC, fix memory leak in new packing, various fixes // // Full history can be found at the end of this file. // // LICENSE // -// This software is dual-licensed to the public domain and under the following -// license: you are granted a perpetual, irrevocable license to copy, modify, -// publish, and distribute this file as you see fit. +// See end of file for license information. // // USAGE // @@ -93,14 +97,15 @@ // Improved 3D API (more shippable): // #include "stb_rect_pack.h" -- optional, but you really want it // stbtt_PackBegin() -// stbtt_PackSetOversample() -- for improved quality on small fonts +// stbtt_PackSetOversampling() -- for improved quality on small fonts // stbtt_PackFontRanges() -- pack and renders // stbtt_PackEnd() // stbtt_GetPackedQuad() // // "Load" a font file from a memory buffer (you have to keep the buffer loaded) // stbtt_InitFont() -// stbtt_GetFontOffsetForIndex() -- use for TTC font collections +// stbtt_GetFontOffsetForIndex() -- indexing for TTC font collections +// stbtt_GetNumberOfFonts() -- number of fonts for TTC font collections // // Render a unicode codepoint to a bitmap // stbtt_GetCodepointBitmap() -- allocates and returns a bitmap @@ -110,6 +115,7 @@ // Character advance/positioning // stbtt_GetCodepointHMetrics() // stbtt_GetFontVMetrics() +// stbtt_GetFontVMetricsOS2() // stbtt_GetCodepointKernAdvance() // // Starting with version 1.06, the rasterizer was replaced with a new, @@ -218,7 +224,7 @@ // Curve tesselation 120 LOC \__ 550 LOC Bitmap creation // Bitmap management 100 LOC / // Baked bitmap interface 70 LOC / -// Font name matching & access 150 LOC ---- 150 +// Font name matching & access 150 LOC ---- 150 // C runtime library abstraction 60 LOC ---- 60 // // @@ -311,7 +317,7 @@ int main(int argc, char **argv) } return 0; } -#endif +#endif // // Output: // @@ -325,9 +331,9 @@ int main(int argc, char **argv) // :@@. M@M // @@@o@@@@ // :M@@V:@@. -// +// ////////////////////////////////////////////////////////////////////////////// -// +// // Complete program: print "Hello World!" banner, with bugs // #if 0 @@ -407,6 +413,18 @@ int main(int arg, char **argv) #ifndef STBTT_sqrt #include <math.h> #define STBTT_sqrt(x) sqrt(x) + #define STBTT_pow(x,y) pow(x,y) + #endif + + #ifndef STBTT_cos + #include <math.h> + #define STBTT_cos(x) cos(x) + #define STBTT_acos(x) acos(x) + #endif + + #ifndef STBTT_fabs + #include <math.h> + #define STBTT_fabs(x) fabs(x) #endif #ifndef STBTT_fabs @@ -432,7 +450,7 @@ int main(int arg, char **argv) #endif #ifndef STBTT_memcpy - #include <memory.h> + #include <string.h> #define STBTT_memcpy memcpy #define STBTT_memset memset #endif @@ -458,6 +476,14 @@ int main(int arg, char **argv) extern "C" { #endif +// private structure +typedef struct +{ + unsigned char *data; + int cursor; + int size; +} stbtt__buf; + ////////////////////////////////////////////////////////////////////////////// // // TEXTURE BAKING API @@ -487,7 +513,7 @@ typedef struct float x1,y1,s1,t1; // bottom-right } stbtt_aligned_quad; -STBTT_DEF void stbtt_GetBakedQuad(stbtt_bakedchar *chardata, int pw, int ph, // same data as above +STBTT_DEF void stbtt_GetBakedQuad(const stbtt_bakedchar *chardata, int pw, int ph, // same data as above int char_index, // character to display float *xpos, float *ypos, // pointers to current position in screen pixel space stbtt_aligned_quad *q, // output: quad to draw @@ -527,7 +553,7 @@ typedef struct stbrp_rect stbrp_rect; STBTT_DEF int stbtt_PackBegin(stbtt_pack_context *spc, unsigned char *pixels, int width, int height, int stride_in_bytes, int padding, void *alloc_context); // Initializes a packing context stored in the passed-in stbtt_pack_context. // Future calls using this context will pack characters into the bitmap passed -// in here: a 1-channel bitmap that is weight x height. stride_in_bytes is +// in here: a 1-channel bitmap that is width * height. stride_in_bytes is // the distance from one row to the next (or 0 to mean they are packed tightly // together). "padding" is the amount of padding to leave between each // character (normally you want '1' for bitmaps you'll use as textures with @@ -540,7 +566,7 @@ STBTT_DEF void stbtt_PackEnd (stbtt_pack_context *spc); #define STBTT_POINT_SIZE(x) (-(x)) -STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, unsigned char *fontdata, int font_index, float font_size, +STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, float font_size, int first_unicode_char_in_range, int num_chars_in_range, stbtt_packedchar *chardata_for_range); // Creates character bitmaps from the font_index'th font found in fontdata (use // font_index=0 if you don't know what that is). It creates num_chars_in_range @@ -565,7 +591,7 @@ typedef struct unsigned char h_oversample, v_oversample; // don't set these, they're used internally } stbtt_pack_range; -STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, unsigned char *fontdata, int font_index, stbtt_pack_range *ranges, int num_ranges); +STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, stbtt_pack_range *ranges, int num_ranges); // Creates character bitmaps from multiple ranges of characters stored in // ranges. This will usually create a better-packed bitmap than multiple // calls to stbtt_PackFontRange. Note that you can call this multiple @@ -587,19 +613,19 @@ STBTT_DEF void stbtt_PackSetOversampling(stbtt_pack_context *spc, unsigned int h // To use with PackFontRangesGather etc., you must set it before calls // call to PackFontRangesGatherRects. -STBTT_DEF void stbtt_GetPackedQuad(stbtt_packedchar *chardata, int pw, int ph, // same data as above +STBTT_DEF void stbtt_GetPackedQuad(const stbtt_packedchar *chardata, int pw, int ph, // same data as above int char_index, // character to display float *xpos, float *ypos, // pointers to current position in screen pixel space stbtt_aligned_quad *q, // output: quad to draw int align_to_integer); -STBTT_DEF int stbtt_PackFontRangesGatherRects(stbtt_pack_context *spc, stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects); +STBTT_DEF int stbtt_PackFontRangesGatherRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects); STBTT_DEF void stbtt_PackFontRangesPackRects(stbtt_pack_context *spc, stbrp_rect *rects, int num_rects); -STBTT_DEF int stbtt_PackFontRangesRenderIntoRects(stbtt_pack_context *spc, stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects); +STBTT_DEF int stbtt_PackFontRangesRenderIntoRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects); // Calling these functions in sequence is roughly equivalent to calling // stbtt_PackFontRanges(). If you more control over the packing of multiple // fonts, or if you want to pack custom data into a font texture, take a look -// at the source to of stbtt_PackFontRanges() and create a custom version +// at the source to of stbtt_PackFontRanges() and create a custom version // using these functions, e.g. call GatherRects multiple times, // building up a single array of rects, then call PackRects once, // then call RenderIntoRects repeatedly. This may result in a @@ -626,14 +652,19 @@ struct stbtt_pack_context { // // +STBTT_DEF int stbtt_GetNumberOfFonts(const unsigned char *data); +// This function will determine the number of fonts in a font file. TrueType +// collection (.ttc) files may contain multiple fonts, while TrueType font +// (.ttf) files only contain one font. The number of fonts can be used for +// indexing with the previous function where the index is between zero and one +// less than the total fonts. If an error occurs, -1 is returned. + STBTT_DEF int stbtt_GetFontOffsetForIndex(const unsigned char *data, int index); // Each .ttf/.ttc file may have more than one font. Each font has a sequential // index number starting from 0. Call this function to get the font offset for // a given index; it returns -1 if the index is out of range. A regular .ttf // file will only define one font and it always be at offset 0, so it will -// return '0' for index 0, and -1 for all other indices. You can just skip -// this step if you know it's that kind of font. - +// return '0' for index 0, and -1 for all other indices. // The following structure is defined publically so you can declare one on // the stack or as a global or etc, but you should treat it as opaque. @@ -648,6 +679,13 @@ struct stbtt_fontinfo int loca,head,glyf,hhea,hmtx,kern; // table locations as offset from start of .ttf int index_map; // a cmap mapping for our chosen character encoding int indexToLocFormat; // format needed to map from glyph index to glyph + + stbtt__buf cff; // cff font data + stbtt__buf charstrings; // the charstring index + stbtt__buf gsubrs; // global charstring subroutines index + stbtt__buf subrs; // private charstring subroutines index + stbtt__buf fontdicts; // array of font dicts + stbtt__buf fdselect; // map from glyph to fontdict }; STBTT_DEF int stbtt_InitFont(stbtt_fontinfo *info, const unsigned char *data, int offset); @@ -695,6 +733,12 @@ STBTT_DEF void stbtt_GetFontVMetrics(const stbtt_fontinfo *info, int *ascent, in // these are expressed in unscaled coordinates, so you must multiply by // the scale factor for a given size +STBTT_DEF int stbtt_GetFontVMetricsOS2(const stbtt_fontinfo *info, int *typoAscent, int *typoDescent, int *typoLineGap); +// analogous to GetFontVMetrics, but returns the "typographic" values from the OS/2 +// table (specific to MS/Windows TTF files). +// +// Returns 1 on success (table present), 0 on failure. + STBTT_DEF void stbtt_GetFontBoundingBox(const stbtt_fontinfo *info, int *x0, int *y0, int *x1, int *y1); // the bounding box around all possible characters @@ -725,7 +769,8 @@ STBTT_DEF int stbtt_GetGlyphBox(const stbtt_fontinfo *info, int glyph_index, in enum { STBTT_vmove=1, STBTT_vline, - STBTT_vcurve + STBTT_vcurve, + STBTT_vcubic }; #endif @@ -734,7 +779,7 @@ STBTT_DEF int stbtt_GetGlyphBox(const stbtt_fontinfo *info, int glyph_index, in #define stbtt_vertex_type short // can't use stbtt_int16 because that's not visible in the header file typedef struct { - stbtt_vertex_type x,y,cx,cy; + stbtt_vertex_type x,y,cx,cy,cx1,cy1; unsigned char type,padding; } stbtt_vertex; #endif @@ -788,6 +833,10 @@ STBTT_DEF void stbtt_MakeCodepointBitmapSubpixel(const stbtt_fontinfo *info, uns // same as stbtt_MakeCodepointBitmap, but you can specify a subpixel // shift for the character +STBTT_DEF void stbtt_MakeCodepointBitmapSubpixelPrefilter(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int oversample_x, int oversample_y, float *sub_x, float *sub_y, int codepoint); +// same as stbtt_MakeCodepointBitmapSubpixel, but prefiltering +// is performed (see stbtt_PackSetOversampling) + STBTT_DEF void stbtt_GetCodepointBitmapBox(const stbtt_fontinfo *font, int codepoint, float scale_x, float scale_y, int *ix0, int *iy0, int *ix1, int *iy1); // get the bbox of the bitmap centered around the glyph origin; so the // bitmap width is ix1-ix0, height is iy1-iy0, and location to place @@ -805,6 +854,7 @@ STBTT_DEF unsigned char *stbtt_GetGlyphBitmap(const stbtt_fontinfo *info, float STBTT_DEF unsigned char *stbtt_GetGlyphBitmapSubpixel(const stbtt_fontinfo *info, float scale_x, float scale_y, float shift_x, float shift_y, int glyph, int *width, int *height, int *xoff, int *yoff); STBTT_DEF void stbtt_MakeGlyphBitmap(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, int glyph); STBTT_DEF void stbtt_MakeGlyphBitmapSubpixel(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int glyph); +STBTT_DEF void stbtt_MakeGlyphBitmapSubpixelPrefilter(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int oversample_x, int oversample_y, float *sub_x, float *sub_y, int glyph); STBTT_DEF void stbtt_GetGlyphBitmapBox(const stbtt_fontinfo *font, int glyph, float scale_x, float scale_y, int *ix0, int *iy0, int *ix1, int *iy1); STBTT_DEF void stbtt_GetGlyphBitmapBoxSubpixel(const stbtt_fontinfo *font, int glyph, float scale_x, float scale_y,float shift_x, float shift_y, int *ix0, int *iy0, int *ix1, int *iy1); @@ -829,6 +879,64 @@ STBTT_DEF void stbtt_Rasterize(stbtt__bitmap *result, // 1-channel bitmap ////////////////////////////////////////////////////////////////////////////// // +// Signed Distance Function (or Field) rendering + +STBTT_DEF void stbtt_FreeSDF(unsigned char *bitmap, void *userdata); +// frees the SDF bitmap allocated below + +STBTT_DEF unsigned char * stbtt_GetGlyphSDF(const stbtt_fontinfo *info, float scale, int glyph, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff); +STBTT_DEF unsigned char * stbtt_GetCodepointSDF(const stbtt_fontinfo *info, float scale, int codepoint, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff); +// These functions compute a discretized SDF field for a single character, suitable for storing +// in a single-channel texture, sampling with bilinear filtering, and testing against +// larger than some threshhold to produce scalable fonts. +// info -- the font +// scale -- controls the size of the resulting SDF bitmap, same as it would be creating a regular bitmap +// glyph/codepoint -- the character to generate the SDF for +// padding -- extra "pixels" around the character which are filled with the distance to the character (not 0), +// which allows effects like bit outlines +// onedge_value -- value 0-255 to test the SDF against to reconstruct the character (i.e. the isocontour of the character) +// pixel_dist_scale -- what value the SDF should increase by when moving one SDF "pixel" away from the edge (on the 0..255 scale) +// if positive, > onedge_value is inside; if negative, < onedge_value is inside +// width,height -- output height & width of the SDF bitmap (including padding) +// xoff,yoff -- output origin of the character +// return value -- a 2D array of bytes 0..255, width*height in size +// +// pixel_dist_scale & onedge_value are a scale & bias that allows you to make +// optimal use of the limited 0..255 for your application, trading off precision +// and special effects. SDF values outside the range 0..255 are clamped to 0..255. +// +// Example: +// scale = stbtt_ScaleForPixelHeight(22) +// padding = 5 +// onedge_value = 180 +// pixel_dist_scale = 180/5.0 = 36.0 +// +// This will create an SDF bitmap in which the character is about 22 pixels +// high but the whole bitmap is about 22+5+5=32 pixels high. To produce a filled +// shape, sample the SDF at each pixel and fill the pixel if the SDF value +// is greater than or equal to 180/255. (You'll actually want to antialias, +// which is beyond the scope of this example.) Additionally, you can compute +// offset outlines (e.g. to stroke the character border inside & outside, +// or only outside). For example, to fill outside the character up to 3 SDF +// pixels, you would compare against (180-36.0*3)/255 = 72/255. The above +// choice of variables maps a range from 5 pixels outside the shape to +// 2 pixels inside the shape to 0..255; this is intended primarily for apply +// outside effects only (the interior range is needed to allow proper +// antialiasing of the font at *smaller* sizes) +// +// The function computes the SDF analytically at each SDF pixel, not by e.g. +// building a higher-res bitmap and approximating it. In theory the quality +// should be as high as possible for an SDF of this size & representation, but +// unclear if this is true in practice (perhaps building a higher-res bitmap +// and computing from that can allow drop-out prevention). +// +// The algorithm has not been optimized at all, so expect it to be slow +// if computing lots of characters or very large sizes. + + + +////////////////////////////////////////////////////////////////////////////// +// // Finding the right font... // // You should really just solve this offline, keep your own tables @@ -958,6 +1066,152 @@ typedef int stbtt__test_oversample_pow2[(STBTT_MAX_OVERSAMPLE & (STBTT_MAX_OVERS ////////////////////////////////////////////////////////////////////////// // +// stbtt__buf helpers to parse data from file +// + +static stbtt_uint8 stbtt__buf_get8(stbtt__buf *b) +{ + if (b->cursor >= b->size) + return 0; + return b->data[b->cursor++]; +} + +static stbtt_uint8 stbtt__buf_peek8(stbtt__buf *b) +{ + if (b->cursor >= b->size) + return 0; + return b->data[b->cursor]; +} + +static void stbtt__buf_seek(stbtt__buf *b, int o) +{ + STBTT_assert(!(o > b->size || o < 0)); + b->cursor = (o > b->size || o < 0) ? b->size : o; +} + +static void stbtt__buf_skip(stbtt__buf *b, int o) +{ + stbtt__buf_seek(b, b->cursor + o); +} + +static stbtt_uint32 stbtt__buf_get(stbtt__buf *b, int n) +{ + stbtt_uint32 v = 0; + int i; + STBTT_assert(n >= 1 && n <= 4); + for (i = 0; i < n; i++) + v = (v << 8) | stbtt__buf_get8(b); + return v; +} + +static stbtt__buf stbtt__new_buf(const void *p, size_t size) +{ + stbtt__buf r; + STBTT_assert(size < 0x40000000); + r.data = (stbtt_uint8*) p; + r.size = (int) size; + r.cursor = 0; + return r; +} + +#define stbtt__buf_get16(b) stbtt__buf_get((b), 2) +#define stbtt__buf_get32(b) stbtt__buf_get((b), 4) + +static stbtt__buf stbtt__buf_range(const stbtt__buf *b, int o, int s) +{ + stbtt__buf r = stbtt__new_buf(NULL, 0); + if (o < 0 || s < 0 || o > b->size || s > b->size - o) return r; + r.data = b->data + o; + r.size = s; + return r; +} + +static stbtt__buf stbtt__cff_get_index(stbtt__buf *b) +{ + int count, start, offsize; + start = b->cursor; + count = stbtt__buf_get16(b); + if (count) { + offsize = stbtt__buf_get8(b); + STBTT_assert(offsize >= 1 && offsize <= 4); + stbtt__buf_skip(b, offsize * count); + stbtt__buf_skip(b, stbtt__buf_get(b, offsize) - 1); + } + return stbtt__buf_range(b, start, b->cursor - start); +} + +static stbtt_uint32 stbtt__cff_int(stbtt__buf *b) +{ + int b0 = stbtt__buf_get8(b); + if (b0 >= 32 && b0 <= 246) return b0 - 139; + else if (b0 >= 247 && b0 <= 250) return (b0 - 247)*256 + stbtt__buf_get8(b) + 108; + else if (b0 >= 251 && b0 <= 254) return -(b0 - 251)*256 - stbtt__buf_get8(b) - 108; + else if (b0 == 28) return stbtt__buf_get16(b); + else if (b0 == 29) return stbtt__buf_get32(b); + STBTT_assert(0); + return 0; +} + +static void stbtt__cff_skip_operand(stbtt__buf *b) { + int v, b0 = stbtt__buf_peek8(b); + STBTT_assert(b0 >= 28); + if (b0 == 30) { + stbtt__buf_skip(b, 1); + while (b->cursor < b->size) { + v = stbtt__buf_get8(b); + if ((v & 0xF) == 0xF || (v >> 4) == 0xF) + break; + } + } else { + stbtt__cff_int(b); + } +} + +static stbtt__buf stbtt__dict_get(stbtt__buf *b, int key) +{ + stbtt__buf_seek(b, 0); + while (b->cursor < b->size) { + int start = b->cursor, end, op; + while (stbtt__buf_peek8(b) >= 28) + stbtt__cff_skip_operand(b); + end = b->cursor; + op = stbtt__buf_get8(b); + if (op == 12) op = stbtt__buf_get8(b) | 0x100; + if (op == key) return stbtt__buf_range(b, start, end-start); + } + return stbtt__buf_range(b, 0, 0); +} + +static void stbtt__dict_get_ints(stbtt__buf *b, int key, int outcount, stbtt_uint32 *out) +{ + int i; + stbtt__buf operands = stbtt__dict_get(b, key); + for (i = 0; i < outcount && operands.cursor < operands.size; i++) + out[i] = stbtt__cff_int(&operands); +} + +static int stbtt__cff_index_count(stbtt__buf *b) +{ + stbtt__buf_seek(b, 0); + return stbtt__buf_get16(b); +} + +static stbtt__buf stbtt__cff_index_get(stbtt__buf b, int i) +{ + int count, offsize, start, end; + stbtt__buf_seek(&b, 0); + count = stbtt__buf_get16(&b); + offsize = stbtt__buf_get8(&b); + STBTT_assert(i >= 0 && i < count); + STBTT_assert(offsize >= 1 && offsize <= 4); + stbtt__buf_skip(&b, i*offsize); + start = stbtt__buf_get(&b, offsize); + end = stbtt__buf_get(&b, offsize); + return stbtt__buf_range(&b, 2+(count+1)*offsize+start, end - start); +} + +////////////////////////////////////////////////////////////////////////// +// // accessors to parse data from file // @@ -968,32 +1222,22 @@ typedef int stbtt__test_oversample_pow2[(STBTT_MAX_OVERSAMPLE & (STBTT_MAX_OVERS #define ttCHAR(p) (* (stbtt_int8 *) (p)) #define ttFixed(p) ttLONG(p) -#if defined(STB_TRUETYPE_BIGENDIAN) && !defined(ALLOW_UNALIGNED_TRUETYPE) - - #define ttUSHORT(p) (* (stbtt_uint16 *) (p)) - #define ttSHORT(p) (* (stbtt_int16 *) (p)) - #define ttULONG(p) (* (stbtt_uint32 *) (p)) - #define ttLONG(p) (* (stbtt_int32 *) (p)) - -#else - - static stbtt_uint16 ttUSHORT(const stbtt_uint8 *p) { return p[0]*256 + p[1]; } - static stbtt_int16 ttSHORT(const stbtt_uint8 *p) { return p[0]*256 + p[1]; } - static stbtt_uint32 ttULONG(const stbtt_uint8 *p) { return (p[0]<<24) + (p[1]<<16) + (p[2]<<8) + p[3]; } - static stbtt_int32 ttLONG(const stbtt_uint8 *p) { return (p[0]<<24) + (p[1]<<16) + (p[2]<<8) + p[3]; } - -#endif +static stbtt_uint16 ttUSHORT(stbtt_uint8 *p) { return p[0]*256 + p[1]; } +static stbtt_int16 ttSHORT(stbtt_uint8 *p) { return p[0]*256 + p[1]; } +static stbtt_uint32 ttULONG(stbtt_uint8 *p) { return (p[0]<<24) + (p[1]<<16) + (p[2]<<8) + p[3]; } +static stbtt_int32 ttLONG(stbtt_uint8 *p) { return (p[0]<<24) + (p[1]<<16) + (p[2]<<8) + p[3]; } #define stbtt_tag4(p,c0,c1,c2,c3) ((p)[0] == (c0) && (p)[1] == (c1) && (p)[2] == (c2) && (p)[3] == (c3)) #define stbtt_tag(p,str) stbtt_tag4(p,str[0],str[1],str[2],str[3]) -static int stbtt__isfont(const stbtt_uint8 *font) +static int stbtt__isfont(stbtt_uint8 *font) { // check the version number if (stbtt_tag4(font, '1',0,0,0)) return 1; // TrueType 1 if (stbtt_tag(font, "typ1")) return 1; // TrueType with type 1 font -- we don't support this! if (stbtt_tag(font, "OTTO")) return 1; // OpenType with CFF if (stbtt_tag4(font, 0,1,0,0)) return 1; // OpenType 1.0 + if (stbtt_tag(font, "true")) return 1; // Apple specification for TrueType fonts return 0; } @@ -1011,7 +1255,7 @@ static stbtt_uint32 stbtt__find_table(stbtt_uint8 *data, stbtt_uint32 fontstart, return 0; } -STBTT_DEF int stbtt_GetFontOffsetForIndex(const unsigned char *font_collection, int index) +static int stbtt_GetFontOffsetForIndex_internal(unsigned char *font_collection, int index) { // if it's just a font, there's only one valid index if (stbtt__isfont(font_collection)) @@ -1030,14 +1274,43 @@ STBTT_DEF int stbtt_GetFontOffsetForIndex(const unsigned char *font_collection, return -1; } -STBTT_DEF int stbtt_InitFont(stbtt_fontinfo *info, const unsigned char *data2, int fontstart) +static int stbtt_GetNumberOfFonts_internal(unsigned char *font_collection) +{ + // if it's just a font, there's only one valid font + if (stbtt__isfont(font_collection)) + return 1; + + // check if it's a TTC + if (stbtt_tag(font_collection, "ttcf")) { + // version 1? + if (ttULONG(font_collection+4) == 0x00010000 || ttULONG(font_collection+4) == 0x00020000) { + return ttLONG(font_collection+8); + } + } + return 0; +} + +static stbtt__buf stbtt__get_subrs(stbtt__buf cff, stbtt__buf fontdict) +{ + stbtt_uint32 subrsoff = 0, private_loc[2] = { 0, 0 }; + stbtt__buf pdict; + stbtt__dict_get_ints(&fontdict, 18, 2, private_loc); + if (!private_loc[1] || !private_loc[0]) return stbtt__new_buf(NULL, 0); + pdict = stbtt__buf_range(&cff, private_loc[1], private_loc[0]); + stbtt__dict_get_ints(&pdict, 19, 1, &subrsoff); + if (!subrsoff) return stbtt__new_buf(NULL, 0); + stbtt__buf_seek(&cff, private_loc[1]+subrsoff); + return stbtt__cff_get_index(&cff); +} + +static int stbtt_InitFont_internal(stbtt_fontinfo *info, unsigned char *data, int fontstart) { - stbtt_uint8 *data = (stbtt_uint8 *) data2; stbtt_uint32 cmap, t; stbtt_int32 i,numTables; info->data = data; info->fontstart = fontstart; + info->cff = stbtt__new_buf(NULL, 0); cmap = stbtt__find_table(data, fontstart, "cmap"); // required info->loca = stbtt__find_table(data, fontstart, "loca"); // required @@ -1046,8 +1319,61 @@ STBTT_DEF int stbtt_InitFont(stbtt_fontinfo *info, const unsigned char *data2, i info->hhea = stbtt__find_table(data, fontstart, "hhea"); // required info->hmtx = stbtt__find_table(data, fontstart, "hmtx"); // required info->kern = stbtt__find_table(data, fontstart, "kern"); // not required - if (!cmap || !info->loca || !info->head || !info->glyf || !info->hhea || !info->hmtx) + + if (!cmap || !info->head || !info->hhea || !info->hmtx) return 0; + if (info->glyf) { + // required for truetype + if (!info->loca) return 0; + } else { + // initialization for CFF / Type2 fonts (OTF) + stbtt__buf b, topdict, topdictidx; + stbtt_uint32 cstype = 2, charstrings = 0, fdarrayoff = 0, fdselectoff = 0; + stbtt_uint32 cff; + + cff = stbtt__find_table(data, fontstart, "CFF "); + if (!cff) return 0; + + info->fontdicts = stbtt__new_buf(NULL, 0); + info->fdselect = stbtt__new_buf(NULL, 0); + + // @TODO this should use size from table (not 512MB) + info->cff = stbtt__new_buf(data+cff, 512*1024*1024); + b = info->cff; + + // read the header + stbtt__buf_skip(&b, 2); + stbtt__buf_seek(&b, stbtt__buf_get8(&b)); // hdrsize + + // @TODO the name INDEX could list multiple fonts, + // but we just use the first one. + stbtt__cff_get_index(&b); // name INDEX + topdictidx = stbtt__cff_get_index(&b); + topdict = stbtt__cff_index_get(topdictidx, 0); + stbtt__cff_get_index(&b); // string INDEX + info->gsubrs = stbtt__cff_get_index(&b); + + stbtt__dict_get_ints(&topdict, 17, 1, &charstrings); + stbtt__dict_get_ints(&topdict, 0x100 | 6, 1, &cstype); + stbtt__dict_get_ints(&topdict, 0x100 | 36, 1, &fdarrayoff); + stbtt__dict_get_ints(&topdict, 0x100 | 37, 1, &fdselectoff); + info->subrs = stbtt__get_subrs(b, topdict); + + // we only support Type 2 charstrings + if (cstype != 2) return 0; + if (charstrings == 0) return 0; + + if (fdarrayoff) { + // looks like a CID font + if (!fdselectoff) return 0; + stbtt__buf_seek(&b, fdarrayoff); + info->fontdicts = stbtt__cff_get_index(&b); + info->fdselect = stbtt__buf_range(&b, fdselectoff, b.size-fdselectoff); + } + + stbtt__buf_seek(&b, charstrings); + info->charstrings = stbtt__cff_get_index(&b); + } t = stbtt__find_table(data, fontstart, "maxp"); if (t) @@ -1198,6 +1524,8 @@ static int stbtt__GetGlyfOffset(const stbtt_fontinfo *info, int glyph_index) { int g1,g2; + STBTT_assert(!info->cff.size); + if (glyph_index >= info->numGlyphs) return -1; // glyph index out of range if (info->indexToLocFormat >= 2) return -1; // unknown index->glyph map format @@ -1212,15 +1540,21 @@ static int stbtt__GetGlyfOffset(const stbtt_fontinfo *info, int glyph_index) return g1==g2 ? -1 : g1; // if length is 0, return -1 } +static int stbtt__GetGlyphInfoT2(const stbtt_fontinfo *info, int glyph_index, int *x0, int *y0, int *x1, int *y1); + STBTT_DEF int stbtt_GetGlyphBox(const stbtt_fontinfo *info, int glyph_index, int *x0, int *y0, int *x1, int *y1) { - int g = stbtt__GetGlyfOffset(info, glyph_index); - if (g < 0) return 0; + if (info->cff.size) { + stbtt__GetGlyphInfoT2(info, glyph_index, x0, y0, x1, y1); + } else { + int g = stbtt__GetGlyfOffset(info, glyph_index); + if (g < 0) return 0; - if (x0) *x0 = ttSHORT(info->data + g + 2); - if (y0) *y0 = ttSHORT(info->data + g + 4); - if (x1) *x1 = ttSHORT(info->data + g + 6); - if (y1) *y1 = ttSHORT(info->data + g + 8); + if (x0) *x0 = ttSHORT(info->data + g + 2); + if (y0) *y0 = ttSHORT(info->data + g + 4); + if (x1) *x1 = ttSHORT(info->data + g + 6); + if (y1) *y1 = ttSHORT(info->data + g + 8); + } return 1; } @@ -1232,7 +1566,10 @@ STBTT_DEF int stbtt_GetCodepointBox(const stbtt_fontinfo *info, int codepoint, i STBTT_DEF int stbtt_IsGlyphEmpty(const stbtt_fontinfo *info, int glyph_index) { stbtt_int16 numberOfContours; - int g = stbtt__GetGlyfOffset(info, glyph_index); + int g; + if (info->cff.size) + return stbtt__GetGlyphInfoT2(info, glyph_index, NULL, NULL, NULL, NULL) == 0; + g = stbtt__GetGlyfOffset(info, glyph_index); if (g < 0) return 1; numberOfContours = ttSHORT(info->data + g); return numberOfContours == 0; @@ -1254,7 +1591,7 @@ static int stbtt__close_shape(stbtt_vertex *vertices, int num_vertices, int was_ return num_vertices; } -STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **pvertices) +static int stbtt__GetGlyphShapeTT(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **pvertices) { stbtt_int16 numberOfContours; stbtt_uint8 *endPtsOfContours; @@ -1350,7 +1687,7 @@ STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, s if (i != 0) num_vertices = stbtt__close_shape(vertices, num_vertices, was_off, start_off, sx,sy,scx,scy,cx,cy); - // now start the new one + // now start the new one start_off = !(flags & 1); if (start_off) { // if we start off with an off-curve point, then when we need to find a point on the curve @@ -1403,7 +1740,7 @@ STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, s int comp_num_verts = 0, i; stbtt_vertex *comp_verts = 0, *tmp = 0; float mtx[6] = {1,0,0,1,0,0}, m, n; - + flags = ttSHORT(comp); comp+=2; gidx = ttSHORT(comp); comp+=2; @@ -1433,7 +1770,7 @@ STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, s mtx[2] = ttSHORT(comp)/16384.0f; comp+=2; mtx[3] = ttSHORT(comp)/16384.0f; comp+=2; } - + // Find transformation scales. m = (float) STBTT_sqrt(mtx[0]*mtx[0] + mtx[1]*mtx[1]); n = (float) STBTT_sqrt(mtx[2]*mtx[2] + mtx[3]*mtx[3]); @@ -1480,6 +1817,416 @@ STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, s return num_vertices; } +typedef struct +{ + int bounds; + int started; + float first_x, first_y; + float x, y; + stbtt_int32 min_x, max_x, min_y, max_y; + + stbtt_vertex *pvertices; + int num_vertices; +} stbtt__csctx; + +#define STBTT__CSCTX_INIT(bounds) {bounds,0, 0,0, 0,0, 0,0,0,0, NULL, 0} + +static void stbtt__track_vertex(stbtt__csctx *c, stbtt_int32 x, stbtt_int32 y) +{ + if (x > c->max_x || !c->started) c->max_x = x; + if (y > c->max_y || !c->started) c->max_y = y; + if (x < c->min_x || !c->started) c->min_x = x; + if (y < c->min_y || !c->started) c->min_y = y; + c->started = 1; +} + +static void stbtt__csctx_v(stbtt__csctx *c, stbtt_uint8 type, stbtt_int32 x, stbtt_int32 y, stbtt_int32 cx, stbtt_int32 cy, stbtt_int32 cx1, stbtt_int32 cy1) +{ + if (c->bounds) { + stbtt__track_vertex(c, x, y); + if (type == STBTT_vcubic) { + stbtt__track_vertex(c, cx, cy); + stbtt__track_vertex(c, cx1, cy1); + } + } else { + stbtt_setvertex(&c->pvertices[c->num_vertices], type, x, y, cx, cy); + c->pvertices[c->num_vertices].cx1 = (stbtt_int16) cx1; + c->pvertices[c->num_vertices].cy1 = (stbtt_int16) cy1; + } + c->num_vertices++; +} + +static void stbtt__csctx_close_shape(stbtt__csctx *ctx) +{ + if (ctx->first_x != ctx->x || ctx->first_y != ctx->y) + stbtt__csctx_v(ctx, STBTT_vline, (int)ctx->first_x, (int)ctx->first_y, 0, 0, 0, 0); +} + +static void stbtt__csctx_rmove_to(stbtt__csctx *ctx, float dx, float dy) +{ + stbtt__csctx_close_shape(ctx); + ctx->first_x = ctx->x = ctx->x + dx; + ctx->first_y = ctx->y = ctx->y + dy; + stbtt__csctx_v(ctx, STBTT_vmove, (int)ctx->x, (int)ctx->y, 0, 0, 0, 0); +} + +static void stbtt__csctx_rline_to(stbtt__csctx *ctx, float dx, float dy) +{ + ctx->x += dx; + ctx->y += dy; + stbtt__csctx_v(ctx, STBTT_vline, (int)ctx->x, (int)ctx->y, 0, 0, 0, 0); +} + +static void stbtt__csctx_rccurve_to(stbtt__csctx *ctx, float dx1, float dy1, float dx2, float dy2, float dx3, float dy3) +{ + float cx1 = ctx->x + dx1; + float cy1 = ctx->y + dy1; + float cx2 = cx1 + dx2; + float cy2 = cy1 + dy2; + ctx->x = cx2 + dx3; + ctx->y = cy2 + dy3; + stbtt__csctx_v(ctx, STBTT_vcubic, (int)ctx->x, (int)ctx->y, (int)cx1, (int)cy1, (int)cx2, (int)cy2); +} + +static stbtt__buf stbtt__get_subr(stbtt__buf idx, int n) +{ + int count = stbtt__cff_index_count(&idx); + int bias = 107; + if (count >= 33900) + bias = 32768; + else if (count >= 1240) + bias = 1131; + n += bias; + if (n < 0 || n >= count) + return stbtt__new_buf(NULL, 0); + return stbtt__cff_index_get(idx, n); +} + +static stbtt__buf stbtt__cid_get_glyph_subrs(const stbtt_fontinfo *info, int glyph_index) +{ + stbtt__buf fdselect = info->fdselect; + int nranges, start, end, v, fmt, fdselector = -1, i; + + stbtt__buf_seek(&fdselect, 0); + fmt = stbtt__buf_get8(&fdselect); + if (fmt == 0) { + // untested + stbtt__buf_skip(&fdselect, glyph_index); + fdselector = stbtt__buf_get8(&fdselect); + } else if (fmt == 3) { + nranges = stbtt__buf_get16(&fdselect); + start = stbtt__buf_get16(&fdselect); + for (i = 0; i < nranges; i++) { + v = stbtt__buf_get8(&fdselect); + end = stbtt__buf_get16(&fdselect); + if (glyph_index >= start && glyph_index < end) { + fdselector = v; + break; + } + start = end; + } + } + if (fdselector == -1) stbtt__new_buf(NULL, 0); + return stbtt__get_subrs(info->cff, stbtt__cff_index_get(info->fontdicts, fdselector)); +} + +static int stbtt__run_charstring(const stbtt_fontinfo *info, int glyph_index, stbtt__csctx *c) +{ + int in_header = 1, maskbits = 0, subr_stack_height = 0, sp = 0, v, i, b0; + int has_subrs = 0, clear_stack; + float s[48]; + stbtt__buf subr_stack[10], subrs = info->subrs, b; + float f; + +#define STBTT__CSERR(s) (0) + + // this currently ignores the initial width value, which isn't needed if we have hmtx + b = stbtt__cff_index_get(info->charstrings, glyph_index); + while (b.cursor < b.size) { + i = 0; + clear_stack = 1; + b0 = stbtt__buf_get8(&b); + switch (b0) { + // @TODO implement hinting + case 0x13: // hintmask + case 0x14: // cntrmask + if (in_header) + maskbits += (sp / 2); // implicit "vstem" + in_header = 0; + stbtt__buf_skip(&b, (maskbits + 7) / 8); + break; + + case 0x01: // hstem + case 0x03: // vstem + case 0x12: // hstemhm + case 0x17: // vstemhm + maskbits += (sp / 2); + break; + + case 0x15: // rmoveto + in_header = 0; + if (sp < 2) return STBTT__CSERR("rmoveto stack"); + stbtt__csctx_rmove_to(c, s[sp-2], s[sp-1]); + break; + case 0x04: // vmoveto + in_header = 0; + if (sp < 1) return STBTT__CSERR("vmoveto stack"); + stbtt__csctx_rmove_to(c, 0, s[sp-1]); + break; + case 0x16: // hmoveto + in_header = 0; + if (sp < 1) return STBTT__CSERR("hmoveto stack"); + stbtt__csctx_rmove_to(c, s[sp-1], 0); + break; + + case 0x05: // rlineto + if (sp < 2) return STBTT__CSERR("rlineto stack"); + for (; i + 1 < sp; i += 2) + stbtt__csctx_rline_to(c, s[i], s[i+1]); + break; + + // hlineto/vlineto and vhcurveto/hvcurveto alternate horizontal and vertical + // starting from a different place. + + case 0x07: // vlineto + if (sp < 1) return STBTT__CSERR("vlineto stack"); + goto vlineto; + case 0x06: // hlineto + if (sp < 1) return STBTT__CSERR("hlineto stack"); + for (;;) { + if (i >= sp) break; + stbtt__csctx_rline_to(c, s[i], 0); + i++; + vlineto: + if (i >= sp) break; + stbtt__csctx_rline_to(c, 0, s[i]); + i++; + } + break; + + case 0x1F: // hvcurveto + if (sp < 4) return STBTT__CSERR("hvcurveto stack"); + goto hvcurveto; + case 0x1E: // vhcurveto + if (sp < 4) return STBTT__CSERR("vhcurveto stack"); + for (;;) { + if (i + 3 >= sp) break; + stbtt__csctx_rccurve_to(c, 0, s[i], s[i+1], s[i+2], s[i+3], (sp - i == 5) ? s[i + 4] : 0.0f); + i += 4; + hvcurveto: + if (i + 3 >= sp) break; + stbtt__csctx_rccurve_to(c, s[i], 0, s[i+1], s[i+2], (sp - i == 5) ? s[i+4] : 0.0f, s[i+3]); + i += 4; + } + break; + + case 0x08: // rrcurveto + if (sp < 6) return STBTT__CSERR("rcurveline stack"); + for (; i + 5 < sp; i += 6) + stbtt__csctx_rccurve_to(c, s[i], s[i+1], s[i+2], s[i+3], s[i+4], s[i+5]); + break; + + case 0x18: // rcurveline + if (sp < 8) return STBTT__CSERR("rcurveline stack"); + for (; i + 5 < sp - 2; i += 6) + stbtt__csctx_rccurve_to(c, s[i], s[i+1], s[i+2], s[i+3], s[i+4], s[i+5]); + if (i + 1 >= sp) return STBTT__CSERR("rcurveline stack"); + stbtt__csctx_rline_to(c, s[i], s[i+1]); + break; + + case 0x19: // rlinecurve + if (sp < 8) return STBTT__CSERR("rlinecurve stack"); + for (; i + 1 < sp - 6; i += 2) + stbtt__csctx_rline_to(c, s[i], s[i+1]); + if (i + 5 >= sp) return STBTT__CSERR("rlinecurve stack"); + stbtt__csctx_rccurve_to(c, s[i], s[i+1], s[i+2], s[i+3], s[i+4], s[i+5]); + break; + + case 0x1A: // vvcurveto + case 0x1B: // hhcurveto + if (sp < 4) return STBTT__CSERR("(vv|hh)curveto stack"); + f = 0.0; + if (sp & 1) { f = s[i]; i++; } + for (; i + 3 < sp; i += 4) { + if (b0 == 0x1B) + stbtt__csctx_rccurve_to(c, s[i], f, s[i+1], s[i+2], s[i+3], 0.0); + else + stbtt__csctx_rccurve_to(c, f, s[i], s[i+1], s[i+2], 0.0, s[i+3]); + f = 0.0; + } + break; + + case 0x0A: // callsubr + if (!has_subrs) { + if (info->fdselect.size) + subrs = stbtt__cid_get_glyph_subrs(info, glyph_index); + has_subrs = 1; + } + // fallthrough + case 0x1D: // callgsubr + if (sp < 1) return STBTT__CSERR("call(g|)subr stack"); + v = (int) s[--sp]; + if (subr_stack_height >= 10) return STBTT__CSERR("recursion limit"); + subr_stack[subr_stack_height++] = b; + b = stbtt__get_subr(b0 == 0x0A ? subrs : info->gsubrs, v); + if (b.size == 0) return STBTT__CSERR("subr not found"); + b.cursor = 0; + clear_stack = 0; + break; + + case 0x0B: // return + if (subr_stack_height <= 0) return STBTT__CSERR("return outside subr"); + b = subr_stack[--subr_stack_height]; + clear_stack = 0; + break; + + case 0x0E: // endchar + stbtt__csctx_close_shape(c); + return 1; + + case 0x0C: { // two-byte escape + float dx1, dx2, dx3, dx4, dx5, dx6, dy1, dy2, dy3, dy4, dy5, dy6; + float dx, dy; + int b1 = stbtt__buf_get8(&b); + switch (b1) { + // @TODO These "flex" implementations ignore the flex-depth and resolution, + // and always draw beziers. + case 0x22: // hflex + if (sp < 7) return STBTT__CSERR("hflex stack"); + dx1 = s[0]; + dx2 = s[1]; + dy2 = s[2]; + dx3 = s[3]; + dx4 = s[4]; + dx5 = s[5]; + dx6 = s[6]; + stbtt__csctx_rccurve_to(c, dx1, 0, dx2, dy2, dx3, 0); + stbtt__csctx_rccurve_to(c, dx4, 0, dx5, -dy2, dx6, 0); + break; + + case 0x23: // flex + if (sp < 13) return STBTT__CSERR("flex stack"); + dx1 = s[0]; + dy1 = s[1]; + dx2 = s[2]; + dy2 = s[3]; + dx3 = s[4]; + dy3 = s[5]; + dx4 = s[6]; + dy4 = s[7]; + dx5 = s[8]; + dy5 = s[9]; + dx6 = s[10]; + dy6 = s[11]; + //fd is s[12] + stbtt__csctx_rccurve_to(c, dx1, dy1, dx2, dy2, dx3, dy3); + stbtt__csctx_rccurve_to(c, dx4, dy4, dx5, dy5, dx6, dy6); + break; + + case 0x24: // hflex1 + if (sp < 9) return STBTT__CSERR("hflex1 stack"); + dx1 = s[0]; + dy1 = s[1]; + dx2 = s[2]; + dy2 = s[3]; + dx3 = s[4]; + dx4 = s[5]; + dx5 = s[6]; + dy5 = s[7]; + dx6 = s[8]; + stbtt__csctx_rccurve_to(c, dx1, dy1, dx2, dy2, dx3, 0); + stbtt__csctx_rccurve_to(c, dx4, 0, dx5, dy5, dx6, -(dy1+dy2+dy5)); + break; + + case 0x25: // flex1 + if (sp < 11) return STBTT__CSERR("flex1 stack"); + dx1 = s[0]; + dy1 = s[1]; + dx2 = s[2]; + dy2 = s[3]; + dx3 = s[4]; + dy3 = s[5]; + dx4 = s[6]; + dy4 = s[7]; + dx5 = s[8]; + dy5 = s[9]; + dx6 = dy6 = s[10]; + dx = dx1+dx2+dx3+dx4+dx5; + dy = dy1+dy2+dy3+dy4+dy5; + if (STBTT_fabs(dx) > STBTT_fabs(dy)) + dy6 = -dy; + else + dx6 = -dx; + stbtt__csctx_rccurve_to(c, dx1, dy1, dx2, dy2, dx3, dy3); + stbtt__csctx_rccurve_to(c, dx4, dy4, dx5, dy5, dx6, dy6); + break; + + default: + return STBTT__CSERR("unimplemented"); + } + } break; + + default: + if (b0 != 255 && b0 != 28 && (b0 < 32 || b0 > 254)) + return STBTT__CSERR("reserved operator"); + + // push immediate + if (b0 == 255) { + f = (float)stbtt__buf_get32(&b) / 0x10000; + } else { + stbtt__buf_skip(&b, -1); + f = (float)(stbtt_int16)stbtt__cff_int(&b); + } + if (sp >= 48) return STBTT__CSERR("push stack overflow"); + s[sp++] = f; + clear_stack = 0; + break; + } + if (clear_stack) sp = 0; + } + return STBTT__CSERR("no endchar"); + +#undef STBTT__CSERR +} + +static int stbtt__GetGlyphShapeT2(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **pvertices) +{ + // runs the charstring twice, once to count and once to output (to avoid realloc) + stbtt__csctx count_ctx = STBTT__CSCTX_INIT(1); + stbtt__csctx output_ctx = STBTT__CSCTX_INIT(0); + if (stbtt__run_charstring(info, glyph_index, &count_ctx)) { + *pvertices = (stbtt_vertex*)STBTT_malloc(count_ctx.num_vertices*sizeof(stbtt_vertex), info->userdata); + output_ctx.pvertices = *pvertices; + if (stbtt__run_charstring(info, glyph_index, &output_ctx)) { + STBTT_assert(output_ctx.num_vertices == count_ctx.num_vertices); + return output_ctx.num_vertices; + } + } + *pvertices = NULL; + return 0; +} + +static int stbtt__GetGlyphInfoT2(const stbtt_fontinfo *info, int glyph_index, int *x0, int *y0, int *x1, int *y1) +{ + stbtt__csctx c = STBTT__CSCTX_INIT(1); + int r = stbtt__run_charstring(info, glyph_index, &c); + if (x0) { + *x0 = r ? c.min_x : 0; + *y0 = r ? c.min_y : 0; + *x1 = r ? c.max_x : 0; + *y1 = r ? c.max_y : 0; + } + return r ? c.num_vertices : 0; +} + +STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **pvertices) +{ + if (!info->cff.size) + return stbtt__GetGlyphShapeTT(info, glyph_index, pvertices); + else + return stbtt__GetGlyphShapeT2(info, glyph_index, pvertices); +} + STBTT_DEF void stbtt_GetGlyphHMetrics(const stbtt_fontinfo *info, int glyph_index, int *advanceWidth, int *leftSideBearing) { stbtt_uint16 numOfLongHorMetrics = ttUSHORT(info->data+info->hhea + 34); @@ -1541,6 +2288,17 @@ STBTT_DEF void stbtt_GetFontVMetrics(const stbtt_fontinfo *info, int *ascent, in if (lineGap) *lineGap = ttSHORT(info->data+info->hhea + 8); } +STBTT_DEF int stbtt_GetFontVMetricsOS2(const stbtt_fontinfo *info, int *typoAscent, int *typoDescent, int *typoLineGap) +{ + int tab = stbtt__find_table(info->data, info->fontstart, "OS/2"); + if (!tab) + return 0; + if (typoAscent ) *typoAscent = ttSHORT(info->data+tab + 68); + if (typoDescent) *typoDescent = ttSHORT(info->data+tab + 70); + if (typoLineGap) *typoLineGap = ttSHORT(info->data+tab + 72); + return 1; +} + STBTT_DEF void stbtt_GetFontBoundingBox(const stbtt_fontinfo *info, int *x0, int *y0, int *x1, int *y1) { *x0 = ttSHORT(info->data + info->head + 36); @@ -1691,7 +2449,7 @@ static stbtt__active_edge *stbtt__new_active(stbtt__hheap *hh, stbtt__edge *e, i float dxdy = (e->x1 - e->x0) / (e->y1 - e->y0); STBTT_assert(z != NULL); if (!z) return z; - + // round dx down to avoid overshooting if (dxdy < 0) z->dx = -STBTT_ifloor(STBTT_FIX * -dxdy); @@ -1769,7 +2527,7 @@ static void stbtt__fill_active_edges(unsigned char *scanline, int len, stbtt__ac } } } - + e = e->next; } } @@ -2033,19 +2791,18 @@ static void stbtt__fill_active_edges_new(float *scanline, float *scanline_fill, // from the other y segment, and it might ignored as an empty segment. to avoid // that, we need to explicitly produce segments based on x positions. - // rename variables to clear pairs + // rename variables to clearly-defined pairs float y0 = y_top; float x1 = (float) (x); float x2 = (float) (x+1); float x3 = xb; float y3 = y_bottom; - float y1,y2; // x = e->x + e->dx * (y-y_top) // (y-y_top) = (x - e->x) / e->dx // y = (x - e->x) / e->dx + y_top - y1 = (x - x0) / dx + y_top; - y2 = (x+1 - x0) / dx + y_top; + float y1 = (x - x0) / dx + y_top; + float y2 = (x+1 - x0) / dx + y_top; if (x0 < x1 && x3 > x2) { // three segments descending down-right stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x1,y1); @@ -2350,6 +3107,48 @@ static int stbtt__tesselate_curve(stbtt__point *points, int *num_points, float x return 1; } +static void stbtt__tesselate_cubic(stbtt__point *points, int *num_points, float x0, float y0, float x1, float y1, float x2, float y2, float x3, float y3, float objspace_flatness_squared, int n) +{ + // @TODO this "flatness" calculation is just made-up nonsense that seems to work well enough + float dx0 = x1-x0; + float dy0 = y1-y0; + float dx1 = x2-x1; + float dy1 = y2-y1; + float dx2 = x3-x2; + float dy2 = y3-y2; + float dx = x3-x0; + float dy = y3-y0; + float longlen = (float) (STBTT_sqrt(dx0*dx0+dy0*dy0)+STBTT_sqrt(dx1*dx1+dy1*dy1)+STBTT_sqrt(dx2*dx2+dy2*dy2)); + float shortlen = (float) STBTT_sqrt(dx*dx+dy*dy); + float flatness_squared = longlen*longlen-shortlen*shortlen; + + if (n > 16) // 65536 segments on one curve better be enough! + return; + + if (flatness_squared > objspace_flatness_squared) { + float x01 = (x0+x1)/2; + float y01 = (y0+y1)/2; + float x12 = (x1+x2)/2; + float y12 = (y1+y2)/2; + float x23 = (x2+x3)/2; + float y23 = (y2+y3)/2; + + float xa = (x01+x12)/2; + float ya = (y01+y12)/2; + float xb = (x12+x23)/2; + float yb = (y12+y23)/2; + + float mx = (xa+xb)/2; + float my = (ya+yb)/2; + + stbtt__tesselate_cubic(points, num_points, x0,y0, x01,y01, xa,ya, mx,my, objspace_flatness_squared,n+1); + stbtt__tesselate_cubic(points, num_points, mx,my, xb,yb, x23,y23, x3,y3, objspace_flatness_squared,n+1); + } else { + stbtt__add_point(points, *num_points,x3,y3); + *num_points = *num_points+1; + } +} + // returns number of contours static stbtt__point *stbtt_FlattenCurves(stbtt_vertex *vertices, int num_verts, float objspace_flatness, int **contour_lengths, int *num_contours, void *userdata) { @@ -2406,6 +3205,14 @@ static stbtt__point *stbtt_FlattenCurves(stbtt_vertex *vertices, int num_verts, objspace_flatness_squared, 0); x = vertices[i].x, y = vertices[i].y; break; + case STBTT_vcubic: + stbtt__tesselate_cubic(points, &num_points, x,y, + vertices[i].cx, vertices[i].cy, + vertices[i].cx1, vertices[i].cy1, + vertices[i].x, vertices[i].y, + objspace_flatness_squared, 0); + x = vertices[i].x, y = vertices[i].y; + break; } } (*contour_lengths)[n] = num_points - start; @@ -2441,7 +3248,7 @@ STBTT_DEF unsigned char *stbtt_GetGlyphBitmapSubpixel(const stbtt_fontinfo *info { int ix0,iy0,ix1,iy1; stbtt__bitmap gbm; - stbtt_vertex *vertices; + stbtt_vertex *vertices; int num_verts = stbtt_GetGlyphShape(info, glyph, &vertices); if (scale_x == 0) scale_x = scale_y; @@ -2464,7 +3271,7 @@ STBTT_DEF unsigned char *stbtt_GetGlyphBitmapSubpixel(const stbtt_fontinfo *info if (height) *height = gbm.h; if (xoff ) *xoff = ix0; if (yoff ) *yoff = iy0; - + if (gbm.w && gbm.h) { gbm.pixels = (unsigned char *) STBTT_malloc(gbm.w * gbm.h, info->userdata); if (gbm.pixels) { @@ -2475,7 +3282,7 @@ STBTT_DEF unsigned char *stbtt_GetGlyphBitmapSubpixel(const stbtt_fontinfo *info } STBTT_free(vertices, info->userdata); return gbm.pixels; -} +} STBTT_DEF unsigned char *stbtt_GetGlyphBitmap(const stbtt_fontinfo *info, float scale_x, float scale_y, int glyph, int *width, int *height, int *xoff, int *yoff) { @@ -2487,7 +3294,7 @@ STBTT_DEF void stbtt_MakeGlyphBitmapSubpixel(const stbtt_fontinfo *info, unsigne int ix0,iy0; stbtt_vertex *vertices; int num_verts = stbtt_GetGlyphShape(info, glyph, &vertices); - stbtt__bitmap gbm; + stbtt__bitmap gbm; stbtt_GetGlyphBitmapBoxSubpixel(info, glyph, scale_x, scale_y, shift_x, shift_y, &ix0,&iy0,0,0); gbm.pixels = output; @@ -2509,7 +3316,7 @@ STBTT_DEF void stbtt_MakeGlyphBitmap(const stbtt_fontinfo *info, unsigned char * STBTT_DEF unsigned char *stbtt_GetCodepointBitmapSubpixel(const stbtt_fontinfo *info, float scale_x, float scale_y, float shift_x, float shift_y, int codepoint, int *width, int *height, int *xoff, int *yoff) { return stbtt_GetGlyphBitmapSubpixel(info, scale_x, scale_y,shift_x,shift_y, stbtt_FindGlyphIndex(info,codepoint), width,height,xoff,yoff); -} +} STBTT_DEF void stbtt_MakeCodepointBitmapSubpixel(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int codepoint) { @@ -2519,7 +3326,7 @@ STBTT_DEF void stbtt_MakeCodepointBitmapSubpixel(const stbtt_fontinfo *info, uns STBTT_DEF unsigned char *stbtt_GetCodepointBitmap(const stbtt_fontinfo *info, float scale_x, float scale_y, int codepoint, int *width, int *height, int *xoff, int *yoff) { return stbtt_GetCodepointBitmapSubpixel(info, scale_x, scale_y, 0.0f,0.0f, codepoint, width,height,xoff,yoff); -} +} STBTT_DEF void stbtt_MakeCodepointBitmap(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, int codepoint) { @@ -2532,7 +3339,7 @@ STBTT_DEF void stbtt_MakeCodepointBitmap(const stbtt_fontinfo *info, unsigned ch // // This is SUPER-CRAPPY packing to keep source code small -STBTT_DEF int stbtt_BakeFontBitmap(const unsigned char *data, int offset, // font location (use offset=0 for plain .ttf) +static int stbtt_BakeFontBitmap_internal(unsigned char *data, int offset, // font location (use offset=0 for plain .ttf) float pixel_height, // height of font in pixels unsigned char *pixels, int pw, int ph, // bitmap to be filled in int first_char, int num_chars, // characters to bake @@ -2578,11 +3385,11 @@ STBTT_DEF int stbtt_BakeFontBitmap(const unsigned char *data, int offset, // fo return bottom_y; } -STBTT_DEF void stbtt_GetBakedQuad(stbtt_bakedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int opengl_fillrule) +STBTT_DEF void stbtt_GetBakedQuad(const stbtt_bakedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int opengl_fillrule) { float d3d_bias = opengl_fillrule ? 0 : -0.5f; float ipw = 1.0f / pw, iph = 1.0f / ph; - stbtt_bakedchar *b = chardata + char_index; + const stbtt_bakedchar *b = chardata + char_index; int round_x = STBTT_ifloor((*xpos + b->xoff) + 0.5f); int round_y = STBTT_ifloor((*ypos + b->yoff) + 0.5f); @@ -2644,7 +3451,7 @@ static void stbrp_init_target(stbrp_context *con, int pw, int ph, stbrp_node *no con->y = 0; con->bottom_y = 0; STBTT__NOTUSED(nodes); - STBTT__NOTUSED(num_nodes); + STBTT__NOTUSED(num_nodes); } static void stbrp_pack_rects(stbrp_context *con, stbrp_rect *rects, int num_rects) @@ -2862,7 +3669,7 @@ static float stbtt__oversample_shift(int oversample) } // rects array must be big enough to accommodate all characters in the given ranges -STBTT_DEF int stbtt_PackFontRangesGatherRects(stbtt_pack_context *spc, stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects) +STBTT_DEF int stbtt_PackFontRangesGatherRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects) { int i,j,k; @@ -2890,8 +3697,31 @@ STBTT_DEF int stbtt_PackFontRangesGatherRects(stbtt_pack_context *spc, stbtt_fon return k; } +STBTT_DEF void stbtt_MakeGlyphBitmapSubpixelPrefilter(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int prefilter_x, int prefilter_y, float *sub_x, float *sub_y, int glyph) +{ + stbtt_MakeGlyphBitmapSubpixel(info, + output, + out_w - (prefilter_x - 1), + out_h - (prefilter_y - 1), + out_stride, + scale_x, + scale_y, + shift_x, + shift_y, + glyph); + + if (prefilter_x > 1) + stbtt__h_prefilter(output, out_w, out_h, out_stride, prefilter_x); + + if (prefilter_y > 1) + stbtt__v_prefilter(output, out_w, out_h, out_stride, prefilter_y); + + *sub_x = stbtt__oversample_shift(prefilter_x); + *sub_y = stbtt__oversample_shift(prefilter_y); +} + // rects array must be big enough to accommodate all characters in the given ranges -STBTT_DEF int stbtt_PackFontRangesRenderIntoRects(stbtt_pack_context *spc, stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects) +STBTT_DEF int stbtt_PackFontRangesRenderIntoRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects) { int i,j,k, return_value = 1; @@ -2978,7 +3808,7 @@ STBTT_DEF void stbtt_PackFontRangesPackRects(stbtt_pack_context *spc, stbrp_rect stbrp_pack_rects((stbrp_context *) spc->pack_info, rects, num_rects); } -STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, unsigned char *fontdata, int font_index, stbtt_pack_range *ranges, int num_ranges) +STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, stbtt_pack_range *ranges, int num_ranges) { stbtt_fontinfo info; int i,j,n, return_value = 1; @@ -2996,7 +3826,7 @@ STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, unsigned char *fontd n = 0; for (i=0; i < num_ranges; ++i) n += ranges[i].num_chars; - + rects = (stbrp_rect *) STBTT_malloc(sizeof(*rects) * n, spc->user_allocator_context); if (rects == NULL) return 0; @@ -3007,14 +3837,14 @@ STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, unsigned char *fontd n = stbtt_PackFontRangesGatherRects(spc, &info, ranges, num_ranges, rects); stbtt_PackFontRangesPackRects(spc, rects, n); - + return_value = stbtt_PackFontRangesRenderIntoRects(spc, &info, ranges, num_ranges, rects); STBTT_free(rects, spc->user_allocator_context); return return_value; } -STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, unsigned char *fontdata, int font_index, float font_size, +STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, float font_size, int first_unicode_codepoint_in_range, int num_chars_in_range, stbtt_packedchar *chardata_for_range) { stbtt_pack_range range; @@ -3026,10 +3856,10 @@ STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, unsigned char *fontda return stbtt_PackFontRanges(spc, fontdata, font_index, &range, 1); } -STBTT_DEF void stbtt_GetPackedQuad(stbtt_packedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int align_to_integer) +STBTT_DEF void stbtt_GetPackedQuad(const stbtt_packedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int align_to_integer) { float ipw = 1.0f / pw, iph = 1.0f / ph; - stbtt_packedchar *b = chardata + char_index; + const stbtt_packedchar *b = chardata + char_index; if (align_to_integer) { float x = (float) STBTT_ifloor((*xpos + b->xoff) + 0.5f); @@ -3053,6 +3883,387 @@ STBTT_DEF void stbtt_GetPackedQuad(stbtt_packedchar *chardata, int pw, int ph, i *xpos += b->xadvance; } +////////////////////////////////////////////////////////////////////////////// +// +// sdf computation +// + +#define STBTT_min(a,b) ((a) < (b) ? (a) : (b)) +#define STBTT_max(a,b) ((a) < (b) ? (b) : (a)) + +static int stbtt__ray_intersect_bezier(float orig[2], float ray[2], float q0[2], float q1[2], float q2[2], float hits[2][2]) +{ + float q0perp = q0[1]*ray[0] - q0[0]*ray[1]; + float q1perp = q1[1]*ray[0] - q1[0]*ray[1]; + float q2perp = q2[1]*ray[0] - q2[0]*ray[1]; + float roperp = orig[1]*ray[0] - orig[0]*ray[1]; + + float a = q0perp - 2*q1perp + q2perp; + float b = q1perp - q0perp; + float c = q0perp - roperp; + + float s0 = 0., s1 = 0.; + int num_s = 0; + + if (a != 0.0) { + float discr = b*b - a*c; + if (discr > 0.0) { + float rcpna = -1 / a; + float d = (float) sqrt(discr); + s0 = (b+d) * rcpna; + s1 = (b-d) * rcpna; + if (s0 >= 0.0 && s0 <= 1.0) + num_s = 1; + if (d > 0.0 && s1 >= 0.0 && s1 <= 1.0) { + if (num_s == 0) s0 = s1; + ++num_s; + } + } + } else { + // 2*b*s + c = 0 + // s = -c / (2*b) + s0 = c / (-2 * b); + if (s0 >= 0.0 && s0 <= 1.0) + num_s = 1; + } + + if (num_s == 0) + return 0; + else { + float rcp_len2 = 1 / (ray[0]*ray[0] + ray[1]*ray[1]); + float rayn_x = ray[0] * rcp_len2, rayn_y = ray[1] * rcp_len2; + + float q0d = q0[0]*rayn_x + q0[1]*rayn_y; + float q1d = q1[0]*rayn_x + q1[1]*rayn_y; + float q2d = q2[0]*rayn_x + q2[1]*rayn_y; + float rod = orig[0]*rayn_x + orig[1]*rayn_y; + + float q10d = q1d - q0d; + float q20d = q2d - q0d; + float q0rd = q0d - rod; + + hits[0][0] = q0rd + s0*(2.0f - 2.0f*s0)*q10d + s0*s0*q20d; + hits[0][1] = a*s0+b; + + if (num_s > 1) { + hits[1][0] = q0rd + s1*(2.0f - 2.0f*s1)*q10d + s1*s1*q20d; + hits[1][1] = a*s1+b; + return 2; + } else { + return 1; + } + } +} + +static int equal(float *a, float *b) +{ + return (a[0] == b[0] && a[1] == b[1]); +} + +static int stbtt__compute_crossings_x(float x, float y, int nverts, stbtt_vertex *verts) +{ + int i; + float orig[2], ray[2] = { 1, 0 }; + float y_frac; + int winding = 0; + + orig[0] = x; + orig[1] = y; + + // make sure y never passes through a vertex of the shape + y_frac = (float) fmod(y, 1.0f); + if (y_frac < 0.01f) + y += 0.01f; + else if (y_frac > 0.99f) + y -= 0.01f; + orig[1] = y; + + // test a ray from (-infinity,y) to (x,y) + for (i=0; i < nverts; ++i) { + if (verts[i].type == STBTT_vline) { + int x0 = (int) verts[i-1].x, y0 = (int) verts[i-1].y; + int x1 = (int) verts[i ].x, y1 = (int) verts[i ].y; + if (y > STBTT_min(y0,y1) && y < STBTT_max(y0,y1) && x > STBTT_min(x0,x1)) { + float x_inter = (y - y0) / (y1 - y0) * (x1-x0) + x0; + if (x_inter < x) + winding += (y0 < y1) ? 1 : -1; + } + } + if (verts[i].type == STBTT_vcurve) { + int x0 = (int) verts[i-1].x , y0 = (int) verts[i-1].y ; + int x1 = (int) verts[i ].cx, y1 = (int) verts[i ].cy; + int x2 = (int) verts[i ].x , y2 = (int) verts[i ].y ; + int ax = STBTT_min(x0,STBTT_min(x1,x2)), ay = STBTT_min(y0,STBTT_min(y1,y2)); + int by = STBTT_max(y0,STBTT_max(y1,y2)); + if (y > ay && y < by && x > ax) { + float q0[2],q1[2],q2[2]; + float hits[2][2]; + q0[0] = (float)x0; + q0[1] = (float)y0; + q1[0] = (float)x1; + q1[1] = (float)y1; + q2[0] = (float)x2; + q2[1] = (float)y2; + if (equal(q0,q1) || equal(q1,q2)) { + x0 = (int)verts[i-1].x; + y0 = (int)verts[i-1].y; + x1 = (int)verts[i ].x; + y1 = (int)verts[i ].y; + if (y > STBTT_min(y0,y1) && y < STBTT_max(y0,y1) && x > STBTT_min(x0,x1)) { + float x_inter = (y - y0) / (y1 - y0) * (x1-x0) + x0; + if (x_inter < x) + winding += (y0 < y1) ? 1 : -1; + } + } else { + int num_hits = stbtt__ray_intersect_bezier(orig, ray, q0, q1, q2, hits); + if (num_hits >= 1) + if (hits[0][0] < 0) + winding += (hits[0][1] < 0 ? -1 : 1); + if (num_hits >= 2) + if (hits[1][0] < 0) + winding += (hits[1][1] < 0 ? -1 : 1); + } + } + } + } + return winding; +} + +static float stbtt__cuberoot( float x ) +{ + if (x<0) + return -(float) STBTT_pow(-x,1.0f/3.0f); + else + return (float) STBTT_pow( x,1.0f/3.0f); +} + +// x^3 + c*x^2 + b*x + a = 0 +static int stbtt__solve_cubic(float a, float b, float c, float* r) +{ + float s = -a / 3; + float p = b - a*a / 3; + float q = a * (2*a*a - 9*b) / 27 + c; + float p3 = p*p*p; + float d = q*q + 4*p3 / 27; + if (d >= 0) { + float z = (float) STBTT_sqrt(d); + float u = (-q + z) / 2; + float v = (-q - z) / 2; + u = stbtt__cuberoot(u); + v = stbtt__cuberoot(v); + r[0] = s + u + v; + return 1; + } else { + float u = (float) STBTT_sqrt(-p/3); + float v = (float) STBTT_acos(-STBTT_sqrt(-27/p3) * q / 2) / 3; // p3 must be negative, since d is negative + float m = (float) STBTT_cos(v); + float n = (float) STBTT_cos(v-3.141592/2)*1.732050808f; + r[0] = s + u * 2 * m; + r[1] = s - u * (m + n); + r[2] = s - u * (m - n); + + //STBTT_assert( STBTT_fabs(((r[0]+a)*r[0]+b)*r[0]+c) < 0.05f); // these asserts may not be safe at all scales, though they're in bezier t parameter units so maybe? + //STBTT_assert( STBTT_fabs(((r[1]+a)*r[1]+b)*r[1]+c) < 0.05f); + //STBTT_assert( STBTT_fabs(((r[2]+a)*r[2]+b)*r[2]+c) < 0.05f); + return 3; + } +} + +STBTT_DEF unsigned char * stbtt_GetGlyphSDF(const stbtt_fontinfo *info, float scale, int glyph, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff) +{ + float scale_x = scale, scale_y = scale; + int ix0,iy0,ix1,iy1; + int w,h; + unsigned char *data; + + // if one scale is 0, use same scale for both + if (scale_x == 0) scale_x = scale_y; + if (scale_y == 0) { + if (scale_x == 0) return NULL; // if both scales are 0, return NULL + scale_y = scale_x; + } + + stbtt_GetGlyphBitmapBoxSubpixel(info, glyph, scale, scale, 0.0f,0.0f, &ix0,&iy0,&ix1,&iy1); + + // if empty, return NULL + if (ix0 == ix1 || iy0 == iy1) + return NULL; + + ix0 -= padding; + iy0 -= padding; + ix1 += padding; + iy1 += padding; + + w = (ix1 - ix0); + h = (iy1 - iy0); + + if (width ) *width = w; + if (height) *height = h; + if (xoff ) *xoff = ix0; + if (yoff ) *yoff = iy0; + + // invert for y-downwards bitmaps + scale_y = -scale_y; + + { + int x,y,i,j; + float *precompute; + stbtt_vertex *verts; + int num_verts = stbtt_GetGlyphShape(info, glyph, &verts); + data = (unsigned char *) STBTT_malloc(w * h, info->userdata); + precompute = (float *) STBTT_malloc(num_verts * sizeof(float), info->userdata); + + for (i=0,j=num_verts-1; i < num_verts; j=i++) { + if (verts[i].type == STBTT_vline) { + float x0 = verts[i].x*scale_x, y0 = verts[i].y*scale_y; + float x1 = verts[j].x*scale_x, y1 = verts[j].y*scale_y; + float dist = (float) STBTT_sqrt((x1-x0)*(x1-x0) + (y1-y0)*(y1-y0)); + precompute[i] = (dist == 0) ? 0.0f : 1.0f / dist; + } else if (verts[i].type == STBTT_vcurve) { + float x2 = verts[j].x *scale_x, y2 = verts[j].y *scale_y; + float x1 = verts[i].cx*scale_x, y1 = verts[i].cy*scale_y; + float x0 = verts[i].x *scale_x, y0 = verts[i].y *scale_y; + float bx = x0 - 2*x1 + x2, by = y0 - 2*y1 + y2; + float len2 = bx*bx + by*by; + if (len2 != 0.0f) + precompute[i] = 1.0f / (bx*bx + by*by); + else + precompute[i] = 0.0f; + } else + precompute[i] = 0.0f; + } + + for (y=iy0; y < iy1; ++y) { + for (x=ix0; x < ix1; ++x) { + float val; + float min_dist = 999999.0f; + float sx = (float) x + 0.5f; + float sy = (float) y + 0.5f; + float x_gspace = (sx / scale_x); + float y_gspace = (sy / scale_y); + + int winding = stbtt__compute_crossings_x(x_gspace, y_gspace, num_verts, verts); // @OPTIMIZE: this could just be a rasterization, but needs to be line vs. non-tesselated curves so a new path + + for (i=0; i < num_verts; ++i) { + float x0 = verts[i].x*scale_x, y0 = verts[i].y*scale_y; + + // check against every point here rather than inside line/curve primitives -- @TODO: wrong if multiple 'moves' in a row produce a garbage point, and given culling, probably more efficient to do within line/curve + float dist2 = (x0-sx)*(x0-sx) + (y0-sy)*(y0-sy); + if (dist2 < min_dist*min_dist) + min_dist = (float) STBTT_sqrt(dist2); + + if (verts[i].type == STBTT_vline) { + float x1 = verts[i-1].x*scale_x, y1 = verts[i-1].y*scale_y; + + // coarse culling against bbox + //if (sx > STBTT_min(x0,x1)-min_dist && sx < STBTT_max(x0,x1)+min_dist && + // sy > STBTT_min(y0,y1)-min_dist && sy < STBTT_max(y0,y1)+min_dist) + float dist = (float) STBTT_fabs((x1-x0)*(y0-sy) - (y1-y0)*(x0-sx)) * precompute[i]; + STBTT_assert(i != 0); + if (dist < min_dist) { + // check position along line + // x' = x0 + t*(x1-x0), y' = y0 + t*(y1-y0) + // minimize (x'-sx)*(x'-sx)+(y'-sy)*(y'-sy) + float dx = x1-x0, dy = y1-y0; + float px = x0-sx, py = y0-sy; + // minimize (px+t*dx)^2 + (py+t*dy)^2 = px*px + 2*px*dx*t + t^2*dx*dx + py*py + 2*py*dy*t + t^2*dy*dy + // derivative: 2*px*dx + 2*py*dy + (2*dx*dx+2*dy*dy)*t, set to 0 and solve + float t = -(px*dx + py*dy) / (dx*dx + dy*dy); + if (t >= 0.0f && t <= 1.0f) + min_dist = dist; + } + } else if (verts[i].type == STBTT_vcurve) { + float x2 = verts[i-1].x *scale_x, y2 = verts[i-1].y *scale_y; + float x1 = verts[i ].cx*scale_x, y1 = verts[i ].cy*scale_y; + float box_x0 = STBTT_min(STBTT_min(x0,x1),x2); + float box_y0 = STBTT_min(STBTT_min(y0,y1),y2); + float box_x1 = STBTT_max(STBTT_max(x0,x1),x2); + float box_y1 = STBTT_max(STBTT_max(y0,y1),y2); + // coarse culling against bbox to avoid computing cubic unnecessarily + if (sx > box_x0-min_dist && sx < box_x1+min_dist && sy > box_y0-min_dist && sy < box_y1+min_dist) { + int num=0; + float ax = x1-x0, ay = y1-y0; + float bx = x0 - 2*x1 + x2, by = y0 - 2*y1 + y2; + float mx = x0 - sx, my = y0 - sy; + float res[3],px,py,t,it; + float a_inv = precompute[i]; + if (a_inv == 0.0) { // if a_inv is 0, it's 2nd degree so use quadratic formula + float a = 3*(ax*bx + ay*by); + float b = 2*(ax*ax + ay*ay) + (mx*bx+my*by); + float c = mx*ax+my*ay; + if (a == 0.0) { // if a is 0, it's linear + if (b != 0.0) { + res[num++] = -c/b; + } + } else { + float discriminant = b*b - 4*a*c; + if (discriminant < 0) + num = 0; + else { + float root = (float) STBTT_sqrt(discriminant); + res[0] = (-b - root)/(2*a); + res[1] = (-b + root)/(2*a); + num = 2; // don't bother distinguishing 1-solution case, as code below will still work + } + } + } else { + float b = 3*(ax*bx + ay*by) * a_inv; // could precompute this as it doesn't depend on sample point + float c = (2*(ax*ax + ay*ay) + (mx*bx+my*by)) * a_inv; + float d = (mx*ax+my*ay) * a_inv; + num = stbtt__solve_cubic(b, c, d, res); + } + if (num >= 1 && res[0] >= 0.0f && res[0] <= 1.0f) { + t = res[0], it = 1.0f - t; + px = it*it*x0 + 2*t*it*x1 + t*t*x2; + py = it*it*y0 + 2*t*it*y1 + t*t*y2; + dist2 = (px-sx)*(px-sx) + (py-sy)*(py-sy); + if (dist2 < min_dist * min_dist) + min_dist = (float) STBTT_sqrt(dist2); + } + if (num >= 2 && res[1] >= 0.0f && res[1] <= 1.0f) { + t = res[1], it = 1.0f - t; + px = it*it*x0 + 2*t*it*x1 + t*t*x2; + py = it*it*y0 + 2*t*it*y1 + t*t*y2; + dist2 = (px-sx)*(px-sx) + (py-sy)*(py-sy); + if (dist2 < min_dist * min_dist) + min_dist = (float) STBTT_sqrt(dist2); + } + if (num >= 3 && res[2] >= 0.0f && res[2] <= 1.0f) { + t = res[2], it = 1.0f - t; + px = it*it*x0 + 2*t*it*x1 + t*t*x2; + py = it*it*y0 + 2*t*it*y1 + t*t*y2; + dist2 = (px-sx)*(px-sx) + (py-sy)*(py-sy); + if (dist2 < min_dist * min_dist) + min_dist = (float) STBTT_sqrt(dist2); + } + } + } + } + if (winding == 0) + min_dist = -min_dist; // if outside the shape, value is negative + val = onedge_value + pixel_dist_scale * min_dist; + if (val < 0) + val = 0; + else if (val > 255) + val = 255; + data[(y-iy0)*w+(x-ix0)] = (unsigned char) val; + } + } + STBTT_free(precompute, info->userdata); + STBTT_free(verts, info->userdata); + } + return data; +} + +STBTT_DEF unsigned char * stbtt_GetCodepointSDF(const stbtt_fontinfo *info, float scale, int codepoint, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff) +{ + return stbtt_GetGlyphSDF(info, scale, stbtt_FindGlyphIndex(info, codepoint), padding, onedge_value, pixel_dist_scale, width, height, xoff, yoff); +} + +STBTT_DEF void stbtt_FreeSDF(unsigned char *bitmap, void *userdata) +{ + STBTT_free(bitmap, userdata); +} ////////////////////////////////////////////////////////////////////////////// // @@ -3060,7 +4271,7 @@ STBTT_DEF void stbtt_GetPackedQuad(stbtt_packedchar *chardata, int pw, int ph, i // // check if a utf8 string contains a prefix which is the utf16 string; if so return length of matching utf8 string -static stbtt_int32 stbtt__CompareUTF8toUTF16_bigendian_prefix(const stbtt_uint8 *s1, stbtt_int32 len1, const stbtt_uint8 *s2, stbtt_int32 len2) +static stbtt_int32 stbtt__CompareUTF8toUTF16_bigendian_prefix(stbtt_uint8 *s1, stbtt_int32 len1, stbtt_uint8 *s2, stbtt_int32 len2) { stbtt_int32 i=0; @@ -3099,9 +4310,9 @@ static stbtt_int32 stbtt__CompareUTF8toUTF16_bigendian_prefix(const stbtt_uint8 return i; } -STBTT_DEF int stbtt_CompareUTF8toUTF16_bigendian(const char *s1, int len1, const char *s2, int len2) +static int stbtt_CompareUTF8toUTF16_bigendian_internal(char *s1, int len1, char *s2, int len2) { - return len1 == stbtt__CompareUTF8toUTF16_bigendian_prefix((const stbtt_uint8*) s1, len1, (const stbtt_uint8*) s2, len2); + return len1 == stbtt__CompareUTF8toUTF16_bigendian_prefix((stbtt_uint8*) s1, len1, (stbtt_uint8*) s2, len2); } // returns results in whatever encoding you request... but note that 2-byte encodings @@ -3157,7 +4368,7 @@ static int stbtt__matchpair(stbtt_uint8 *fc, stbtt_uint32 nm, stbtt_uint8 *name, return 1; } else if (matchlen < nlen && name[matchlen] == ' ') { ++matchlen; - if (stbtt_CompareUTF8toUTF16_bigendian((char*) (name+matchlen), nlen-matchlen, (char*)(fc+stringOffset+off),slen)) + if (stbtt_CompareUTF8toUTF16_bigendian_internal((char*) (name+matchlen), nlen-matchlen, (char*)(fc+stringOffset+off),slen)) return 1; } } else { @@ -3203,7 +4414,7 @@ static int stbtt__matches(stbtt_uint8 *fc, stbtt_uint32 offset, stbtt_uint8 *nam return 0; } -STBTT_DEF int stbtt_FindMatchingFont(const unsigned char *font_collection, const char *name_utf8, stbtt_int32 flags) +static int stbtt_FindMatchingFont_internal(unsigned char *font_collection, char *name_utf8, stbtt_int32 flags) { stbtt_int32 i; for (i=0;;++i) { @@ -3214,11 +4425,57 @@ STBTT_DEF int stbtt_FindMatchingFont(const unsigned char *font_collection, const } } +#if defined(__GNUC__) || defined(__clang__) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wcast-qual" +#endif + +STBTT_DEF int stbtt_BakeFontBitmap(const unsigned char *data, int offset, + float pixel_height, unsigned char *pixels, int pw, int ph, + int first_char, int num_chars, stbtt_bakedchar *chardata) +{ + return stbtt_BakeFontBitmap_internal((unsigned char *) data, offset, pixel_height, pixels, pw, ph, first_char, num_chars, chardata); +} + +STBTT_DEF int stbtt_GetFontOffsetForIndex(const unsigned char *data, int index) +{ + return stbtt_GetFontOffsetForIndex_internal((unsigned char *) data, index); +} + +STBTT_DEF int stbtt_GetNumberOfFonts(const unsigned char *data) +{ + return stbtt_GetNumberOfFonts_internal((unsigned char *) data); +} + +STBTT_DEF int stbtt_InitFont(stbtt_fontinfo *info, const unsigned char *data, int offset) +{ + return stbtt_InitFont_internal(info, (unsigned char *) data, offset); +} + +STBTT_DEF int stbtt_FindMatchingFont(const unsigned char *fontdata, const char *name, int flags) +{ + return stbtt_FindMatchingFont_internal((unsigned char *) fontdata, (char *) name, flags); +} + +STBTT_DEF int stbtt_CompareUTF8toUTF16_bigendian(const char *s1, int len1, const char *s2, int len2) +{ + return stbtt_CompareUTF8toUTF16_bigendian_internal((char *) s1, len1, (char *) s2, len2); +} + +#if defined(__GNUC__) || defined(__clang__) +#pragma GCC diagnostic pop +#endif + #endif // STB_TRUETYPE_IMPLEMENTATION // FULL VERSION HISTORY // +// 1.16 (2017-07-12) SDF support +// 1.15 (2017-03-03) make more arguments const +// 1.14 (2017-01-16) num-fonts-in-TTC function +// 1.13 (2017-01-02) support OpenType fonts, certain Apple fonts +// 1.12 (2016-10-25) suppress warnings about casting away const with -Wcast-qual // 1.11 (2016-04-02) fix unused-variable warning // 1.10 (2016-04-02) allow user-defined fabs() replacement // fix memory leak if fontsize=0.0 @@ -3265,3 +4522,45 @@ STBTT_DEF int stbtt_FindMatchingFont(const unsigned char *font_collection, const // 0.2 (2009-03-11) Fix unsigned/signed char warnings // 0.1 (2009-03-09) First public release // + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/thirdparty/misc/stb_vorbis.c b/thirdparty/misc/stb_vorbis.c index c4f24d5898..14cebbf87e 100644 --- a/thirdparty/misc/stb_vorbis.c +++ b/thirdparty/misc/stb_vorbis.c @@ -1,4 +1,4 @@ -// Ogg Vorbis audio decoder - v1.09 - public domain +// Ogg Vorbis audio decoder - v1.11 - public domain // http://nothings.org/stb_vorbis/ // // Original version written by Sean Barrett in 2007. @@ -9,12 +9,7 @@ // // LICENSE // -// This software is dual-licensed to the public domain and under the following -// license: you are granted a perpetual, irrevocable license to copy, modify, -// publish, and distribute this file as you see fit. -// -// No warranty for any purpose is expressed or implied by the author (nor -// by RAD Game Tools). Report bugs and send enhancements to the author. +// See end of file for license information. // // Limitations: // @@ -34,9 +29,11 @@ // Bernhard Wodo Evan Balster alxprd@github // Tom Beaumont Ingo Leitgeb Nicolas Guillemot // Phillip Bennefall Rohit Thiago Goulart -// manxorist@github saga musix +// manxorist@github saga musix github:infatum // // Partial history: +// 1.11 - 2017/07/23 - fix MinGW compilation +// 1.10 - 2017/03/03 - more robust seeking; fix negative ilog(); clear error in open_memory // 1.09 - 2016/04/04 - back out 'truncation of last frame' fix from previous version // 1.08 - 2016/04/02 - warnings; setup memory leaks; truncation of last frame // 1.07 - 2015/01/16 - fixes for crashes on invalid files; warning fixes; const @@ -275,7 +272,7 @@ extern int stb_vorbis_seek(stb_vorbis *f, unsigned int sample_number); // do not need to seek to EXACTLY the target sample when using get_samples_*, // you can also use seek_frame(). -extern void stb_vorbis_seek_start(stb_vorbis *f); +extern int stb_vorbis_seek_start(stb_vorbis *f); // this function is equivalent to stb_vorbis_seek(f,0) extern unsigned int stb_vorbis_stream_length_in_samples(stb_vorbis *f); @@ -555,7 +552,7 @@ enum STBVorbisError #include <math.h> // find definition of alloca if it's not in stdlib.h: - #ifdef _MSC_VER + #if defined(_MSC_VER) || defined(__MINGW32__) #include <malloc.h> #endif #if defined(__linux__) || defined(__linux) || defined(__EMSCRIPTEN__) @@ -580,6 +577,7 @@ enum STBVorbisError #undef __forceinline #endif #define __forceinline + #define alloca __builtin_alloca #elif !defined(_MSC_VER) #if __GNUC__ #define __forceinline inline @@ -986,17 +984,18 @@ static int ilog(int32 n) { static signed char log2_4[16] = { 0,1,2,2,3,3,3,3,4,4,4,4,4,4,4,4 }; + if (n < 0) return 0; // signed n returns 0 + // 2 compares if n < 16, 3 compares otherwise (4 if signed or n > 1<<29) if (n < (1 << 14)) - if (n < (1 << 4)) return 0 + log2_4[n ]; - else if (n < (1 << 9)) return 5 + log2_4[n >> 5]; + if (n < (1 << 4)) return 0 + log2_4[n ]; + else if (n < (1 << 9)) return 5 + log2_4[n >> 5]; else return 10 + log2_4[n >> 10]; else if (n < (1 << 24)) - if (n < (1 << 19)) return 15 + log2_4[n >> 15]; + if (n < (1 << 19)) return 15 + log2_4[n >> 15]; else return 20 + log2_4[n >> 20]; - else if (n < (1 << 29)) return 25 + log2_4[n >> 25]; - else if (n < (1 << 31)) return 30 + log2_4[n >> 30]; - else return 0; // signed n returns 0 + else if (n < (1 << 29)) return 25 + log2_4[n >> 25]; + else return 30 + log2_4[n >> 30]; } #ifndef M_PI @@ -1269,13 +1268,13 @@ static void neighbors(uint16 *x, int n, int *plow, int *phigh) // this has been repurposed so y is now the original index instead of y typedef struct { - uint16 x,y; -} Point; + uint16 x,id; +} stbv__floor_ordering; static int STBV_CDECL point_compare(const void *p, const void *q) { - Point *a = (Point *) p; - Point *b = (Point *) q; + stbv__floor_ordering *a = (stbv__floor_ordering *) p; + stbv__floor_ordering *b = (stbv__floor_ordering *) q; return a->x < b->x ? -1 : a->x > b->x; } @@ -3484,11 +3483,13 @@ static int vorbis_finish_frame(stb_vorbis *f, int len, int left, int right) return right - left; } -static void vorbis_pump_first_frame(stb_vorbis *f) +static int vorbis_pump_first_frame(stb_vorbis *f) { - int len, right, left; - if (vorbis_decode_packet(f, &len, &left, &right)) + int len, right, left, res; + res = vorbis_decode_packet(f, &len, &left, &right); + if (res) vorbis_finish_frame(f, len, left, right); + return res; } #ifndef STB_VORBIS_NO_PUSHDATA_API @@ -3871,7 +3872,7 @@ static int start_decoder(vorb *f) g->book_list[j] = get_bits(f,8); return error(f, VORBIS_feature_not_supported); } else { - Point p[31*8+2]; + stbv__floor_ordering p[31*8+2]; Floor1 *g = &f->floor_config[i].floor1; int max_class = -1; g->partitions = get_bits(f, 5); @@ -3907,11 +3908,11 @@ static int start_decoder(vorb *f) // precompute the sorting for (j=0; j < g->values; ++j) { p[j].x = g->Xlist[j]; - p[j].y = j; + p[j].id = j; } qsort(p, g->values, sizeof(p[0]), point_compare); for (j=0; j < g->values; ++j) - g->sorted_order[j] = (uint8) p[j].y; + g->sorted_order[j] = (uint8) p[j].id; // precompute the neighbors for (j=2; j < g->values; ++j) { int low,hi; @@ -4615,8 +4616,9 @@ static int seek_to_sample_coarse(stb_vorbis *f, uint32 sample_number) // starting from the start is handled differently if (sample_number <= left.last_decoded_sample) { - stb_vorbis_seek_start(f); - return 1; + if (stb_vorbis_seek_start(f)) + return 1; + return 0; } while (left.page_end != right.page_start) { @@ -4717,7 +4719,10 @@ static int seek_to_sample_coarse(stb_vorbis *f, uint32 sample_number) skip(f, f->segments[i]); // start decoding (optimizable - this frame is generally discarded) - vorbis_pump_first_frame(f); + if (!vorbis_pump_first_frame(f)) + return 0; + if (f->current_loc > sample_number) + return error(f, VORBIS_seek_failed); return 1; error: @@ -4808,14 +4813,14 @@ int stb_vorbis_seek(stb_vorbis *f, unsigned int sample_number) return 1; } -void stb_vorbis_seek_start(stb_vorbis *f) +int stb_vorbis_seek_start(stb_vorbis *f) { - if (IS_PUSH_MODE(f)) { error(f, VORBIS_invalid_api_mixing); return; } + if (IS_PUSH_MODE(f)) { return error(f, VORBIS_invalid_api_mixing); } set_file_offset(f, f->first_audio_page_offset); f->previous_length = 0; f->first_decode = TRUE; f->next_seg = -1; - vorbis_pump_first_frame(f); + return vorbis_pump_first_frame(f); } unsigned int stb_vorbis_stream_length_in_samples(stb_vorbis *f) @@ -4980,6 +4985,7 @@ stb_vorbis * stb_vorbis_open_memory(const unsigned char *data, int len, int *err if (f) { *f = p; vorbis_pump_first_frame(f); + if (error) *error = VORBIS__no_error; return f; } } @@ -5345,6 +5351,7 @@ int stb_vorbis_get_samples_float(stb_vorbis *f, int channels, float **buffer, in #endif // STB_VORBIS_NO_PULLDATA_API /* Version history + 1.10 - 2017/03/03 - more robust seeking; fix negative ilog(); clear error in open_memory 1.09 - 2016/04/04 - back out 'avoid discarding last frame' fix from previous version 1.08 - 2016/04/02 - fixed multiple warnings; fix setup memory leaks; avoid discarding last frame of audio data @@ -5397,3 +5404,46 @@ int stb_vorbis_get_samples_float(stb_vorbis *f, int channels, float **buffer, in */ #endif // STB_VORBIS_HEADER_ONLY + + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/thirdparty/nanosvg/LICENSE.txt b/thirdparty/nanosvg/LICENSE.txt new file mode 100644 index 0000000000..6fde401cb2 --- /dev/null +++ b/thirdparty/nanosvg/LICENSE.txt @@ -0,0 +1,18 @@ +Copyright (c) 2013-14 Mikko Mononen memon@inside.org
+
+This software is provided 'as-is', without any express or implied
+warranty. In no event will the authors be held liable for any damages
+arising from the use of this software.
+
+Permission is granted to anyone to use this software for any purpose,
+including commercial applications, and to alter it and redistribute it
+freely, subject to the following restrictions:
+
+1. The origin of this software must not be misrepresented; you must not
+claim that you wrote the original software. If you use this software
+in a product, an acknowledgment in the product documentation would be
+appreciated but is not required.
+2. Altered source versions must be plainly marked as such, and must not be
+misrepresented as being the original software.
+3. This notice may not be removed or altered from any source distribution.
+
diff --git a/thirdparty/nanosvg/nanosvg.cc b/thirdparty/nanosvg/nanosvg.cc new file mode 100644 index 0000000000..3e8e86c792 --- /dev/null +++ b/thirdparty/nanosvg/nanosvg.cc @@ -0,0 +1,8 @@ +#include "stdio.h" +#include "string.h" +#include "math.h" +#define NANOSVG_ALL_COLOR_KEYWORDS +#define NANOSVG_IMPLEMENTATION +#include "nanosvg.h" +#define NANOSVGRAST_IMPLEMENTATION +#include "nanosvgrast.h" diff --git a/thirdparty/nanosvg/nanosvg.h b/thirdparty/nanosvg/nanosvg.h new file mode 100644 index 0000000000..2321c56fd2 --- /dev/null +++ b/thirdparty/nanosvg/nanosvg.h @@ -0,0 +1,2925 @@ +/* + * Copyright (c) 2013-14 Mikko Mononen memon@inside.org + * + * This software is provided 'as-is', without any express or implied + * warranty. In no event will the authors be held liable for any damages + * arising from the use of this software. + * + * Permission is granted to anyone to use this software for any purpose, + * including commercial applications, and to alter it and redistribute it + * freely, subject to the following restrictions: + * + * 1. The origin of this software must not be misrepresented; you must not + * claim that you wrote the original software. If you use this software + * in a product, an acknowledgment in the product documentation would be + * appreciated but is not required. + * 2. Altered source versions must be plainly marked as such, and must not be + * misrepresented as being the original software. + * 3. This notice may not be removed or altered from any source distribution. + * + * The SVG parser is based on Anti-Grain Geometry 2.4 SVG example + * Copyright (C) 2002-2004 Maxim Shemanarev (McSeem) (http://www.antigrain.com/) + * + * Arc calculation code based on canvg (https://code.google.com/p/canvg/) + * + * Bounding box calculation based on http://blog.hackers-cafe.net/2009/06/how-to-calculate-bezier-curves-bounding.html + * + */ + +#ifndef NANOSVG_H +#define NANOSVG_H + +#ifdef __cplusplus +extern "C" { +#endif + +// NanoSVG is a simple stupid single-header-file SVG parse. The output of the parser is a list of cubic bezier shapes. +// +// The library suits well for anything from rendering scalable icons in your editor application to prototyping a game. +// +// NanoSVG supports a wide range of SVG features, but something may be missing, feel free to create a pull request! +// +// The shapes in the SVG images are transformed by the viewBox and converted to specified units. +// That is, you should get the same looking data as your designed in your favorite app. +// +// NanoSVG can return the paths in few different units. For example if you want to render an image, you may choose +// to get the paths in pixels, or if you are feeding the data into a CNC-cutter, you may want to use millimeters. +// +// The units passed to NanoVG should be one of: 'px', 'pt', 'pc' 'mm', 'cm', or 'in'. +// DPI (dots-per-inch) controls how the unit conversion is done. +// +// If you don't know or care about the units stuff, "px" and 96 should get you going. + + +/* Example Usage: + // Load + NSVGImage* image; + image = nsvgParseFromFile("test.svg", "px", 96); + printf("size: %f x %f\n", image->width, image->height); + // Use... + for (NSVGshape *shape = image->shapes; shape != NULL; shape = shape->next) { + for (NSVGpath *path = shape->paths; path != NULL; path = path->next) { + for (int i = 0; i < path->npts-1; i += 3) { + float* p = &path->pts[i*2]; + drawCubicBez(p[0],p[1], p[2],p[3], p[4],p[5], p[6],p[7]); + } + } + } + // Delete + nsvgDelete(image); +*/ + +enum NSVGpaintType { + NSVG_PAINT_NONE = 0, + NSVG_PAINT_COLOR = 1, + NSVG_PAINT_LINEAR_GRADIENT = 2, + NSVG_PAINT_RADIAL_GRADIENT = 3 +}; + +enum NSVGspreadType { + NSVG_SPREAD_PAD = 0, + NSVG_SPREAD_REFLECT = 1, + NSVG_SPREAD_REPEAT = 2 +}; + +enum NSVGlineJoin { + NSVG_JOIN_MITER = 0, + NSVG_JOIN_ROUND = 1, + NSVG_JOIN_BEVEL = 2 +}; + +enum NSVGlineCap { + NSVG_CAP_BUTT = 0, + NSVG_CAP_ROUND = 1, + NSVG_CAP_SQUARE = 2 +}; + +enum NSVGfillRule { + NSVG_FILLRULE_NONZERO = 0, + NSVG_FILLRULE_EVENODD = 1 +}; + +enum NSVGflags { + NSVG_FLAGS_VISIBLE = 0x01 +}; + +typedef struct NSVGgradientStop { + unsigned int color; + float offset; +} NSVGgradientStop; + +typedef struct NSVGgradient { + float xform[6]; + char spread; + float fx, fy; + int nstops; + NSVGgradientStop stops[1]; +} NSVGgradient; + +typedef struct NSVGpaint { + char type; + union { + unsigned int color; + NSVGgradient* gradient; + }; +} NSVGpaint; + +typedef struct NSVGpath +{ + float* pts; // Cubic bezier points: x0,y0, [cpx1,cpx1,cpx2,cpy2,x1,y1], ... + int npts; // Total number of bezier points. + char closed; // Flag indicating if shapes should be treated as closed. + float bounds[4]; // Tight bounding box of the shape [minx,miny,maxx,maxy]. + struct NSVGpath* next; // Pointer to next path, or NULL if last element. +} NSVGpath; + +typedef struct NSVGshape +{ + char id[64]; // Optional 'id' attr of the shape or its group + NSVGpaint fill; // Fill paint + NSVGpaint stroke; // Stroke paint + float opacity; // Opacity of the shape. + float strokeWidth; // Stroke width (scaled). + float strokeDashOffset; // Stroke dash offset (scaled). + float strokeDashArray[8]; // Stroke dash array (scaled). + char strokeDashCount; // Number of dash values in dash array. + char strokeLineJoin; // Stroke join type. + char strokeLineCap; // Stroke cap type. + float miterLimit; // Miter limit + char fillRule; // Fill rule, see NSVGfillRule. + unsigned char flags; // Logical or of NSVG_FLAGS_* flags + float bounds[4]; // Tight bounding box of the shape [minx,miny,maxx,maxy]. + NSVGpath* paths; // Linked list of paths in the image. + struct NSVGshape* next; // Pointer to next shape, or NULL if last element. +} NSVGshape; + +typedef struct NSVGimage +{ + float width; // Width of the image. + float height; // Height of the image. + NSVGshape* shapes; // Linked list of shapes in the image. +} NSVGimage; + +// Parses SVG file from a file, returns SVG image as paths. +NSVGimage* nsvgParseFromFile(const char* filename, const char* units, float dpi); + +// Parses SVG file from a null terminated string, returns SVG image as paths. +// Important note: changes the string. +NSVGimage* nsvgParse(char* input, const char* units, float dpi); + +// Deletes list of paths. +void nsvgDelete(NSVGimage* image); + +#ifdef __cplusplus +} +#endif + +#endif // NANOSVG_H + +#ifdef NANOSVG_IMPLEMENTATION + +#include <string.h> +#include <stdlib.h> +#include <math.h> + +#define NSVG_PI (3.14159265358979323846264338327f) +#define NSVG_KAPPA90 (0.5522847493f) // Length proportional to radius of a cubic bezier handle for 90deg arcs. + +#define NSVG_ALIGN_MIN 0 +#define NSVG_ALIGN_MID 1 +#define NSVG_ALIGN_MAX 2 +#define NSVG_ALIGN_NONE 0 +#define NSVG_ALIGN_MEET 1 +#define NSVG_ALIGN_SLICE 2 + +#define NSVG_NOTUSED(v) do { (void)(1 ? (void)0 : ( (void)(v) ) ); } while(0) +#define NSVG_RGB(r, g, b) (((unsigned int)r) | ((unsigned int)g << 8) | ((unsigned int)b << 16)) + +#ifdef _MSC_VER + #pragma warning (disable: 4996) // Switch off security warnings + #pragma warning (disable: 4100) // Switch off unreferenced formal parameter warnings + #ifdef __cplusplus + #define NSVG_INLINE inline + #else + #define NSVG_INLINE + #endif +#else + #define NSVG_INLINE inline +#endif + + +static int nsvg__isspace(char c) +{ + return strchr(" \t\n\v\f\r", c) != 0; +} + +static int nsvg__isdigit(char c) +{ + return c >= '0' && c <= '9'; +} + +static int nsvg__isnum(char c) +{ + return strchr("0123456789+-.eE", c) != 0; +} + +static NSVG_INLINE float nsvg__minf(float a, float b) { return a < b ? a : b; } +static NSVG_INLINE float nsvg__maxf(float a, float b) { return a > b ? a : b; } + + +// Simple XML parser + +#define NSVG_XML_TAG 1 +#define NSVG_XML_CONTENT 2 +#define NSVG_XML_MAX_ATTRIBS 256 + +static void nsvg__parseContent(char* s, + void (*contentCb)(void* ud, const char* s), + void* ud) +{ + // Trim start white spaces + while (*s && nsvg__isspace(*s)) s++; + if (!*s) return; + + if (contentCb) + (*contentCb)(ud, s); +} + +static void nsvg__parseElement(char* s, + void (*startelCb)(void* ud, const char* el, const char** attr), + void (*endelCb)(void* ud, const char* el), + void* ud) +{ + const char* attr[NSVG_XML_MAX_ATTRIBS]; + int nattr = 0; + char* name; + int start = 0; + int end = 0; + char quote; + + // Skip white space after the '<' + while (*s && nsvg__isspace(*s)) s++; + + // Check if the tag is end tag + if (*s == '/') { + s++; + end = 1; + } else { + start = 1; + } + + // Skip comments, data and preprocessor stuff. + if (!*s || *s == '?' || *s == '!') + return; + + // Get tag name + name = s; + while (*s && !nsvg__isspace(*s)) s++; + if (*s) { *s++ = '\0'; } + + // Get attribs + while (!end && *s && nattr < NSVG_XML_MAX_ATTRIBS-3) { + char* name = NULL; + char* value = NULL; + + // Skip white space before the attrib name + while (*s && nsvg__isspace(*s)) s++; + if (!*s) break; + if (*s == '/') { + end = 1; + break; + } + name = s; + // Find end of the attrib name. + while (*s && !nsvg__isspace(*s) && *s != '=') s++; + if (*s) { *s++ = '\0'; } + // Skip until the beginning of the value. + while (*s && *s != '\"' && *s != '\'') s++; + if (!*s) break; + quote = *s; + s++; + // Store value and find the end of it. + value = s; + while (*s && *s != quote) s++; + if (*s) { *s++ = '\0'; } + + // Store only well formed attributes + if (name && value) { + attr[nattr++] = name; + attr[nattr++] = value; + } + } + + // List terminator + attr[nattr++] = 0; + attr[nattr++] = 0; + + // Call callbacks. + if (start && startelCb) + (*startelCb)(ud, name, attr); + if (end && endelCb) + (*endelCb)(ud, name); +} + +int nsvg__parseXML(char* input, + void (*startelCb)(void* ud, const char* el, const char** attr), + void (*endelCb)(void* ud, const char* el), + void (*contentCb)(void* ud, const char* s), + void* ud) +{ + char* s = input; + char* mark = s; + int state = NSVG_XML_CONTENT; + while (*s) { + if (*s == '<' && state == NSVG_XML_CONTENT) { + // Start of a tag + *s++ = '\0'; + nsvg__parseContent(mark, contentCb, ud); + mark = s; + state = NSVG_XML_TAG; + } else if (*s == '>' && state == NSVG_XML_TAG) { + // Start of a content or new tag. + *s++ = '\0'; + nsvg__parseElement(mark, startelCb, endelCb, ud); + mark = s; + state = NSVG_XML_CONTENT; + } else { + s++; + } + } + + return 1; +} + + +/* Simple SVG parser. */ + +#define NSVG_MAX_ATTR 128 + +enum NSVGgradientUnits { + NSVG_USER_SPACE = 0, + NSVG_OBJECT_SPACE = 1 +}; + +#define NSVG_MAX_DASHES 8 + +enum NSVGunits { + NSVG_UNITS_USER, + NSVG_UNITS_PX, + NSVG_UNITS_PT, + NSVG_UNITS_PC, + NSVG_UNITS_MM, + NSVG_UNITS_CM, + NSVG_UNITS_IN, + NSVG_UNITS_PERCENT, + NSVG_UNITS_EM, + NSVG_UNITS_EX +}; + +typedef struct NSVGcoordinate { + float value; + int units; +} NSVGcoordinate; + +typedef struct NSVGlinearData { + NSVGcoordinate x1, y1, x2, y2; +} NSVGlinearData; + +typedef struct NSVGradialData { + NSVGcoordinate cx, cy, r, fx, fy; +} NSVGradialData; + +typedef struct NSVGgradientData +{ + char id[64]; + char ref[64]; + char type; + union { + NSVGlinearData linear; + NSVGradialData radial; + }; + char spread; + char units; + float xform[6]; + int nstops; + NSVGgradientStop* stops; + struct NSVGgradientData* next; +} NSVGgradientData; + +typedef struct NSVGattrib +{ + char id[64]; + float xform[6]; + unsigned int fillColor; + unsigned int strokeColor; + float opacity; + float fillOpacity; + float strokeOpacity; + char fillGradient[64]; + char strokeGradient[64]; + float strokeWidth; + float strokeDashOffset; + float strokeDashArray[NSVG_MAX_DASHES]; + int strokeDashCount; + char strokeLineJoin; + char strokeLineCap; + float miterLimit; + char fillRule; + float fontSize; + unsigned int stopColor; + float stopOpacity; + float stopOffset; + char hasFill; + char hasStroke; + char visible; +} NSVGattrib; + +typedef struct NSVGparser +{ + NSVGattrib attr[NSVG_MAX_ATTR]; + int attrHead; + float* pts; + int npts; + int cpts; + NSVGpath* plist; + NSVGimage* image; + NSVGgradientData* gradients; + NSVGshape* shapesTail; + float viewMinx, viewMiny, viewWidth, viewHeight; + int alignX, alignY, alignType; + float dpi; + char pathFlag; + char defsFlag; +} NSVGparser; + +static void nsvg__xformIdentity(float* t) +{ + t[0] = 1.0f; t[1] = 0.0f; + t[2] = 0.0f; t[3] = 1.0f; + t[4] = 0.0f; t[5] = 0.0f; +} + +static void nsvg__xformSetTranslation(float* t, float tx, float ty) +{ + t[0] = 1.0f; t[1] = 0.0f; + t[2] = 0.0f; t[3] = 1.0f; + t[4] = tx; t[5] = ty; +} + +static void nsvg__xformSetScale(float* t, float sx, float sy) +{ + t[0] = sx; t[1] = 0.0f; + t[2] = 0.0f; t[3] = sy; + t[4] = 0.0f; t[5] = 0.0f; +} + +static void nsvg__xformSetSkewX(float* t, float a) +{ + t[0] = 1.0f; t[1] = 0.0f; + t[2] = tanf(a); t[3] = 1.0f; + t[4] = 0.0f; t[5] = 0.0f; +} + +static void nsvg__xformSetSkewY(float* t, float a) +{ + t[0] = 1.0f; t[1] = tanf(a); + t[2] = 0.0f; t[3] = 1.0f; + t[4] = 0.0f; t[5] = 0.0f; +} + +static void nsvg__xformSetRotation(float* t, float a) +{ + float cs = cosf(a), sn = sinf(a); + t[0] = cs; t[1] = sn; + t[2] = -sn; t[3] = cs; + t[4] = 0.0f; t[5] = 0.0f; +} + +static void nsvg__xformMultiply(float* t, float* s) +{ + float t0 = t[0] * s[0] + t[1] * s[2]; + float t2 = t[2] * s[0] + t[3] * s[2]; + float t4 = t[4] * s[0] + t[5] * s[2] + s[4]; + t[1] = t[0] * s[1] + t[1] * s[3]; + t[3] = t[2] * s[1] + t[3] * s[3]; + t[5] = t[4] * s[1] + t[5] * s[3] + s[5]; + t[0] = t0; + t[2] = t2; + t[4] = t4; +} + +static void nsvg__xformInverse(float* inv, float* t) +{ + double invdet, det = (double)t[0] * t[3] - (double)t[2] * t[1]; + if (det > -1e-6 && det < 1e-6) { + nsvg__xformIdentity(t); + return; + } + invdet = 1.0 / det; + inv[0] = (float)(t[3] * invdet); + inv[2] = (float)(-t[2] * invdet); + inv[4] = (float)(((double)t[2] * t[5] - (double)t[3] * t[4]) * invdet); + inv[1] = (float)(-t[1] * invdet); + inv[3] = (float)(t[0] * invdet); + inv[5] = (float)(((double)t[1] * t[4] - (double)t[0] * t[5]) * invdet); +} + +static void nsvg__xformPremultiply(float* t, float* s) +{ + float s2[6]; + memcpy(s2, s, sizeof(float)*6); + nsvg__xformMultiply(s2, t); + memcpy(t, s2, sizeof(float)*6); +} + +static void nsvg__xformPoint(float* dx, float* dy, float x, float y, float* t) +{ + *dx = x*t[0] + y*t[2] + t[4]; + *dy = x*t[1] + y*t[3] + t[5]; +} + +static void nsvg__xformVec(float* dx, float* dy, float x, float y, float* t) +{ + *dx = x*t[0] + y*t[2]; + *dy = x*t[1] + y*t[3]; +} + +#define NSVG_EPSILON (1e-12) + +static int nsvg__ptInBounds(float* pt, float* bounds) +{ + return pt[0] >= bounds[0] && pt[0] <= bounds[2] && pt[1] >= bounds[1] && pt[1] <= bounds[3]; +} + + +static double nsvg__evalBezier(double t, double p0, double p1, double p2, double p3) +{ + double it = 1.0-t; + return it*it*it*p0 + 3.0*it*it*t*p1 + 3.0*it*t*t*p2 + t*t*t*p3; +} + +static void nsvg__curveBounds(float* bounds, float* curve) +{ + int i, j, count; + double roots[2], a, b, c, b2ac, t, v; + float* v0 = &curve[0]; + float* v1 = &curve[2]; + float* v2 = &curve[4]; + float* v3 = &curve[6]; + + // Start the bounding box by end points + bounds[0] = nsvg__minf(v0[0], v3[0]); + bounds[1] = nsvg__minf(v0[1], v3[1]); + bounds[2] = nsvg__maxf(v0[0], v3[0]); + bounds[3] = nsvg__maxf(v0[1], v3[1]); + + // Bezier curve fits inside the convex hull of it's control points. + // If control points are inside the bounds, we're done. + if (nsvg__ptInBounds(v1, bounds) && nsvg__ptInBounds(v2, bounds)) + return; + + // Add bezier curve inflection points in X and Y. + for (i = 0; i < 2; i++) { + a = -3.0 * v0[i] + 9.0 * v1[i] - 9.0 * v2[i] + 3.0 * v3[i]; + b = 6.0 * v0[i] - 12.0 * v1[i] + 6.0 * v2[i]; + c = 3.0 * v1[i] - 3.0 * v0[i]; + count = 0; + if (fabs(a) < NSVG_EPSILON) { + if (fabs(b) > NSVG_EPSILON) { + t = -c / b; + if (t > NSVG_EPSILON && t < 1.0-NSVG_EPSILON) + roots[count++] = t; + } + } else { + b2ac = b*b - 4.0*c*a; + if (b2ac > NSVG_EPSILON) { + t = (-b + sqrt(b2ac)) / (2.0 * a); + if (t > NSVG_EPSILON && t < 1.0-NSVG_EPSILON) + roots[count++] = t; + t = (-b - sqrt(b2ac)) / (2.0 * a); + if (t > NSVG_EPSILON && t < 1.0-NSVG_EPSILON) + roots[count++] = t; + } + } + for (j = 0; j < count; j++) { + v = nsvg__evalBezier(roots[j], v0[i], v1[i], v2[i], v3[i]); + bounds[0+i] = nsvg__minf(bounds[0+i], (float)v); + bounds[2+i] = nsvg__maxf(bounds[2+i], (float)v); + } + } +} + +static NSVGparser* nsvg__createParser() +{ + NSVGparser* p; + p = (NSVGparser*)malloc(sizeof(NSVGparser)); + if (p == NULL) goto error; + memset(p, 0, sizeof(NSVGparser)); + + p->image = (NSVGimage*)malloc(sizeof(NSVGimage)); + if (p->image == NULL) goto error; + memset(p->image, 0, sizeof(NSVGimage)); + + // Init style + nsvg__xformIdentity(p->attr[0].xform); + memset(p->attr[0].id, 0, sizeof p->attr[0].id); + p->attr[0].fillColor = NSVG_RGB(0,0,0); + p->attr[0].strokeColor = NSVG_RGB(0,0,0); + p->attr[0].opacity = 1; + p->attr[0].fillOpacity = 1; + p->attr[0].strokeOpacity = 1; + p->attr[0].stopOpacity = 1; + p->attr[0].strokeWidth = 1; + p->attr[0].strokeLineJoin = NSVG_JOIN_MITER; + p->attr[0].strokeLineCap = NSVG_CAP_BUTT; + p->attr[0].miterLimit = 4; + p->attr[0].fillRule = NSVG_FILLRULE_NONZERO; + p->attr[0].hasFill = 1; + p->attr[0].visible = 1; + + return p; + +error: + if (p) { + if (p->image) free(p->image); + free(p); + } + return NULL; +} + +static void nsvg__deletePaths(NSVGpath* path) +{ + while (path) { + NSVGpath *next = path->next; + if (path->pts != NULL) + free(path->pts); + free(path); + path = next; + } +} + +static void nsvg__deletePaint(NSVGpaint* paint) +{ + if (paint->type == NSVG_PAINT_LINEAR_GRADIENT || paint->type == NSVG_PAINT_RADIAL_GRADIENT) + free(paint->gradient); +} + +static void nsvg__deleteGradientData(NSVGgradientData* grad) +{ + NSVGgradientData* next; + while (grad != NULL) { + next = grad->next; + free(grad->stops); + free(grad); + grad = next; + } +} + +static void nsvg__deleteParser(NSVGparser* p) +{ + if (p != NULL) { + nsvg__deletePaths(p->plist); + nsvg__deleteGradientData(p->gradients); + nsvgDelete(p->image); + free(p->pts); + free(p); + } +} + +static void nsvg__resetPath(NSVGparser* p) +{ + p->npts = 0; +} + +static void nsvg__addPoint(NSVGparser* p, float x, float y) +{ + if (p->npts+1 > p->cpts) { + p->cpts = p->cpts ? p->cpts*2 : 8; + p->pts = (float*)realloc(p->pts, p->cpts*2*sizeof(float)); + if (!p->pts) return; + } + p->pts[p->npts*2+0] = x; + p->pts[p->npts*2+1] = y; + p->npts++; +} + +static void nsvg__moveTo(NSVGparser* p, float x, float y) +{ + if (p->npts > 0) { + p->pts[(p->npts-1)*2+0] = x; + p->pts[(p->npts-1)*2+1] = y; + } else { + nsvg__addPoint(p, x, y); + } +} + +static void nsvg__lineTo(NSVGparser* p, float x, float y) +{ + float px,py, dx,dy; + if (p->npts > 0) { + px = p->pts[(p->npts-1)*2+0]; + py = p->pts[(p->npts-1)*2+1]; + dx = x - px; + dy = y - py; + nsvg__addPoint(p, px + dx/3.0f, py + dy/3.0f); + nsvg__addPoint(p, x - dx/3.0f, y - dy/3.0f); + nsvg__addPoint(p, x, y); + } +} + +static void nsvg__cubicBezTo(NSVGparser* p, float cpx1, float cpy1, float cpx2, float cpy2, float x, float y) +{ + nsvg__addPoint(p, cpx1, cpy1); + nsvg__addPoint(p, cpx2, cpy2); + nsvg__addPoint(p, x, y); +} + +static NSVGattrib* nsvg__getAttr(NSVGparser* p) +{ + return &p->attr[p->attrHead]; +} + +static void nsvg__pushAttr(NSVGparser* p) +{ + if (p->attrHead < NSVG_MAX_ATTR-1) { + p->attrHead++; + memcpy(&p->attr[p->attrHead], &p->attr[p->attrHead-1], sizeof(NSVGattrib)); + } +} + +static void nsvg__popAttr(NSVGparser* p) +{ + if (p->attrHead > 0) + p->attrHead--; +} + +static float nsvg__actualOrigX(NSVGparser* p) +{ + return p->viewMinx; +} + +static float nsvg__actualOrigY(NSVGparser* p) +{ + return p->viewMiny; +} + +static float nsvg__actualWidth(NSVGparser* p) +{ + return p->viewWidth; +} + +static float nsvg__actualHeight(NSVGparser* p) +{ + return p->viewHeight; +} + +static float nsvg__actualLength(NSVGparser* p) +{ + float w = nsvg__actualWidth(p), h = nsvg__actualHeight(p); + return sqrtf(w*w + h*h) / sqrtf(2.0f); +} + +static float nsvg__convertToPixels(NSVGparser* p, NSVGcoordinate c, float orig, float length) +{ + NSVGattrib* attr = nsvg__getAttr(p); + switch (c.units) { + case NSVG_UNITS_USER: return c.value; + case NSVG_UNITS_PX: return c.value; + case NSVG_UNITS_PT: return c.value / 72.0f * p->dpi; + case NSVG_UNITS_PC: return c.value / 6.0f * p->dpi; + case NSVG_UNITS_MM: return c.value / 25.4f * p->dpi; + case NSVG_UNITS_CM: return c.value / 2.54f * p->dpi; + case NSVG_UNITS_IN: return c.value * p->dpi; + case NSVG_UNITS_EM: return c.value * attr->fontSize; + case NSVG_UNITS_EX: return c.value * attr->fontSize * 0.52f; // x-height of Helvetica. + case NSVG_UNITS_PERCENT: return orig + c.value / 100.0f * length; + default: return c.value; + } + return c.value; +} + +static NSVGgradientData* nsvg__findGradientData(NSVGparser* p, const char* id) +{ + NSVGgradientData* grad = p->gradients; + while (grad) { + if (strcmp(grad->id, id) == 0) + return grad; + grad = grad->next; + } + return NULL; +} + +static NSVGgradient* nsvg__createGradient(NSVGparser* p, const char* id, const float* localBounds, char* paintType) +{ + NSVGattrib* attr = nsvg__getAttr(p); + NSVGgradientData* data = NULL; + NSVGgradientData* ref = NULL; + NSVGgradientStop* stops = NULL; + NSVGgradient* grad; + float ox, oy, sw, sh, sl; + int nstops = 0; + + data = nsvg__findGradientData(p, id); + if (data == NULL) return NULL; + + // TODO: use ref to fill in all unset values too. + ref = data; + while (ref != NULL) { + if (stops == NULL && ref->stops != NULL) { + stops = ref->stops; + nstops = ref->nstops; + break; + } + ref = nsvg__findGradientData(p, ref->ref); + } + if (stops == NULL) return NULL; + + grad = (NSVGgradient*)malloc(sizeof(NSVGgradient) + sizeof(NSVGgradientStop)*(nstops-1)); + if (grad == NULL) return NULL; + + // The shape width and height. + if (data->units == NSVG_OBJECT_SPACE) { + ox = localBounds[0]; + oy = localBounds[1]; + sw = localBounds[2] - localBounds[0]; + sh = localBounds[3] - localBounds[1]; + } else { + ox = nsvg__actualOrigX(p); + oy = nsvg__actualOrigY(p); + sw = nsvg__actualWidth(p); + sh = nsvg__actualHeight(p); + } + sl = sqrtf(sw*sw + sh*sh) / sqrtf(2.0f); + + if (data->type == NSVG_PAINT_LINEAR_GRADIENT) { + float x1, y1, x2, y2, dx, dy; + x1 = nsvg__convertToPixels(p, data->linear.x1, ox, sw); + y1 = nsvg__convertToPixels(p, data->linear.y1, oy, sh); + x2 = nsvg__convertToPixels(p, data->linear.x2, ox, sw); + y2 = nsvg__convertToPixels(p, data->linear.y2, oy, sh); + // Calculate transform aligned to the line + dx = x2 - x1; + dy = y2 - y1; + grad->xform[0] = dy; grad->xform[1] = -dx; + grad->xform[2] = dx; grad->xform[3] = dy; + grad->xform[4] = x1; grad->xform[5] = y1; + } else { + float cx, cy, fx, fy, r; + cx = nsvg__convertToPixels(p, data->radial.cx, ox, sw); + cy = nsvg__convertToPixels(p, data->radial.cy, oy, sh); + fx = nsvg__convertToPixels(p, data->radial.fx, ox, sw); + fy = nsvg__convertToPixels(p, data->radial.fy, oy, sh); + r = nsvg__convertToPixels(p, data->radial.r, 0, sl); + // Calculate transform aligned to the circle + grad->xform[0] = r; grad->xform[1] = 0; + grad->xform[2] = 0; grad->xform[3] = r; + grad->xform[4] = cx; grad->xform[5] = cy; + grad->fx = fx / r; + grad->fy = fy / r; + } + + nsvg__xformMultiply(grad->xform, data->xform); + nsvg__xformMultiply(grad->xform, attr->xform); + + grad->spread = data->spread; + memcpy(grad->stops, stops, nstops*sizeof(NSVGgradientStop)); + grad->nstops = nstops; + + *paintType = data->type; + + return grad; +} + +static float nsvg__getAverageScale(float* t) +{ + float sx = sqrtf(t[0]*t[0] + t[2]*t[2]); + float sy = sqrtf(t[1]*t[1] + t[3]*t[3]); + return (sx + sy) * 0.5f; +} + +static void nsvg__getLocalBounds(float* bounds, NSVGshape *shape, float* xform) +{ + NSVGpath* path; + float curve[4*2], curveBounds[4]; + int i, first = 1; + for (path = shape->paths; path != NULL; path = path->next) { + nsvg__xformPoint(&curve[0], &curve[1], path->pts[0], path->pts[1], xform); + for (i = 0; i < path->npts-1; i += 3) { + nsvg__xformPoint(&curve[2], &curve[3], path->pts[(i+1)*2], path->pts[(i+1)*2+1], xform); + nsvg__xformPoint(&curve[4], &curve[5], path->pts[(i+2)*2], path->pts[(i+2)*2+1], xform); + nsvg__xformPoint(&curve[6], &curve[7], path->pts[(i+3)*2], path->pts[(i+3)*2+1], xform); + nsvg__curveBounds(curveBounds, curve); + if (first) { + bounds[0] = curveBounds[0]; + bounds[1] = curveBounds[1]; + bounds[2] = curveBounds[2]; + bounds[3] = curveBounds[3]; + first = 0; + } else { + bounds[0] = nsvg__minf(bounds[0], curveBounds[0]); + bounds[1] = nsvg__minf(bounds[1], curveBounds[1]); + bounds[2] = nsvg__maxf(bounds[2], curveBounds[2]); + bounds[3] = nsvg__maxf(bounds[3], curveBounds[3]); + } + curve[0] = curve[6]; + curve[1] = curve[7]; + } + } +} + +static void nsvg__addShape(NSVGparser* p) +{ + NSVGattrib* attr = nsvg__getAttr(p); + float scale = 1.0f; + NSVGshape* shape; + NSVGpath* path; + int i; + + if (p->plist == NULL) + return; + + shape = (NSVGshape*)malloc(sizeof(NSVGshape)); + if (shape == NULL) goto error; + memset(shape, 0, sizeof(NSVGshape)); + + memcpy(shape->id, attr->id, sizeof shape->id); + scale = nsvg__getAverageScale(attr->xform); + shape->strokeWidth = attr->strokeWidth * scale; + shape->strokeDashOffset = attr->strokeDashOffset * scale; + shape->strokeDashCount = (char)attr->strokeDashCount; + for (i = 0; i < attr->strokeDashCount; i++) + shape->strokeDashArray[i] = attr->strokeDashArray[i] * scale; + shape->strokeLineJoin = attr->strokeLineJoin; + shape->strokeLineCap = attr->strokeLineCap; + shape->miterLimit = attr->miterLimit; + shape->fillRule = attr->fillRule; + shape->opacity = attr->opacity; + + shape->paths = p->plist; + p->plist = NULL; + + // Calculate shape bounds + shape->bounds[0] = shape->paths->bounds[0]; + shape->bounds[1] = shape->paths->bounds[1]; + shape->bounds[2] = shape->paths->bounds[2]; + shape->bounds[3] = shape->paths->bounds[3]; + for (path = shape->paths->next; path != NULL; path = path->next) { + shape->bounds[0] = nsvg__minf(shape->bounds[0], path->bounds[0]); + shape->bounds[1] = nsvg__minf(shape->bounds[1], path->bounds[1]); + shape->bounds[2] = nsvg__maxf(shape->bounds[2], path->bounds[2]); + shape->bounds[3] = nsvg__maxf(shape->bounds[3], path->bounds[3]); + } + + // Set fill + if (attr->hasFill == 0) { + shape->fill.type = NSVG_PAINT_NONE; + } else if (attr->hasFill == 1) { + shape->fill.type = NSVG_PAINT_COLOR; + shape->fill.color = attr->fillColor; + shape->fill.color |= (unsigned int)(attr->fillOpacity*255) << 24; + } else if (attr->hasFill == 2) { + float inv[6], localBounds[4]; + nsvg__xformInverse(inv, attr->xform); + nsvg__getLocalBounds(localBounds, shape, inv); + shape->fill.gradient = nsvg__createGradient(p, attr->fillGradient, localBounds, &shape->fill.type); + if (shape->fill.gradient == NULL) { + shape->fill.type = NSVG_PAINT_NONE; + } + } + + // Set stroke + if (attr->hasStroke == 0) { + shape->stroke.type = NSVG_PAINT_NONE; + } else if (attr->hasStroke == 1) { + shape->stroke.type = NSVG_PAINT_COLOR; + shape->stroke.color = attr->strokeColor; + shape->stroke.color |= (unsigned int)(attr->strokeOpacity*255) << 24; + } else if (attr->hasStroke == 2) { + float inv[6], localBounds[4]; + nsvg__xformInverse(inv, attr->xform); + nsvg__getLocalBounds(localBounds, shape, inv); + shape->stroke.gradient = nsvg__createGradient(p, attr->strokeGradient, localBounds, &shape->stroke.type); + if (shape->stroke.gradient == NULL) + shape->stroke.type = NSVG_PAINT_NONE; + } + + // Set flags + shape->flags = (attr->visible ? NSVG_FLAGS_VISIBLE : 0x00); + + // Add to tail + if (p->image->shapes == NULL) + p->image->shapes = shape; + else + p->shapesTail->next = shape; + p->shapesTail = shape; + + return; + +error: + if (shape) free(shape); +} + +static void nsvg__addPath(NSVGparser* p, char closed) +{ + NSVGattrib* attr = nsvg__getAttr(p); + NSVGpath* path = NULL; + float bounds[4]; + float* curve; + int i; + + if (p->npts < 4) + return; + + if (closed) + nsvg__lineTo(p, p->pts[0], p->pts[1]); + + path = (NSVGpath*)malloc(sizeof(NSVGpath)); + if (path == NULL) goto error; + memset(path, 0, sizeof(NSVGpath)); + + path->pts = (float*)malloc(p->npts*2*sizeof(float)); + if (path->pts == NULL) goto error; + path->closed = closed; + path->npts = p->npts; + + // Transform path. + for (i = 0; i < p->npts; ++i) + nsvg__xformPoint(&path->pts[i*2], &path->pts[i*2+1], p->pts[i*2], p->pts[i*2+1], attr->xform); + + // Find bounds + for (i = 0; i < path->npts-1; i += 3) { + curve = &path->pts[i*2]; + nsvg__curveBounds(bounds, curve); + if (i == 0) { + path->bounds[0] = bounds[0]; + path->bounds[1] = bounds[1]; + path->bounds[2] = bounds[2]; + path->bounds[3] = bounds[3]; + } else { + path->bounds[0] = nsvg__minf(path->bounds[0], bounds[0]); + path->bounds[1] = nsvg__minf(path->bounds[1], bounds[1]); + path->bounds[2] = nsvg__maxf(path->bounds[2], bounds[2]); + path->bounds[3] = nsvg__maxf(path->bounds[3], bounds[3]); + } + } + + path->next = p->plist; + p->plist = path; + + return; + +error: + if (path != NULL) { + if (path->pts != NULL) free(path->pts); + free(path); + } +} + +// We roll our own string to float because the std library one uses locale and messes things up. +static double nsvg__atof(const char* s) +{ + char* cur = (char*)s; + char* end = NULL; + double res = 0.0, sign = 1.0; + long long intPart = 0, fracPart = 0; + char hasIntPart = 0, hasFracPart = 0; + + // Parse optional sign + if (*cur == '+') { + cur++; + } else if (*cur == '-') { + sign = -1; + cur++; + } + + // Parse integer part + if (nsvg__isdigit(*cur)) { + // Parse digit sequence + intPart = (double)strtoll(cur, &end, 10); + if (cur != end) { + res = (double)intPart; + hasIntPart = 1; + cur = end; + } + } + + // Parse fractional part. + if (*cur == '.') { + cur++; // Skip '.' + if (nsvg__isdigit(*cur)) { + // Parse digit sequence + fracPart = strtoll(cur, &end, 10); + if (cur != end) { + res += (double)fracPart / pow(10.0, (double)(end - cur)); + hasFracPart = 1; + cur = end; + } + } + } + + // A valid number should have integer or fractional part. + if (!hasIntPart && !hasFracPart) + return 0.0; + + // Parse optional exponent + if (*cur == 'e' || *cur == 'E') { + int expPart = 0; + cur++; // skip 'E' + expPart = strtol(cur, &end, 10); // Parse digit sequence with sign + if (cur != end) { + res *= pow(10.0, (double)expPart); + } + } + + return res * sign; +} + + +static const char* nsvg__parseNumber(const char* s, char* it, const int size) +{ + const int last = size-1; + int i = 0; + + // sign + if (*s == '-' || *s == '+') { + if (i < last) it[i++] = *s; + s++; + } + // integer part + while (*s && nsvg__isdigit(*s)) { + if (i < last) it[i++] = *s; + s++; + } + if (*s == '.') { + // decimal point + if (i < last) it[i++] = *s; + s++; + // fraction part + while (*s && nsvg__isdigit(*s)) { + if (i < last) it[i++] = *s; + s++; + } + } + // exponent + if (*s == 'e' || *s == 'E') { + if (i < last) it[i++] = *s; + s++; + if (*s == '-' || *s == '+') { + if (i < last) it[i++] = *s; + s++; + } + while (*s && nsvg__isdigit(*s)) { + if (i < last) it[i++] = *s; + s++; + } + } + it[i] = '\0'; + + return s; +} + +static const char* nsvg__getNextPathItem(const char* s, char* it) +{ + it[0] = '\0'; + // Skip white spaces and commas + while (*s && (nsvg__isspace(*s) || *s == ',')) s++; + if (!*s) return s; + if (*s == '-' || *s == '+' || *s == '.' || nsvg__isdigit(*s)) { + s = nsvg__parseNumber(s, it, 64); + } else { + // Parse command + it[0] = *s++; + it[1] = '\0'; + return s; + } + + return s; +} + +static unsigned int nsvg__parseColorHex(const char* str) +{ + unsigned int c = 0, r = 0, g = 0, b = 0; + int n = 0; + str++; // skip # + // Calculate number of characters. + while(str[n] && !nsvg__isspace(str[n])) + n++; + if (n == 6) { + sscanf(str, "%x", &c); + } else if (n == 3) { + sscanf(str, "%x", &c); + c = (c&0xf) | ((c&0xf0) << 4) | ((c&0xf00) << 8); + c |= c<<4; + } + r = (c >> 16) & 0xff; + g = (c >> 8) & 0xff; + b = c & 0xff; + return NSVG_RGB(r,g,b); +} + +static unsigned int nsvg__parseColorRGB(const char* str) +{ + int r = -1, g = -1, b = -1; + char s1[32]="", s2[32]=""; + sscanf(str + 4, "%d%[%%, \t]%d%[%%, \t]%d", &r, s1, &g, s2, &b); + if (strchr(s1, '%')) { + return NSVG_RGB((r*255)/100,(g*255)/100,(b*255)/100); + } else { + return NSVG_RGB(r,g,b); + } +} + +typedef struct NSVGNamedColor { + const char* name; + unsigned int color; +} NSVGNamedColor; + +NSVGNamedColor nsvg__colors[] = { + + { "red", NSVG_RGB(255, 0, 0) }, + { "green", NSVG_RGB( 0, 128, 0) }, + { "blue", NSVG_RGB( 0, 0, 255) }, + { "yellow", NSVG_RGB(255, 255, 0) }, + { "cyan", NSVG_RGB( 0, 255, 255) }, + { "magenta", NSVG_RGB(255, 0, 255) }, + { "black", NSVG_RGB( 0, 0, 0) }, + { "grey", NSVG_RGB(128, 128, 128) }, + { "gray", NSVG_RGB(128, 128, 128) }, + { "white", NSVG_RGB(255, 255, 255) }, + +#ifdef NANOSVG_ALL_COLOR_KEYWORDS + { "aliceblue", NSVG_RGB(240, 248, 255) }, + { "antiquewhite", NSVG_RGB(250, 235, 215) }, + { "aqua", NSVG_RGB( 0, 255, 255) }, + { "aquamarine", NSVG_RGB(127, 255, 212) }, + { "azure", NSVG_RGB(240, 255, 255) }, + { "beige", NSVG_RGB(245, 245, 220) }, + { "bisque", NSVG_RGB(255, 228, 196) }, + { "blanchedalmond", NSVG_RGB(255, 235, 205) }, + { "blueviolet", NSVG_RGB(138, 43, 226) }, + { "brown", NSVG_RGB(165, 42, 42) }, + { "burlywood", NSVG_RGB(222, 184, 135) }, + { "cadetblue", NSVG_RGB( 95, 158, 160) }, + { "chartreuse", NSVG_RGB(127, 255, 0) }, + { "chocolate", NSVG_RGB(210, 105, 30) }, + { "coral", NSVG_RGB(255, 127, 80) }, + { "cornflowerblue", NSVG_RGB(100, 149, 237) }, + { "cornsilk", NSVG_RGB(255, 248, 220) }, + { "crimson", NSVG_RGB(220, 20, 60) }, + { "darkblue", NSVG_RGB( 0, 0, 139) }, + { "darkcyan", NSVG_RGB( 0, 139, 139) }, + { "darkgoldenrod", NSVG_RGB(184, 134, 11) }, + { "darkgray", NSVG_RGB(169, 169, 169) }, + { "darkgreen", NSVG_RGB( 0, 100, 0) }, + { "darkgrey", NSVG_RGB(169, 169, 169) }, + { "darkkhaki", NSVG_RGB(189, 183, 107) }, + { "darkmagenta", NSVG_RGB(139, 0, 139) }, + { "darkolivegreen", NSVG_RGB( 85, 107, 47) }, + { "darkorange", NSVG_RGB(255, 140, 0) }, + { "darkorchid", NSVG_RGB(153, 50, 204) }, + { "darkred", NSVG_RGB(139, 0, 0) }, + { "darksalmon", NSVG_RGB(233, 150, 122) }, + { "darkseagreen", NSVG_RGB(143, 188, 143) }, + { "darkslateblue", NSVG_RGB( 72, 61, 139) }, + { "darkslategray", NSVG_RGB( 47, 79, 79) }, + { "darkslategrey", NSVG_RGB( 47, 79, 79) }, + { "darkturquoise", NSVG_RGB( 0, 206, 209) }, + { "darkviolet", NSVG_RGB(148, 0, 211) }, + { "deeppink", NSVG_RGB(255, 20, 147) }, + { "deepskyblue", NSVG_RGB( 0, 191, 255) }, + { "dimgray", NSVG_RGB(105, 105, 105) }, + { "dimgrey", NSVG_RGB(105, 105, 105) }, + { "dodgerblue", NSVG_RGB( 30, 144, 255) }, + { "firebrick", NSVG_RGB(178, 34, 34) }, + { "floralwhite", NSVG_RGB(255, 250, 240) }, + { "forestgreen", NSVG_RGB( 34, 139, 34) }, + { "fuchsia", NSVG_RGB(255, 0, 255) }, + { "gainsboro", NSVG_RGB(220, 220, 220) }, + { "ghostwhite", NSVG_RGB(248, 248, 255) }, + { "gold", NSVG_RGB(255, 215, 0) }, + { "goldenrod", NSVG_RGB(218, 165, 32) }, + { "greenyellow", NSVG_RGB(173, 255, 47) }, + { "honeydew", NSVG_RGB(240, 255, 240) }, + { "hotpink", NSVG_RGB(255, 105, 180) }, + { "indianred", NSVG_RGB(205, 92, 92) }, + { "indigo", NSVG_RGB( 75, 0, 130) }, + { "ivory", NSVG_RGB(255, 255, 240) }, + { "khaki", NSVG_RGB(240, 230, 140) }, + { "lavender", NSVG_RGB(230, 230, 250) }, + { "lavenderblush", NSVG_RGB(255, 240, 245) }, + { "lawngreen", NSVG_RGB(124, 252, 0) }, + { "lemonchiffon", NSVG_RGB(255, 250, 205) }, + { "lightblue", NSVG_RGB(173, 216, 230) }, + { "lightcoral", NSVG_RGB(240, 128, 128) }, + { "lightcyan", NSVG_RGB(224, 255, 255) }, + { "lightgoldenrodyellow", NSVG_RGB(250, 250, 210) }, + { "lightgray", NSVG_RGB(211, 211, 211) }, + { "lightgreen", NSVG_RGB(144, 238, 144) }, + { "lightgrey", NSVG_RGB(211, 211, 211) }, + { "lightpink", NSVG_RGB(255, 182, 193) }, + { "lightsalmon", NSVG_RGB(255, 160, 122) }, + { "lightseagreen", NSVG_RGB( 32, 178, 170) }, + { "lightskyblue", NSVG_RGB(135, 206, 250) }, + { "lightslategray", NSVG_RGB(119, 136, 153) }, + { "lightslategrey", NSVG_RGB(119, 136, 153) }, + { "lightsteelblue", NSVG_RGB(176, 196, 222) }, + { "lightyellow", NSVG_RGB(255, 255, 224) }, + { "lime", NSVG_RGB( 0, 255, 0) }, + { "limegreen", NSVG_RGB( 50, 205, 50) }, + { "linen", NSVG_RGB(250, 240, 230) }, + { "maroon", NSVG_RGB(128, 0, 0) }, + { "mediumaquamarine", NSVG_RGB(102, 205, 170) }, + { "mediumblue", NSVG_RGB( 0, 0, 205) }, + { "mediumorchid", NSVG_RGB(186, 85, 211) }, + { "mediumpurple", NSVG_RGB(147, 112, 219) }, + { "mediumseagreen", NSVG_RGB( 60, 179, 113) }, + { "mediumslateblue", NSVG_RGB(123, 104, 238) }, + { "mediumspringgreen", NSVG_RGB( 0, 250, 154) }, + { "mediumturquoise", NSVG_RGB( 72, 209, 204) }, + { "mediumvioletred", NSVG_RGB(199, 21, 133) }, + { "midnightblue", NSVG_RGB( 25, 25, 112) }, + { "mintcream", NSVG_RGB(245, 255, 250) }, + { "mistyrose", NSVG_RGB(255, 228, 225) }, + { "moccasin", NSVG_RGB(255, 228, 181) }, + { "navajowhite", NSVG_RGB(255, 222, 173) }, + { "navy", NSVG_RGB( 0, 0, 128) }, + { "oldlace", NSVG_RGB(253, 245, 230) }, + { "olive", NSVG_RGB(128, 128, 0) }, + { "olivedrab", NSVG_RGB(107, 142, 35) }, + { "orange", NSVG_RGB(255, 165, 0) }, + { "orangered", NSVG_RGB(255, 69, 0) }, + { "orchid", NSVG_RGB(218, 112, 214) }, + { "palegoldenrod", NSVG_RGB(238, 232, 170) }, + { "palegreen", NSVG_RGB(152, 251, 152) }, + { "paleturquoise", NSVG_RGB(175, 238, 238) }, + { "palevioletred", NSVG_RGB(219, 112, 147) }, + { "papayawhip", NSVG_RGB(255, 239, 213) }, + { "peachpuff", NSVG_RGB(255, 218, 185) }, + { "peru", NSVG_RGB(205, 133, 63) }, + { "pink", NSVG_RGB(255, 192, 203) }, + { "plum", NSVG_RGB(221, 160, 221) }, + { "powderblue", NSVG_RGB(176, 224, 230) }, + { "purple", NSVG_RGB(128, 0, 128) }, + { "rosybrown", NSVG_RGB(188, 143, 143) }, + { "royalblue", NSVG_RGB( 65, 105, 225) }, + { "saddlebrown", NSVG_RGB(139, 69, 19) }, + { "salmon", NSVG_RGB(250, 128, 114) }, + { "sandybrown", NSVG_RGB(244, 164, 96) }, + { "seagreen", NSVG_RGB( 46, 139, 87) }, + { "seashell", NSVG_RGB(255, 245, 238) }, + { "sienna", NSVG_RGB(160, 82, 45) }, + { "silver", NSVG_RGB(192, 192, 192) }, + { "skyblue", NSVG_RGB(135, 206, 235) }, + { "slateblue", NSVG_RGB(106, 90, 205) }, + { "slategray", NSVG_RGB(112, 128, 144) }, + { "slategrey", NSVG_RGB(112, 128, 144) }, + { "snow", NSVG_RGB(255, 250, 250) }, + { "springgreen", NSVG_RGB( 0, 255, 127) }, + { "steelblue", NSVG_RGB( 70, 130, 180) }, + { "tan", NSVG_RGB(210, 180, 140) }, + { "teal", NSVG_RGB( 0, 128, 128) }, + { "thistle", NSVG_RGB(216, 191, 216) }, + { "tomato", NSVG_RGB(255, 99, 71) }, + { "turquoise", NSVG_RGB( 64, 224, 208) }, + { "violet", NSVG_RGB(238, 130, 238) }, + { "wheat", NSVG_RGB(245, 222, 179) }, + { "whitesmoke", NSVG_RGB(245, 245, 245) }, + { "yellowgreen", NSVG_RGB(154, 205, 50) }, +#endif +}; + +static unsigned int nsvg__parseColorName(const char* str) +{ + int i, ncolors = sizeof(nsvg__colors) / sizeof(NSVGNamedColor); + + for (i = 0; i < ncolors; i++) { + if (strcmp(nsvg__colors[i].name, str) == 0) { + return nsvg__colors[i].color; + } + } + + return NSVG_RGB(128, 128, 128); +} + +static unsigned int nsvg__parseColor(const char* str) +{ + size_t len = 0; + while(*str == ' ') ++str; + len = strlen(str); + if (len >= 1 && *str == '#') + return nsvg__parseColorHex(str); + else if (len >= 4 && str[0] == 'r' && str[1] == 'g' && str[2] == 'b' && str[3] == '(') + return nsvg__parseColorRGB(str); + return nsvg__parseColorName(str); +} + +static float nsvg__parseOpacity(const char* str) +{ + float val = 0; + sscanf(str, "%f", &val); + if (val < 0.0f) val = 0.0f; + if (val > 1.0f) val = 1.0f; + return val; +} + +static float nsvg__parseMiterLimit(const char* str) +{ + float val = 0; + sscanf(str, "%f", &val); + if (val < 0.0f) val = 0.0f; + return val; +} + +static int nsvg__parseUnits(const char* units) +{ + if (units[0] == 'p' && units[1] == 'x') + return NSVG_UNITS_PX; + else if (units[0] == 'p' && units[1] == 't') + return NSVG_UNITS_PT; + else if (units[0] == 'p' && units[1] == 'c') + return NSVG_UNITS_PC; + else if (units[0] == 'm' && units[1] == 'm') + return NSVG_UNITS_MM; + else if (units[0] == 'c' && units[1] == 'm') + return NSVG_UNITS_CM; + else if (units[0] == 'i' && units[1] == 'n') + return NSVG_UNITS_IN; + else if (units[0] == '%') + return NSVG_UNITS_PERCENT; + else if (units[0] == 'e' && units[1] == 'm') + return NSVG_UNITS_EM; + else if (units[0] == 'e' && units[1] == 'x') + return NSVG_UNITS_EX; + return NSVG_UNITS_USER; +} + +static NSVGcoordinate nsvg__parseCoordinateRaw(const char* str) +{ + NSVGcoordinate coord = {0, NSVG_UNITS_USER}; + char units[32]=""; + sscanf(str, "%f%31s", &coord.value, units); + coord.units = nsvg__parseUnits(units); + return coord; +} + +static NSVGcoordinate nsvg__coord(float v, int units) +{ + NSVGcoordinate coord = {v, units}; + return coord; +} + +static float nsvg__parseCoordinate(NSVGparser* p, const char* str, float orig, float length) +{ + NSVGcoordinate coord = nsvg__parseCoordinateRaw(str); + return nsvg__convertToPixels(p, coord, orig, length); +} + +static int nsvg__parseTransformArgs(const char* str, float* args, int maxNa, int* na) +{ + const char* end; + const char* ptr; + char it[64]; + + *na = 0; + ptr = str; + while (*ptr && *ptr != '(') ++ptr; + if (*ptr == 0) + return 1; + end = ptr; + while (*end && *end != ')') ++end; + if (*end == 0) + return 1; + + while (ptr < end) { + if (*ptr == '-' || *ptr == '+' || *ptr == '.' || nsvg__isdigit(*ptr)) { + if (*na >= maxNa) return 0; + ptr = nsvg__parseNumber(ptr, it, 64); + args[(*na)++] = (float)nsvg__atof(it); + } else { + ++ptr; + } + } + return (int)(end - str); +} + + +static int nsvg__parseMatrix(float* xform, const char* str) +{ + float t[6]; + int na = 0; + int len = nsvg__parseTransformArgs(str, t, 6, &na); + if (na != 6) return len; + memcpy(xform, t, sizeof(float)*6); + return len; +} + +static int nsvg__parseTranslate(float* xform, const char* str) +{ + float args[2]; + float t[6]; + int na = 0; + int len = nsvg__parseTransformArgs(str, args, 2, &na); + if (na == 1) args[1] = 0.0; + + nsvg__xformSetTranslation(t, args[0], args[1]); + memcpy(xform, t, sizeof(float)*6); + return len; +} + +static int nsvg__parseScale(float* xform, const char* str) +{ + float args[2]; + int na = 0; + float t[6]; + int len = nsvg__parseTransformArgs(str, args, 2, &na); + if (na == 1) args[1] = args[0]; + nsvg__xformSetScale(t, args[0], args[1]); + memcpy(xform, t, sizeof(float)*6); + return len; +} + +static int nsvg__parseSkewX(float* xform, const char* str) +{ + float args[1]; + int na = 0; + float t[6]; + int len = nsvg__parseTransformArgs(str, args, 1, &na); + nsvg__xformSetSkewX(t, args[0]/180.0f*NSVG_PI); + memcpy(xform, t, sizeof(float)*6); + return len; +} + +static int nsvg__parseSkewY(float* xform, const char* str) +{ + float args[1]; + int na = 0; + float t[6]; + int len = nsvg__parseTransformArgs(str, args, 1, &na); + nsvg__xformSetSkewY(t, args[0]/180.0f*NSVG_PI); + memcpy(xform, t, sizeof(float)*6); + return len; +} + +static int nsvg__parseRotate(float* xform, const char* str) +{ + float args[3]; + int na = 0; + float m[6]; + float t[6]; + int len = nsvg__parseTransformArgs(str, args, 3, &na); + if (na == 1) + args[1] = args[2] = 0.0f; + nsvg__xformIdentity(m); + + if (na > 1) { + nsvg__xformSetTranslation(t, -args[1], -args[2]); + nsvg__xformMultiply(m, t); + } + + nsvg__xformSetRotation(t, args[0]/180.0f*NSVG_PI); + nsvg__xformMultiply(m, t); + + if (na > 1) { + nsvg__xformSetTranslation(t, args[1], args[2]); + nsvg__xformMultiply(m, t); + } + + memcpy(xform, m, sizeof(float)*6); + + return len; +} + +static void nsvg__parseTransform(float* xform, const char* str) +{ + float t[6]; + nsvg__xformIdentity(xform); + while (*str) + { + if (strncmp(str, "matrix", 6) == 0) + str += nsvg__parseMatrix(t, str); + else if (strncmp(str, "translate", 9) == 0) + str += nsvg__parseTranslate(t, str); + else if (strncmp(str, "scale", 5) == 0) + str += nsvg__parseScale(t, str); + else if (strncmp(str, "rotate", 6) == 0) + str += nsvg__parseRotate(t, str); + else if (strncmp(str, "skewX", 5) == 0) + str += nsvg__parseSkewX(t, str); + else if (strncmp(str, "skewY", 5) == 0) + str += nsvg__parseSkewY(t, str); + else{ + ++str; + continue; + } + + nsvg__xformPremultiply(xform, t); + } +} + +static void nsvg__parseUrl(char* id, const char* str) +{ + int i = 0; + str += 4; // "url("; + if (*str == '#') + str++; + while (i < 63 && *str != ')') { + id[i] = *str++; + i++; + } + id[i] = '\0'; +} + +static char nsvg__parseLineCap(const char* str) +{ + if (strcmp(str, "butt") == 0) + return NSVG_CAP_BUTT; + else if (strcmp(str, "round") == 0) + return NSVG_CAP_ROUND; + else if (strcmp(str, "square") == 0) + return NSVG_CAP_SQUARE; + // TODO: handle inherit. + return NSVG_CAP_BUTT; +} + +static char nsvg__parseLineJoin(const char* str) +{ + if (strcmp(str, "miter") == 0) + return NSVG_JOIN_MITER; + else if (strcmp(str, "round") == 0) + return NSVG_JOIN_ROUND; + else if (strcmp(str, "bevel") == 0) + return NSVG_JOIN_BEVEL; + // TODO: handle inherit. + return NSVG_CAP_BUTT; +} + +static char nsvg__parseFillRule(const char* str) +{ + if (strcmp(str, "nonzero") == 0) + return NSVG_FILLRULE_NONZERO; + else if (strcmp(str, "evenodd") == 0) + return NSVG_FILLRULE_EVENODD; + // TODO: handle inherit. + return NSVG_FILLRULE_NONZERO; +} + +static const char* nsvg__getNextDashItem(const char* s, char* it) +{ + int n = 0; + it[0] = '\0'; + // Skip white spaces and commas + while (*s && (nsvg__isspace(*s) || *s == ',')) s++; + // Advance until whitespace, comma or end. + while (*s && (!nsvg__isspace(*s) && *s != ',')) { + if (n < 63) + it[n++] = *s; + s++; + } + it[n++] = '\0'; + return s; +} + +static int nsvg__parseStrokeDashArray(NSVGparser* p, const char* str, float* strokeDashArray) +{ + char item[64]; + int count = 0, i; + float sum = 0.0f; + + // Handle "none" + if (str[0] == 'n') + return 0; + + // Parse dashes + while (*str) { + str = nsvg__getNextDashItem(str, item); + if (!*item) break; + if (count < NSVG_MAX_DASHES) + strokeDashArray[count++] = fabsf(nsvg__parseCoordinate(p, item, 0.0f, nsvg__actualLength(p))); + } + + for (i = 0; i < count; i++) + sum += strokeDashArray[i]; + if (sum <= 1e-6f) + count = 0; + + return count; +} + +static void nsvg__parseStyle(NSVGparser* p, const char* str); + +static int nsvg__parseAttr(NSVGparser* p, const char* name, const char* value) +{ + float xform[6]; + NSVGattrib* attr = nsvg__getAttr(p); + if (!attr) return 0; + + if (strcmp(name, "style") == 0) { + nsvg__parseStyle(p, value); + } else if (strcmp(name, "display") == 0) { + if (strcmp(value, "none") == 0) + attr->visible = 0; + // Don't reset ->visible on display:inline, one display:none hides the whole subtree + + } else if (strcmp(name, "fill") == 0) { + if (strcmp(value, "none") == 0) { + attr->hasFill = 0; + } else if (strncmp(value, "url(", 4) == 0) { + attr->hasFill = 2; + nsvg__parseUrl(attr->fillGradient, value); + } else { + attr->hasFill = 1; + attr->fillColor = nsvg__parseColor(value); + } + } else if (strcmp(name, "opacity") == 0) { + attr->opacity = nsvg__parseOpacity(value); + } else if (strcmp(name, "fill-opacity") == 0) { + attr->fillOpacity = nsvg__parseOpacity(value); + } else if (strcmp(name, "stroke") == 0) { + if (strcmp(value, "none") == 0) { + attr->hasStroke = 0; + } else if (strncmp(value, "url(", 4) == 0) { + attr->hasStroke = 2; + nsvg__parseUrl(attr->strokeGradient, value); + } else { + attr->hasStroke = 1; + attr->strokeColor = nsvg__parseColor(value); + } + } else if (strcmp(name, "stroke-width") == 0) { + attr->strokeWidth = nsvg__parseCoordinate(p, value, 0.0f, nsvg__actualLength(p)); + } else if (strcmp(name, "stroke-dasharray") == 0) { + attr->strokeDashCount = nsvg__parseStrokeDashArray(p, value, attr->strokeDashArray); + } else if (strcmp(name, "stroke-dashoffset") == 0) { + attr->strokeDashOffset = nsvg__parseCoordinate(p, value, 0.0f, nsvg__actualLength(p)); + } else if (strcmp(name, "stroke-opacity") == 0) { + attr->strokeOpacity = nsvg__parseOpacity(value); + } else if (strcmp(name, "stroke-linecap") == 0) { + attr->strokeLineCap = nsvg__parseLineCap(value); + } else if (strcmp(name, "stroke-linejoin") == 0) { + attr->strokeLineJoin = nsvg__parseLineJoin(value); + } else if (strcmp(name, "stroke-miterlimit") == 0) { + attr->miterLimit = nsvg__parseMiterLimit(value); + } else if (strcmp(name, "fill-rule") == 0) { + attr->fillRule = nsvg__parseFillRule(value); + } else if (strcmp(name, "font-size") == 0) { + attr->fontSize = nsvg__parseCoordinate(p, value, 0.0f, nsvg__actualLength(p)); + } else if (strcmp(name, "transform") == 0) { + nsvg__parseTransform(xform, value); + nsvg__xformPremultiply(attr->xform, xform); + } else if (strcmp(name, "stop-color") == 0) { + attr->stopColor = nsvg__parseColor(value); + } else if (strcmp(name, "stop-opacity") == 0) { + attr->stopOpacity = nsvg__parseOpacity(value); + } else if (strcmp(name, "offset") == 0) { + attr->stopOffset = nsvg__parseCoordinate(p, value, 0.0f, 1.0f); + } else if (strcmp(name, "id") == 0) { + strncpy(attr->id, value, 63); + attr->id[63] = '\0'; + } else { + return 0; + } + return 1; +} + +static int nsvg__parseNameValue(NSVGparser* p, const char* start, const char* end) +{ + const char* str; + const char* val; + char name[512]; + char value[512]; + int n; + + str = start; + while (str < end && *str != ':') ++str; + + val = str; + + // Right Trim + while (str > start && (*str == ':' || nsvg__isspace(*str))) --str; + ++str; + + n = (int)(str - start); + if (n > 511) n = 511; + if (n) memcpy(name, start, n); + name[n] = 0; + + while (val < end && (*val == ':' || nsvg__isspace(*val))) ++val; + + n = (int)(end - val); + if (n > 511) n = 511; + if (n) memcpy(value, val, n); + value[n] = 0; + + return nsvg__parseAttr(p, name, value); +} + +static void nsvg__parseStyle(NSVGparser* p, const char* str) +{ + const char* start; + const char* end; + + while (*str) { + // Left Trim + while(*str && nsvg__isspace(*str)) ++str; + start = str; + while(*str && *str != ';') ++str; + end = str; + + // Right Trim + while (end > start && (*end == ';' || nsvg__isspace(*end))) --end; + ++end; + + nsvg__parseNameValue(p, start, end); + if (*str) ++str; + } +} + +static void nsvg__parseAttribs(NSVGparser* p, const char** attr) +{ + int i; + for (i = 0; attr[i]; i += 2) + { + if (strcmp(attr[i], "style") == 0) + nsvg__parseStyle(p, attr[i + 1]); + else + nsvg__parseAttr(p, attr[i], attr[i + 1]); + } +} + +static int nsvg__getArgsPerElement(char cmd) +{ + switch (cmd) { + case 'v': + case 'V': + case 'h': + case 'H': + return 1; + case 'm': + case 'M': + case 'l': + case 'L': + case 't': + case 'T': + return 2; + case 'q': + case 'Q': + case 's': + case 'S': + return 4; + case 'c': + case 'C': + return 6; + case 'a': + case 'A': + return 7; + } + return 0; +} + +static void nsvg__pathMoveTo(NSVGparser* p, float* cpx, float* cpy, float* args, int rel) +{ + if (rel) { + *cpx += args[0]; + *cpy += args[1]; + } else { + *cpx = args[0]; + *cpy = args[1]; + } + nsvg__moveTo(p, *cpx, *cpy); +} + +static void nsvg__pathLineTo(NSVGparser* p, float* cpx, float* cpy, float* args, int rel) +{ + if (rel) { + *cpx += args[0]; + *cpy += args[1]; + } else { + *cpx = args[0]; + *cpy = args[1]; + } + nsvg__lineTo(p, *cpx, *cpy); +} + +static void nsvg__pathHLineTo(NSVGparser* p, float* cpx, float* cpy, float* args, int rel) +{ + if (rel) + *cpx += args[0]; + else + *cpx = args[0]; + nsvg__lineTo(p, *cpx, *cpy); +} + +static void nsvg__pathVLineTo(NSVGparser* p, float* cpx, float* cpy, float* args, int rel) +{ + if (rel) + *cpy += args[0]; + else + *cpy = args[0]; + nsvg__lineTo(p, *cpx, *cpy); +} + +static void nsvg__pathCubicBezTo(NSVGparser* p, float* cpx, float* cpy, + float* cpx2, float* cpy2, float* args, int rel) +{ + float x2, y2, cx1, cy1, cx2, cy2; + + if (rel) { + cx1 = *cpx + args[0]; + cy1 = *cpy + args[1]; + cx2 = *cpx + args[2]; + cy2 = *cpy + args[3]; + x2 = *cpx + args[4]; + y2 = *cpy + args[5]; + } else { + cx1 = args[0]; + cy1 = args[1]; + cx2 = args[2]; + cy2 = args[3]; + x2 = args[4]; + y2 = args[5]; + } + + nsvg__cubicBezTo(p, cx1,cy1, cx2,cy2, x2,y2); + + *cpx2 = cx2; + *cpy2 = cy2; + *cpx = x2; + *cpy = y2; +} + +static void nsvg__pathCubicBezShortTo(NSVGparser* p, float* cpx, float* cpy, + float* cpx2, float* cpy2, float* args, int rel) +{ + float x1, y1, x2, y2, cx1, cy1, cx2, cy2; + + x1 = *cpx; + y1 = *cpy; + if (rel) { + cx2 = *cpx + args[0]; + cy2 = *cpy + args[1]; + x2 = *cpx + args[2]; + y2 = *cpy + args[3]; + } else { + cx2 = args[0]; + cy2 = args[1]; + x2 = args[2]; + y2 = args[3]; + } + + cx1 = 2*x1 - *cpx2; + cy1 = 2*y1 - *cpy2; + + nsvg__cubicBezTo(p, cx1,cy1, cx2,cy2, x2,y2); + + *cpx2 = cx2; + *cpy2 = cy2; + *cpx = x2; + *cpy = y2; +} + +static void nsvg__pathQuadBezTo(NSVGparser* p, float* cpx, float* cpy, + float* cpx2, float* cpy2, float* args, int rel) +{ + float x1, y1, x2, y2, cx, cy; + float cx1, cy1, cx2, cy2; + + x1 = *cpx; + y1 = *cpy; + if (rel) { + cx = *cpx + args[0]; + cy = *cpy + args[1]; + x2 = *cpx + args[2]; + y2 = *cpy + args[3]; + } else { + cx = args[0]; + cy = args[1]; + x2 = args[2]; + y2 = args[3]; + } + + // Convert to cubic bezier + cx1 = x1 + 2.0f/3.0f*(cx - x1); + cy1 = y1 + 2.0f/3.0f*(cy - y1); + cx2 = x2 + 2.0f/3.0f*(cx - x2); + cy2 = y2 + 2.0f/3.0f*(cy - y2); + + nsvg__cubicBezTo(p, cx1,cy1, cx2,cy2, x2,y2); + + *cpx2 = cx; + *cpy2 = cy; + *cpx = x2; + *cpy = y2; +} + +static void nsvg__pathQuadBezShortTo(NSVGparser* p, float* cpx, float* cpy, + float* cpx2, float* cpy2, float* args, int rel) +{ + float x1, y1, x2, y2, cx, cy; + float cx1, cy1, cx2, cy2; + + x1 = *cpx; + y1 = *cpy; + if (rel) { + x2 = *cpx + args[0]; + y2 = *cpy + args[1]; + } else { + x2 = args[0]; + y2 = args[1]; + } + + cx = 2*x1 - *cpx2; + cy = 2*y1 - *cpy2; + + // Convert to cubix bezier + cx1 = x1 + 2.0f/3.0f*(cx - x1); + cy1 = y1 + 2.0f/3.0f*(cy - y1); + cx2 = x2 + 2.0f/3.0f*(cx - x2); + cy2 = y2 + 2.0f/3.0f*(cy - y2); + + nsvg__cubicBezTo(p, cx1,cy1, cx2,cy2, x2,y2); + + *cpx2 = cx; + *cpy2 = cy; + *cpx = x2; + *cpy = y2; +} + +static float nsvg__sqr(float x) { return x*x; } +static float nsvg__vmag(float x, float y) { return sqrtf(x*x + y*y); } + +static float nsvg__vecrat(float ux, float uy, float vx, float vy) +{ + return (ux*vx + uy*vy) / (nsvg__vmag(ux,uy) * nsvg__vmag(vx,vy)); +} + +static float nsvg__vecang(float ux, float uy, float vx, float vy) +{ + float r = nsvg__vecrat(ux,uy, vx,vy); + if (r < -1.0f) r = -1.0f; + if (r > 1.0f) r = 1.0f; + return ((ux*vy < uy*vx) ? -1.0f : 1.0f) * acosf(r); +} + +static void nsvg__pathArcTo(NSVGparser* p, float* cpx, float* cpy, float* args, int rel) +{ + // Ported from canvg (https://code.google.com/p/canvg/) + float rx, ry, rotx; + float x1, y1, x2, y2, cx, cy, dx, dy, d; + float x1p, y1p, cxp, cyp, s, sa, sb; + float ux, uy, vx, vy, a1, da; + float x, y, tanx, tany, a, px = 0, py = 0, ptanx = 0, ptany = 0, t[6]; + float sinrx, cosrx; + int fa, fs; + int i, ndivs; + float hda, kappa; + + rx = fabsf(args[0]); // y radius + ry = fabsf(args[1]); // x radius + rotx = args[2] / 180.0f * NSVG_PI; // x rotation angle + fa = fabsf(args[3]) > 1e-6 ? 1 : 0; // Large arc + fs = fabsf(args[4]) > 1e-6 ? 1 : 0; // Sweep direction + x1 = *cpx; // start point + y1 = *cpy; + if (rel) { // end point + x2 = *cpx + args[5]; + y2 = *cpy + args[6]; + } else { + x2 = args[5]; + y2 = args[6]; + } + + dx = x1 - x2; + dy = y1 - y2; + d = sqrtf(dx*dx + dy*dy); + if (d < 1e-6f || rx < 1e-6f || ry < 1e-6f) { + // The arc degenerates to a line + nsvg__lineTo(p, x2, y2); + *cpx = x2; + *cpy = y2; + return; + } + + sinrx = sinf(rotx); + cosrx = cosf(rotx); + + // Convert to center point parameterization. + // http://www.w3.org/TR/SVG11/implnote.html#ArcImplementationNotes + // 1) Compute x1', y1' + x1p = cosrx * dx / 2.0f + sinrx * dy / 2.0f; + y1p = -sinrx * dx / 2.0f + cosrx * dy / 2.0f; + d = nsvg__sqr(x1p)/nsvg__sqr(rx) + nsvg__sqr(y1p)/nsvg__sqr(ry); + if (d > 1) { + d = sqrtf(d); + rx *= d; + ry *= d; + } + // 2) Compute cx', cy' + s = 0.0f; + sa = nsvg__sqr(rx)*nsvg__sqr(ry) - nsvg__sqr(rx)*nsvg__sqr(y1p) - nsvg__sqr(ry)*nsvg__sqr(x1p); + sb = nsvg__sqr(rx)*nsvg__sqr(y1p) + nsvg__sqr(ry)*nsvg__sqr(x1p); + if (sa < 0.0f) sa = 0.0f; + if (sb > 0.0f) + s = sqrtf(sa / sb); + if (fa == fs) + s = -s; + cxp = s * rx * y1p / ry; + cyp = s * -ry * x1p / rx; + + // 3) Compute cx,cy from cx',cy' + cx = (x1 + x2)/2.0f + cosrx*cxp - sinrx*cyp; + cy = (y1 + y2)/2.0f + sinrx*cxp + cosrx*cyp; + + // 4) Calculate theta1, and delta theta. + ux = (x1p - cxp) / rx; + uy = (y1p - cyp) / ry; + vx = (-x1p - cxp) / rx; + vy = (-y1p - cyp) / ry; + a1 = nsvg__vecang(1.0f,0.0f, ux,uy); // Initial angle + da = nsvg__vecang(ux,uy, vx,vy); // Delta angle + +// if (vecrat(ux,uy,vx,vy) <= -1.0f) da = NSVG_PI; +// if (vecrat(ux,uy,vx,vy) >= 1.0f) da = 0; + + if (fs == 0 && da > 0) + da -= 2 * NSVG_PI; + else if (fs == 1 && da < 0) + da += 2 * NSVG_PI; + + // Approximate the arc using cubic spline segments. + t[0] = cosrx; t[1] = sinrx; + t[2] = -sinrx; t[3] = cosrx; + t[4] = cx; t[5] = cy; + + // Split arc into max 90 degree segments. + // The loop assumes an iteration per end point (including start and end), this +1. + ndivs = (int)(fabsf(da) / (NSVG_PI*0.5f) + 1.0f); + hda = (da / (float)ndivs) / 2.0f; + kappa = fabsf(4.0f / 3.0f * (1.0f - cosf(hda)) / sinf(hda)); + if (da < 0.0f) + kappa = -kappa; + + for (i = 0; i <= ndivs; i++) { + a = a1 + da * ((float)i/(float)ndivs); + dx = cosf(a); + dy = sinf(a); + nsvg__xformPoint(&x, &y, dx*rx, dy*ry, t); // position + nsvg__xformVec(&tanx, &tany, -dy*rx * kappa, dx*ry * kappa, t); // tangent + if (i > 0) + nsvg__cubicBezTo(p, px+ptanx,py+ptany, x-tanx, y-tany, x, y); + px = x; + py = y; + ptanx = tanx; + ptany = tany; + } + + *cpx = x2; + *cpy = y2; +} + +static void nsvg__parsePath(NSVGparser* p, const char** attr) +{ + const char* s = NULL; + char cmd = '\0'; + float args[10]; + int nargs; + int rargs = 0; + float cpx, cpy, cpx2, cpy2; + const char* tmp[4]; + char closedFlag; + int i; + char item[64]; + + for (i = 0; attr[i]; i += 2) { + if (strcmp(attr[i], "d") == 0) { + s = attr[i + 1]; + } else { + tmp[0] = attr[i]; + tmp[1] = attr[i + 1]; + tmp[2] = 0; + tmp[3] = 0; + nsvg__parseAttribs(p, tmp); + } + } + + if (s) { + nsvg__resetPath(p); + cpx = 0; cpy = 0; + cpx2 = 0; cpy2 = 0; + closedFlag = 0; + nargs = 0; + + while (*s) { + s = nsvg__getNextPathItem(s, item); + if (!*item) break; + if (nsvg__isnum(item[0])) { + if (nargs < 10) + args[nargs++] = (float)nsvg__atof(item); + if (nargs >= rargs) { + switch (cmd) { + case 'm': + case 'M': + nsvg__pathMoveTo(p, &cpx, &cpy, args, cmd == 'm' ? 1 : 0); + // Moveto can be followed by multiple coordinate pairs, + // which should be treated as linetos. + cmd = (cmd == 'm') ? 'l' : 'L'; + rargs = nsvg__getArgsPerElement(cmd); + cpx2 = cpx; cpy2 = cpy; + break; + case 'l': + case 'L': + nsvg__pathLineTo(p, &cpx, &cpy, args, cmd == 'l' ? 1 : 0); + cpx2 = cpx; cpy2 = cpy; + break; + case 'H': + case 'h': + nsvg__pathHLineTo(p, &cpx, &cpy, args, cmd == 'h' ? 1 : 0); + cpx2 = cpx; cpy2 = cpy; + break; + case 'V': + case 'v': + nsvg__pathVLineTo(p, &cpx, &cpy, args, cmd == 'v' ? 1 : 0); + cpx2 = cpx; cpy2 = cpy; + break; + case 'C': + case 'c': + nsvg__pathCubicBezTo(p, &cpx, &cpy, &cpx2, &cpy2, args, cmd == 'c' ? 1 : 0); + break; + case 'S': + case 's': + nsvg__pathCubicBezShortTo(p, &cpx, &cpy, &cpx2, &cpy2, args, cmd == 's' ? 1 : 0); + break; + case 'Q': + case 'q': + nsvg__pathQuadBezTo(p, &cpx, &cpy, &cpx2, &cpy2, args, cmd == 'q' ? 1 : 0); + break; + case 'T': + case 't': + nsvg__pathQuadBezShortTo(p, &cpx, &cpy, &cpx2, &cpy2, args, cmd == 't' ? 1 : 0); + break; + case 'A': + case 'a': + nsvg__pathArcTo(p, &cpx, &cpy, args, cmd == 'a' ? 1 : 0); + cpx2 = cpx; cpy2 = cpy; + break; + default: + if (nargs >= 2) { + cpx = args[nargs-2]; + cpy = args[nargs-1]; + cpx2 = cpx; cpy2 = cpy; + } + break; + } + nargs = 0; + } + } else { + cmd = item[0]; + rargs = nsvg__getArgsPerElement(cmd); + if (cmd == 'M' || cmd == 'm') { + // Commit path. + if (p->npts > 0) + nsvg__addPath(p, closedFlag); + // Start new subpath. + nsvg__resetPath(p); + closedFlag = 0; + nargs = 0; + } else if (cmd == 'Z' || cmd == 'z') { + closedFlag = 1; + // Commit path. + if (p->npts > 0) { + // Move current point to first point + cpx = p->pts[0]; + cpy = p->pts[1]; + cpx2 = cpx; cpy2 = cpy; + nsvg__addPath(p, closedFlag); + } + // Start new subpath. + nsvg__resetPath(p); + nsvg__moveTo(p, cpx, cpy); + closedFlag = 0; + nargs = 0; + } + } + } + // Commit path. + if (p->npts) + nsvg__addPath(p, closedFlag); + } + + nsvg__addShape(p); +} + +static void nsvg__parseRect(NSVGparser* p, const char** attr) +{ + float x = 0.0f; + float y = 0.0f; + float w = 0.0f; + float h = 0.0f; + float rx = -1.0f; // marks not set + float ry = -1.0f; + int i; + + for (i = 0; attr[i]; i += 2) { + if (!nsvg__parseAttr(p, attr[i], attr[i + 1])) { + if (strcmp(attr[i], "x") == 0) x = nsvg__parseCoordinate(p, attr[i+1], nsvg__actualOrigX(p), nsvg__actualWidth(p)); + if (strcmp(attr[i], "y") == 0) y = nsvg__parseCoordinate(p, attr[i+1], nsvg__actualOrigY(p), nsvg__actualHeight(p)); + if (strcmp(attr[i], "width") == 0) w = nsvg__parseCoordinate(p, attr[i+1], 0.0f, nsvg__actualWidth(p)); + if (strcmp(attr[i], "height") == 0) h = nsvg__parseCoordinate(p, attr[i+1], 0.0f, nsvg__actualHeight(p)); + if (strcmp(attr[i], "rx") == 0) rx = fabsf(nsvg__parseCoordinate(p, attr[i+1], 0.0f, nsvg__actualWidth(p))); + if (strcmp(attr[i], "ry") == 0) ry = fabsf(nsvg__parseCoordinate(p, attr[i+1], 0.0f, nsvg__actualHeight(p))); + } + } + + if (rx < 0.0f && ry > 0.0f) rx = ry; + if (ry < 0.0f && rx > 0.0f) ry = rx; + if (rx < 0.0f) rx = 0.0f; + if (ry < 0.0f) ry = 0.0f; + if (rx > w/2.0f) rx = w/2.0f; + if (ry > h/2.0f) ry = h/2.0f; + + if (w != 0.0f && h != 0.0f) { + nsvg__resetPath(p); + + if (rx < 0.00001f || ry < 0.0001f) { + nsvg__moveTo(p, x, y); + nsvg__lineTo(p, x+w, y); + nsvg__lineTo(p, x+w, y+h); + nsvg__lineTo(p, x, y+h); + } else { + // Rounded rectangle + nsvg__moveTo(p, x+rx, y); + nsvg__lineTo(p, x+w-rx, y); + nsvg__cubicBezTo(p, x+w-rx*(1-NSVG_KAPPA90), y, x+w, y+ry*(1-NSVG_KAPPA90), x+w, y+ry); + nsvg__lineTo(p, x+w, y+h-ry); + nsvg__cubicBezTo(p, x+w, y+h-ry*(1-NSVG_KAPPA90), x+w-rx*(1-NSVG_KAPPA90), y+h, x+w-rx, y+h); + nsvg__lineTo(p, x+rx, y+h); + nsvg__cubicBezTo(p, x+rx*(1-NSVG_KAPPA90), y+h, x, y+h-ry*(1-NSVG_KAPPA90), x, y+h-ry); + nsvg__lineTo(p, x, y+ry); + nsvg__cubicBezTo(p, x, y+ry*(1-NSVG_KAPPA90), x+rx*(1-NSVG_KAPPA90), y, x+rx, y); + } + + nsvg__addPath(p, 1); + + nsvg__addShape(p); + } +} + +static void nsvg__parseCircle(NSVGparser* p, const char** attr) +{ + float cx = 0.0f; + float cy = 0.0f; + float r = 0.0f; + int i; + + for (i = 0; attr[i]; i += 2) { + if (!nsvg__parseAttr(p, attr[i], attr[i + 1])) { + if (strcmp(attr[i], "cx") == 0) cx = nsvg__parseCoordinate(p, attr[i+1], nsvg__actualOrigX(p), nsvg__actualWidth(p)); + if (strcmp(attr[i], "cy") == 0) cy = nsvg__parseCoordinate(p, attr[i+1], nsvg__actualOrigY(p), nsvg__actualHeight(p)); + if (strcmp(attr[i], "r") == 0) r = fabsf(nsvg__parseCoordinate(p, attr[i+1], 0.0f, nsvg__actualLength(p))); + } + } + + if (r > 0.0f) { + nsvg__resetPath(p); + + nsvg__moveTo(p, cx+r, cy); + nsvg__cubicBezTo(p, cx+r, cy+r*NSVG_KAPPA90, cx+r*NSVG_KAPPA90, cy+r, cx, cy+r); + nsvg__cubicBezTo(p, cx-r*NSVG_KAPPA90, cy+r, cx-r, cy+r*NSVG_KAPPA90, cx-r, cy); + nsvg__cubicBezTo(p, cx-r, cy-r*NSVG_KAPPA90, cx-r*NSVG_KAPPA90, cy-r, cx, cy-r); + nsvg__cubicBezTo(p, cx+r*NSVG_KAPPA90, cy-r, cx+r, cy-r*NSVG_KAPPA90, cx+r, cy); + + nsvg__addPath(p, 1); + + nsvg__addShape(p); + } +} + +static void nsvg__parseEllipse(NSVGparser* p, const char** attr) +{ + float cx = 0.0f; + float cy = 0.0f; + float rx = 0.0f; + float ry = 0.0f; + int i; + + for (i = 0; attr[i]; i += 2) { + if (!nsvg__parseAttr(p, attr[i], attr[i + 1])) { + if (strcmp(attr[i], "cx") == 0) cx = nsvg__parseCoordinate(p, attr[i+1], nsvg__actualOrigX(p), nsvg__actualWidth(p)); + if (strcmp(attr[i], "cy") == 0) cy = nsvg__parseCoordinate(p, attr[i+1], nsvg__actualOrigY(p), nsvg__actualHeight(p)); + if (strcmp(attr[i], "rx") == 0) rx = fabsf(nsvg__parseCoordinate(p, attr[i+1], 0.0f, nsvg__actualWidth(p))); + if (strcmp(attr[i], "ry") == 0) ry = fabsf(nsvg__parseCoordinate(p, attr[i+1], 0.0f, nsvg__actualHeight(p))); + } + } + + if (rx > 0.0f && ry > 0.0f) { + + nsvg__resetPath(p); + + nsvg__moveTo(p, cx+rx, cy); + nsvg__cubicBezTo(p, cx+rx, cy+ry*NSVG_KAPPA90, cx+rx*NSVG_KAPPA90, cy+ry, cx, cy+ry); + nsvg__cubicBezTo(p, cx-rx*NSVG_KAPPA90, cy+ry, cx-rx, cy+ry*NSVG_KAPPA90, cx-rx, cy); + nsvg__cubicBezTo(p, cx-rx, cy-ry*NSVG_KAPPA90, cx-rx*NSVG_KAPPA90, cy-ry, cx, cy-ry); + nsvg__cubicBezTo(p, cx+rx*NSVG_KAPPA90, cy-ry, cx+rx, cy-ry*NSVG_KAPPA90, cx+rx, cy); + + nsvg__addPath(p, 1); + + nsvg__addShape(p); + } +} + +static void nsvg__parseLine(NSVGparser* p, const char** attr) +{ + float x1 = 0.0; + float y1 = 0.0; + float x2 = 0.0; + float y2 = 0.0; + int i; + + for (i = 0; attr[i]; i += 2) { + if (!nsvg__parseAttr(p, attr[i], attr[i + 1])) { + if (strcmp(attr[i], "x1") == 0) x1 = nsvg__parseCoordinate(p, attr[i + 1], nsvg__actualOrigX(p), nsvg__actualWidth(p)); + if (strcmp(attr[i], "y1") == 0) y1 = nsvg__parseCoordinate(p, attr[i + 1], nsvg__actualOrigY(p), nsvg__actualHeight(p)); + if (strcmp(attr[i], "x2") == 0) x2 = nsvg__parseCoordinate(p, attr[i + 1], nsvg__actualOrigX(p), nsvg__actualWidth(p)); + if (strcmp(attr[i], "y2") == 0) y2 = nsvg__parseCoordinate(p, attr[i + 1], nsvg__actualOrigY(p), nsvg__actualHeight(p)); + } + } + + nsvg__resetPath(p); + + nsvg__moveTo(p, x1, y1); + nsvg__lineTo(p, x2, y2); + + nsvg__addPath(p, 0); + + nsvg__addShape(p); +} + +static void nsvg__parsePoly(NSVGparser* p, const char** attr, int closeFlag) +{ + int i; + const char* s; + float args[2]; + int nargs, npts = 0; + char item[64]; + + nsvg__resetPath(p); + + for (i = 0; attr[i]; i += 2) { + if (!nsvg__parseAttr(p, attr[i], attr[i + 1])) { + if (strcmp(attr[i], "points") == 0) { + s = attr[i + 1]; + nargs = 0; + while (*s) { + s = nsvg__getNextPathItem(s, item); + args[nargs++] = (float)nsvg__atof(item); + if (nargs >= 2) { + if (npts == 0) + nsvg__moveTo(p, args[0], args[1]); + else + nsvg__lineTo(p, args[0], args[1]); + nargs = 0; + npts++; + } + } + } + } + } + + nsvg__addPath(p, (char)closeFlag); + + nsvg__addShape(p); +} + +static void nsvg__parseSVG(NSVGparser* p, const char** attr) +{ + int i; + for (i = 0; attr[i]; i += 2) { + if (!nsvg__parseAttr(p, attr[i], attr[i + 1])) { + if (strcmp(attr[i], "width") == 0) { + p->image->width = nsvg__parseCoordinate(p, attr[i + 1], 0.0f, 1.0f); + } else if (strcmp(attr[i], "height") == 0) { + p->image->height = nsvg__parseCoordinate(p, attr[i + 1], 0.0f, 1.0f); + } else if (strcmp(attr[i], "viewBox") == 0) { + sscanf(attr[i + 1], "%f%*[%%, \t]%f%*[%%, \t]%f%*[%%, \t]%f", &p->viewMinx, &p->viewMiny, &p->viewWidth, &p->viewHeight); + } else if (strcmp(attr[i], "preserveAspectRatio") == 0) { + if (strstr(attr[i + 1], "none") != 0) { + // No uniform scaling + p->alignType = NSVG_ALIGN_NONE; + } else { + // Parse X align + if (strstr(attr[i + 1], "xMin") != 0) + p->alignX = NSVG_ALIGN_MIN; + else if (strstr(attr[i + 1], "xMid") != 0) + p->alignX = NSVG_ALIGN_MID; + else if (strstr(attr[i + 1], "xMax") != 0) + p->alignX = NSVG_ALIGN_MAX; + // Parse X align + if (strstr(attr[i + 1], "yMin") != 0) + p->alignY = NSVG_ALIGN_MIN; + else if (strstr(attr[i + 1], "yMid") != 0) + p->alignY = NSVG_ALIGN_MID; + else if (strstr(attr[i + 1], "yMax") != 0) + p->alignY = NSVG_ALIGN_MAX; + // Parse meet/slice + p->alignType = NSVG_ALIGN_MEET; + if (strstr(attr[i + 1], "slice") != 0) + p->alignType = NSVG_ALIGN_SLICE; + } + } + } + } +} + +static void nsvg__parseGradient(NSVGparser* p, const char** attr, char type) +{ + int i; + NSVGgradientData* grad = (NSVGgradientData*)malloc(sizeof(NSVGgradientData)); + if (grad == NULL) return; + memset(grad, 0, sizeof(NSVGgradientData)); + grad->units = NSVG_OBJECT_SPACE; + grad->type = type; + if (grad->type == NSVG_PAINT_LINEAR_GRADIENT) { + grad->linear.x1 = nsvg__coord(0.0f, NSVG_UNITS_PERCENT); + grad->linear.y1 = nsvg__coord(0.0f, NSVG_UNITS_PERCENT); + grad->linear.x2 = nsvg__coord(100.0f, NSVG_UNITS_PERCENT); + grad->linear.y2 = nsvg__coord(0.0f, NSVG_UNITS_PERCENT); + } else if (grad->type == NSVG_PAINT_RADIAL_GRADIENT) { + grad->radial.cx = nsvg__coord(50.0f, NSVG_UNITS_PERCENT); + grad->radial.cy = nsvg__coord(50.0f, NSVG_UNITS_PERCENT); + grad->radial.r = nsvg__coord(50.0f, NSVG_UNITS_PERCENT); + } + + nsvg__xformIdentity(grad->xform); + + for (i = 0; attr[i]; i += 2) { + if (strcmp(attr[i], "id") == 0) { + strncpy(grad->id, attr[i+1], 63); + grad->id[63] = '\0'; + } else if (!nsvg__parseAttr(p, attr[i], attr[i + 1])) { + if (strcmp(attr[i], "gradientUnits") == 0) { + if (strcmp(attr[i+1], "objectBoundingBox") == 0) + grad->units = NSVG_OBJECT_SPACE; + else + grad->units = NSVG_USER_SPACE; + } else if (strcmp(attr[i], "gradientTransform") == 0) { + nsvg__parseTransform(grad->xform, attr[i + 1]); + } else if (strcmp(attr[i], "cx") == 0) { + grad->radial.cx = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "cy") == 0) { + grad->radial.cy = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "r") == 0) { + grad->radial.r = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "fx") == 0) { + grad->radial.fx = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "fy") == 0) { + grad->radial.fy = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "x1") == 0) { + grad->linear.x1 = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "y1") == 0) { + grad->linear.y1 = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "x2") == 0) { + grad->linear.x2 = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "y2") == 0) { + grad->linear.y2 = nsvg__parseCoordinateRaw(attr[i + 1]); + } else if (strcmp(attr[i], "spreadMethod") == 0) { + if (strcmp(attr[i+1], "pad") == 0) + grad->spread = NSVG_SPREAD_PAD; + else if (strcmp(attr[i+1], "reflect") == 0) + grad->spread = NSVG_SPREAD_REFLECT; + else if (strcmp(attr[i+1], "repeat") == 0) + grad->spread = NSVG_SPREAD_REPEAT; + } else if (strcmp(attr[i], "xlink:href") == 0) { + const char *href = attr[i+1]; + strncpy(grad->ref, href+1, 62); + grad->ref[62] = '\0'; + } + } + } + + grad->next = p->gradients; + p->gradients = grad; +} + +static void nsvg__parseGradientStop(NSVGparser* p, const char** attr) +{ + NSVGattrib* curAttr = nsvg__getAttr(p); + NSVGgradientData* grad; + NSVGgradientStop* stop; + int i, idx; + + curAttr->stopOffset = 0; + curAttr->stopColor = 0; + curAttr->stopOpacity = 1.0f; + + for (i = 0; attr[i]; i += 2) { + nsvg__parseAttr(p, attr[i], attr[i + 1]); + } + + // Add stop to the last gradient. + grad = p->gradients; + if (grad == NULL) return; + + grad->nstops++; + grad->stops = (NSVGgradientStop*)realloc(grad->stops, sizeof(NSVGgradientStop)*grad->nstops); + if (grad->stops == NULL) return; + + // Insert + idx = grad->nstops-1; + for (i = 0; i < grad->nstops-1; i++) { + if (curAttr->stopOffset < grad->stops[i].offset) { + idx = i; + break; + } + } + if (idx != grad->nstops-1) { + for (i = grad->nstops-1; i > idx; i--) + grad->stops[i] = grad->stops[i-1]; + } + + stop = &grad->stops[idx]; + stop->color = curAttr->stopColor; + stop->color |= (unsigned int)(curAttr->stopOpacity*255) << 24; + stop->offset = curAttr->stopOffset; +} + +static void nsvg__startElement(void* ud, const char* el, const char** attr) +{ + NSVGparser* p = (NSVGparser*)ud; + + if (p->defsFlag) { + // Skip everything but gradients in defs + if (strcmp(el, "linearGradient") == 0) { + nsvg__parseGradient(p, attr, NSVG_PAINT_LINEAR_GRADIENT); + } else if (strcmp(el, "radialGradient") == 0) { + nsvg__parseGradient(p, attr, NSVG_PAINT_RADIAL_GRADIENT); + } else if (strcmp(el, "stop") == 0) { + nsvg__parseGradientStop(p, attr); + } + return; + } + + if (strcmp(el, "g") == 0) { + nsvg__pushAttr(p); + nsvg__parseAttribs(p, attr); + } else if (strcmp(el, "path") == 0) { + if (p->pathFlag) // Do not allow nested paths. + return; + nsvg__pushAttr(p); + nsvg__parsePath(p, attr); + nsvg__popAttr(p); + } else if (strcmp(el, "rect") == 0) { + nsvg__pushAttr(p); + nsvg__parseRect(p, attr); + nsvg__popAttr(p); + } else if (strcmp(el, "circle") == 0) { + nsvg__pushAttr(p); + nsvg__parseCircle(p, attr); + nsvg__popAttr(p); + } else if (strcmp(el, "ellipse") == 0) { + nsvg__pushAttr(p); + nsvg__parseEllipse(p, attr); + nsvg__popAttr(p); + } else if (strcmp(el, "line") == 0) { + nsvg__pushAttr(p); + nsvg__parseLine(p, attr); + nsvg__popAttr(p); + } else if (strcmp(el, "polyline") == 0) { + nsvg__pushAttr(p); + nsvg__parsePoly(p, attr, 0); + nsvg__popAttr(p); + } else if (strcmp(el, "polygon") == 0) { + nsvg__pushAttr(p); + nsvg__parsePoly(p, attr, 1); + nsvg__popAttr(p); + } else if (strcmp(el, "linearGradient") == 0) { + nsvg__parseGradient(p, attr, NSVG_PAINT_LINEAR_GRADIENT); + } else if (strcmp(el, "radialGradient") == 0) { + nsvg__parseGradient(p, attr, NSVG_PAINT_RADIAL_GRADIENT); + } else if (strcmp(el, "stop") == 0) { + nsvg__parseGradientStop(p, attr); + } else if (strcmp(el, "defs") == 0) { + p->defsFlag = 1; + } else if (strcmp(el, "svg") == 0) { + nsvg__parseSVG(p, attr); + } +} + +static void nsvg__endElement(void* ud, const char* el) +{ + NSVGparser* p = (NSVGparser*)ud; + + if (strcmp(el, "g") == 0) { + nsvg__popAttr(p); + } else if (strcmp(el, "path") == 0) { + p->pathFlag = 0; + } else if (strcmp(el, "defs") == 0) { + p->defsFlag = 0; + } +} + +static void nsvg__content(void* ud, const char* s) +{ + NSVG_NOTUSED(ud); + NSVG_NOTUSED(s); + // empty +} + +static void nsvg__imageBounds(NSVGparser* p, float* bounds) +{ + NSVGshape* shape; + shape = p->image->shapes; + if (shape == NULL) { + bounds[0] = bounds[1] = bounds[2] = bounds[3] = 0.0; + return; + } + bounds[0] = shape->bounds[0]; + bounds[1] = shape->bounds[1]; + bounds[2] = shape->bounds[2]; + bounds[3] = shape->bounds[3]; + for (shape = shape->next; shape != NULL; shape = shape->next) { + bounds[0] = nsvg__minf(bounds[0], shape->bounds[0]); + bounds[1] = nsvg__minf(bounds[1], shape->bounds[1]); + bounds[2] = nsvg__maxf(bounds[2], shape->bounds[2]); + bounds[3] = nsvg__maxf(bounds[3], shape->bounds[3]); + } +} + +static float nsvg__viewAlign(float content, float container, int type) +{ + if (type == NSVG_ALIGN_MIN) + return 0; + else if (type == NSVG_ALIGN_MAX) + return container - content; + // mid + return (container - content) * 0.5f; +} + +static void nsvg__scaleGradient(NSVGgradient* grad, float tx, float ty, float sx, float sy) +{ + float t[6]; + nsvg__xformSetTranslation(t, tx, ty); + nsvg__xformMultiply (grad->xform, t); + + nsvg__xformSetScale(t, sx, sy); + nsvg__xformMultiply (grad->xform, t); +} + +static void nsvg__scaleToViewbox(NSVGparser* p, const char* units) +{ + NSVGshape* shape; + NSVGpath* path; + float tx, ty, sx, sy, us, bounds[4], t[6], avgs; + int i; + float* pt; + + // Guess image size if not set completely. + nsvg__imageBounds(p, bounds); + + if (p->viewWidth == 0) { + if (p->image->width > 0) { + p->viewWidth = p->image->width; + } else { + p->viewMinx = bounds[0]; + p->viewWidth = bounds[2] - bounds[0]; + } + } + if (p->viewHeight == 0) { + if (p->image->height > 0) { + p->viewHeight = p->image->height; + } else { + p->viewMiny = bounds[1]; + p->viewHeight = bounds[3] - bounds[1]; + } + } + if (p->image->width == 0) + p->image->width = p->viewWidth; + if (p->image->height == 0) + p->image->height = p->viewHeight; + + tx = -p->viewMinx; + ty = -p->viewMiny; + sx = p->viewWidth > 0 ? p->image->width / p->viewWidth : 0; + sy = p->viewHeight > 0 ? p->image->height / p->viewHeight : 0; + // Unit scaling + us = 1.0f / nsvg__convertToPixels(p, nsvg__coord(1.0f, nsvg__parseUnits(units)), 0.0f, 1.0f); + + // Fix aspect ratio + if (p->alignType == NSVG_ALIGN_MEET) { + // fit whole image into viewbox + sx = sy = nsvg__minf(sx, sy); + tx += nsvg__viewAlign(p->viewWidth*sx, p->image->width, p->alignX) / sx; + ty += nsvg__viewAlign(p->viewHeight*sy, p->image->height, p->alignY) / sy; + } else if (p->alignType == NSVG_ALIGN_SLICE) { + // fill whole viewbox with image + sx = sy = nsvg__maxf(sx, sy); + tx += nsvg__viewAlign(p->viewWidth*sx, p->image->width, p->alignX) / sx; + ty += nsvg__viewAlign(p->viewHeight*sy, p->image->height, p->alignY) / sy; + } + + // Transform + sx *= us; + sy *= us; + avgs = (sx+sy) / 2.0f; + for (shape = p->image->shapes; shape != NULL; shape = shape->next) { + shape->bounds[0] = (shape->bounds[0] + tx) * sx; + shape->bounds[1] = (shape->bounds[1] + ty) * sy; + shape->bounds[2] = (shape->bounds[2] + tx) * sx; + shape->bounds[3] = (shape->bounds[3] + ty) * sy; + for (path = shape->paths; path != NULL; path = path->next) { + path->bounds[0] = (path->bounds[0] + tx) * sx; + path->bounds[1] = (path->bounds[1] + ty) * sy; + path->bounds[2] = (path->bounds[2] + tx) * sx; + path->bounds[3] = (path->bounds[3] + ty) * sy; + for (i =0; i < path->npts; i++) { + pt = &path->pts[i*2]; + pt[0] = (pt[0] + tx) * sx; + pt[1] = (pt[1] + ty) * sy; + } + } + + if (shape->fill.type == NSVG_PAINT_LINEAR_GRADIENT || shape->fill.type == NSVG_PAINT_RADIAL_GRADIENT) { + nsvg__scaleGradient(shape->fill.gradient, tx,ty, sx,sy); + memcpy(t, shape->fill.gradient->xform, sizeof(float)*6); + nsvg__xformInverse(shape->fill.gradient->xform, t); + } + if (shape->stroke.type == NSVG_PAINT_LINEAR_GRADIENT || shape->stroke.type == NSVG_PAINT_RADIAL_GRADIENT) { + nsvg__scaleGradient(shape->stroke.gradient, tx,ty, sx,sy); + memcpy(t, shape->stroke.gradient->xform, sizeof(float)*6); + nsvg__xformInverse(shape->stroke.gradient->xform, t); + } + + shape->strokeWidth *= avgs; + shape->strokeDashOffset *= avgs; + for (i = 0; i < shape->strokeDashCount; i++) + shape->strokeDashArray[i] *= avgs; + } +} + +NSVGimage* nsvgParse(char* input, const char* units, float dpi) +{ + NSVGparser* p; + NSVGimage* ret = 0; + + p = nsvg__createParser(); + if (p == NULL) { + return NULL; + } + p->dpi = dpi; + + nsvg__parseXML(input, nsvg__startElement, nsvg__endElement, nsvg__content, p); + + // Scale to viewBox + nsvg__scaleToViewbox(p, units); + + ret = p->image; + p->image = NULL; + + nsvg__deleteParser(p); + + return ret; +} + +NSVGimage* nsvgParseFromFile(const char* filename, const char* units, float dpi) +{ + FILE* fp = NULL; + size_t size; + char* data = NULL; + NSVGimage* image = NULL; + + fp = fopen(filename, "rb"); + if (!fp) goto error; + fseek(fp, 0, SEEK_END); + size = ftell(fp); + fseek(fp, 0, SEEK_SET); + data = (char*)malloc(size+1); + if (data == NULL) goto error; + if (fread(data, 1, size, fp) != size) goto error; + data[size] = '\0'; // Must be null terminated. + fclose(fp); + image = nsvgParse(data, units, dpi); + free(data); + + return image; + +error: + if (fp) fclose(fp); + if (data) free(data); + if (image) nsvgDelete(image); + return NULL; +} + +void nsvgDelete(NSVGimage* image) +{ + NSVGshape *snext, *shape; + if (image == NULL) return; + shape = image->shapes; + while (shape != NULL) { + snext = shape->next; + nsvg__deletePaths(shape->paths); + nsvg__deletePaint(&shape->fill); + nsvg__deletePaint(&shape->stroke); + free(shape); + shape = snext; + } + free(image); +} + +#endif diff --git a/thirdparty/nanosvg/nanosvgrast.h b/thirdparty/nanosvg/nanosvgrast.h new file mode 100644 index 0000000000..2940c1f916 --- /dev/null +++ b/thirdparty/nanosvg/nanosvgrast.h @@ -0,0 +1,1447 @@ +/* + * Copyright (c) 2013-14 Mikko Mononen memon@inside.org + * + * This software is provided 'as-is', without any express or implied + * warranty. In no event will the authors be held liable for any damages + * arising from the use of this software. + * + * Permission is granted to anyone to use this software for any purpose, + * including commercial applications, and to alter it and redistribute it + * freely, subject to the following restrictions: + * + * 1. The origin of this software must not be misrepresented; you must not + * claim that you wrote the original software. If you use this software + * in a product, an acknowledgment in the product documentation would be + * appreciated but is not required. + * 2. Altered source versions must be plainly marked as such, and must not be + * misrepresented as being the original software. + * 3. This notice may not be removed or altered from any source distribution. + * + * The polygon rasterization is heavily based on stb_truetype rasterizer + * by Sean Barrett - http://nothings.org/ + * + */ + +#ifndef NANOSVGRAST_H +#define NANOSVGRAST_H + +#ifdef __cplusplus +extern "C" { +#endif + +typedef struct NSVGrasterizer NSVGrasterizer; + +/* Example Usage: + // Load SVG + struct SNVGImage* image = nsvgParseFromFile("test.svg."); + + // Create rasterizer (can be used to render multiple images). + struct NSVGrasterizer* rast = nsvgCreateRasterizer(); + // Allocate memory for image + unsigned char* img = malloc(w*h*4); + // Rasterize + nsvgRasterize(rast, image, 0,0,1, img, w, h, w*4); +*/ + +// Allocated rasterizer context. +NSVGrasterizer* nsvgCreateRasterizer(); + +// Rasterizes SVG image, returns RGBA image (non-premultiplied alpha) +// r - pointer to rasterizer context +// image - pointer to image to rasterize +// tx,ty - image offset (applied after scaling) +// scale - image scale +// dst - pointer to destination image data, 4 bytes per pixel (RGBA) +// w - width of the image to render +// h - height of the image to render +// stride - number of bytes per scaleline in the destination buffer +void nsvgRasterize(NSVGrasterizer* r, + NSVGimage* image, float tx, float ty, float scale, + unsigned char* dst, int w, int h, int stride); + +// Deletes rasterizer context. +void nsvgDeleteRasterizer(NSVGrasterizer*); + + +#ifdef __cplusplus +} +#endif + +#endif // NANOSVGRAST_H + +#ifdef NANOSVGRAST_IMPLEMENTATION + +#include <math.h> + +#define NSVG__SUBSAMPLES 5 +#define NSVG__FIXSHIFT 10 +#define NSVG__FIX (1 << NSVG__FIXSHIFT) +#define NSVG__FIXMASK (NSVG__FIX-1) +#define NSVG__MEMPAGE_SIZE 1024 + +typedef struct NSVGedge { + float x0,y0, x1,y1; + int dir; + struct NSVGedge* next; +} NSVGedge; + +typedef struct NSVGpoint { + float x, y; + float dx, dy; + float len; + float dmx, dmy; + unsigned char flags; +} NSVGpoint; + +typedef struct NSVGactiveEdge { + int x,dx; + float ey; + int dir; + struct NSVGactiveEdge *next; +} NSVGactiveEdge; + +typedef struct NSVGmemPage { + unsigned char mem[NSVG__MEMPAGE_SIZE]; + int size; + struct NSVGmemPage* next; +} NSVGmemPage; + +typedef struct NSVGcachedPaint { + char type; + char spread; + float xform[6]; + unsigned int colors[256]; +} NSVGcachedPaint; + +struct NSVGrasterizer +{ + float px, py; + + float tessTol; + float distTol; + + NSVGedge* edges; + int nedges; + int cedges; + + NSVGpoint* points; + int npoints; + int cpoints; + + NSVGpoint* points2; + int npoints2; + int cpoints2; + + NSVGactiveEdge* freelist; + NSVGmemPage* pages; + NSVGmemPage* curpage; + + unsigned char* scanline; + int cscanline; + + unsigned char* bitmap; + int width, height, stride; +}; + +NSVGrasterizer* nsvgCreateRasterizer() +{ + NSVGrasterizer* r = (NSVGrasterizer*)malloc(sizeof(NSVGrasterizer)); + if (r == NULL) goto error; + memset(r, 0, sizeof(NSVGrasterizer)); + + r->tessTol = 0.25f; + r->distTol = 0.01f; + + return r; + +error: + nsvgDeleteRasterizer(r); + return NULL; +} + +void nsvgDeleteRasterizer(NSVGrasterizer* r) +{ + NSVGmemPage* p; + + if (r == NULL) return; + + p = r->pages; + while (p != NULL) { + NSVGmemPage* next = p->next; + free(p); + p = next; + } + + if (r->edges) free(r->edges); + if (r->points) free(r->points); + if (r->points2) free(r->points2); + if (r->scanline) free(r->scanline); + + free(r); +} + +static NSVGmemPage* nsvg__nextPage(NSVGrasterizer* r, NSVGmemPage* cur) +{ + NSVGmemPage *newp; + + // If using existing chain, return the next page in chain + if (cur != NULL && cur->next != NULL) { + return cur->next; + } + + // Alloc new page + newp = (NSVGmemPage*)malloc(sizeof(NSVGmemPage)); + if (newp == NULL) return NULL; + memset(newp, 0, sizeof(NSVGmemPage)); + + // Add to linked list + if (cur != NULL) + cur->next = newp; + else + r->pages = newp; + + return newp; +} + +static void nsvg__resetPool(NSVGrasterizer* r) +{ + NSVGmemPage* p = r->pages; + while (p != NULL) { + p->size = 0; + p = p->next; + } + r->curpage = r->pages; +} + +static unsigned char* nsvg__alloc(NSVGrasterizer* r, int size) +{ + unsigned char* buf; + if (size > NSVG__MEMPAGE_SIZE) return NULL; + if (r->curpage == NULL || r->curpage->size+size > NSVG__MEMPAGE_SIZE) { + r->curpage = nsvg__nextPage(r, r->curpage); + } + buf = &r->curpage->mem[r->curpage->size]; + r->curpage->size += size; + return buf; +} + +static int nsvg__ptEquals(float x1, float y1, float x2, float y2, float tol) +{ + float dx = x2 - x1; + float dy = y2 - y1; + return dx*dx + dy*dy < tol*tol; +} + +static void nsvg__addPathPoint(NSVGrasterizer* r, float x, float y, int flags) +{ + NSVGpoint* pt; + + if (r->npoints > 0) { + pt = &r->points[r->npoints-1]; + if (nsvg__ptEquals(pt->x,pt->y, x,y, r->distTol)) { + pt->flags = (unsigned char)(pt->flags | flags); + return; + } + } + + if (r->npoints+1 > r->cpoints) { + r->cpoints = r->cpoints > 0 ? r->cpoints * 2 : 64; + r->points = (NSVGpoint*)realloc(r->points, sizeof(NSVGpoint) * r->cpoints); + if (r->points == NULL) return; + } + + pt = &r->points[r->npoints]; + pt->x = x; + pt->y = y; + pt->flags = (unsigned char)flags; + r->npoints++; +} + +static void nsvg__appendPathPoint(NSVGrasterizer* r, NSVGpoint pt) +{ + if (r->npoints+1 > r->cpoints) { + r->cpoints = r->cpoints > 0 ? r->cpoints * 2 : 64; + r->points = (NSVGpoint*)realloc(r->points, sizeof(NSVGpoint) * r->cpoints); + if (r->points == NULL) return; + } + r->points[r->npoints] = pt; + r->npoints++; +} + +static void nsvg__duplicatePoints(NSVGrasterizer* r) +{ + if (r->npoints > r->cpoints2) { + r->cpoints2 = r->npoints; + r->points2 = (NSVGpoint*)realloc(r->points2, sizeof(NSVGpoint) * r->cpoints2); + if (r->points2 == NULL) return; + } + + memcpy(r->points2, r->points, sizeof(NSVGpoint) * r->npoints); + r->npoints2 = r->npoints; +} + +static void nsvg__addEdge(NSVGrasterizer* r, float x0, float y0, float x1, float y1) +{ + NSVGedge* e; + + // Skip horizontal edges + if (y0 == y1) + return; + + if (r->nedges+1 > r->cedges) { + r->cedges = r->cedges > 0 ? r->cedges * 2 : 64; + r->edges = (NSVGedge*)realloc(r->edges, sizeof(NSVGedge) * r->cedges); + if (r->edges == NULL) return; + } + + e = &r->edges[r->nedges]; + r->nedges++; + + if (y0 < y1) { + e->x0 = x0; + e->y0 = y0; + e->x1 = x1; + e->y1 = y1; + e->dir = 1; + } else { + e->x0 = x1; + e->y0 = y1; + e->x1 = x0; + e->y1 = y0; + e->dir = -1; + } +} + +static float nsvg__normalize(float *x, float* y) +{ + float d = sqrtf((*x)*(*x) + (*y)*(*y)); + if (d > 1e-6f) { + float id = 1.0f / d; + *x *= id; + *y *= id; + } + return d; +} + +static float nsvg__absf(float x) { return x < 0 ? -x : x; } + +static void nsvg__flattenCubicBez(NSVGrasterizer* r, + float x1, float y1, float x2, float y2, + float x3, float y3, float x4, float y4, + int level, int type) +{ + float x12,y12,x23,y23,x34,y34,x123,y123,x234,y234,x1234,y1234; + float dx,dy,d2,d3; + + if (level > 10) return; + + x12 = (x1+x2)*0.5f; + y12 = (y1+y2)*0.5f; + x23 = (x2+x3)*0.5f; + y23 = (y2+y3)*0.5f; + x34 = (x3+x4)*0.5f; + y34 = (y3+y4)*0.5f; + x123 = (x12+x23)*0.5f; + y123 = (y12+y23)*0.5f; + + dx = x4 - x1; + dy = y4 - y1; + d2 = nsvg__absf(((x2 - x4) * dy - (y2 - y4) * dx)); + d3 = nsvg__absf(((x3 - x4) * dy - (y3 - y4) * dx)); + + if ((d2 + d3)*(d2 + d3) < r->tessTol * (dx*dx + dy*dy)) { + nsvg__addPathPoint(r, x4, y4, type); + return; + } + + x234 = (x23+x34)*0.5f; + y234 = (y23+y34)*0.5f; + x1234 = (x123+x234)*0.5f; + y1234 = (y123+y234)*0.5f; + + nsvg__flattenCubicBez(r, x1,y1, x12,y12, x123,y123, x1234,y1234, level+1, 0); + nsvg__flattenCubicBez(r, x1234,y1234, x234,y234, x34,y34, x4,y4, level+1, type); +} + +static void nsvg__flattenShape(NSVGrasterizer* r, NSVGshape* shape, float scale) +{ + int i, j; + NSVGpath* path; + + for (path = shape->paths; path != NULL; path = path->next) { + r->npoints = 0; + // Flatten path + nsvg__addPathPoint(r, path->pts[0]*scale, path->pts[1]*scale, 0); + for (i = 0; i < path->npts-1; i += 3) { + float* p = &path->pts[i*2]; + nsvg__flattenCubicBez(r, p[0]*scale,p[1]*scale, p[2]*scale,p[3]*scale, p[4]*scale,p[5]*scale, p[6]*scale,p[7]*scale, 0, 0); + } + // Close path + nsvg__addPathPoint(r, path->pts[0]*scale, path->pts[1]*scale, 0); + // Build edges + for (i = 0, j = r->npoints-1; i < r->npoints; j = i++) + nsvg__addEdge(r, r->points[j].x, r->points[j].y, r->points[i].x, r->points[i].y); + } +} + +enum NSVGpointFlags +{ + NSVG_PT_CORNER = 0x01, + NSVG_PT_BEVEL = 0x02, + NSVG_PT_LEFT = 0x04 +}; + +static void nsvg__initClosed(NSVGpoint* left, NSVGpoint* right, NSVGpoint* p0, NSVGpoint* p1, float lineWidth) +{ + float w = lineWidth * 0.5f; + float dx = p1->x - p0->x; + float dy = p1->y - p0->y; + float len = nsvg__normalize(&dx, &dy); + float px = p0->x + dx*len*0.5f, py = p0->y + dy*len*0.5f; + float dlx = dy, dly = -dx; + float lx = px - dlx*w, ly = py - dly*w; + float rx = px + dlx*w, ry = py + dly*w; + left->x = lx; left->y = ly; + right->x = rx; right->y = ry; +} + +static void nsvg__buttCap(NSVGrasterizer* r, NSVGpoint* left, NSVGpoint* right, NSVGpoint* p, float dx, float dy, float lineWidth, int connect) +{ + float w = lineWidth * 0.5f; + float px = p->x, py = p->y; + float dlx = dy, dly = -dx; + float lx = px - dlx*w, ly = py - dly*w; + float rx = px + dlx*w, ry = py + dly*w; + + nsvg__addEdge(r, lx, ly, rx, ry); + + if (connect) { + nsvg__addEdge(r, left->x, left->y, lx, ly); + nsvg__addEdge(r, rx, ry, right->x, right->y); + } + left->x = lx; left->y = ly; + right->x = rx; right->y = ry; +} + +static void nsvg__squareCap(NSVGrasterizer* r, NSVGpoint* left, NSVGpoint* right, NSVGpoint* p, float dx, float dy, float lineWidth, int connect) +{ + float w = lineWidth * 0.5f; + float px = p->x - dx*w, py = p->y - dy*w; + float dlx = dy, dly = -dx; + float lx = px - dlx*w, ly = py - dly*w; + float rx = px + dlx*w, ry = py + dly*w; + + nsvg__addEdge(r, lx, ly, rx, ry); + + if (connect) { + nsvg__addEdge(r, left->x, left->y, lx, ly); + nsvg__addEdge(r, rx, ry, right->x, right->y); + } + left->x = lx; left->y = ly; + right->x = rx; right->y = ry; +} + +#ifndef NSVG_PI +#define NSVG_PI (3.14159265358979323846264338327f) +#endif + +static void nsvg__roundCap(NSVGrasterizer* r, NSVGpoint* left, NSVGpoint* right, NSVGpoint* p, float dx, float dy, float lineWidth, int ncap, int connect) +{ + int i; + float w = lineWidth * 0.5f; + float px = p->x, py = p->y; + float dlx = dy, dly = -dx; + float lx = 0, ly = 0, rx = 0, ry = 0, prevx = 0, prevy = 0; + + for (i = 0; i < ncap; i++) { + float a = (float)i/(float)(ncap-1)*NSVG_PI; + float ax = cosf(a) * w, ay = sinf(a) * w; + float x = px - dlx*ax - dx*ay; + float y = py - dly*ax - dy*ay; + + if (i > 0) + nsvg__addEdge(r, prevx, prevy, x, y); + + prevx = x; + prevy = y; + + if (i == 0) { + lx = x; ly = y; + } else if (i == ncap-1) { + rx = x; ry = y; + } + } + + if (connect) { + nsvg__addEdge(r, left->x, left->y, lx, ly); + nsvg__addEdge(r, rx, ry, right->x, right->y); + } + + left->x = lx; left->y = ly; + right->x = rx; right->y = ry; +} + +static void nsvg__bevelJoin(NSVGrasterizer* r, NSVGpoint* left, NSVGpoint* right, NSVGpoint* p0, NSVGpoint* p1, float lineWidth) +{ + float w = lineWidth * 0.5f; + float dlx0 = p0->dy, dly0 = -p0->dx; + float dlx1 = p1->dy, dly1 = -p1->dx; + float lx0 = p1->x - (dlx0 * w), ly0 = p1->y - (dly0 * w); + float rx0 = p1->x + (dlx0 * w), ry0 = p1->y + (dly0 * w); + float lx1 = p1->x - (dlx1 * w), ly1 = p1->y - (dly1 * w); + float rx1 = p1->x + (dlx1 * w), ry1 = p1->y + (dly1 * w); + + nsvg__addEdge(r, lx0, ly0, left->x, left->y); + nsvg__addEdge(r, lx1, ly1, lx0, ly0); + + nsvg__addEdge(r, right->x, right->y, rx0, ry0); + nsvg__addEdge(r, rx0, ry0, rx1, ry1); + + left->x = lx1; left->y = ly1; + right->x = rx1; right->y = ry1; +} + +static void nsvg__miterJoin(NSVGrasterizer* r, NSVGpoint* left, NSVGpoint* right, NSVGpoint* p0, NSVGpoint* p1, float lineWidth) +{ + float w = lineWidth * 0.5f; + float dlx0 = p0->dy, dly0 = -p0->dx; + float dlx1 = p1->dy, dly1 = -p1->dx; + float lx0, rx0, lx1, rx1; + float ly0, ry0, ly1, ry1; + + if (p1->flags & NSVG_PT_LEFT) { + lx0 = lx1 = p1->x - p1->dmx * w; + ly0 = ly1 = p1->y - p1->dmy * w; + nsvg__addEdge(r, lx1, ly1, left->x, left->y); + + rx0 = p1->x + (dlx0 * w); + ry0 = p1->y + (dly0 * w); + rx1 = p1->x + (dlx1 * w); + ry1 = p1->y + (dly1 * w); + nsvg__addEdge(r, right->x, right->y, rx0, ry0); + nsvg__addEdge(r, rx0, ry0, rx1, ry1); + } else { + lx0 = p1->x - (dlx0 * w); + ly0 = p1->y - (dly0 * w); + lx1 = p1->x - (dlx1 * w); + ly1 = p1->y - (dly1 * w); + nsvg__addEdge(r, lx0, ly0, left->x, left->y); + nsvg__addEdge(r, lx1, ly1, lx0, ly0); + + rx0 = rx1 = p1->x + p1->dmx * w; + ry0 = ry1 = p1->y + p1->dmy * w; + nsvg__addEdge(r, right->x, right->y, rx1, ry1); + } + + left->x = lx1; left->y = ly1; + right->x = rx1; right->y = ry1; +} + +static void nsvg__roundJoin(NSVGrasterizer* r, NSVGpoint* left, NSVGpoint* right, NSVGpoint* p0, NSVGpoint* p1, float lineWidth, int ncap) +{ + int i, n; + float w = lineWidth * 0.5f; + float dlx0 = p0->dy, dly0 = -p0->dx; + float dlx1 = p1->dy, dly1 = -p1->dx; + float a0 = atan2f(dly0, dlx0); + float a1 = atan2f(dly1, dlx1); + float da = a1 - a0; + float lx, ly, rx, ry; + + if (da < NSVG_PI) da += NSVG_PI*2; + if (da > NSVG_PI) da -= NSVG_PI*2; + + n = (int)ceilf((nsvg__absf(da) / NSVG_PI) * (float)ncap); + if (n < 2) n = 2; + if (n > ncap) n = ncap; + + lx = left->x; + ly = left->y; + rx = right->x; + ry = right->y; + + for (i = 0; i < n; i++) { + float u = (float)i/(float)(n-1); + float a = a0 + u*da; + float ax = cosf(a) * w, ay = sinf(a) * w; + float lx1 = p1->x - ax, ly1 = p1->y - ay; + float rx1 = p1->x + ax, ry1 = p1->y + ay; + + nsvg__addEdge(r, lx1, ly1, lx, ly); + nsvg__addEdge(r, rx, ry, rx1, ry1); + + lx = lx1; ly = ly1; + rx = rx1; ry = ry1; + } + + left->x = lx; left->y = ly; + right->x = rx; right->y = ry; +} + +static void nsvg__straightJoin(NSVGrasterizer* r, NSVGpoint* left, NSVGpoint* right, NSVGpoint* p1, float lineWidth) +{ + float w = lineWidth * 0.5f; + float lx = p1->x - (p1->dmx * w), ly = p1->y - (p1->dmy * w); + float rx = p1->x + (p1->dmx * w), ry = p1->y + (p1->dmy * w); + + nsvg__addEdge(r, lx, ly, left->x, left->y); + nsvg__addEdge(r, right->x, right->y, rx, ry); + + left->x = lx; left->y = ly; + right->x = rx; right->y = ry; +} + +static int nsvg__curveDivs(float r, float arc, float tol) +{ + float da = acosf(r / (r + tol)) * 2.0f; + int divs = (int)ceilf(arc / da); + if (divs < 2) divs = 2; + return divs; +} + +static void nsvg__expandStroke(NSVGrasterizer* r, NSVGpoint* points, int npoints, int closed, int lineJoin, int lineCap, float lineWidth) +{ + int ncap = nsvg__curveDivs(lineWidth*0.5f, NSVG_PI, r->tessTol); // Calculate divisions per half circle. + NSVGpoint left = {0,0,0,0,0,0,0,0}, right = {0,0,0,0,0,0,0,0}, firstLeft = {0,0,0,0,0,0,0,0}, firstRight = {0,0,0,0,0,0,0,0}; + NSVGpoint* p0, *p1; + int j, s, e; + + // Build stroke edges + if (closed) { + // Looping + p0 = &points[npoints-1]; + p1 = &points[0]; + s = 0; + e = npoints; + } else { + // Add cap + p0 = &points[0]; + p1 = &points[1]; + s = 1; + e = npoints-1; + } + + if (closed) { + nsvg__initClosed(&left, &right, p0, p1, lineWidth); + firstLeft = left; + firstRight = right; + } else { + // Add cap + float dx = p1->x - p0->x; + float dy = p1->y - p0->y; + nsvg__normalize(&dx, &dy); + if (lineCap == NSVG_CAP_BUTT) + nsvg__buttCap(r, &left, &right, p0, dx, dy, lineWidth, 0); + else if (lineCap == NSVG_CAP_SQUARE) + nsvg__squareCap(r, &left, &right, p0, dx, dy, lineWidth, 0); + else if (lineCap == NSVG_CAP_ROUND) + nsvg__roundCap(r, &left, &right, p0, dx, dy, lineWidth, ncap, 0); + } + + for (j = s; j < e; ++j) { + if (p1->flags & NSVG_PT_CORNER) { + if (lineJoin == NSVG_JOIN_ROUND) + nsvg__roundJoin(r, &left, &right, p0, p1, lineWidth, ncap); + else if (lineJoin == NSVG_JOIN_BEVEL || (p1->flags & NSVG_PT_BEVEL)) + nsvg__bevelJoin(r, &left, &right, p0, p1, lineWidth); + else + nsvg__miterJoin(r, &left, &right, p0, p1, lineWidth); + } else { + nsvg__straightJoin(r, &left, &right, p1, lineWidth); + } + p0 = p1++; + } + + if (closed) { + // Loop it + nsvg__addEdge(r, firstLeft.x, firstLeft.y, left.x, left.y); + nsvg__addEdge(r, right.x, right.y, firstRight.x, firstRight.y); + } else { + // Add cap + float dx = p1->x - p0->x; + float dy = p1->y - p0->y; + nsvg__normalize(&dx, &dy); + if (lineCap == NSVG_CAP_BUTT) + nsvg__buttCap(r, &right, &left, p1, -dx, -dy, lineWidth, 1); + else if (lineCap == NSVG_CAP_SQUARE) + nsvg__squareCap(r, &right, &left, p1, -dx, -dy, lineWidth, 1); + else if (lineCap == NSVG_CAP_ROUND) + nsvg__roundCap(r, &right, &left, p1, -dx, -dy, lineWidth, ncap, 1); + } +} + +static void nsvg__prepareStroke(NSVGrasterizer* r, float miterLimit, int lineJoin) +{ + int i, j; + NSVGpoint* p0, *p1; + + p0 = &r->points[r->npoints-1]; + p1 = &r->points[0]; + for (i = 0; i < r->npoints; i++) { + // Calculate segment direction and length + p0->dx = p1->x - p0->x; + p0->dy = p1->y - p0->y; + p0->len = nsvg__normalize(&p0->dx, &p0->dy); + // Advance + p0 = p1++; + } + + // calculate joins + p0 = &r->points[r->npoints-1]; + p1 = &r->points[0]; + for (j = 0; j < r->npoints; j++) { + float dlx0, dly0, dlx1, dly1, dmr2, cross; + dlx0 = p0->dy; + dly0 = -p0->dx; + dlx1 = p1->dy; + dly1 = -p1->dx; + // Calculate extrusions + p1->dmx = (dlx0 + dlx1) * 0.5f; + p1->dmy = (dly0 + dly1) * 0.5f; + dmr2 = p1->dmx*p1->dmx + p1->dmy*p1->dmy; + if (dmr2 > 0.000001f) { + float s2 = 1.0f / dmr2; + if (s2 > 600.0f) { + s2 = 600.0f; + } + p1->dmx *= s2; + p1->dmy *= s2; + } + + // Clear flags, but keep the corner. + p1->flags = (p1->flags & NSVG_PT_CORNER) ? NSVG_PT_CORNER : 0; + + // Keep track of left turns. + cross = p1->dx * p0->dy - p0->dx * p1->dy; + if (cross > 0.0f) + p1->flags |= NSVG_PT_LEFT; + + // Check to see if the corner needs to be beveled. + if (p1->flags & NSVG_PT_CORNER) { + if ((dmr2 * miterLimit*miterLimit) < 1.0f || lineJoin == NSVG_JOIN_BEVEL || lineJoin == NSVG_JOIN_ROUND) { + p1->flags |= NSVG_PT_BEVEL; + } + } + + p0 = p1++; + } +} + +static void nsvg__flattenShapeStroke(NSVGrasterizer* r, NSVGshape* shape, float scale) +{ + int i, j, closed; + NSVGpath* path; + NSVGpoint* p0, *p1; + float miterLimit = shape->miterLimit; + int lineJoin = shape->strokeLineJoin; + int lineCap = shape->strokeLineCap; + float lineWidth = shape->strokeWidth * scale; + + for (path = shape->paths; path != NULL; path = path->next) { + // Flatten path + r->npoints = 0; + nsvg__addPathPoint(r, path->pts[0]*scale, path->pts[1]*scale, NSVG_PT_CORNER); + for (i = 0; i < path->npts-1; i += 3) { + float* p = &path->pts[i*2]; + nsvg__flattenCubicBez(r, p[0]*scale,p[1]*scale, p[2]*scale,p[3]*scale, p[4]*scale,p[5]*scale, p[6]*scale,p[7]*scale, 0, NSVG_PT_CORNER); + } + if (r->npoints < 2) + continue; + + closed = path->closed; + + // If the first and last points are the same, remove the last, mark as closed path. + p0 = &r->points[r->npoints-1]; + p1 = &r->points[0]; + if (nsvg__ptEquals(p0->x,p0->y, p1->x,p1->y, r->distTol)) { + r->npoints--; + p0 = &r->points[r->npoints-1]; + closed = 1; + } + + if (shape->strokeDashCount > 0) { + int idash = 0, dashState = 1; + float totalDist = 0, dashLen, allDashLen, dashOffset; + NSVGpoint cur; + + if (closed) + nsvg__appendPathPoint(r, r->points[0]); + + // Duplicate points -> points2. + nsvg__duplicatePoints(r); + + r->npoints = 0; + cur = r->points2[0]; + nsvg__appendPathPoint(r, cur); + + // Figure out dash offset. + allDashLen = 0; + for (j = 0; j < shape->strokeDashCount; j++) + allDashLen += shape->strokeDashArray[j]; + if (shape->strokeDashCount & 1) + allDashLen *= 2.0f; + // Find location inside pattern + dashOffset = fmodf(shape->strokeDashOffset, allDashLen); + if (dashOffset < 0.0f) + dashOffset += allDashLen; + + while (dashOffset > shape->strokeDashArray[idash]) { + dashOffset -= shape->strokeDashArray[idash]; + idash = (idash + 1) % shape->strokeDashCount; + } + dashLen = (shape->strokeDashArray[idash] - dashOffset) * scale; + + for (j = 1; j < r->npoints2; ) { + float dx = r->points2[j].x - cur.x; + float dy = r->points2[j].y - cur.y; + float dist = sqrtf(dx*dx + dy*dy); + + if ((totalDist + dist) > dashLen) { + // Calculate intermediate point + float d = (dashLen - totalDist) / dist; + float x = cur.x + dx * d; + float y = cur.y + dy * d; + nsvg__addPathPoint(r, x, y, NSVG_PT_CORNER); + + // Stroke + if (r->npoints > 1 && dashState) { + nsvg__prepareStroke(r, miterLimit, lineJoin); + nsvg__expandStroke(r, r->points, r->npoints, 0, lineJoin, lineCap, lineWidth); + } + // Advance dash pattern + dashState = !dashState; + idash = (idash+1) % shape->strokeDashCount; + dashLen = shape->strokeDashArray[idash] * scale; + // Restart + cur.x = x; + cur.y = y; + cur.flags = NSVG_PT_CORNER; + totalDist = 0.0f; + r->npoints = 0; + nsvg__appendPathPoint(r, cur); + } else { + totalDist += dist; + cur = r->points2[j]; + nsvg__appendPathPoint(r, cur); + j++; + } + } + // Stroke any leftover path + if (r->npoints > 1 && dashState) + nsvg__expandStroke(r, r->points, r->npoints, 0, lineJoin, lineCap, lineWidth); + } else { + nsvg__prepareStroke(r, miterLimit, lineJoin); + nsvg__expandStroke(r, r->points, r->npoints, closed, lineJoin, lineCap, lineWidth); + } + } +} + +static int nsvg__cmpEdge(const void *p, const void *q) +{ + const NSVGedge* a = (const NSVGedge*)p; + const NSVGedge* b = (const NSVGedge*)q; + + if (a->y0 < b->y0) return -1; + if (a->y0 > b->y0) return 1; + return 0; +} + + +static NSVGactiveEdge* nsvg__addActive(NSVGrasterizer* r, NSVGedge* e, float startPoint) +{ + NSVGactiveEdge* z; + + if (r->freelist != NULL) { + // Restore from freelist. + z = r->freelist; + r->freelist = z->next; + } else { + // Alloc new edge. + z = (NSVGactiveEdge*)nsvg__alloc(r, sizeof(NSVGactiveEdge)); + if (z == NULL) return NULL; + } + + float dxdy = (e->x1 - e->x0) / (e->y1 - e->y0); +// STBTT_assert(e->y0 <= start_point); + // round dx down to avoid going too far + if (dxdy < 0) + z->dx = (int)(-floorf(NSVG__FIX * -dxdy)); + else + z->dx = (int)floorf(NSVG__FIX * dxdy); + z->x = (int)floorf(NSVG__FIX * (e->x0 + dxdy * (startPoint - e->y0))); +// z->x -= off_x * FIX; + z->ey = e->y1; + z->next = 0; + z->dir = e->dir; + + return z; +} + +static void nsvg__freeActive(NSVGrasterizer* r, NSVGactiveEdge* z) +{ + z->next = r->freelist; + r->freelist = z; +} + +static void nsvg__fillScanline(unsigned char* scanline, int len, int x0, int x1, int maxWeight, int* xmin, int* xmax) +{ + int i = x0 >> NSVG__FIXSHIFT; + int j = x1 >> NSVG__FIXSHIFT; + if (i < *xmin) *xmin = i; + if (j > *xmax) *xmax = j; + if (i < len && j >= 0) { + if (i == j) { + // x0,x1 are the same pixel, so compute combined coverage + scanline[i] = (unsigned char)(scanline[i] + ((x1 - x0) * maxWeight >> NSVG__FIXSHIFT)); + } else { + if (i >= 0) // add antialiasing for x0 + scanline[i] = (unsigned char)(scanline[i] + (((NSVG__FIX - (x0 & NSVG__FIXMASK)) * maxWeight) >> NSVG__FIXSHIFT)); + else + i = -1; // clip + + if (j < len) // add antialiasing for x1 + scanline[j] = (unsigned char)(scanline[j] + (((x1 & NSVG__FIXMASK) * maxWeight) >> NSVG__FIXSHIFT)); + else + j = len; // clip + + for (++i; i < j; ++i) // fill pixels between x0 and x1 + scanline[i] = (unsigned char)(scanline[i] + maxWeight); + } + } +} + +// note: this routine clips fills that extend off the edges... ideally this +// wouldn't happen, but it could happen if the truetype glyph bounding boxes +// are wrong, or if the user supplies a too-small bitmap +static void nsvg__fillActiveEdges(unsigned char* scanline, int len, NSVGactiveEdge* e, int maxWeight, int* xmin, int* xmax, char fillRule) +{ + // non-zero winding fill + int x0 = 0, w = 0; + + if (fillRule == NSVG_FILLRULE_NONZERO) { + // Non-zero + while (e != NULL) { + if (w == 0) { + // if we're currently at zero, we need to record the edge start point + x0 = e->x; w += e->dir; + } else { + int x1 = e->x; w += e->dir; + // if we went to zero, we need to draw + if (w == 0) + nsvg__fillScanline(scanline, len, x0, x1, maxWeight, xmin, xmax); + } + e = e->next; + } + } else if (fillRule == NSVG_FILLRULE_EVENODD) { + // Even-odd + while (e != NULL) { + if (w == 0) { + // if we're currently at zero, we need to record the edge start point + x0 = e->x; w = 1; + } else { + int x1 = e->x; w = 0; + nsvg__fillScanline(scanline, len, x0, x1, maxWeight, xmin, xmax); + } + e = e->next; + } + } +} + +static float nsvg__clampf(float a, float mn, float mx) { return a < mn ? mn : (a > mx ? mx : a); } + +static unsigned int nsvg__RGBA(unsigned char r, unsigned char g, unsigned char b, unsigned char a) +{ + return (r) | (g << 8) | (b << 16) | (a << 24); +} + +static unsigned int nsvg__lerpRGBA(unsigned int c0, unsigned int c1, float u) +{ + int iu = (int)(nsvg__clampf(u, 0.0f, 1.0f) * 256.0f); + int r = (((c0) & 0xff)*(256-iu) + (((c1) & 0xff)*iu)) >> 8; + int g = (((c0>>8) & 0xff)*(256-iu) + (((c1>>8) & 0xff)*iu)) >> 8; + int b = (((c0>>16) & 0xff)*(256-iu) + (((c1>>16) & 0xff)*iu)) >> 8; + int a = (((c0>>24) & 0xff)*(256-iu) + (((c1>>24) & 0xff)*iu)) >> 8; + return nsvg__RGBA((unsigned char)r, (unsigned char)g, (unsigned char)b, (unsigned char)a); +} + +static unsigned int nsvg__applyOpacity(unsigned int c, float u) +{ + int iu = (int)(nsvg__clampf(u, 0.0f, 1.0f) * 256.0f); + int r = (c) & 0xff; + int g = (c>>8) & 0xff; + int b = (c>>16) & 0xff; + int a = (((c>>24) & 0xff)*iu) >> 8; + return nsvg__RGBA((unsigned char)r, (unsigned char)g, (unsigned char)b, (unsigned char)a); +} + +static inline int nsvg__div255(int x) +{ + return ((x+1) * 257) >> 16; +} + +static void nsvg__scanlineSolid(unsigned char* dst, int count, unsigned char* cover, int x, int y, + float tx, float ty, float scale, NSVGcachedPaint* cache) +{ + + if (cache->type == NSVG_PAINT_COLOR) { + int i, cr, cg, cb, ca; + cr = cache->colors[0] & 0xff; + cg = (cache->colors[0] >> 8) & 0xff; + cb = (cache->colors[0] >> 16) & 0xff; + ca = (cache->colors[0] >> 24) & 0xff; + + for (i = 0; i < count; i++) { + int r,g,b; + int a = nsvg__div255((int)cover[0] * ca); + int ia = 255 - a; + // Premultiply + r = nsvg__div255(cr * a); + g = nsvg__div255(cg * a); + b = nsvg__div255(cb * a); + + // Blend over + r += nsvg__div255(ia * (int)dst[0]); + g += nsvg__div255(ia * (int)dst[1]); + b += nsvg__div255(ia * (int)dst[2]); + a += nsvg__div255(ia * (int)dst[3]); + + dst[0] = (unsigned char)r; + dst[1] = (unsigned char)g; + dst[2] = (unsigned char)b; + dst[3] = (unsigned char)a; + + cover++; + dst += 4; + } + } else if (cache->type == NSVG_PAINT_LINEAR_GRADIENT) { + // TODO: spread modes. + // TODO: plenty of opportunities to optimize. + float fx, fy, dx, gy; + float* t = cache->xform; + int i, cr, cg, cb, ca; + unsigned int c; + + fx = ((float)x - tx) / scale; + fy = ((float)y - ty) / scale; + dx = 1.0f / scale; + + for (i = 0; i < count; i++) { + int r,g,b,a,ia; + gy = fx*t[1] + fy*t[3] + t[5]; + c = cache->colors[(int)nsvg__clampf(gy*255.0f, 0, 255.0f)]; + cr = (c) & 0xff; + cg = (c >> 8) & 0xff; + cb = (c >> 16) & 0xff; + ca = (c >> 24) & 0xff; + + a = nsvg__div255((int)cover[0] * ca); + ia = 255 - a; + + // Premultiply + r = nsvg__div255(cr * a); + g = nsvg__div255(cg * a); + b = nsvg__div255(cb * a); + + // Blend over + r += nsvg__div255(ia * (int)dst[0]); + g += nsvg__div255(ia * (int)dst[1]); + b += nsvg__div255(ia * (int)dst[2]); + a += nsvg__div255(ia * (int)dst[3]); + + dst[0] = (unsigned char)r; + dst[1] = (unsigned char)g; + dst[2] = (unsigned char)b; + dst[3] = (unsigned char)a; + + cover++; + dst += 4; + fx += dx; + } + } else if (cache->type == NSVG_PAINT_RADIAL_GRADIENT) { + // TODO: spread modes. + // TODO: plenty of opportunities to optimize. + // TODO: focus (fx,fy) + float fx, fy, dx, gx, gy, gd; + float* t = cache->xform; + int i, cr, cg, cb, ca; + unsigned int c; + + fx = ((float)x - tx) / scale; + fy = ((float)y - ty) / scale; + dx = 1.0f / scale; + + for (i = 0; i < count; i++) { + int r,g,b,a,ia; + gx = fx*t[0] + fy*t[2] + t[4]; + gy = fx*t[1] + fy*t[3] + t[5]; + gd = sqrtf(gx*gx + gy*gy); + c = cache->colors[(int)nsvg__clampf(gd*255.0f, 0, 255.0f)]; + cr = (c) & 0xff; + cg = (c >> 8) & 0xff; + cb = (c >> 16) & 0xff; + ca = (c >> 24) & 0xff; + + a = nsvg__div255((int)cover[0] * ca); + ia = 255 - a; + + // Premultiply + r = nsvg__div255(cr * a); + g = nsvg__div255(cg * a); + b = nsvg__div255(cb * a); + + // Blend over + r += nsvg__div255(ia * (int)dst[0]); + g += nsvg__div255(ia * (int)dst[1]); + b += nsvg__div255(ia * (int)dst[2]); + a += nsvg__div255(ia * (int)dst[3]); + + dst[0] = (unsigned char)r; + dst[1] = (unsigned char)g; + dst[2] = (unsigned char)b; + dst[3] = (unsigned char)a; + + cover++; + dst += 4; + fx += dx; + } + } +} + +static void nsvg__rasterizeSortedEdges(NSVGrasterizer *r, float tx, float ty, float scale, NSVGcachedPaint* cache, char fillRule) +{ + NSVGactiveEdge *active = NULL; + int y, s; + int e = 0; + int maxWeight = (255 / NSVG__SUBSAMPLES); // weight per vertical scanline + int xmin, xmax; + + for (y = 0; y < r->height; y++) { + memset(r->scanline, 0, r->width); + xmin = r->width; + xmax = 0; + for (s = 0; s < NSVG__SUBSAMPLES; ++s) { + // find center of pixel for this scanline + float scany = (float)(y*NSVG__SUBSAMPLES + s) + 0.5f; + NSVGactiveEdge **step = &active; + + // update all active edges; + // remove all active edges that terminate before the center of this scanline + while (*step) { + NSVGactiveEdge *z = *step; + if (z->ey <= scany) { + *step = z->next; // delete from list +// NSVG__assert(z->valid); + nsvg__freeActive(r, z); + } else { + z->x += z->dx; // advance to position for current scanline + step = &((*step)->next); // advance through list + } + } + + // resort the list if needed + for (;;) { + int changed = 0; + step = &active; + while (*step && (*step)->next) { + if ((*step)->x > (*step)->next->x) { + NSVGactiveEdge* t = *step; + NSVGactiveEdge* q = t->next; + t->next = q->next; + q->next = t; + *step = q; + changed = 1; + } + step = &(*step)->next; + } + if (!changed) break; + } + + // insert all edges that start before the center of this scanline -- omit ones that also end on this scanline + while (e < r->nedges && r->edges[e].y0 <= scany) { + if (r->edges[e].y1 > scany) { + NSVGactiveEdge* z = nsvg__addActive(r, &r->edges[e], scany); + if (z == NULL) break; + // find insertion point + if (active == NULL) { + active = z; + } else if (z->x < active->x) { + // insert at front + z->next = active; + active = z; + } else { + // find thing to insert AFTER + NSVGactiveEdge* p = active; + while (p->next && p->next->x < z->x) + p = p->next; + // at this point, p->next->x is NOT < z->x + z->next = p->next; + p->next = z; + } + } + e++; + } + + // now process all active edges in non-zero fashion + if (active != NULL) + nsvg__fillActiveEdges(r->scanline, r->width, active, maxWeight, &xmin, &xmax, fillRule); + } + // Blit + if (xmin < 0) xmin = 0; + if (xmax > r->width-1) xmax = r->width-1; + if (xmin <= xmax) { + nsvg__scanlineSolid(&r->bitmap[y * r->stride] + xmin*4, xmax-xmin+1, &r->scanline[xmin], xmin, y, tx,ty, scale, cache); + } + } + +} + +static void nsvg__unpremultiplyAlpha(unsigned char* image, int w, int h, int stride) +{ + int x,y; + + // Unpremultiply + for (y = 0; y < h; y++) { + unsigned char *row = &image[y*stride]; + for (x = 0; x < w; x++) { + int r = row[0], g = row[1], b = row[2], a = row[3]; + if (a != 0) { + row[0] = (unsigned char)(r*255/a); + row[1] = (unsigned char)(g*255/a); + row[2] = (unsigned char)(b*255/a); + } + row += 4; + } + } + + // Defringe + for (y = 0; y < h; y++) { + unsigned char *row = &image[y*stride]; + for (x = 0; x < w; x++) { + int r = 0, g = 0, b = 0, a = row[3], n = 0; + if (a == 0) { + if (x-1 > 0 && row[-1] != 0) { + r += row[-4]; + g += row[-3]; + b += row[-2]; + n++; + } + if (x+1 < w && row[7] != 0) { + r += row[4]; + g += row[5]; + b += row[6]; + n++; + } + if (y-1 > 0 && row[-stride+3] != 0) { + r += row[-stride]; + g += row[-stride+1]; + b += row[-stride+2]; + n++; + } + if (y+1 < h && row[stride+3] != 0) { + r += row[stride]; + g += row[stride+1]; + b += row[stride+2]; + n++; + } + if (n > 0) { + row[0] = (unsigned char)(r/n); + row[1] = (unsigned char)(g/n); + row[2] = (unsigned char)(b/n); + } + } + row += 4; + } + } +} + + +static void nsvg__initPaint(NSVGcachedPaint* cache, NSVGpaint* paint, float opacity) +{ + int i, j; + NSVGgradient* grad; + + cache->type = paint->type; + + if (paint->type == NSVG_PAINT_COLOR) { + cache->colors[0] = nsvg__applyOpacity(paint->color, opacity); + return; + } + + grad = paint->gradient; + + cache->spread = grad->spread; + memcpy(cache->xform, grad->xform, sizeof(float)*6); + + if (grad->nstops == 0) { + for (i = 0; i < 256; i++) + cache->colors[i] = 0; + } if (grad->nstops == 1) { + for (i = 0; i < 256; i++) + cache->colors[i] = nsvg__applyOpacity(grad->stops[i].color, opacity); + } else { + unsigned int ca, cb = 0; + float ua, ub, du, u; + int ia, ib, count; + + ca = nsvg__applyOpacity(grad->stops[0].color, opacity); + ua = nsvg__clampf(grad->stops[0].offset, 0, 1); + ub = nsvg__clampf(grad->stops[grad->nstops-1].offset, ua, 1); + ia = (int)(ua * 255.0f); + ib = (int)(ub * 255.0f); + for (i = 0; i < ia; i++) { + cache->colors[i] = ca; + } + + for (i = 0; i < grad->nstops-1; i++) { + ca = nsvg__applyOpacity(grad->stops[i].color, opacity); + cb = nsvg__applyOpacity(grad->stops[i+1].color, opacity); + ua = nsvg__clampf(grad->stops[i].offset, 0, 1); + ub = nsvg__clampf(grad->stops[i+1].offset, 0, 1); + ia = (int)(ua * 255.0f); + ib = (int)(ub * 255.0f); + count = ib - ia; + if (count <= 0) continue; + u = 0; + du = 1.0f / (float)count; + for (j = 0; j < count; j++) { + cache->colors[ia+j] = nsvg__lerpRGBA(ca,cb,u); + u += du; + } + } + + for (i = ib; i < 256; i++) + cache->colors[i] = cb; + } + +} + +/* +static void dumpEdges(NSVGrasterizer* r, const char* name) +{ + float xmin = 0, xmax = 0, ymin = 0, ymax = 0; + NSVGedge *e = NULL; + int i; + if (r->nedges == 0) return; + FILE* fp = fopen(name, "w"); + if (fp == NULL) return; + + xmin = xmax = r->edges[0].x0; + ymin = ymax = r->edges[0].y0; + for (i = 0; i < r->nedges; i++) { + e = &r->edges[i]; + xmin = nsvg__minf(xmin, e->x0); + xmin = nsvg__minf(xmin, e->x1); + xmax = nsvg__maxf(xmax, e->x0); + xmax = nsvg__maxf(xmax, e->x1); + ymin = nsvg__minf(ymin, e->y0); + ymin = nsvg__minf(ymin, e->y1); + ymax = nsvg__maxf(ymax, e->y0); + ymax = nsvg__maxf(ymax, e->y1); + } + + fprintf(fp, "<svg viewBox=\"%f %f %f %f\" xmlns=\"http://www.w3.org/2000/svg\">", xmin, ymin, (xmax - xmin), (ymax - ymin)); + + for (i = 0; i < r->nedges; i++) { + e = &r->edges[i]; + fprintf(fp ,"<line x1=\"%f\" y1=\"%f\" x2=\"%f\" y2=\"%f\" style=\"stroke:#000;\" />", e->x0,e->y0, e->x1,e->y1); + } + + for (i = 0; i < r->npoints; i++) { + if (i+1 < r->npoints) + fprintf(fp ,"<line x1=\"%f\" y1=\"%f\" x2=\"%f\" y2=\"%f\" style=\"stroke:#f00;\" />", r->points[i].x, r->points[i].y, r->points[i+1].x, r->points[i+1].y); + fprintf(fp ,"<circle cx=\"%f\" cy=\"%f\" r=\"1\" style=\"fill:%s;\" />", r->points[i].x, r->points[i].y, r->points[i].flags == 0 ? "#f00" : "#0f0"); + } + + fprintf(fp, "</svg>"); + fclose(fp); +} +*/ + +void nsvgRasterize(NSVGrasterizer* r, + NSVGimage* image, float tx, float ty, float scale, + unsigned char* dst, int w, int h, int stride) +{ + NSVGshape *shape = NULL; + NSVGedge *e = NULL; + NSVGcachedPaint cache; + int i; + + r->bitmap = dst; + r->width = w; + r->height = h; + r->stride = stride; + + if (w > r->cscanline) { + r->cscanline = w; + r->scanline = (unsigned char*)realloc(r->scanline, w); + if (r->scanline == NULL) return; + } + + for (i = 0; i < h; i++) + memset(&dst[i*stride], 0, w*4); + + for (shape = image->shapes; shape != NULL; shape = shape->next) { + if (!(shape->flags & NSVG_FLAGS_VISIBLE)) + continue; + + if (shape->fill.type != NSVG_PAINT_NONE) { + nsvg__resetPool(r); + r->freelist = NULL; + r->nedges = 0; + + nsvg__flattenShape(r, shape, scale); + + // Scale and translate edges + for (i = 0; i < r->nedges; i++) { + e = &r->edges[i]; + e->x0 = tx + e->x0; + e->y0 = (ty + e->y0) * NSVG__SUBSAMPLES; + e->x1 = tx + e->x1; + e->y1 = (ty + e->y1) * NSVG__SUBSAMPLES; + } + + // Rasterize edges + qsort(r->edges, r->nedges, sizeof(NSVGedge), nsvg__cmpEdge); + + // now, traverse the scanlines and find the intersections on each scanline, use non-zero rule + nsvg__initPaint(&cache, &shape->fill, shape->opacity); + + nsvg__rasterizeSortedEdges(r, tx,ty,scale, &cache, shape->fillRule); + } + if (shape->stroke.type != NSVG_PAINT_NONE && (shape->strokeWidth * scale) > 0.01f) { + nsvg__resetPool(r); + r->freelist = NULL; + r->nedges = 0; + + nsvg__flattenShapeStroke(r, shape, scale); + +// dumpEdges(r, "edge.svg"); + + // Scale and translate edges + for (i = 0; i < r->nedges; i++) { + e = &r->edges[i]; + e->x0 = tx + e->x0; + e->y0 = (ty + e->y0) * NSVG__SUBSAMPLES; + e->x1 = tx + e->x1; + e->y1 = (ty + e->y1) * NSVG__SUBSAMPLES; + } + + // Rasterize edges + qsort(r->edges, r->nedges, sizeof(NSVGedge), nsvg__cmpEdge); + + // now, traverse the scanlines and find the intersections on each scanline, use non-zero rule + nsvg__initPaint(&cache, &shape->stroke, shape->opacity); + + nsvg__rasterizeSortedEdges(r, tx,ty,scale, &cache, NSVG_FILLRULE_NONZERO); + } + } + + nsvg__unpremultiplyAlpha(dst, w, h, stride); + + r->bitmap = NULL; + r->width = 0; + r->height = 0; + r->stride = 0; +} + +#endif diff --git a/thirdparty/zstd/PATENTS b/thirdparty/zstd/PATENTS deleted file mode 100644 index 15b4a2ea5c..0000000000 --- a/thirdparty/zstd/PATENTS +++ /dev/null @@ -1,33 +0,0 @@ -Additional Grant of Patent Rights Version 2 - -"Software" means the Zstandard software distributed by Facebook, Inc. - -Facebook, Inc. ("Facebook") hereby grants to each recipient of the Software -("you") a perpetual, worldwide, royalty-free, non-exclusive, irrevocable -(subject to the termination provision below) license under any Necessary -Claims, to make, have made, use, sell, offer to sell, import, and otherwise -transfer the Software. For avoidance of doubt, no license is granted under -Facebook’s rights in any patent claims that are infringed by (i) modifications -to the Software made by you or any third party or (ii) the Software in -combination with any software or other technology. - -The license granted hereunder will terminate, automatically and without notice, -if you (or any of your subsidiaries, corporate affiliates or agents) initiate -directly or indirectly, or take a direct financial interest in, any Patent -Assertion: (i) against Facebook or any of its subsidiaries or corporate -affiliates, (ii) against any party if such Patent Assertion arises in whole or -in part from any software, technology, product or service of Facebook or any of -its subsidiaries or corporate affiliates, or (iii) against any party relating -to the Software. Notwithstanding the foregoing, if Facebook or any of its -subsidiaries or corporate affiliates files a lawsuit alleging patent -infringement against you in the first instance, and you respond by filing a -patent infringement counterclaim in that lawsuit against that party that is -unrelated to the Software, the license granted hereunder will not terminate -under section (i) of this paragraph due to such counterclaim. - -A "Necessary Claim" is a claim of a patent owned by Facebook that is -necessarily infringed by the Software standing alone. - -A "Patent Assertion" is any lawsuit or other action alleging direct, indirect, -or contributory infringement or inducement to infringe any patent, including a -cross-claim or counterclaim. diff --git a/thirdparty/zstd/README.md b/thirdparty/zstd/README.md deleted file mode 100644 index f37be4542c..0000000000 --- a/thirdparty/zstd/README.md +++ /dev/null @@ -1,153 +0,0 @@ - __Zstandard__, or `zstd` as short version, is a fast lossless compression algorithm, - targeting real-time compression scenarios at zlib-level and better compression ratios. - -It is provided as an open-source BSD-licensed **C** library, -and a command line utility producing and decoding `.zst` and `.gz` files. -For other programming languages, -you can consult a list of known ports on [Zstandard homepage](http://www.zstd.net/#other-languages). - -| dev branch status | -|-------------------| -| [![Build Status][travisDevBadge]][travisLink] [![Build status][AppveyorDevBadge]][AppveyorLink] [![Build status][CircleDevBadge]][CircleLink] - -[travisDevBadge]: https://travis-ci.org/facebook/zstd.svg?branch=dev "Continuous Integration test suite" -[travisLink]: https://travis-ci.org/facebook/zstd -[AppveyorDevBadge]: https://ci.appveyor.com/api/projects/status/xt38wbdxjk5mrbem/branch/dev?svg=true "Windows test suite" -[AppveyorLink]: https://ci.appveyor.com/project/YannCollet/zstd-p0yf0 -[CircleDevBadge]: https://circleci.com/gh/facebook/zstd/tree/dev.svg?style=shield "Short test suite" -[CircleLink]: https://circleci.com/gh/facebook/zstd - - -As a reference, several fast compression algorithms were tested and compared -on a server running Linux Debian (`Linux version 4.8.0-1-amd64`), -with a Core i7-6700K CPU @ 4.0GHz, -using [lzbench], an open-source in-memory benchmark by @inikep -compiled with GCC 6.3.0, -on the [Silesia compression corpus]. - -[lzbench]: https://github.com/inikep/lzbench -[Silesia compression corpus]: http://sun.aei.polsl.pl/~sdeor/index.php?page=silesia - -| Compressor name | Ratio | Compression| Decompress.| -| --------------- | ------| -----------| ---------- | -| **zstd 1.1.3 -1** | 2.877 | 430 MB/s | 1110 MB/s | -| zlib 1.2.8 -1 | 2.743 | 110 MB/s | 400 MB/s | -| brotli 0.5.2 -0 | 2.708 | 400 MB/s | 430 MB/s | -| quicklz 1.5.0 -1 | 2.238 | 550 MB/s | 710 MB/s | -| lzo1x 2.09 -1 | 2.108 | 650 MB/s | 830 MB/s | -| lz4 1.7.5 | 2.101 | 720 MB/s | 3600 MB/s | -| snappy 1.1.3 | 2.091 | 500 MB/s | 1650 MB/s | -| lzf 3.6 -1 | 2.077 | 400 MB/s | 860 MB/s | - -[zlib]:http://www.zlib.net/ -[LZ4]: http://www.lz4.org/ - -Zstd can also offer stronger compression ratios at the cost of compression speed. -Speed vs Compression trade-off is configurable by small increments. Decompression speed is preserved and remains roughly the same at all settings, a property shared by most LZ compression algorithms, such as [zlib] or lzma. - -The following tests were run -on a server running Linux Debian (`Linux version 4.8.0-1-amd64`) -with a Core i7-6700K CPU @ 4.0GHz, -using [lzbench], an open-source in-memory benchmark by @inikep -compiled with GCC 6.3.0, -on the [Silesia compression corpus]. - -Compression Speed vs Ratio | Decompression Speed ----------------------------|-------------------- - |  - -Several algorithms can produce higher compression ratios, but at slower speeds, falling outside of the graph. -For a larger picture including very slow modes, [click on this link](doc/images/DCspeed5.png) . - - -### The case for Small Data compression - -Previous charts provide results applicable to typical file and stream scenarios (several MB). Small data comes with different perspectives. - -The smaller the amount of data to compress, the more difficult it is to compress. This problem is common to all compression algorithms, and reason is, compression algorithms learn from past data how to compress future data. But at the beginning of a new data set, there is no "past" to build upon. - -To solve this situation, Zstd offers a __training mode__, which can be used to tune the algorithm for a selected type of data. -Training Zstandard is achieved by providing it with a few samples (one file per sample). The result of this training is stored in a file called "dictionary", which must be loaded before compression and decompression. -Using this dictionary, the compression ratio achievable on small data improves dramatically. - -The following example uses the `github-users` [sample set](https://github.com/facebook/zstd/releases/tag/v1.1.3), created from [github public API](https://developer.github.com/v3/users/#get-all-users). -It consists of roughly 10K records weighing about 1KB each. - -Compression Ratio | Compression Speed | Decompression Speed -------------------|-------------------|-------------------- - |  |  - - -These compression gains are achieved while simultaneously providing _faster_ compression and decompression speeds. - -Training works if there is some correlation in a family of small data samples. The more data-specific a dictionary is, the more efficient it is (there is no _universal dictionary_). -Hence, deploying one dictionary per type of data will provide the greatest benefits. -Dictionary gains are mostly effective in the first few KB. Then, the compression algorithm will gradually use previously decoded content to better compress the rest of the file. - -#### Dictionary compression How To : - -1) Create the dictionary - -`zstd --train FullPathToTrainingSet/* -o dictionaryName` - -2) Compress with dictionary - -`zstd -D dictionaryName FILE` - -3) Decompress with dictionary - -`zstd -D dictionaryName --decompress FILE.zst` - - -### Build - -Once you have the repository cloned, there are multiple ways provided to build Zstandard. - -#### Makefile - -If your system is compatible with a standard `make` (or `gmake`) binary generator, -you can simply run it at the root directory. -It will generate `zstd` within root directory. - -Other available options include : -- `make install` : create and install zstd binary, library and man page -- `make test` : create and run `zstd` and test tools on local platform - -#### cmake - -A `cmake` project generator is provided within `build/cmake`. -It can generate Makefiles or other build scripts -to create `zstd` binary, and `libzstd` dynamic and static libraries. - -#### Meson - -A Meson project is provided within `contrib/meson`. - -#### Visual Studio (Windows) - -Going into `build` directory, you will find additional possibilities : -- Projects for Visual Studio 2005, 2008 and 2010 - + VS2010 project is compatible with VS2012, VS2013 and VS2015 -- Automated build scripts for Visual compiler by @KrzysFR , in `build/VS_scripts`, - which will build `zstd` cli and `libzstd` library without any need to open Visual Studio solution. - - -### Status - -Zstandard is currently deployed within Facebook. It is used daily to compress and decompress very large amounts of data in multiple formats and use cases. -Zstandard is considered safe for production environments. - -### License - -Zstandard is [BSD-licensed](LICENSE). We also provide an [additional patent grant](PATENTS). - -### Contributing - -The "dev" branch is the one where all contributions will be merged before reaching "master". -If you plan to propose a patch, please commit into the "dev" branch or its own feature branch. -Direct commit to "master" are not permitted. -For more information, please read [CONTRIBUTING](CONTRIBUTING.md). - -### Miscellaneous - -Zstd entropy stage is provided by [Huff0 and FSE, from Finite State Entropy library](https://github.com/Cyan4973/FiniteStateEntropy). diff --git a/thirdparty/zstd/common/bitstream.h b/thirdparty/zstd/common/bitstream.h index 07b85026c9..06121f21c5 100644 --- a/thirdparty/zstd/common/bitstream.h +++ b/thirdparty/zstd/common/bitstream.h @@ -80,9 +80,9 @@ extern "C" { * bitStream encoding API (write forward) ********************************************/ /* bitStream can mix input from multiple sources. -* A critical property of these streams is that they encode and decode in **reverse** direction. -* So the first bit sequence you add will be the last to be read, like a LIFO stack. -*/ + * A critical property of these streams is that they encode and decode in **reverse** direction. + * So the first bit sequence you add will be the last to be read, like a LIFO stack. + */ typedef struct { size_t bitContainer; @@ -203,7 +203,7 @@ static const unsigned BIT_mask[] = { 0, 1, 3, 7, 0xF, 0x1F, 0x3F, 0x7F, /*! BIT_initCStream() : * `dstCapacity` must be > sizeof(size_t) * @return : 0 if success, - otherwise an error code (can be tested using ERR_isError() ) */ + * otherwise an error code (can be tested using ERR_isError()) */ MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC, void* startPtr, size_t dstCapacity) { @@ -217,8 +217,8 @@ MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC, } /*! BIT_addBits() : - can add up to 26 bits into `bitC`. - Does not check for register overflow ! */ + * can add up to 26 bits into `bitC`. + * Note : does not check for register overflow ! */ MEM_STATIC void BIT_addBits(BIT_CStream_t* bitC, size_t value, unsigned nbBits) { @@ -268,7 +268,7 @@ MEM_STATIC void BIT_flushBits(BIT_CStream_t* bitC) /*! BIT_closeCStream() : * @return : size of CStream, in bytes, - or 0 if it could not fit into dstBuffer */ + * or 0 if it could not fit into dstBuffer */ MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC) { BIT_addBitsFast(bitC, 1, 1); /* endMark */ @@ -279,14 +279,14 @@ MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC) /*-******************************************************** -* bitStream decoding +* bitStream decoding **********************************************************/ /*! BIT_initDStream() : -* Initialize a BIT_DStream_t. -* `bitD` : a pointer to an already allocated BIT_DStream_t structure. -* `srcSize` must be the *exact* size of the bitStream, in bytes. -* @return : size of stream (== srcSize) or an errorCode if a problem is detected -*/ + * Initialize a BIT_DStream_t. + * `bitD` : a pointer to an already allocated BIT_DStream_t structure. + * `srcSize` must be the *exact* size of the bitStream, in bytes. + * @return : size of stream (== srcSize), or an errorCode if a problem is detected + */ MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize) { if (srcSize < 1) { memset(bitD, 0, sizeof(*bitD)); return ERROR(srcSize_wrong); } @@ -305,29 +305,30 @@ MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, si bitD->bitContainer = *(const BYTE*)(bitD->start); switch(srcSize) { - case 7: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[6]) << (sizeof(bitD->bitContainer)*8 - 16); - /* fall-through */ + case 7: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[6]) << (sizeof(bitD->bitContainer)*8 - 16); + /* fall-through */ - case 6: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[5]) << (sizeof(bitD->bitContainer)*8 - 24); - /* fall-through */ + case 6: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[5]) << (sizeof(bitD->bitContainer)*8 - 24); + /* fall-through */ - case 5: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[4]) << (sizeof(bitD->bitContainer)*8 - 32); - /* fall-through */ + case 5: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[4]) << (sizeof(bitD->bitContainer)*8 - 32); + /* fall-through */ - case 4: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[3]) << 24; - /* fall-through */ + case 4: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[3]) << 24; + /* fall-through */ - case 3: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[2]) << 16; - /* fall-through */ + case 3: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[2]) << 16; + /* fall-through */ - case 2: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[1]) << 8; - /* fall-through */ + case 2: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[1]) << 8; + /* fall-through */ - default: break; + default: break; + } + { BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1]; + bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0; + if (lastByte == 0) return ERROR(corruption_detected); /* endMark not present */ } - { BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1]; - bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0; - if (lastByte == 0) return ERROR(GENERIC); /* endMark not present */ } bitD->bitsConsumed += (U32)(sizeof(bitD->bitContainer) - srcSize)*8; } @@ -363,9 +364,8 @@ MEM_STATIC size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits) * local register is not modified. * On 32-bits, maxNbBits==24. * On 64-bits, maxNbBits==56. - * @return : value extracted - */ - MEM_STATIC size_t BIT_lookBits(const BIT_DStream_t* bitD, U32 nbBits) + * @return : value extracted */ +MEM_STATIC size_t BIT_lookBits(const BIT_DStream_t* bitD, U32 nbBits) { #if defined(__BMI__) && defined(__GNUC__) /* experimental; fails if bitD->bitsConsumed + nbBits > sizeof(bitD->bitContainer)*8 */ return BIT_getMiddleBits(bitD->bitContainer, (sizeof(bitD->bitContainer)*8) - bitD->bitsConsumed - nbBits, nbBits); @@ -392,8 +392,7 @@ MEM_STATIC void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits) /*! BIT_readBits() : * Read (consume) next n bits from local register and update. * Pay attention to not read more than nbBits contained into local register. - * @return : extracted value. - */ + * @return : extracted value. */ MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, U32 nbBits) { size_t const value = BIT_lookBits(bitD, nbBits); @@ -402,7 +401,7 @@ MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, U32 nbBits) } /*! BIT_readBitsFast() : -* unsafe version; only works only if nbBits >= 1 */ + * unsafe version; only works only if nbBits >= 1 */ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, U32 nbBits) { size_t const value = BIT_lookBitsFast(bitD, nbBits); @@ -412,10 +411,10 @@ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, U32 nbBits) } /*! BIT_reloadDStream() : -* Refill `bitD` from buffer previously set in BIT_initDStream() . -* This function is safe, it guarantees it will not read beyond src buffer. -* @return : status of `BIT_DStream_t` internal register. - if status == BIT_DStream_unfinished, internal register is filled with >= (sizeof(bitD->bitContainer)*8 - 7) bits */ + * Refill `bitD` from buffer previously set in BIT_initDStream() . + * This function is safe, it guarantees it will not read beyond src buffer. + * @return : status of `BIT_DStream_t` internal register. + * when status == BIT_DStream_unfinished, internal register is filled with at least 25 or 57 bits */ MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD) { if (bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8)) /* overflow detected, like end of stream */ @@ -446,8 +445,8 @@ MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD) } /*! BIT_endOfDStream() : -* @return Tells if DStream has exactly reached its end (all bits consumed). -*/ + * @return : 1 if DStream has _exactly_ reached its end (all bits consumed). + */ MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* DStream) { return ((DStream->ptr == DStream->start) && (DStream->bitsConsumed == sizeof(DStream->bitContainer)*8)); diff --git a/thirdparty/zstd/common/compiler.h b/thirdparty/zstd/common/compiler.h new file mode 100644 index 0000000000..d7225c443e --- /dev/null +++ b/thirdparty/zstd/common/compiler.h @@ -0,0 +1,85 @@ +/* + * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +#ifndef ZSTD_COMPILER_H +#define ZSTD_COMPILER_H + +/*-******************************************************* +* Compiler specifics +*********************************************************/ +/* force inlining */ +#if defined (__GNUC__) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ +# define INLINE_KEYWORD inline +#else +# define INLINE_KEYWORD +#endif + +#if defined(__GNUC__) +# define FORCE_INLINE_ATTR __attribute__((always_inline)) +#elif defined(_MSC_VER) +# define FORCE_INLINE_ATTR __forceinline +#else +# define FORCE_INLINE_ATTR +#endif + +/** + * FORCE_INLINE_TEMPLATE is used to define C "templates", which take constant + * parameters. They must be inlined for the compiler to elimininate the constant + * branches. + */ +#define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR +/** + * HINT_INLINE is used to help the compiler generate better code. It is *not* + * used for "templates", so it can be tweaked based on the compilers + * performance. + * + * gcc-4.8 and gcc-4.9 have been shown to benefit from leaving off the + * always_inline attribute. + * + * clang up to 5.0.0 (trunk) benefit tremendously from the always_inline + * attribute. + */ +#if !defined(__clang__) && defined(__GNUC__) && __GNUC__ >= 4 && __GNUC_MINOR__ >= 8 && __GNUC__ < 5 +# define HINT_INLINE static INLINE_KEYWORD +#else +# define HINT_INLINE static INLINE_KEYWORD FORCE_INLINE_ATTR +#endif + +/* force no inlining */ +#ifdef _MSC_VER +# define FORCE_NOINLINE static __declspec(noinline) +#else +# ifdef __GNUC__ +# define FORCE_NOINLINE static __attribute__((__noinline__)) +# else +# define FORCE_NOINLINE static +# endif +#endif + +/* prefetch */ +#if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_I86)) /* _mm_prefetch() is not defined outside of x86/x64 */ +# include <mmintrin.h> /* https://msdn.microsoft.com/fr-fr/library/84szxsww(v=vs.90).aspx */ +# define PREFETCH(ptr) _mm_prefetch((const char*)ptr, _MM_HINT_T0) +#elif defined(__GNUC__) +# define PREFETCH(ptr) __builtin_prefetch(ptr, 0, 0) +#else +# define PREFETCH(ptr) /* disabled */ +#endif + +/* disable warnings */ +#ifdef _MSC_VER /* Visual Studio */ +# include <intrin.h> /* For Visual 2005 */ +# pragma warning(disable : 4100) /* disable: C4100: unreferenced formal parameter */ +# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ +# pragma warning(disable : 4204) /* disable: C4204: non-constant aggregate initializer */ +# pragma warning(disable : 4214) /* disable: C4214: non-int bitfields */ +# pragma warning(disable : 4324) /* disable: C4324: padded structure */ +#endif + +#endif /* ZSTD_COMPILER_H */ diff --git a/thirdparty/zstd/common/error_private.c b/thirdparty/zstd/common/error_private.c index 2d752cd23a..b5b14b509c 100644 --- a/thirdparty/zstd/common/error_private.c +++ b/thirdparty/zstd/common/error_private.c @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ /* The purpose of this file is to have a single list of error strings embedded in binary */ @@ -20,19 +20,17 @@ const char* ERR_getErrorString(ERR_enum code) case PREFIX(GENERIC): return "Error (generic)"; case PREFIX(prefix_unknown): return "Unknown frame descriptor"; case PREFIX(version_unsupported): return "Version not supported"; - case PREFIX(parameter_unknown): return "Unknown parameter type"; case PREFIX(frameParameter_unsupported): return "Unsupported frame parameter"; - case PREFIX(frameParameter_unsupportedBy32bits): return "Frame parameter unsupported in 32-bits mode"; case PREFIX(frameParameter_windowTooLarge): return "Frame requires too much memory for decoding"; - case PREFIX(compressionParameter_unsupported): return "Compression parameter is not supported"; - case PREFIX(compressionParameter_outOfBound): return "Compression parameter is out of bound"; + case PREFIX(corruption_detected): return "Corrupted block detected"; + case PREFIX(checksum_wrong): return "Restored data doesn't match checksum"; + case PREFIX(parameter_unsupported): return "Unsupported parameter"; + case PREFIX(parameter_outOfBound): return "Parameter is out of bound"; case PREFIX(init_missing): return "Context should be init first"; case PREFIX(memory_allocation): return "Allocation error : not enough memory"; case PREFIX(stage_wrong): return "Operation not authorized at current processing stage"; case PREFIX(dstSize_tooSmall): return "Destination buffer is too small"; case PREFIX(srcSize_wrong): return "Src size is incorrect"; - case PREFIX(corruption_detected): return "Corrupted block detected"; - case PREFIX(checksum_wrong): return "Restored data doesn't match checksum"; case PREFIX(tableLog_tooLarge): return "tableLog requires too much memory : unsupported"; case PREFIX(maxSymbolValue_tooLarge): return "Unsupported max Symbol Value : too large"; case PREFIX(maxSymbolValue_tooSmall): return "Specified maxSymbolValue is too small"; diff --git a/thirdparty/zstd/common/error_private.h b/thirdparty/zstd/common/error_private.h index 1bc2e49548..9dd9a87cfa 100644 --- a/thirdparty/zstd/common/error_private.h +++ b/thirdparty/zstd/common/error_private.h @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ /* Note : this module is expected to remain private, do not expose it */ diff --git a/thirdparty/zstd/common/fse.h b/thirdparty/zstd/common/fse.h index 6d5d41def1..1c44f83750 100644 --- a/thirdparty/zstd/common/fse.h +++ b/thirdparty/zstd/common/fse.h @@ -31,13 +31,14 @@ You can contact the author at : - Source repository : https://github.com/Cyan4973/FiniteStateEntropy ****************************************************************** */ -#ifndef FSE_H -#define FSE_H #if defined (__cplusplus) extern "C" { #endif +#ifndef FSE_H +#define FSE_H + /*-***************************************** * Dependencies @@ -297,8 +298,10 @@ FSE_decompress_usingDTable() result will tell how many bytes were regenerated (< If there is an error, the function will return an error code, which can be tested using FSE_isError(). (ex: dst buffer too small) */ +#endif /* FSE_H */ -#ifdef FSE_STATIC_LINKING_ONLY +#if defined(FSE_STATIC_LINKING_ONLY) && !defined(FSE_H_FSE_STATIC_LINKING_ONLY) +#define FSE_H_FSE_STATIC_LINKING_ONLY /* *** Dependency *** */ #include "bitstream.h" @@ -381,6 +384,11 @@ size_t FSE_buildDTable_rle (FSE_DTable* dt, unsigned char symbolValue); size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, FSE_DTable* workSpace, unsigned maxLog); /**< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DTABLE_SIZE_U32(maxLog)` */ +typedef enum { + FSE_repeat_none, /**< Cannot use the previous table */ + FSE_repeat_check, /**< Can use the previous table but it must be checked */ + FSE_repeat_valid /**< Can use the previous table and it is asumed to be valid */ + } FSE_repeat; /* ***************************************** * FSE symbol compression API @@ -694,5 +702,3 @@ MEM_STATIC unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr) #if defined (__cplusplus) } #endif - -#endif /* FSE_H */ diff --git a/thirdparty/zstd/common/fse_decompress.c b/thirdparty/zstd/common/fse_decompress.c index 8474a4c079..8e3f0035f6 100644 --- a/thirdparty/zstd/common/fse_decompress.c +++ b/thirdparty/zstd/common/fse_decompress.c @@ -34,34 +34,15 @@ /* ************************************************************** -* Compiler specifics -****************************************************************/ -#ifdef _MSC_VER /* Visual Studio */ -# define FORCE_INLINE static __forceinline -# include <intrin.h> /* For Visual 2005 */ -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ -# pragma warning(disable : 4214) /* disable: C4214: non-int bitfields */ -#else -# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ -# ifdef __GNUC__ -# define FORCE_INLINE static inline __attribute__((always_inline)) -# else -# define FORCE_INLINE static inline -# endif -# else -# define FORCE_INLINE static -# endif /* __STDC_VERSION__ */ -#endif - - -/* ************************************************************** * Includes ****************************************************************/ #include <stdlib.h> /* malloc, free, qsort */ #include <string.h> /* memcpy, memset */ #include "bitstream.h" +#include "compiler.h" #define FSE_STATIC_LINKING_ONLY #include "fse.h" +#include "error_private.h" /* ************************************************************** @@ -216,7 +197,7 @@ size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits) return 0; } -FORCE_INLINE size_t FSE_decompress_usingDTable_generic( +FORCE_INLINE_TEMPLATE size_t FSE_decompress_usingDTable_generic( void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const FSE_DTable* dt, const unsigned fast) diff --git a/thirdparty/zstd/common/huf.h b/thirdparty/zstd/common/huf.h index dabd359915..2b3015a84c 100644 --- a/thirdparty/zstd/common/huf.h +++ b/thirdparty/zstd/common/huf.h @@ -31,13 +31,13 @@ You can contact the author at : - Source repository : https://github.com/Cyan4973/FiniteStateEntropy ****************************************************************** */ -#ifndef HUF_H_298734234 -#define HUF_H_298734234 #if defined (__cplusplus) extern "C" { #endif +#ifndef HUF_H_298734234 +#define HUF_H_298734234 /* *** Dependencies *** */ #include <stddef.h> /* size_t */ @@ -124,6 +124,7 @@ HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity, const #define HUF_DECOMPRESS_WORKSPACE_SIZE (2 << 10) #define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32)) +#endif /* HUF_H_298734234 */ /* ****************************************************************** * WARNING !! @@ -132,7 +133,8 @@ HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity, const * because they are not guaranteed to remain stable in the future. * Only consider them in association with static linking. *******************************************************************/ -#ifdef HUF_STATIC_LINKING_ONLY +#if defined(HUF_STATIC_LINKING_ONLY) && !defined(HUF_H_HUF_STATIC_LINKING_ONLY) +#define HUF_H_HUF_STATIC_LINKING_ONLY /* *** Dependencies *** */ #include "mem.h" /* U32 */ @@ -295,9 +297,6 @@ size_t HUF_decompress1X4_usingDTable(void* dst, size_t maxDstSize, const void* c #endif /* HUF_STATIC_LINKING_ONLY */ - #if defined (__cplusplus) } #endif - -#endif /* HUF_H_298734234 */ diff --git a/thirdparty/zstd/common/mem.h b/thirdparty/zstd/common/mem.h index b0e5bf60b4..df85404fb8 100644 --- a/thirdparty/zstd/common/mem.h +++ b/thirdparty/zstd/common/mem.h @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ #ifndef MEM_H_MODULE @@ -110,7 +110,7 @@ Only use if no other choice to achieve best performance on target platform */ MEM_STATIC U16 MEM_read16(const void* memPtr) { return *(const U16*) memPtr; } MEM_STATIC U32 MEM_read32(const void* memPtr) { return *(const U32*) memPtr; } MEM_STATIC U64 MEM_read64(const void* memPtr) { return *(const U64*) memPtr; } -MEM_STATIC U64 MEM_readST(const void* memPtr) { return *(const size_t*) memPtr; } +MEM_STATIC size_t MEM_readST(const void* memPtr) { return *(const size_t*) memPtr; } MEM_STATIC void MEM_write16(void* memPtr, U16 value) { *(U16*)memPtr = value; } MEM_STATIC void MEM_write32(void* memPtr, U32 value) { *(U32*)memPtr = value; } @@ -131,7 +131,7 @@ MEM_STATIC void MEM_write64(void* memPtr, U64 value) { *(U64*)memPtr = value; } MEM_STATIC U16 MEM_read16(const void* ptr) { return ((const unalign*)ptr)->u16; } MEM_STATIC U32 MEM_read32(const void* ptr) { return ((const unalign*)ptr)->u32; } MEM_STATIC U64 MEM_read64(const void* ptr) { return ((const unalign*)ptr)->u64; } -MEM_STATIC U64 MEM_readST(const void* ptr) { return ((const unalign*)ptr)->st; } +MEM_STATIC size_t MEM_readST(const void* ptr) { return ((const unalign*)ptr)->st; } MEM_STATIC void MEM_write16(void* memPtr, U16 value) { ((unalign*)memPtr)->u16 = value; } MEM_STATIC void MEM_write32(void* memPtr, U32 value) { ((unalign*)memPtr)->u32 = value; } diff --git a/thirdparty/zstd/common/pool.c b/thirdparty/zstd/common/pool.c index 749fa4f2f7..a227044f7f 100644 --- a/thirdparty/zstd/common/pool.c +++ b/thirdparty/zstd/common/pool.c @@ -1,10 +1,10 @@ -/** - * Copyright (c) 2016-present, Facebook, Inc. +/* + * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ @@ -39,6 +39,12 @@ struct POOL_ctx_s { size_t queueHead; size_t queueTail; size_t queueSize; + + /* The number of threads working on jobs */ + size_t numThreadsBusy; + /* Indicates if the queue is empty */ + int queueEmpty; + /* The mutex protects the queue */ pthread_mutex_t queueMutex; /* Condition variable for pushers to wait on when the queue is full */ @@ -60,30 +66,41 @@ static void* POOL_thread(void* opaque) { for (;;) { /* Lock the mutex and wait for a non-empty queue or until shutdown */ pthread_mutex_lock(&ctx->queueMutex); - while (ctx->queueHead == ctx->queueTail && !ctx->shutdown) { + + while (ctx->queueEmpty && !ctx->shutdown) { pthread_cond_wait(&ctx->queuePopCond, &ctx->queueMutex); } /* empty => shutting down: so stop */ - if (ctx->queueHead == ctx->queueTail) { + if (ctx->queueEmpty) { pthread_mutex_unlock(&ctx->queueMutex); return opaque; } /* Pop a job off the queue */ { POOL_job const job = ctx->queue[ctx->queueHead]; ctx->queueHead = (ctx->queueHead + 1) % ctx->queueSize; + ctx->numThreadsBusy++; + ctx->queueEmpty = ctx->queueHead == ctx->queueTail; /* Unlock the mutex, signal a pusher, and run the job */ pthread_mutex_unlock(&ctx->queueMutex); pthread_cond_signal(&ctx->queuePushCond); + job.function(job.opaque); - } - } + + /* If the intended queue size was 0, signal after finishing job */ + if (ctx->queueSize == 1) { + pthread_mutex_lock(&ctx->queueMutex); + ctx->numThreadsBusy--; + pthread_mutex_unlock(&ctx->queueMutex); + pthread_cond_signal(&ctx->queuePushCond); + } } + } /* for (;;) */ /* Unreachable */ } POOL_ctx *POOL_create(size_t numThreads, size_t queueSize) { POOL_ctx *ctx; /* Check the parameters */ - if (!numThreads || !queueSize) { return NULL; } + if (!numThreads) { return NULL; } /* Allocate the context and zero initialize */ ctx = (POOL_ctx *)calloc(1, sizeof(POOL_ctx)); if (!ctx) { return NULL; } @@ -92,15 +109,17 @@ POOL_ctx *POOL_create(size_t numThreads, size_t queueSize) { * and full queues. */ ctx->queueSize = queueSize + 1; - ctx->queue = (POOL_job *)malloc(ctx->queueSize * sizeof(POOL_job)); + ctx->queue = (POOL_job*) malloc(ctx->queueSize * sizeof(POOL_job)); ctx->queueHead = 0; ctx->queueTail = 0; - pthread_mutex_init(&ctx->queueMutex, NULL); - pthread_cond_init(&ctx->queuePushCond, NULL); - pthread_cond_init(&ctx->queuePopCond, NULL); + ctx->numThreadsBusy = 0; + ctx->queueEmpty = 1; + (void)pthread_mutex_init(&ctx->queueMutex, NULL); + (void)pthread_cond_init(&ctx->queuePushCond, NULL); + (void)pthread_cond_init(&ctx->queuePopCond, NULL); ctx->shutdown = 0; /* Allocate space for the thread handles */ - ctx->threads = (pthread_t *)malloc(numThreads * sizeof(pthread_t)); + ctx->threads = (pthread_t*)malloc(numThreads * sizeof(pthread_t)); ctx->numThreads = 0; /* Check for errors */ if (!ctx->threads || !ctx->queue) { POOL_free(ctx); return NULL; } @@ -153,22 +172,37 @@ size_t POOL_sizeof(POOL_ctx *ctx) { + ctx->numThreads * sizeof(pthread_t); } -void POOL_add(void *ctxVoid, POOL_function function, void *opaque) { - POOL_ctx *ctx = (POOL_ctx *)ctxVoid; +/** + * Returns 1 if the queue is full and 0 otherwise. + * + * If the queueSize is 1 (the pool was created with an intended queueSize of 0), + * then a queue is empty if there is a thread free and no job is waiting. + */ +static int isQueueFull(POOL_ctx const* ctx) { + if (ctx->queueSize > 1) { + return ctx->queueHead == ((ctx->queueTail + 1) % ctx->queueSize); + } else { + return ctx->numThreadsBusy == ctx->numThreads || + !ctx->queueEmpty; + } +} + +void POOL_add(void* ctxVoid, POOL_function function, void *opaque) { + POOL_ctx* const ctx = (POOL_ctx*)ctxVoid; if (!ctx) { return; } pthread_mutex_lock(&ctx->queueMutex); { POOL_job const job = {function, opaque}; + /* Wait until there is space in the queue for the new job */ - size_t newTail = (ctx->queueTail + 1) % ctx->queueSize; - while (ctx->queueHead == newTail && !ctx->shutdown) { + while (isQueueFull(ctx) && !ctx->shutdown) { pthread_cond_wait(&ctx->queuePushCond, &ctx->queueMutex); - newTail = (ctx->queueTail + 1) % ctx->queueSize; } /* The queue is still going => there is space */ if (!ctx->shutdown) { + ctx->queueEmpty = 0; ctx->queue[ctx->queueTail] = job; - ctx->queueTail = newTail; + ctx->queueTail = (ctx->queueTail + 1) % ctx->queueSize; } } pthread_mutex_unlock(&ctx->queueMutex); @@ -183,22 +217,22 @@ struct POOL_ctx_s { int data; }; -POOL_ctx *POOL_create(size_t numThreads, size_t queueSize) { +POOL_ctx* POOL_create(size_t numThreads, size_t queueSize) { (void)numThreads; (void)queueSize; - return (POOL_ctx *)malloc(sizeof(POOL_ctx)); + return (POOL_ctx*)malloc(sizeof(POOL_ctx)); } -void POOL_free(POOL_ctx *ctx) { - if (ctx) free(ctx); +void POOL_free(POOL_ctx* ctx) { + free(ctx); } -void POOL_add(void *ctx, POOL_function function, void *opaque) { +void POOL_add(void* ctx, POOL_function function, void* opaque) { (void)ctx; function(opaque); } -size_t POOL_sizeof(POOL_ctx *ctx) { +size_t POOL_sizeof(POOL_ctx* ctx) { if (ctx==NULL) return 0; /* supports sizeof NULL */ return sizeof(*ctx); } diff --git a/thirdparty/zstd/common/pool.h b/thirdparty/zstd/common/pool.h index 386cd674b7..264c5c9ca7 100644 --- a/thirdparty/zstd/common/pool.h +++ b/thirdparty/zstd/common/pool.h @@ -1,11 +1,12 @@ -/** - * Copyright (c) 2016-present, Facebook, Inc. +/* + * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ + #ifndef POOL_H #define POOL_H @@ -19,11 +20,10 @@ extern "C" { typedef struct POOL_ctx_s POOL_ctx; /*! POOL_create() : - Create a thread pool with at most `numThreads` threads. - `numThreads` must be at least 1. - The maximum number of queued jobs before blocking is `queueSize`. - `queueSize` must be at least 1. - @return : The POOL_ctx pointer on success else NULL. + * Create a thread pool with at most `numThreads` threads. + * `numThreads` must be at least 1. + * The maximum number of queued jobs before blocking is `queueSize`. + * @return : POOL_ctx pointer on success, else NULL. */ POOL_ctx *POOL_create(size_t numThreads, size_t queueSize); diff --git a/thirdparty/zstd/common/threading.h b/thirdparty/zstd/common/threading.h index c0086139ea..ab09977a86 100644 --- a/thirdparty/zstd/common/threading.h +++ b/thirdparty/zstd/common/threading.h @@ -1,4 +1,3 @@ - /** * Copyright (c) 2016 Tino Reichardt * All rights reserved. @@ -42,14 +41,14 @@ extern "C" { /* mutex */ #define pthread_mutex_t CRITICAL_SECTION -#define pthread_mutex_init(a,b) InitializeCriticalSection((a)) +#define pthread_mutex_init(a,b) (InitializeCriticalSection((a)), 0) #define pthread_mutex_destroy(a) DeleteCriticalSection((a)) #define pthread_mutex_lock(a) EnterCriticalSection((a)) #define pthread_mutex_unlock(a) LeaveCriticalSection((a)) /* condition variable */ #define pthread_cond_t CONDITION_VARIABLE -#define pthread_cond_init(a, b) InitializeConditionVariable((a)) +#define pthread_cond_init(a, b) (InitializeConditionVariable((a)), 0) #define pthread_cond_destroy(a) /* No delete */ #define pthread_cond_wait(a, b) SleepConditionVariableCS((a), (b), INFINITE) #define pthread_cond_signal(a) WakeConditionVariable((a)) @@ -80,14 +79,14 @@ int _pthread_join(pthread_t* thread, void** value_ptr); #else /* ZSTD_MULTITHREAD not defined */ /* No multithreading support */ -#define pthread_mutex_t int /* #define rather than typedef, as sometimes pthread support is implicit, resulting in duplicated symbols */ -#define pthread_mutex_init(a,b) +#define pthread_mutex_t int /* #define rather than typedef, because sometimes pthread support is implicit, resulting in duplicated symbols */ +#define pthread_mutex_init(a,b) ((void)a, 0) #define pthread_mutex_destroy(a) #define pthread_mutex_lock(a) #define pthread_mutex_unlock(a) #define pthread_cond_t int -#define pthread_cond_init(a,b) +#define pthread_cond_init(a,b) ((void)a, 0) #define pthread_cond_destroy(a) #define pthread_cond_wait(a,b) #define pthread_cond_signal(a) diff --git a/thirdparty/zstd/common/xxhash.c b/thirdparty/zstd/common/xxhash.c index eb44222c5f..9d9c0e963c 100644 --- a/thirdparty/zstd/common/xxhash.c +++ b/thirdparty/zstd/common/xxhash.c @@ -113,19 +113,25 @@ static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcp /* ************************************* * Compiler Specific Options ***************************************/ -#ifdef _MSC_VER /* Visual Studio */ -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ -# define FORCE_INLINE static __forceinline +#if defined (__GNUC__) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ +# define INLINE_KEYWORD inline #else -# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ -# ifdef __GNUC__ -# define FORCE_INLINE static inline __attribute__((always_inline)) -# else -# define FORCE_INLINE static inline -# endif -# else -# define FORCE_INLINE static -# endif /* __STDC_VERSION__ */ +# define INLINE_KEYWORD +#endif + +#if defined(__GNUC__) +# define FORCE_INLINE_ATTR __attribute__((always_inline)) +#elif defined(_MSC_VER) +# define FORCE_INLINE_ATTR __forceinline +#else +# define FORCE_INLINE_ATTR +#endif + +#define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR + + +#ifdef _MSC_VER +# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ #endif @@ -248,7 +254,7 @@ typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess; *****************************/ typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment; -FORCE_INLINE U32 XXH_readLE32_align(const void* ptr, XXH_endianess endian, XXH_alignment align) +FORCE_INLINE_TEMPLATE U32 XXH_readLE32_align(const void* ptr, XXH_endianess endian, XXH_alignment align) { if (align==XXH_unaligned) return endian==XXH_littleEndian ? XXH_read32(ptr) : XXH_swap32(XXH_read32(ptr)); @@ -256,7 +262,7 @@ FORCE_INLINE U32 XXH_readLE32_align(const void* ptr, XXH_endianess endian, XXH_a return endian==XXH_littleEndian ? *(const U32*)ptr : XXH_swap32(*(const U32*)ptr); } -FORCE_INLINE U32 XXH_readLE32(const void* ptr, XXH_endianess endian) +FORCE_INLINE_TEMPLATE U32 XXH_readLE32(const void* ptr, XXH_endianess endian) { return XXH_readLE32_align(ptr, endian, XXH_unaligned); } @@ -266,7 +272,7 @@ static U32 XXH_readBE32(const void* ptr) return XXH_CPU_LITTLE_ENDIAN ? XXH_swap32(XXH_read32(ptr)) : XXH_read32(ptr); } -FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align) +FORCE_INLINE_TEMPLATE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align) { if (align==XXH_unaligned) return endian==XXH_littleEndian ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr)); @@ -274,7 +280,7 @@ FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_a return endian==XXH_littleEndian ? *(const U64*)ptr : XXH_swap64(*(const U64*)ptr); } -FORCE_INLINE U64 XXH_readLE64(const void* ptr, XXH_endianess endian) +FORCE_INLINE_TEMPLATE U64 XXH_readLE64(const void* ptr, XXH_endianess endian) { return XXH_readLE64_align(ptr, endian, XXH_unaligned); } @@ -335,7 +341,7 @@ static U32 XXH32_round(U32 seed, U32 input) return seed; } -FORCE_INLINE U32 XXH32_endian_align(const void* input, size_t len, U32 seed, XXH_endianess endian, XXH_alignment align) +FORCE_INLINE_TEMPLATE U32 XXH32_endian_align(const void* input, size_t len, U32 seed, XXH_endianess endian, XXH_alignment align) { const BYTE* p = (const BYTE*)input; const BYTE* bEnd = p + len; @@ -435,7 +441,7 @@ static U64 XXH64_mergeRound(U64 acc, U64 val) return acc; } -FORCE_INLINE U64 XXH64_endian_align(const void* input, size_t len, U64 seed, XXH_endianess endian, XXH_alignment align) +FORCE_INLINE_TEMPLATE U64 XXH64_endian_align(const void* input, size_t len, U64 seed, XXH_endianess endian, XXH_alignment align) { const BYTE* p = (const BYTE*)input; const BYTE* const bEnd = p + len; @@ -584,7 +590,7 @@ XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long } -FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state, const void* input, size_t len, XXH_endianess endian) +FORCE_INLINE_TEMPLATE XXH_errorcode XXH32_update_endian (XXH32_state_t* state, const void* input, size_t len, XXH_endianess endian) { const BYTE* p = (const BYTE*)input; const BYTE* const bEnd = p + len; @@ -654,7 +660,7 @@ XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* state_in, const void* -FORCE_INLINE U32 XXH32_digest_endian (const XXH32_state_t* state, XXH_endianess endian) +FORCE_INLINE_TEMPLATE U32 XXH32_digest_endian (const XXH32_state_t* state, XXH_endianess endian) { const BYTE * p = (const BYTE*)state->mem32; const BYTE* const bEnd = (const BYTE*)(state->mem32) + state->memsize; @@ -704,7 +710,7 @@ XXH_PUBLIC_API unsigned int XXH32_digest (const XXH32_state_t* state_in) /* **** XXH64 **** */ -FORCE_INLINE XXH_errorcode XXH64_update_endian (XXH64_state_t* state, const void* input, size_t len, XXH_endianess endian) +FORCE_INLINE_TEMPLATE XXH_errorcode XXH64_update_endian (XXH64_state_t* state, const void* input, size_t len, XXH_endianess endian) { const BYTE* p = (const BYTE*)input; const BYTE* const bEnd = p + len; @@ -771,7 +777,7 @@ XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void* -FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess endian) +FORCE_INLINE_TEMPLATE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess endian) { const BYTE * p = (const BYTE*)state->mem64; const BYTE* const bEnd = (const BYTE*)state->mem64 + state->memsize; diff --git a/thirdparty/zstd/common/zstd_common.c b/thirdparty/zstd/common/zstd_common.c index f681672381..08384cabf5 100644 --- a/thirdparty/zstd/common/zstd_common.c +++ b/thirdparty/zstd/common/zstd_common.c @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ diff --git a/thirdparty/zstd/common/zstd_errors.h b/thirdparty/zstd/common/zstd_errors.h index 19f1597aa3..a69387b714 100644 --- a/thirdparty/zstd/common/zstd_errors.h +++ b/thirdparty/zstd/common/zstd_errors.h @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ #ifndef ZSTD_ERRORS_H_398273423 @@ -37,43 +37,41 @@ extern "C" { /*-**************************************** * error codes list * note : this API is still considered unstable - * it should not be used with a dynamic library + * and shall not be used with a dynamic library. * only static linking is allowed ******************************************/ typedef enum { - ZSTD_error_no_error, - ZSTD_error_GENERIC, - ZSTD_error_prefix_unknown, - ZSTD_error_version_unsupported, - ZSTD_error_parameter_unknown, - ZSTD_error_frameParameter_unsupported, - ZSTD_error_frameParameter_unsupportedBy32bits, - ZSTD_error_frameParameter_windowTooLarge, - ZSTD_error_compressionParameter_unsupported, - ZSTD_error_compressionParameter_outOfBound, - ZSTD_error_init_missing, - ZSTD_error_memory_allocation, - ZSTD_error_stage_wrong, - ZSTD_error_dstSize_tooSmall, - ZSTD_error_srcSize_wrong, - ZSTD_error_corruption_detected, - ZSTD_error_checksum_wrong, - ZSTD_error_tableLog_tooLarge, - ZSTD_error_maxSymbolValue_tooLarge, - ZSTD_error_maxSymbolValue_tooSmall, - ZSTD_error_dictionary_corrupted, - ZSTD_error_dictionary_wrong, - ZSTD_error_dictionaryCreation_failed, - ZSTD_error_frameIndex_tooLarge, - ZSTD_error_seekableIO, - ZSTD_error_maxCode + ZSTD_error_no_error = 0, + ZSTD_error_GENERIC = 1, + ZSTD_error_prefix_unknown = 10, + ZSTD_error_version_unsupported = 12, + ZSTD_error_frameParameter_unsupported = 14, + ZSTD_error_frameParameter_windowTooLarge = 16, + ZSTD_error_corruption_detected = 20, + ZSTD_error_checksum_wrong = 22, + ZSTD_error_dictionary_corrupted = 30, + ZSTD_error_dictionary_wrong = 32, + ZSTD_error_dictionaryCreation_failed = 34, + ZSTD_error_parameter_unsupported = 40, + ZSTD_error_parameter_outOfBound = 42, + ZSTD_error_tableLog_tooLarge = 44, + ZSTD_error_maxSymbolValue_tooLarge = 46, + ZSTD_error_maxSymbolValue_tooSmall = 48, + ZSTD_error_stage_wrong = 60, + ZSTD_error_init_missing = 62, + ZSTD_error_memory_allocation = 64, + ZSTD_error_dstSize_tooSmall = 70, + ZSTD_error_srcSize_wrong = 72, + ZSTD_error_frameIndex_tooLarge = 100, + ZSTD_error_seekableIO = 102, + ZSTD_error_maxCode = 120 /* never EVER use this value directly, it may change in future versions! Use ZSTD_isError() instead */ } ZSTD_ErrorCode; /*! ZSTD_getErrorCode() : convert a `size_t` function result into a `ZSTD_ErrorCode` enum type, which can be used to compare with enum list published above */ ZSTDERRORLIB_API ZSTD_ErrorCode ZSTD_getErrorCode(size_t functionResult); -ZSTDERRORLIB_API const char* ZSTD_getErrorString(ZSTD_ErrorCode code); +ZSTDERRORLIB_API const char* ZSTD_getErrorString(ZSTD_ErrorCode code); /**< Same as ZSTD_getErrorName, but using a `ZSTD_ErrorCode` enum argument */ #if defined (__cplusplus) diff --git a/thirdparty/zstd/common/zstd_internal.h b/thirdparty/zstd/common/zstd_internal.h index 49a4796476..2610528608 100644 --- a/thirdparty/zstd/common/zstd_internal.h +++ b/thirdparty/zstd/common/zstd_internal.h @@ -1,55 +1,28 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ #ifndef ZSTD_CCOMMON_H_MODULE #define ZSTD_CCOMMON_H_MODULE -/*-******************************************************* -* Compiler specifics -*********************************************************/ -#ifdef _MSC_VER /* Visual Studio */ -# define FORCE_INLINE static __forceinline -# include <intrin.h> /* For Visual 2005 */ -# pragma warning(disable : 4100) /* disable: C4100: unreferenced formal parameter */ -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ -# pragma warning(disable : 4204) /* disable: C4204: non-constant aggregate initializer */ -# pragma warning(disable : 4324) /* disable: C4324: padded structure */ -#else -# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ -# ifdef __GNUC__ -# define FORCE_INLINE static inline __attribute__((always_inline)) -# else -# define FORCE_INLINE static inline -# endif -# else -# define FORCE_INLINE static -# endif /* __STDC_VERSION__ */ -#endif - -#ifdef _MSC_VER -# define FORCE_NOINLINE static __declspec(noinline) -#else -# ifdef __GNUC__ -# define FORCE_NOINLINE static __attribute__((__noinline__)) -# else -# define FORCE_NOINLINE static -# endif -#endif - /*-************************************* * Dependencies ***************************************/ +#include "compiler.h" #include "mem.h" #include "error_private.h" #define ZSTD_STATIC_LINKING_ONLY #include "zstd.h" +#define FSE_STATIC_LINKING_ONLY +#include "fse.h" +#define HUF_STATIC_LINKING_ONLY +#include "huf.h" #ifndef XXH_STATIC_LINKING_ONLY # define XXH_STATIC_LINKING_ONLY /* XXH64_state_t */ #endif @@ -211,20 +184,6 @@ MEM_STATIC void ZSTD_wildcopy_e(void* dst, const void* src, void* dstEnd) /* s *********************************************/ typedef struct ZSTD_stats_s ZSTD_stats_t; -typedef struct { - U32 off; - U32 len; -} ZSTD_match_t; - -typedef struct { - U32 price; - U32 off; - U32 mlen; - U32 litlen; - U32 rep[ZSTD_REP_NUM]; -} ZSTD_optimal_t; - - typedef struct seqDef_s { U32 offset; U16 litLength; @@ -242,13 +201,31 @@ typedef struct { BYTE* ofCode; U32 longLengthID; /* 0 == no longLength; 1 == Lit.longLength; 2 == Match.longLength; */ U32 longLengthPos; - /* opt */ - ZSTD_optimal_t* priceTable; - ZSTD_match_t* matchTable; - U32* matchLengthFreq; - U32* litLengthFreq; + U32 rep[ZSTD_REP_NUM]; + U32 repToConfirm[ZSTD_REP_NUM]; +} seqStore_t; + +typedef struct { + U32 off; + U32 len; +} ZSTD_match_t; + +typedef struct { + U32 price; + U32 off; + U32 mlen; + U32 litlen; + U32 rep[ZSTD_REP_NUM]; +} ZSTD_optimal_t; + +typedef struct { U32* litFreq; + U32* litLengthFreq; + U32* matchLengthFreq; U32* offCodeFreq; + ZSTD_match_t* matchTable; + ZSTD_optimal_t* priceTable; + U32 matchLengthSum; U32 matchSum; U32 litLengthSum; @@ -264,7 +241,19 @@ typedef struct { U32 cachedPrice; U32 cachedLitLength; const BYTE* cachedLiterals; -} seqStore_t; +} optState_t; + +typedef struct { + U32 hufCTable[HUF_CTABLE_SIZE_U32(255)]; + FSE_CTable offcodeCTable[FSE_CTABLE_SIZE_U32(OffFSELog, MaxOff)]; + FSE_CTable matchlengthCTable[FSE_CTABLE_SIZE_U32(MLFSELog, MaxML)]; + FSE_CTable litlengthCTable[FSE_CTABLE_SIZE_U32(LLFSELog, MaxLL)]; + U32 workspace[HUF_WORKSPACE_SIZE_U32]; + HUF_repeat hufCTable_repeatMode; + FSE_repeat offcode_repeatMode; + FSE_repeat matchlength_repeatMode; + FSE_repeat litlength_repeatMode; +} ZSTD_entropyCTables_t; const seqStore_t* ZSTD_getSeqStore(const ZSTD_CCtx* ctx); void ZSTD_seqToCodes(const seqStore_t* seqStorePtr); @@ -338,9 +327,9 @@ typedef struct { } blockProperties_t; /*! ZSTD_getcBlockSize() : - * Provides the size of compressed block from block header `src` */ - size_t ZSTD_getcBlockSize(const void* src, size_t srcSize, - blockProperties_t* bpPtr); +* Provides the size of compressed block from block header `src` */ +size_t ZSTD_getcBlockSize(const void* src, size_t srcSize, + blockProperties_t* bpPtr); #endif /* ZSTD_CCOMMON_H_MODULE */ diff --git a/thirdparty/zstd/compress/fse_compress.c b/thirdparty/zstd/compress/fse_compress.c index 26e8052ddc..cc9fa73514 100644 --- a/thirdparty/zstd/compress/fse_compress.c +++ b/thirdparty/zstd/compress/fse_compress.c @@ -33,40 +33,22 @@ ****************************************************************** */ /* ************************************************************** -* Compiler specifics -****************************************************************/ -#ifdef _MSC_VER /* Visual Studio */ -# define FORCE_INLINE static __forceinline -# include <intrin.h> /* For Visual 2005 */ -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ -# pragma warning(disable : 4214) /* disable: C4214: non-int bitfields */ -#else -# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ -# ifdef __GNUC__ -# define FORCE_INLINE static inline __attribute__((always_inline)) -# else -# define FORCE_INLINE static inline -# endif -# else -# define FORCE_INLINE static -# endif /* __STDC_VERSION__ */ -#endif - - -/* ************************************************************** * Includes ****************************************************************/ #include <stdlib.h> /* malloc, free, qsort */ #include <string.h> /* memcpy, memset */ #include <stdio.h> /* printf (debug) */ #include "bitstream.h" +#include "compiler.h" #define FSE_STATIC_LINKING_ONLY #include "fse.h" +#include "error_private.h" /* ************************************************************** * Error Management ****************************************************************/ +#define FSE_isError ERR_isError #define FSE_STATIC_ASSERT(c) { enum { FSE_static_assert = 1/(int)(!!(c)) }; } /* use only *after* variable declarations */ @@ -781,7 +763,7 @@ size_t FSE_compress_usingCTable (void* dst, size_t dstSize, size_t FSE_compressBound(size_t size) { return FSE_COMPRESSBOUND(size); } -#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return f +#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e #define CHECK_F(f) { CHECK_V_F(_var_err__, f); } /* FSE_compress_wksp() : diff --git a/thirdparty/zstd/compress/huf_compress.c b/thirdparty/zstd/compress/huf_compress.c index 7af0789a9c..2a47c18205 100644 --- a/thirdparty/zstd/compress/huf_compress.c +++ b/thirdparty/zstd/compress/huf_compress.c @@ -50,13 +50,15 @@ #include "fse.h" /* header compression */ #define HUF_STATIC_LINKING_ONLY #include "huf.h" +#include "error_private.h" /* ************************************************************** * Error Management ****************************************************************/ +#define HUF_isError ERR_isError #define HUF_STATIC_ASSERT(c) { enum { HUF_static_assert = 1/(int)(!!(c)) }; } /* use only *after* variable declarations */ -#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return f +#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e #define CHECK_F(f) { CHECK_V_F(_var_err__, f); } @@ -436,7 +438,7 @@ static void HUF_encodeSymbol(BIT_CStream_t* bitCPtr, U32 symbol, const HUF_CElt* size_t HUF_compressBound(size_t size) { return HUF_COMPRESSBOUND(size); } -#define HUF_FLUSHBITS(s) (fast ? BIT_flushBitsFast(s) : BIT_flushBits(s)) +#define HUF_FLUSHBITS(s) BIT_flushBits(s) #define HUF_FLUSHBITS_1(stream) \ if (sizeof((stream)->bitContainer)*8 < HUF_TABLELOG_MAX*2+7) HUF_FLUSHBITS(stream) @@ -451,7 +453,6 @@ size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, si BYTE* const oend = ostart + dstSize; BYTE* op = ostart; size_t n; - const unsigned fast = (dstSize >= HUF_BLOCKBOUND(srcSize)); BIT_CStream_t bitC; /* init */ diff --git a/thirdparty/zstd/compress/zstd_compress.c b/thirdparty/zstd/compress/zstd_compress.c index 9300357f2d..0322c03eb3 100644 --- a/thirdparty/zstd/compress/zstd_compress.c +++ b/thirdparty/zstd/compress/zstd_compress.c @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ @@ -36,13 +36,6 @@ static const U32 g_searchStrength = 8; /* control skip over incompressible dat #define HASH_READ_SIZE 8 typedef enum { ZSTDcs_created=0, ZSTDcs_init, ZSTDcs_ongoing, ZSTDcs_ending } ZSTD_compressionStage_e; -/* entropy tables always have same size */ -static size_t const hufCTable_size = HUF_CTABLE_SIZE(255); -static size_t const litlengthCTable_size = FSE_CTABLE_SIZE(LLFSELog, MaxLL); -static size_t const offcodeCTable_size = FSE_CTABLE_SIZE(OffFSELog, MaxOff); -static size_t const matchlengthCTable_size = FSE_CTABLE_SIZE(MLFSELog, MaxML); -static size_t const entropyScratchSpace_size = HUF_WORKSPACE_SIZE; - /*-************************************* * Helper functions @@ -89,8 +82,6 @@ struct ZSTD_CCtx_s { U32 loadedDictEnd; /* index of end of dictionary */ U32 forceWindow; /* force back-references to respect limit of 1<<wLog, even for dictionary */ ZSTD_compressionStage_e stage; - U32 rep[ZSTD_REP_NUM]; - U32 repToConfirm[ZSTD_REP_NUM]; U32 dictID; int compressionLevel; ZSTD_parameters requestedParams; @@ -105,16 +96,11 @@ struct ZSTD_CCtx_s { size_t staticSize; seqStore_t seqStore; /* sequences storage ptrs */ + optState_t optState; U32* hashTable; U32* hashTable3; U32* chainTable; - HUF_repeat hufCTable_repeatMode; - HUF_CElt* hufCTable; - U32 fseCTables_ready; - FSE_CTable* offcodeCTable; - FSE_CTable* matchlengthCTable; - FSE_CTable* litlengthCTable; - unsigned* entropyScratchSpace; + ZSTD_entropyCTables_t* entropy; /* streaming */ char* inBuff; @@ -174,19 +160,9 @@ ZSTD_CCtx* ZSTD_initStaticCCtx(void *workspace, size_t workspaceSize) cctx->workSpaceSize = workspaceSize - sizeof(ZSTD_CCtx); /* entropy space (never moves) */ - /* note : this code should be shared with resetCCtx, rather than copy/pasted */ - { void* ptr = cctx->workSpace; - cctx->hufCTable = (HUF_CElt*)ptr; - ptr = (char*)cctx->hufCTable + hufCTable_size; - cctx->offcodeCTable = (FSE_CTable*) ptr; - ptr = (char*)ptr + offcodeCTable_size; - cctx->matchlengthCTable = (FSE_CTable*) ptr; - ptr = (char*)ptr + matchlengthCTable_size; - cctx->litlengthCTable = (FSE_CTable*) ptr; - ptr = (char*)ptr + litlengthCTable_size; - assert(((size_t)ptr & 3) == 0); /* ensure correct alignment */ - cctx->entropyScratchSpace = (unsigned*) ptr; - } + if (cctx->workSpaceSize < sizeof(ZSTD_entropyCTables_t)) return NULL; + assert(((size_t)cctx->workSpace & 7) == 0); /* ensure correct alignment */ + cctx->entropy = (ZSTD_entropyCTables_t*)cctx->workSpace; return cctx; } @@ -237,7 +213,7 @@ size_t ZSTD_setCCtxParameter(ZSTD_CCtx* cctx, ZSTD_CCtxParameter param, unsigned ZSTD_STATIC_ASSERT(ZSTD_dm_auto==0); ZSTD_STATIC_ASSERT(ZSTD_dm_rawContent==1); case ZSTD_p_forceRawDict : cctx->dictMode = (ZSTD_dictMode_e)(value>0); return 0; - default: return ERROR(parameter_unknown); + default: return ERROR(parameter_unsupported); } } @@ -251,9 +227,9 @@ static void ZSTD_cLevelToCParams(ZSTD_CCtx* cctx) cctx->compressionLevel = ZSTD_CLEVEL_CUSTOM; } -#define CLAMPCHECK(val,min,max) { \ - if (((val)<(min)) | ((val)>(max))) { \ - return ERROR(compressionParameter_outOfBound); \ +#define CLAMPCHECK(val,min,max) { \ + if (((val)<(min)) | ((val)>(max))) { \ + return ERROR(parameter_outOfBound); \ } } size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, unsigned value) @@ -349,7 +325,7 @@ size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, unsigned v /* restrict dictionary mode, to "rawContent" or "fullDict" only */ ZSTD_STATIC_ASSERT((U32)ZSTD_dm_fullDict > (U32)ZSTD_dm_rawContent); if (value > (unsigned)ZSTD_dm_fullDict) - return ERROR(compressionParameter_outOfBound); + return ERROR(parameter_outOfBound); cctx->dictMode = (ZSTD_dictMode_e)value; return 0; @@ -370,31 +346,31 @@ size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, unsigned v if (value==0) return 0; DEBUGLOG(5, " setting nbThreads : %u", value); #ifndef ZSTD_MULTITHREAD - if (value > 1) return ERROR(compressionParameter_unsupported); + if (value > 1) return ERROR(parameter_unsupported); #endif if ((value>1) && (cctx->nbThreads != value)) { if (cctx->staticSize) /* MT not compatible with static alloc */ - return ERROR(compressionParameter_unsupported); + return ERROR(parameter_unsupported); ZSTDMT_freeCCtx(cctx->mtctx); cctx->nbThreads = 1; - cctx->mtctx = ZSTDMT_createCCtx(value); + cctx->mtctx = ZSTDMT_createCCtx_advanced(value, cctx->customMem); if (cctx->mtctx == NULL) return ERROR(memory_allocation); } cctx->nbThreads = value; return 0; case ZSTD_p_jobSize: - if (cctx->nbThreads <= 1) return ERROR(compressionParameter_unsupported); + if (cctx->nbThreads <= 1) return ERROR(parameter_unsupported); assert(cctx->mtctx != NULL); return ZSTDMT_setMTCtxParameter(cctx->mtctx, ZSTDMT_p_sectionSize, value); case ZSTD_p_overlapSizeLog: DEBUGLOG(5, " setting overlap with nbThreads == %u", cctx->nbThreads); - if (cctx->nbThreads <= 1) return ERROR(compressionParameter_unsupported); + if (cctx->nbThreads <= 1) return ERROR(parameter_unsupported); assert(cctx->mtctx != NULL); return ZSTDMT_setMTCtxParameter(cctx->mtctx, ZSTDMT_p_overlapSectionLog, value); - default: return ERROR(parameter_unknown); + default: return ERROR(parameter_unsupported); } } @@ -474,7 +450,8 @@ size_t ZSTD_checkCParams(ZSTD_compressionParameters cParams) CLAMPCHECK(cParams.searchLog, ZSTD_SEARCHLOG_MIN, ZSTD_SEARCHLOG_MAX); CLAMPCHECK(cParams.searchLength, ZSTD_SEARCHLENGTH_MIN, ZSTD_SEARCHLENGTH_MAX); CLAMPCHECK(cParams.targetLength, ZSTD_TARGETLENGTH_MIN, ZSTD_TARGETLENGTH_MAX); - if ((U32)(cParams.strategy) > (U32)ZSTD_btultra) return ERROR(compressionParameter_unsupported); + if ((U32)(cParams.strategy) > (U32)ZSTD_btultra) + return ERROR(parameter_unsupported); return 0; } @@ -551,9 +528,7 @@ size_t ZSTD_estimateCCtxSize_advanced(ZSTD_compressionParameters cParams) size_t const hSize = ((size_t)1) << cParams.hashLog; U32 const hashLog3 = (cParams.searchLength>3) ? 0 : MIN(ZSTD_HASHLOG3_MAX, cParams.windowLog); size_t const h3Size = ((size_t)1) << hashLog3; - size_t const entropySpace = hufCTable_size + litlengthCTable_size - + offcodeCTable_size + matchlengthCTable_size - + entropyScratchSpace_size; + size_t const entropySpace = sizeof(ZSTD_entropyCTables_t); size_t const tableSpace = (chainSize + hSize + h3Size) * sizeof(U32); size_t const optBudget = ((MaxML+1) + (MaxLL+1) + (MaxOff+1) + (1<<Litbits))*sizeof(U32) @@ -620,8 +595,8 @@ static size_t ZSTD_continueCCtx(ZSTD_CCtx* cctx, ZSTD_parameters params, U64 ple cctx->stage = ZSTDcs_init; cctx->dictID = 0; cctx->loadedDictEnd = 0; - { int i; for (i=0; i<ZSTD_REP_NUM; i++) cctx->rep[i] = repStartValue[i]; } - cctx->seqStore.litLengthSum = 0; /* force reset of btopt stats */ + { int i; for (i=0; i<ZSTD_REP_NUM; i++) cctx->seqStore.rep[i] = repStartValue[i]; } + cctx->optState.litLengthSum = 0; /* force reset of btopt stats */ XXH64_reset(&cctx->xxhState, 0); return 0; } @@ -641,8 +616,10 @@ static size_t ZSTD_resetCCtx_internal(ZSTD_CCtx* zc, if (crp == ZSTDcrp_continue) { if (ZSTD_equivalentParams(params.cParams, zc->appliedParams.cParams)) { DEBUGLOG(5, "ZSTD_equivalentParams()==1"); - zc->fseCTables_ready = 0; - zc->hufCTable_repeatMode = HUF_repeat_none; + zc->entropy->hufCTable_repeatMode = HUF_repeat_none; + zc->entropy->offcode_repeatMode = FSE_repeat_none; + zc->entropy->matchlength_repeatMode = FSE_repeat_none; + zc->entropy->litlength_repeatMode = FSE_repeat_none; return ZSTD_continueCCtx(zc, params, pledgedSrcSize); } } @@ -662,9 +639,7 @@ static size_t ZSTD_resetCCtx_internal(ZSTD_CCtx* zc, void* ptr; /* Check if workSpace is large enough, alloc a new one if needed */ - { size_t const entropySpace = hufCTable_size + litlengthCTable_size - + offcodeCTable_size + matchlengthCTable_size - + entropyScratchSpace_size; + { size_t const entropySpace = sizeof(ZSTD_entropyCTables_t); size_t const optPotentialSpace = ((MaxML+1) + (MaxLL+1) + (MaxOff+1) + (1<<Litbits)) * sizeof(U32) + (ZSTD_OPT_NUM+1) * (sizeof(ZSTD_match_t)+sizeof(ZSTD_optimal_t)); size_t const optSpace = ( (params.cParams.strategy == ZSTD_btopt) @@ -689,16 +664,9 @@ static size_t ZSTD_resetCCtx_internal(ZSTD_CCtx* zc, ptr = zc->workSpace; /* entropy space */ - zc->hufCTable = (HUF_CElt*)ptr; - ptr = (char*)zc->hufCTable + hufCTable_size; /* note : HUF_CElt* is incomplete type, size is estimated via macro */ - zc->offcodeCTable = (FSE_CTable*) ptr; - ptr = (char*)ptr + offcodeCTable_size; - zc->matchlengthCTable = (FSE_CTable*) ptr; - ptr = (char*)ptr + matchlengthCTable_size; - zc->litlengthCTable = (FSE_CTable*) ptr; - ptr = (char*)ptr + litlengthCTable_size; - assert(((size_t)ptr & 3) == 0); /* ensure correct alignment */ - zc->entropyScratchSpace = (unsigned*) ptr; + assert(((size_t)zc->workSpace & 3) == 0); /* ensure correct alignment */ + assert(zc->workSpaceSize >= sizeof(ZSTD_entropyCTables_t)); + zc->entropy = (ZSTD_entropyCTables_t*)zc->workSpace; } } /* init params */ @@ -715,39 +683,35 @@ static size_t ZSTD_resetCCtx_internal(ZSTD_CCtx* zc, zc->stage = ZSTDcs_init; zc->dictID = 0; zc->loadedDictEnd = 0; - zc->fseCTables_ready = 0; - zc->hufCTable_repeatMode = HUF_repeat_none; + zc->entropy->hufCTable_repeatMode = HUF_repeat_none; + zc->entropy->offcode_repeatMode = FSE_repeat_none; + zc->entropy->matchlength_repeatMode = FSE_repeat_none; + zc->entropy->litlength_repeatMode = FSE_repeat_none; zc->nextToUpdate = 1; zc->nextSrc = NULL; zc->base = NULL; zc->dictBase = NULL; zc->dictLimit = 0; zc->lowLimit = 0; - { int i; for (i=0; i<ZSTD_REP_NUM; i++) zc->rep[i] = repStartValue[i]; } + { int i; for (i=0; i<ZSTD_REP_NUM; i++) zc->seqStore.rep[i] = repStartValue[i]; } zc->hashLog3 = hashLog3; - zc->seqStore.litLengthSum = 0; + zc->optState.litLengthSum = 0; - /* ensure entropy tables are close together at the beginning */ - assert((void*)zc->hufCTable == zc->workSpace); - assert((char*)zc->offcodeCTable == (char*)zc->hufCTable + hufCTable_size); - assert((char*)zc->matchlengthCTable == (char*)zc->offcodeCTable + offcodeCTable_size); - assert((char*)zc->litlengthCTable == (char*)zc->matchlengthCTable + matchlengthCTable_size); - assert((char*)zc->entropyScratchSpace == (char*)zc->litlengthCTable + litlengthCTable_size); - ptr = (char*)zc->entropyScratchSpace + entropyScratchSpace_size; + ptr = zc->entropy + 1; /* opt parser space */ if ((params.cParams.strategy == ZSTD_btopt) || (params.cParams.strategy == ZSTD_btultra)) { DEBUGLOG(5, "reserving optimal parser space"); assert(((size_t)ptr & 3) == 0); /* ensure ptr is properly aligned */ - zc->seqStore.litFreq = (U32*)ptr; - zc->seqStore.litLengthFreq = zc->seqStore.litFreq + (1<<Litbits); - zc->seqStore.matchLengthFreq = zc->seqStore.litLengthFreq + (MaxLL+1); - zc->seqStore.offCodeFreq = zc->seqStore.matchLengthFreq + (MaxML+1); - ptr = zc->seqStore.offCodeFreq + (MaxOff+1); - zc->seqStore.matchTable = (ZSTD_match_t*)ptr; - ptr = zc->seqStore.matchTable + ZSTD_OPT_NUM+1; - zc->seqStore.priceTable = (ZSTD_optimal_t*)ptr; - ptr = zc->seqStore.priceTable + ZSTD_OPT_NUM+1; + zc->optState.litFreq = (U32*)ptr; + zc->optState.litLengthFreq = zc->optState.litFreq + (1<<Litbits); + zc->optState.matchLengthFreq = zc->optState.litLengthFreq + (MaxLL+1); + zc->optState.offCodeFreq = zc->optState.matchLengthFreq + (MaxML+1); + ptr = zc->optState.offCodeFreq + (MaxOff+1); + zc->optState.matchTable = (ZSTD_match_t*)ptr; + ptr = zc->optState.matchTable + ZSTD_OPT_NUM+1; + zc->optState.priceTable = (ZSTD_optimal_t*)ptr; + ptr = zc->optState.priceTable + ZSTD_OPT_NUM+1; } /* table Space */ @@ -783,7 +747,7 @@ static size_t ZSTD_resetCCtx_internal(ZSTD_CCtx* zc, * do not use with extDict variant ! */ void ZSTD_invalidateRepCodes(ZSTD_CCtx* cctx) { int i; - for (i=0; i<ZSTD_REP_NUM; i++) cctx->rep[i] = 0; + for (i=0; i<ZSTD_REP_NUM; i++) cctx->seqStore.rep[i] = 0; } @@ -830,16 +794,7 @@ static size_t ZSTD_copyCCtx_internal(ZSTD_CCtx* dstCCtx, dstCCtx->dictID = srcCCtx->dictID; /* copy entropy tables */ - dstCCtx->fseCTables_ready = srcCCtx->fseCTables_ready; - if (srcCCtx->fseCTables_ready) { - memcpy(dstCCtx->litlengthCTable, srcCCtx->litlengthCTable, litlengthCTable_size); - memcpy(dstCCtx->matchlengthCTable, srcCCtx->matchlengthCTable, matchlengthCTable_size); - memcpy(dstCCtx->offcodeCTable, srcCCtx->offcodeCTable, offcodeCTable_size); - } - dstCCtx->hufCTable_repeatMode = srcCCtx->hufCTable_repeatMode; - if (srcCCtx->hufCTable_repeatMode) { - memcpy(dstCCtx->hufCTable, srcCCtx->hufCTable, hufCTable_size); - } + memcpy(dstCCtx->entropy, srcCCtx->entropy, sizeof(ZSTD_entropyCTables_t)); return 0; } @@ -956,7 +911,8 @@ static size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, cons static size_t ZSTD_minGain(size_t srcSize) { return (srcSize >> 6) + 2; } -static size_t ZSTD_compressLiterals (ZSTD_CCtx* zc, +static size_t ZSTD_compressLiterals (ZSTD_entropyCTables_t * entropy, + ZSTD_strategy strategy, void* dst, size_t dstCapacity, const void* src, size_t srcSize) { @@ -970,28 +926,28 @@ static size_t ZSTD_compressLiterals (ZSTD_CCtx* zc, /* small ? don't even attempt compression (speed opt) */ # define LITERAL_NOENTROPY 63 - { size_t const minLitSize = zc->hufCTable_repeatMode == HUF_repeat_valid ? 6 : LITERAL_NOENTROPY; + { size_t const minLitSize = entropy->hufCTable_repeatMode == HUF_repeat_valid ? 6 : LITERAL_NOENTROPY; if (srcSize <= minLitSize) return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize); } if (dstCapacity < lhSize+1) return ERROR(dstSize_tooSmall); /* not enough space for compression */ - { HUF_repeat repeat = zc->hufCTable_repeatMode; - int const preferRepeat = zc->appliedParams.cParams.strategy < ZSTD_lazy ? srcSize <= 1024 : 0; + { HUF_repeat repeat = entropy->hufCTable_repeatMode; + int const preferRepeat = strategy < ZSTD_lazy ? srcSize <= 1024 : 0; if (repeat == HUF_repeat_valid && lhSize == 3) singleStream = 1; cLitSize = singleStream ? HUF_compress1X_repeat(ostart+lhSize, dstCapacity-lhSize, src, srcSize, 255, 11, - zc->entropyScratchSpace, entropyScratchSpace_size, zc->hufCTable, &repeat, preferRepeat) + entropy->workspace, sizeof(entropy->workspace), (HUF_CElt*)entropy->hufCTable, &repeat, preferRepeat) : HUF_compress4X_repeat(ostart+lhSize, dstCapacity-lhSize, src, srcSize, 255, 11, - zc->entropyScratchSpace, entropyScratchSpace_size, zc->hufCTable, &repeat, preferRepeat); + entropy->workspace, sizeof(entropy->workspace), (HUF_CElt*)entropy->hufCTable, &repeat, preferRepeat); if (repeat != HUF_repeat_none) { hType = set_repeat; } /* reused the existing table */ - else { zc->hufCTable_repeatMode = HUF_repeat_check; } /* now have a table to reuse */ + else { entropy->hufCTable_repeatMode = HUF_repeat_check; } /* now have a table to reuse */ } - if ((cLitSize==0) | (cLitSize >= srcSize - minGain)) { - zc->hufCTable_repeatMode = HUF_repeat_none; + if ((cLitSize==0) | (cLitSize >= srcSize - minGain) | ERR_isError(cLitSize)) { + entropy->hufCTable_repeatMode = HUF_repeat_none; return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize); } if (cLitSize==1) { - zc->hufCTable_repeatMode = HUF_repeat_none; + entropy->hufCTable_repeatMode = HUF_repeat_none; return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize); } @@ -1062,17 +1018,154 @@ void ZSTD_seqToCodes(const seqStore_t* seqStorePtr) mlCodeTable[seqStorePtr->longLengthPos] = MaxML; } -MEM_STATIC size_t ZSTD_compressSequences (ZSTD_CCtx* zc, - void* dst, size_t dstCapacity, - size_t srcSize) +MEM_STATIC symbolEncodingType_e ZSTD_selectEncodingType(FSE_repeat* repeatMode, + size_t const mostFrequent, size_t nbSeq, U32 defaultNormLog) +{ +#define MIN_SEQ_FOR_DYNAMIC_FSE 64 +#define MAX_SEQ_FOR_STATIC_FSE 1000 + + if ((mostFrequent == nbSeq) && (nbSeq > 2)) { + *repeatMode = FSE_repeat_check; + return set_rle; + } + if ((*repeatMode == FSE_repeat_valid) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) { + return set_repeat; + } + if ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (defaultNormLog-1)))) { + *repeatMode = FSE_repeat_valid; + return set_basic; + } + *repeatMode = FSE_repeat_check; + return set_compressed; +} + +MEM_STATIC size_t ZSTD_buildCTable(void* dst, size_t dstCapacity, + FSE_CTable* CTable, U32 FSELog, symbolEncodingType_e type, + U32* count, U32 max, + BYTE const* codeTable, size_t nbSeq, + S16 const* defaultNorm, U32 defaultNormLog, U32 defaultMax, + void* workspace, size_t workspaceSize) +{ + BYTE* op = (BYTE*)dst; + BYTE const* const oend = op + dstCapacity; + + switch (type) { + case set_rle: + *op = codeTable[0]; + CHECK_F(FSE_buildCTable_rle(CTable, (BYTE)max)); + return 1; + case set_repeat: + return 0; + case set_basic: + CHECK_F(FSE_buildCTable_wksp(CTable, defaultNorm, defaultMax, defaultNormLog, workspace, workspaceSize)); + return 0; + case set_compressed: { + S16 norm[MaxSeq + 1]; + size_t nbSeq_1 = nbSeq; + const U32 tableLog = FSE_optimalTableLog(FSELog, nbSeq, max); + if (count[codeTable[nbSeq-1]] > 1) { + count[codeTable[nbSeq-1]]--; + nbSeq_1--; + } + CHECK_F(FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max)); + { size_t const NCountSize = FSE_writeNCount(op, oend - op, norm, max, tableLog); /* overflow protected */ + if (FSE_isError(NCountSize)) return NCountSize; + CHECK_F(FSE_buildCTable_wksp(CTable, norm, max, tableLog, workspace, workspaceSize)); + return NCountSize; + } + } + default: return assert(0), ERROR(GENERIC); + } +} + +MEM_STATIC size_t ZSTD_encodeSequences(void* dst, size_t dstCapacity, + FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable, + FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable, + FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable, + seqDef const* sequences, size_t nbSeq, int longOffsets) +{ + BIT_CStream_t blockStream; + FSE_CState_t stateMatchLength; + FSE_CState_t stateOffsetBits; + FSE_CState_t stateLitLength; + + CHECK_E(BIT_initCStream(&blockStream, dst, dstCapacity), dstSize_tooSmall); /* not enough space remaining */ + + /* first symbols */ + FSE_initCState2(&stateMatchLength, CTable_MatchLength, mlCodeTable[nbSeq-1]); + FSE_initCState2(&stateOffsetBits, CTable_OffsetBits, ofCodeTable[nbSeq-1]); + FSE_initCState2(&stateLitLength, CTable_LitLength, llCodeTable[nbSeq-1]); + BIT_addBits(&blockStream, sequences[nbSeq-1].litLength, LL_bits[llCodeTable[nbSeq-1]]); + if (MEM_32bits()) BIT_flushBits(&blockStream); + BIT_addBits(&blockStream, sequences[nbSeq-1].matchLength, ML_bits[mlCodeTable[nbSeq-1]]); + if (MEM_32bits()) BIT_flushBits(&blockStream); + if (longOffsets) { + U32 const ofBits = ofCodeTable[nbSeq-1]; + int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1); + if (extraBits) { + BIT_addBits(&blockStream, sequences[nbSeq-1].offset, extraBits); + BIT_flushBits(&blockStream); + } + BIT_addBits(&blockStream, sequences[nbSeq-1].offset >> extraBits, + ofBits - extraBits); + } else { + BIT_addBits(&blockStream, sequences[nbSeq-1].offset, ofCodeTable[nbSeq-1]); + } + BIT_flushBits(&blockStream); + + { size_t n; + for (n=nbSeq-2 ; n<nbSeq ; n--) { /* intentional underflow */ + BYTE const llCode = llCodeTable[n]; + BYTE const ofCode = ofCodeTable[n]; + BYTE const mlCode = mlCodeTable[n]; + U32 const llBits = LL_bits[llCode]; + U32 const ofBits = ofCode; /* 32b*/ /* 64b*/ + U32 const mlBits = ML_bits[mlCode]; + /* (7)*/ /* (7)*/ + FSE_encodeSymbol(&blockStream, &stateOffsetBits, ofCode); /* 15 */ /* 15 */ + FSE_encodeSymbol(&blockStream, &stateMatchLength, mlCode); /* 24 */ /* 24 */ + if (MEM_32bits()) BIT_flushBits(&blockStream); /* (7)*/ + FSE_encodeSymbol(&blockStream, &stateLitLength, llCode); /* 16 */ /* 33 */ + if (MEM_32bits() || (ofBits+mlBits+llBits >= 64-7-(LLFSELog+MLFSELog+OffFSELog))) + BIT_flushBits(&blockStream); /* (7)*/ + BIT_addBits(&blockStream, sequences[n].litLength, llBits); + if (MEM_32bits() && ((llBits+mlBits)>24)) BIT_flushBits(&blockStream); + BIT_addBits(&blockStream, sequences[n].matchLength, mlBits); + if (MEM_32bits()) BIT_flushBits(&blockStream); /* (7)*/ + if (longOffsets) { + int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1); + if (extraBits) { + BIT_addBits(&blockStream, sequences[n].offset, extraBits); + BIT_flushBits(&blockStream); /* (7)*/ + } + BIT_addBits(&blockStream, sequences[n].offset >> extraBits, + ofBits - extraBits); /* 31 */ + } else { + BIT_addBits(&blockStream, sequences[n].offset, ofBits); /* 31 */ + } + BIT_flushBits(&blockStream); /* (7)*/ + } } + + FSE_flushCState(&blockStream, &stateMatchLength); + FSE_flushCState(&blockStream, &stateOffsetBits); + FSE_flushCState(&blockStream, &stateLitLength); + + { size_t const streamSize = BIT_closeCStream(&blockStream); + if (streamSize==0) return ERROR(dstSize_tooSmall); /* not enough space */ + return streamSize; + } +} + +MEM_STATIC size_t ZSTD_compressSequences_internal(seqStore_t* seqStorePtr, + ZSTD_entropyCTables_t* entropy, + ZSTD_compressionParameters const* cParams, + void* dst, size_t dstCapacity) { - const int longOffsets = zc->appliedParams.cParams.windowLog > STREAM_ACCUMULATOR_MIN; - const seqStore_t* seqStorePtr = &(zc->seqStore); + const int longOffsets = cParams->windowLog > STREAM_ACCUMULATOR_MIN; U32 count[MaxSeq+1]; - S16 norm[MaxSeq+1]; - FSE_CTable* CTable_LitLength = zc->litlengthCTable; - FSE_CTable* CTable_OffsetBits = zc->offcodeCTable; - FSE_CTable* CTable_MatchLength = zc->matchlengthCTable; + FSE_CTable* CTable_LitLength = entropy->litlengthCTable; + FSE_CTable* CTable_OffsetBits = entropy->offcodeCTable; + FSE_CTable* CTable_MatchLength = entropy->matchlengthCTable; U32 LLtype, Offtype, MLtype; /* compressed, raw or rle */ const seqDef* const sequences = seqStorePtr->sequencesStart; const BYTE* const ofCodeTable = seqStorePtr->ofCode; @@ -1083,13 +1176,16 @@ MEM_STATIC size_t ZSTD_compressSequences (ZSTD_CCtx* zc, BYTE* op = ostart; size_t const nbSeq = seqStorePtr->sequences - seqStorePtr->sequencesStart; BYTE* seqHead; - BYTE scratchBuffer[1<<MAX(MLFSELog,LLFSELog)]; + + ZSTD_STATIC_ASSERT(sizeof(entropy->workspace) >= (1<<MAX(MLFSELog,LLFSELog))); /* Compress literals */ { const BYTE* const literals = seqStorePtr->litStart; size_t const litSize = seqStorePtr->lit - literals; - size_t const cSize = ZSTD_compressLiterals(zc, op, dstCapacity, literals, litSize); - if (ZSTD_isError(cSize)) return cSize; + size_t const cSize = ZSTD_compressLiterals( + entropy, cParams->strategy, op, dstCapacity, literals, litSize); + if (ZSTD_isError(cSize)) + return cSize; op += cSize; } @@ -1098,177 +1194,89 @@ MEM_STATIC size_t ZSTD_compressSequences (ZSTD_CCtx* zc, if (nbSeq < 0x7F) *op++ = (BYTE)nbSeq; else if (nbSeq < LONGNBSEQ) op[0] = (BYTE)((nbSeq>>8) + 0x80), op[1] = (BYTE)nbSeq, op+=2; else op[0]=0xFF, MEM_writeLE16(op+1, (U16)(nbSeq - LONGNBSEQ)), op+=3; - if (nbSeq==0) goto _check_compressibility; + if (nbSeq==0) return op - ostart; /* seqHead : flags for FSE encoding type */ seqHead = op++; -#define MIN_SEQ_FOR_DYNAMIC_FSE 64 -#define MAX_SEQ_FOR_STATIC_FSE 1000 - /* convert length/distances into codes */ ZSTD_seqToCodes(seqStorePtr); - /* CTable for Literal Lengths */ { U32 max = MaxLL; - size_t const mostFrequent = FSE_countFast_wksp(count, &max, llCodeTable, nbSeq, zc->entropyScratchSpace); - if ((mostFrequent == nbSeq) && (nbSeq > 2)) { - *op++ = llCodeTable[0]; - FSE_buildCTable_rle(CTable_LitLength, (BYTE)max); - LLtype = set_rle; - } else if ((zc->fseCTables_ready) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) { - LLtype = set_repeat; - } else if ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (LL_defaultNormLog-1)))) { - FSE_buildCTable_wksp(CTable_LitLength, LL_defaultNorm, MaxLL, LL_defaultNormLog, scratchBuffer, sizeof(scratchBuffer)); - LLtype = set_basic; - } else { - size_t nbSeq_1 = nbSeq; - const U32 tableLog = FSE_optimalTableLog(LLFSELog, nbSeq, max); - if (count[llCodeTable[nbSeq-1]]>1) { count[llCodeTable[nbSeq-1]]--; nbSeq_1--; } - FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max); - { size_t const NCountSize = FSE_writeNCount(op, oend-op, norm, max, tableLog); /* overflow protected */ - if (FSE_isError(NCountSize)) return NCountSize; - op += NCountSize; } - FSE_buildCTable_wksp(CTable_LitLength, norm, max, tableLog, scratchBuffer, sizeof(scratchBuffer)); - LLtype = set_compressed; + size_t const mostFrequent = FSE_countFast_wksp(count, &max, llCodeTable, nbSeq, entropy->workspace); + LLtype = ZSTD_selectEncodingType(&entropy->litlength_repeatMode, mostFrequent, nbSeq, LL_defaultNormLog); + { size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_LitLength, LLFSELog, (symbolEncodingType_e)LLtype, + count, max, llCodeTable, nbSeq, LL_defaultNorm, LL_defaultNormLog, MaxLL, + entropy->workspace, sizeof(entropy->workspace)); + if (ZSTD_isError(countSize)) return countSize; + op += countSize; } } - /* CTable for Offsets */ { U32 max = MaxOff; - size_t const mostFrequent = FSE_countFast_wksp(count, &max, ofCodeTable, nbSeq, zc->entropyScratchSpace); - if ((mostFrequent == nbSeq) && (nbSeq > 2)) { - *op++ = ofCodeTable[0]; - FSE_buildCTable_rle(CTable_OffsetBits, (BYTE)max); - Offtype = set_rle; - } else if ((zc->fseCTables_ready) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) { - Offtype = set_repeat; - } else if ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (OF_defaultNormLog-1)))) { - FSE_buildCTable_wksp(CTable_OffsetBits, OF_defaultNorm, MaxOff, OF_defaultNormLog, scratchBuffer, sizeof(scratchBuffer)); - Offtype = set_basic; - } else { - size_t nbSeq_1 = nbSeq; - const U32 tableLog = FSE_optimalTableLog(OffFSELog, nbSeq, max); - if (count[ofCodeTable[nbSeq-1]]>1) { count[ofCodeTable[nbSeq-1]]--; nbSeq_1--; } - FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max); - { size_t const NCountSize = FSE_writeNCount(op, oend-op, norm, max, tableLog); /* overflow protected */ - if (FSE_isError(NCountSize)) return NCountSize; - op += NCountSize; } - FSE_buildCTable_wksp(CTable_OffsetBits, norm, max, tableLog, scratchBuffer, sizeof(scratchBuffer)); - Offtype = set_compressed; + size_t const mostFrequent = FSE_countFast_wksp(count, &max, ofCodeTable, nbSeq, entropy->workspace); + Offtype = ZSTD_selectEncodingType(&entropy->offcode_repeatMode, mostFrequent, nbSeq, OF_defaultNormLog); + { size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_OffsetBits, OffFSELog, (symbolEncodingType_e)Offtype, + count, max, ofCodeTable, nbSeq, OF_defaultNorm, OF_defaultNormLog, MaxOff, + entropy->workspace, sizeof(entropy->workspace)); + if (ZSTD_isError(countSize)) return countSize; + op += countSize; } } - /* CTable for MatchLengths */ { U32 max = MaxML; - size_t const mostFrequent = FSE_countFast_wksp(count, &max, mlCodeTable, nbSeq, zc->entropyScratchSpace); - if ((mostFrequent == nbSeq) && (nbSeq > 2)) { - *op++ = *mlCodeTable; - FSE_buildCTable_rle(CTable_MatchLength, (BYTE)max); - MLtype = set_rle; - } else if ((zc->fseCTables_ready) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) { - MLtype = set_repeat; - } else if ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (ML_defaultNormLog-1)))) { - FSE_buildCTable_wksp(CTable_MatchLength, ML_defaultNorm, MaxML, ML_defaultNormLog, scratchBuffer, sizeof(scratchBuffer)); - MLtype = set_basic; - } else { - size_t nbSeq_1 = nbSeq; - const U32 tableLog = FSE_optimalTableLog(MLFSELog, nbSeq, max); - if (count[mlCodeTable[nbSeq-1]]>1) { count[mlCodeTable[nbSeq-1]]--; nbSeq_1--; } - FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max); - { size_t const NCountSize = FSE_writeNCount(op, oend-op, norm, max, tableLog); /* overflow protected */ - if (FSE_isError(NCountSize)) return NCountSize; - op += NCountSize; } - FSE_buildCTable_wksp(CTable_MatchLength, norm, max, tableLog, scratchBuffer, sizeof(scratchBuffer)); - MLtype = set_compressed; + size_t const mostFrequent = FSE_countFast_wksp(count, &max, mlCodeTable, nbSeq, entropy->workspace); + MLtype = ZSTD_selectEncodingType(&entropy->matchlength_repeatMode, mostFrequent, nbSeq, ML_defaultNormLog); + { size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_MatchLength, MLFSELog, (symbolEncodingType_e)MLtype, + count, max, mlCodeTable, nbSeq, ML_defaultNorm, ML_defaultNormLog, MaxML, + entropy->workspace, sizeof(entropy->workspace)); + if (ZSTD_isError(countSize)) return countSize; + op += countSize; } } *seqHead = (BYTE)((LLtype<<6) + (Offtype<<4) + (MLtype<<2)); - zc->fseCTables_ready = 0; - - /* Encoding Sequences */ - { BIT_CStream_t blockStream; - FSE_CState_t stateMatchLength; - FSE_CState_t stateOffsetBits; - FSE_CState_t stateLitLength; - - CHECK_E(BIT_initCStream(&blockStream, op, oend-op), dstSize_tooSmall); /* not enough space remaining */ - - /* first symbols */ - FSE_initCState2(&stateMatchLength, CTable_MatchLength, mlCodeTable[nbSeq-1]); - FSE_initCState2(&stateOffsetBits, CTable_OffsetBits, ofCodeTable[nbSeq-1]); - FSE_initCState2(&stateLitLength, CTable_LitLength, llCodeTable[nbSeq-1]); - BIT_addBits(&blockStream, sequences[nbSeq-1].litLength, LL_bits[llCodeTable[nbSeq-1]]); - if (MEM_32bits()) BIT_flushBits(&blockStream); - BIT_addBits(&blockStream, sequences[nbSeq-1].matchLength, ML_bits[mlCodeTable[nbSeq-1]]); - if (MEM_32bits()) BIT_flushBits(&blockStream); - if (longOffsets) { - U32 const ofBits = ofCodeTable[nbSeq-1]; - int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1); - if (extraBits) { - BIT_addBits(&blockStream, sequences[nbSeq-1].offset, extraBits); - BIT_flushBits(&blockStream); - } - BIT_addBits(&blockStream, sequences[nbSeq-1].offset >> extraBits, - ofBits - extraBits); - } else { - BIT_addBits(&blockStream, sequences[nbSeq-1].offset, ofCodeTable[nbSeq-1]); - } - BIT_flushBits(&blockStream); - - { size_t n; - for (n=nbSeq-2 ; n<nbSeq ; n--) { /* intentional underflow */ - BYTE const llCode = llCodeTable[n]; - BYTE const ofCode = ofCodeTable[n]; - BYTE const mlCode = mlCodeTable[n]; - U32 const llBits = LL_bits[llCode]; - U32 const ofBits = ofCode; /* 32b*/ /* 64b*/ - U32 const mlBits = ML_bits[mlCode]; - /* (7)*/ /* (7)*/ - FSE_encodeSymbol(&blockStream, &stateOffsetBits, ofCode); /* 15 */ /* 15 */ - FSE_encodeSymbol(&blockStream, &stateMatchLength, mlCode); /* 24 */ /* 24 */ - if (MEM_32bits()) BIT_flushBits(&blockStream); /* (7)*/ - FSE_encodeSymbol(&blockStream, &stateLitLength, llCode); /* 16 */ /* 33 */ - if (MEM_32bits() || (ofBits+mlBits+llBits >= 64-7-(LLFSELog+MLFSELog+OffFSELog))) - BIT_flushBits(&blockStream); /* (7)*/ - BIT_addBits(&blockStream, sequences[n].litLength, llBits); - if (MEM_32bits() && ((llBits+mlBits)>24)) BIT_flushBits(&blockStream); - BIT_addBits(&blockStream, sequences[n].matchLength, mlBits); - if (MEM_32bits()) BIT_flushBits(&blockStream); /* (7)*/ - if (longOffsets) { - int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1); - if (extraBits) { - BIT_addBits(&blockStream, sequences[n].offset, extraBits); - BIT_flushBits(&blockStream); /* (7)*/ - } - BIT_addBits(&blockStream, sequences[n].offset >> extraBits, - ofBits - extraBits); /* 31 */ - } else { - BIT_addBits(&blockStream, sequences[n].offset, ofBits); /* 31 */ - } - BIT_flushBits(&blockStream); /* (7)*/ - } } - FSE_flushCState(&blockStream, &stateMatchLength); - FSE_flushCState(&blockStream, &stateOffsetBits); - FSE_flushCState(&blockStream, &stateLitLength); + { size_t const streamSize = ZSTD_encodeSequences(op, oend - op, + CTable_MatchLength, mlCodeTable, + CTable_OffsetBits, ofCodeTable, + CTable_LitLength, llCodeTable, + sequences, nbSeq, longOffsets); + if (ZSTD_isError(streamSize)) return streamSize; + op += streamSize; + } - { size_t const streamSize = BIT_closeCStream(&blockStream); - if (streamSize==0) return ERROR(dstSize_tooSmall); /* not enough space */ - op += streamSize; - } } + return op - ostart; +} - /* check compressibility */ -_check_compressibility: - { size_t const minGain = ZSTD_minGain(srcSize); - size_t const maxCSize = srcSize - minGain; - if ((size_t)(op-ostart) >= maxCSize) { - zc->hufCTable_repeatMode = HUF_repeat_none; - return 0; - } } +MEM_STATIC size_t ZSTD_compressSequences(seqStore_t* seqStorePtr, + ZSTD_entropyCTables_t* entropy, + ZSTD_compressionParameters const* cParams, + void* dst, size_t dstCapacity, + size_t srcSize) +{ + size_t const cSize = ZSTD_compressSequences_internal(seqStorePtr, entropy, cParams, + dst, dstCapacity); + size_t const minGain = ZSTD_minGain(srcSize); + size_t const maxCSize = srcSize - minGain; + /* If the srcSize <= dstCapacity, then there is enough space to write a + * raw uncompressed block. Since we ran out of space, the block must not + * be compressible, so fall back to a raw uncompressed block. + */ + int const uncompressibleError = cSize == ERROR(dstSize_tooSmall) && srcSize <= dstCapacity; + + if (ZSTD_isError(cSize) && !uncompressibleError) + return cSize; + /* Check compressibility */ + if (cSize >= maxCSize || uncompressibleError) { + entropy->hufCTable_repeatMode = HUF_repeat_none; + entropy->offcode_repeatMode = FSE_repeat_none; + entropy->matchlength_repeatMode = FSE_repeat_none; + entropy->litlength_repeatMode = FSE_repeat_none; + return 0; + } + assert(!ZSTD_isError(cSize)); /* confirm repcodes */ - { int i; for (i=0; i<ZSTD_REP_NUM; i++) zc->rep[i] = zc->repToConfirm[i]; } - - return op - ostart; + { int i; for (i=0; i<ZSTD_REP_NUM; i++) seqStorePtr->rep[i] = seqStorePtr->repToConfirm[i]; } + return cSize; } @@ -1475,7 +1483,7 @@ static void ZSTD_fillHashTable (ZSTD_CCtx* zc, const void* end, const U32 mls) } -FORCE_INLINE +FORCE_INLINE_TEMPLATE void ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx, const void* src, size_t srcSize, const U32 mls) @@ -1491,7 +1499,7 @@ void ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx, const BYTE* const lowest = base + lowestIndex; const BYTE* const iend = istart + srcSize; const BYTE* const ilimit = iend - HASH_READ_SIZE; - U32 offset_1=cctx->rep[0], offset_2=cctx->rep[1]; + U32 offset_1=seqStorePtr->rep[0], offset_2=seqStorePtr->rep[1]; U32 offsetSaved = 0; /* init */ @@ -1552,8 +1560,8 @@ void ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx, } } } /* save reps for next block */ - cctx->repToConfirm[0] = offset_1 ? offset_1 : offsetSaved; - cctx->repToConfirm[1] = offset_2 ? offset_2 : offsetSaved; + seqStorePtr->repToConfirm[0] = offset_1 ? offset_1 : offsetSaved; + seqStorePtr->repToConfirm[1] = offset_2 ? offset_2 : offsetSaved; /* Last Literals */ { size_t const lastLLSize = iend - anchor; @@ -1601,7 +1609,7 @@ static void ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx, const BYTE* const dictEnd = dictBase + dictLimit; const BYTE* const iend = istart + srcSize; const BYTE* const ilimit = iend - 8; - U32 offset_1=ctx->rep[0], offset_2=ctx->rep[1]; + U32 offset_1=seqStorePtr->rep[0], offset_2=seqStorePtr->rep[1]; /* Search Loop */ while (ip < ilimit) { /* < instead of <=, because (ip+1) */ @@ -1667,7 +1675,7 @@ static void ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx, } } } /* save reps for next block */ - ctx->repToConfirm[0] = offset_1; ctx->repToConfirm[1] = offset_2; + seqStorePtr->repToConfirm[0] = offset_1; seqStorePtr->repToConfirm[1] = offset_2; /* Last Literals */ { size_t const lastLLSize = iend - anchor; @@ -1718,7 +1726,7 @@ static void ZSTD_fillDoubleHashTable (ZSTD_CCtx* cctx, const void* end, const U3 } -FORCE_INLINE +FORCE_INLINE_TEMPLATE void ZSTD_compressBlock_doubleFast_generic(ZSTD_CCtx* cctx, const void* src, size_t srcSize, const U32 mls) @@ -1736,7 +1744,7 @@ void ZSTD_compressBlock_doubleFast_generic(ZSTD_CCtx* cctx, const BYTE* const lowest = base + lowestIndex; const BYTE* const iend = istart + srcSize; const BYTE* const ilimit = iend - HASH_READ_SIZE; - U32 offset_1=cctx->rep[0], offset_2=cctx->rep[1]; + U32 offset_1=seqStorePtr->rep[0], offset_2=seqStorePtr->rep[1]; U32 offsetSaved = 0; /* init */ @@ -1823,8 +1831,8 @@ void ZSTD_compressBlock_doubleFast_generic(ZSTD_CCtx* cctx, } } } /* save reps for next block */ - cctx->repToConfirm[0] = offset_1 ? offset_1 : offsetSaved; - cctx->repToConfirm[1] = offset_2 ? offset_2 : offsetSaved; + seqStorePtr->repToConfirm[0] = offset_1 ? offset_1 : offsetSaved; + seqStorePtr->repToConfirm[1] = offset_2 ? offset_2 : offsetSaved; /* Last Literals */ { size_t const lastLLSize = iend - anchor; @@ -1873,7 +1881,7 @@ static void ZSTD_compressBlock_doubleFast_extDict_generic(ZSTD_CCtx* ctx, const BYTE* const dictEnd = dictBase + dictLimit; const BYTE* const iend = istart + srcSize; const BYTE* const ilimit = iend - 8; - U32 offset_1=ctx->rep[0], offset_2=ctx->rep[1]; + U32 offset_1=seqStorePtr->rep[0], offset_2=seqStorePtr->rep[1]; /* Search Loop */ while (ip < ilimit) { /* < instead of <=, because (ip+1) */ @@ -1973,7 +1981,7 @@ static void ZSTD_compressBlock_doubleFast_extDict_generic(ZSTD_CCtx* ctx, } } } /* save reps for next block */ - ctx->repToConfirm[0] = offset_1; ctx->repToConfirm[1] = offset_2; + seqStorePtr->repToConfirm[0] = offset_1; seqStorePtr->repToConfirm[1] = offset_2; /* Last Literals */ { size_t const lastLLSize = iend - anchor; @@ -2276,7 +2284,7 @@ static size_t ZSTD_BtFindBestMatch_selectMLS_extDict ( /* Update chains up to ip (excluded) Assumption : always within prefix (i.e. not within extDict) */ -FORCE_INLINE +FORCE_INLINE_TEMPLATE U32 ZSTD_insertAndFindFirstIndex (ZSTD_CCtx* zc, const BYTE* ip, U32 mls) { U32* const hashTable = zc->hashTable; @@ -2300,7 +2308,7 @@ U32 ZSTD_insertAndFindFirstIndex (ZSTD_CCtx* zc, const BYTE* ip, U32 mls) /* inlining is important to hardwire a hot branch (template emulation) */ -FORCE_INLINE +FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_generic ( ZSTD_CCtx* zc, /* Index table will be updated */ const BYTE* const ip, const BYTE* const iLimit, @@ -2352,7 +2360,7 @@ size_t ZSTD_HcFindBestMatch_generic ( } -FORCE_INLINE size_t ZSTD_HcFindBestMatch_selectMLS ( +FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_selectMLS ( ZSTD_CCtx* zc, const BYTE* ip, const BYTE* const iLimit, size_t* offsetPtr, @@ -2369,7 +2377,7 @@ FORCE_INLINE size_t ZSTD_HcFindBestMatch_selectMLS ( } -FORCE_INLINE size_t ZSTD_HcFindBestMatch_extDict_selectMLS ( +FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS ( ZSTD_CCtx* zc, const BYTE* ip, const BYTE* const iLimit, size_t* offsetPtr, @@ -2389,7 +2397,7 @@ FORCE_INLINE size_t ZSTD_HcFindBestMatch_extDict_selectMLS ( /* ******************************* * Common parser - lazy strategy *********************************/ -FORCE_INLINE +FORCE_INLINE_TEMPLATE void ZSTD_compressBlock_lazy_generic(ZSTD_CCtx* ctx, const void* src, size_t srcSize, const U32 searchMethod, const U32 depth) @@ -2409,7 +2417,7 @@ void ZSTD_compressBlock_lazy_generic(ZSTD_CCtx* ctx, size_t* offsetPtr, U32 maxNbAttempts, U32 matchLengthSearch); searchMax_f const searchMax = searchMethod ? ZSTD_BtFindBestMatch_selectMLS : ZSTD_HcFindBestMatch_selectMLS; - U32 offset_1 = ctx->rep[0], offset_2 = ctx->rep[1], savedOffset=0; + U32 offset_1 = seqStorePtr->rep[0], offset_2 = seqStorePtr->rep[1], savedOffset=0; /* init */ ip += (ip==base); @@ -2519,8 +2527,8 @@ _storeSequence: } } /* Save reps for next block */ - ctx->repToConfirm[0] = offset_1 ? offset_1 : savedOffset; - ctx->repToConfirm[1] = offset_2 ? offset_2 : savedOffset; + seqStorePtr->repToConfirm[0] = offset_1 ? offset_1 : savedOffset; + seqStorePtr->repToConfirm[1] = offset_2 ? offset_2 : savedOffset; /* Last Literals */ { size_t const lastLLSize = iend - anchor; @@ -2551,7 +2559,7 @@ static void ZSTD_compressBlock_greedy(ZSTD_CCtx* ctx, const void* src, size_t sr } -FORCE_INLINE +FORCE_INLINE_TEMPLATE void ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx* ctx, const void* src, size_t srcSize, const U32 searchMethod, const U32 depth) @@ -2578,7 +2586,7 @@ void ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx* ctx, U32 maxNbAttempts, U32 matchLengthSearch); searchMax_f searchMax = searchMethod ? ZSTD_BtFindBestMatch_selectMLS_extDict : ZSTD_HcFindBestMatch_extDict_selectMLS; - U32 offset_1 = ctx->rep[0], offset_2 = ctx->rep[1]; + U32 offset_1 = seqStorePtr->rep[0], offset_2 = seqStorePtr->rep[1]; /* init */ ctx->nextToUpdate3 = ctx->nextToUpdate; @@ -2714,7 +2722,7 @@ _storeSequence: } } /* Save reps for next block */ - ctx->repToConfirm[0] = offset_1; ctx->repToConfirm[1] = offset_2; + seqStorePtr->repToConfirm[0] = offset_1; seqStorePtr->repToConfirm[1] = offset_2; /* Last Literals */ { size_t const lastLLSize = iend - anchor; @@ -2823,7 +2831,7 @@ static size_t ZSTD_compressBlock_internal(ZSTD_CCtx* zc, void* dst, size_t dstCa if (current > zc->nextToUpdate + 384) zc->nextToUpdate = current - MIN(192, (U32)(current - zc->nextToUpdate - 384)); /* limited update after finding a very long match */ blockCompressor(zc, src, srcSize); - return ZSTD_compressSequences(zc, dst, dstCapacity, srcSize); + return ZSTD_compressSequences(&zc->seqStore, zc->entropy, &zc->appliedParams.cParams, dst, dstCapacity, srcSize); } @@ -3000,7 +3008,6 @@ static size_t ZSTD_compressContinue_internal (ZSTD_CCtx* cctx, return fhSize; } - size_t ZSTD_compressContinue (ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize) @@ -3106,13 +3113,14 @@ static size_t ZSTD_loadZstdDictionary(ZSTD_CCtx* cctx, const void* dict, size_t const BYTE* const dictEnd = dictPtr + dictSize; short offcodeNCount[MaxOff+1]; unsigned offcodeMaxValue = MaxOff; - BYTE scratchBuffer[1<<MAX(MLFSELog,LLFSELog)]; + + ZSTD_STATIC_ASSERT(sizeof(cctx->entropy->workspace) >= (1<<MAX(MLFSELog,LLFSELog))); dictPtr += 4; /* skip magic number */ cctx->dictID = cctx->appliedParams.fParams.noDictIDFlag ? 0 : MEM_readLE32(dictPtr); dictPtr += 4; - { size_t const hufHeaderSize = HUF_readCTable(cctx->hufCTable, 255, dictPtr, dictEnd-dictPtr); + { size_t const hufHeaderSize = HUF_readCTable((HUF_CElt*)cctx->entropy->hufCTable, 255, dictPtr, dictEnd-dictPtr); if (HUF_isError(hufHeaderSize)) return ERROR(dictionary_corrupted); dictPtr += hufHeaderSize; } @@ -3122,7 +3130,7 @@ static size_t ZSTD_loadZstdDictionary(ZSTD_CCtx* cctx, const void* dict, size_t if (FSE_isError(offcodeHeaderSize)) return ERROR(dictionary_corrupted); if (offcodeLog > OffFSELog) return ERROR(dictionary_corrupted); /* Defer checking offcodeMaxValue because we need to know the size of the dictionary content */ - CHECK_E( FSE_buildCTable_wksp(cctx->offcodeCTable, offcodeNCount, offcodeMaxValue, offcodeLog, scratchBuffer, sizeof(scratchBuffer)), + CHECK_E( FSE_buildCTable_wksp(cctx->entropy->offcodeCTable, offcodeNCount, offcodeMaxValue, offcodeLog, cctx->entropy->workspace, sizeof(cctx->entropy->workspace)), dictionary_corrupted); dictPtr += offcodeHeaderSize; } @@ -3134,7 +3142,7 @@ static size_t ZSTD_loadZstdDictionary(ZSTD_CCtx* cctx, const void* dict, size_t if (matchlengthLog > MLFSELog) return ERROR(dictionary_corrupted); /* Every match length code must have non-zero probability */ CHECK_F( ZSTD_checkDictNCount(matchlengthNCount, matchlengthMaxValue, MaxML)); - CHECK_E( FSE_buildCTable_wksp(cctx->matchlengthCTable, matchlengthNCount, matchlengthMaxValue, matchlengthLog, scratchBuffer, sizeof(scratchBuffer)), + CHECK_E( FSE_buildCTable_wksp(cctx->entropy->matchlengthCTable, matchlengthNCount, matchlengthMaxValue, matchlengthLog, cctx->entropy->workspace, sizeof(cctx->entropy->workspace)), dictionary_corrupted); dictPtr += matchlengthHeaderSize; } @@ -3146,15 +3154,15 @@ static size_t ZSTD_loadZstdDictionary(ZSTD_CCtx* cctx, const void* dict, size_t if (litlengthLog > LLFSELog) return ERROR(dictionary_corrupted); /* Every literal length code must have non-zero probability */ CHECK_F( ZSTD_checkDictNCount(litlengthNCount, litlengthMaxValue, MaxLL)); - CHECK_E( FSE_buildCTable_wksp(cctx->litlengthCTable, litlengthNCount, litlengthMaxValue, litlengthLog, scratchBuffer, sizeof(scratchBuffer)), + CHECK_E( FSE_buildCTable_wksp(cctx->entropy->litlengthCTable, litlengthNCount, litlengthMaxValue, litlengthLog, cctx->entropy->workspace, sizeof(cctx->entropy->workspace)), dictionary_corrupted); dictPtr += litlengthHeaderSize; } if (dictPtr+12 > dictEnd) return ERROR(dictionary_corrupted); - cctx->rep[0] = MEM_readLE32(dictPtr+0); - cctx->rep[1] = MEM_readLE32(dictPtr+4); - cctx->rep[2] = MEM_readLE32(dictPtr+8); + cctx->seqStore.rep[0] = MEM_readLE32(dictPtr+0); + cctx->seqStore.rep[1] = MEM_readLE32(dictPtr+4); + cctx->seqStore.rep[2] = MEM_readLE32(dictPtr+8); dictPtr += 12; { size_t const dictContentSize = (size_t)(dictEnd - dictPtr); @@ -3168,12 +3176,14 @@ static size_t ZSTD_loadZstdDictionary(ZSTD_CCtx* cctx, const void* dict, size_t /* All repCodes must be <= dictContentSize and != 0*/ { U32 u; for (u=0; u<3; u++) { - if (cctx->rep[u] == 0) return ERROR(dictionary_corrupted); - if (cctx->rep[u] > dictContentSize) return ERROR(dictionary_corrupted); + if (cctx->seqStore.rep[u] == 0) return ERROR(dictionary_corrupted); + if (cctx->seqStore.rep[u] > dictContentSize) return ERROR(dictionary_corrupted); } } - cctx->fseCTables_ready = 1; - cctx->hufCTable_repeatMode = HUF_repeat_valid; + cctx->entropy->hufCTable_repeatMode = HUF_repeat_valid; + cctx->entropy->offcode_repeatMode = FSE_repeat_valid; + cctx->entropy->matchlength_repeatMode = FSE_repeat_valid; + cctx->entropy->litlength_repeatMode = FSE_repeat_valid; return ZSTD_loadDictionaryContent(cctx, dictPtr, dictContentSize); } } diff --git a/thirdparty/zstd/compress/zstd_opt.h b/thirdparty/zstd/compress/zstd_opt.h index e8e98915ea..ae24732c7d 100644 --- a/thirdparty/zstd/compress/zstd_opt.h +++ b/thirdparty/zstd/compress/zstd_opt.h @@ -1,10 +1,10 @@ -/** - * Copyright (c) 2016-present, Przemyslaw Skibinski, Yann Collet, Facebook, Inc. +/* + * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ @@ -22,173 +22,173 @@ /*-************************************* * Price functions for optimal parser ***************************************/ -FORCE_INLINE void ZSTD_setLog2Prices(seqStore_t* ssPtr) +static void ZSTD_setLog2Prices(optState_t* optPtr) { - ssPtr->log2matchLengthSum = ZSTD_highbit32(ssPtr->matchLengthSum+1); - ssPtr->log2litLengthSum = ZSTD_highbit32(ssPtr->litLengthSum+1); - ssPtr->log2litSum = ZSTD_highbit32(ssPtr->litSum+1); - ssPtr->log2offCodeSum = ZSTD_highbit32(ssPtr->offCodeSum+1); - ssPtr->factor = 1 + ((ssPtr->litSum>>5) / ssPtr->litLengthSum) + ((ssPtr->litSum<<1) / (ssPtr->litSum + ssPtr->matchSum)); + optPtr->log2matchLengthSum = ZSTD_highbit32(optPtr->matchLengthSum+1); + optPtr->log2litLengthSum = ZSTD_highbit32(optPtr->litLengthSum+1); + optPtr->log2litSum = ZSTD_highbit32(optPtr->litSum+1); + optPtr->log2offCodeSum = ZSTD_highbit32(optPtr->offCodeSum+1); + optPtr->factor = 1 + ((optPtr->litSum>>5) / optPtr->litLengthSum) + ((optPtr->litSum<<1) / (optPtr->litSum + optPtr->matchSum)); } -MEM_STATIC void ZSTD_rescaleFreqs(seqStore_t* ssPtr, const BYTE* src, size_t srcSize) +static void ZSTD_rescaleFreqs(optState_t* optPtr, const BYTE* src, size_t srcSize) { unsigned u; - ssPtr->cachedLiterals = NULL; - ssPtr->cachedPrice = ssPtr->cachedLitLength = 0; - ssPtr->staticPrices = 0; + optPtr->cachedLiterals = NULL; + optPtr->cachedPrice = optPtr->cachedLitLength = 0; + optPtr->staticPrices = 0; - if (ssPtr->litLengthSum == 0) { - if (srcSize <= 1024) ssPtr->staticPrices = 1; + if (optPtr->litLengthSum == 0) { + if (srcSize <= 1024) optPtr->staticPrices = 1; - assert(ssPtr->litFreq!=NULL); + assert(optPtr->litFreq!=NULL); for (u=0; u<=MaxLit; u++) - ssPtr->litFreq[u] = 0; + optPtr->litFreq[u] = 0; for (u=0; u<srcSize; u++) - ssPtr->litFreq[src[u]]++; + optPtr->litFreq[src[u]]++; - ssPtr->litSum = 0; - ssPtr->litLengthSum = MaxLL+1; - ssPtr->matchLengthSum = MaxML+1; - ssPtr->offCodeSum = (MaxOff+1); - ssPtr->matchSum = (ZSTD_LITFREQ_ADD<<Litbits); + optPtr->litSum = 0; + optPtr->litLengthSum = MaxLL+1; + optPtr->matchLengthSum = MaxML+1; + optPtr->offCodeSum = (MaxOff+1); + optPtr->matchSum = (ZSTD_LITFREQ_ADD<<Litbits); for (u=0; u<=MaxLit; u++) { - ssPtr->litFreq[u] = 1 + (ssPtr->litFreq[u]>>ZSTD_FREQ_DIV); - ssPtr->litSum += ssPtr->litFreq[u]; + optPtr->litFreq[u] = 1 + (optPtr->litFreq[u]>>ZSTD_FREQ_DIV); + optPtr->litSum += optPtr->litFreq[u]; } for (u=0; u<=MaxLL; u++) - ssPtr->litLengthFreq[u] = 1; + optPtr->litLengthFreq[u] = 1; for (u=0; u<=MaxML; u++) - ssPtr->matchLengthFreq[u] = 1; + optPtr->matchLengthFreq[u] = 1; for (u=0; u<=MaxOff; u++) - ssPtr->offCodeFreq[u] = 1; + optPtr->offCodeFreq[u] = 1; } else { - ssPtr->matchLengthSum = 0; - ssPtr->litLengthSum = 0; - ssPtr->offCodeSum = 0; - ssPtr->matchSum = 0; - ssPtr->litSum = 0; + optPtr->matchLengthSum = 0; + optPtr->litLengthSum = 0; + optPtr->offCodeSum = 0; + optPtr->matchSum = 0; + optPtr->litSum = 0; for (u=0; u<=MaxLit; u++) { - ssPtr->litFreq[u] = 1 + (ssPtr->litFreq[u]>>(ZSTD_FREQ_DIV+1)); - ssPtr->litSum += ssPtr->litFreq[u]; + optPtr->litFreq[u] = 1 + (optPtr->litFreq[u]>>(ZSTD_FREQ_DIV+1)); + optPtr->litSum += optPtr->litFreq[u]; } for (u=0; u<=MaxLL; u++) { - ssPtr->litLengthFreq[u] = 1 + (ssPtr->litLengthFreq[u]>>(ZSTD_FREQ_DIV+1)); - ssPtr->litLengthSum += ssPtr->litLengthFreq[u]; + optPtr->litLengthFreq[u] = 1 + (optPtr->litLengthFreq[u]>>(ZSTD_FREQ_DIV+1)); + optPtr->litLengthSum += optPtr->litLengthFreq[u]; } for (u=0; u<=MaxML; u++) { - ssPtr->matchLengthFreq[u] = 1 + (ssPtr->matchLengthFreq[u]>>ZSTD_FREQ_DIV); - ssPtr->matchLengthSum += ssPtr->matchLengthFreq[u]; - ssPtr->matchSum += ssPtr->matchLengthFreq[u] * (u + 3); + optPtr->matchLengthFreq[u] = 1 + (optPtr->matchLengthFreq[u]>>ZSTD_FREQ_DIV); + optPtr->matchLengthSum += optPtr->matchLengthFreq[u]; + optPtr->matchSum += optPtr->matchLengthFreq[u] * (u + 3); } - ssPtr->matchSum *= ZSTD_LITFREQ_ADD; + optPtr->matchSum *= ZSTD_LITFREQ_ADD; for (u=0; u<=MaxOff; u++) { - ssPtr->offCodeFreq[u] = 1 + (ssPtr->offCodeFreq[u]>>ZSTD_FREQ_DIV); - ssPtr->offCodeSum += ssPtr->offCodeFreq[u]; + optPtr->offCodeFreq[u] = 1 + (optPtr->offCodeFreq[u]>>ZSTD_FREQ_DIV); + optPtr->offCodeSum += optPtr->offCodeFreq[u]; } } - ZSTD_setLog2Prices(ssPtr); + ZSTD_setLog2Prices(optPtr); } -FORCE_INLINE U32 ZSTD_getLiteralPrice(seqStore_t* ssPtr, U32 litLength, const BYTE* literals) +static U32 ZSTD_getLiteralPrice(optState_t* optPtr, U32 litLength, const BYTE* literals) { U32 price, u; - if (ssPtr->staticPrices) + if (optPtr->staticPrices) return ZSTD_highbit32((U32)litLength+1) + (litLength*6); if (litLength == 0) - return ssPtr->log2litLengthSum - ZSTD_highbit32(ssPtr->litLengthFreq[0]+1); + return optPtr->log2litLengthSum - ZSTD_highbit32(optPtr->litLengthFreq[0]+1); /* literals */ - if (ssPtr->cachedLiterals == literals) { - U32 const additional = litLength - ssPtr->cachedLitLength; - const BYTE* literals2 = ssPtr->cachedLiterals + ssPtr->cachedLitLength; - price = ssPtr->cachedPrice + additional * ssPtr->log2litSum; + if (optPtr->cachedLiterals == literals) { + U32 const additional = litLength - optPtr->cachedLitLength; + const BYTE* literals2 = optPtr->cachedLiterals + optPtr->cachedLitLength; + price = optPtr->cachedPrice + additional * optPtr->log2litSum; for (u=0; u < additional; u++) - price -= ZSTD_highbit32(ssPtr->litFreq[literals2[u]]+1); - ssPtr->cachedPrice = price; - ssPtr->cachedLitLength = litLength; + price -= ZSTD_highbit32(optPtr->litFreq[literals2[u]]+1); + optPtr->cachedPrice = price; + optPtr->cachedLitLength = litLength; } else { - price = litLength * ssPtr->log2litSum; + price = litLength * optPtr->log2litSum; for (u=0; u < litLength; u++) - price -= ZSTD_highbit32(ssPtr->litFreq[literals[u]]+1); + price -= ZSTD_highbit32(optPtr->litFreq[literals[u]]+1); if (litLength >= 12) { - ssPtr->cachedLiterals = literals; - ssPtr->cachedPrice = price; - ssPtr->cachedLitLength = litLength; + optPtr->cachedLiterals = literals; + optPtr->cachedPrice = price; + optPtr->cachedLitLength = litLength; } } /* literal Length */ { const BYTE LL_deltaCode = 19; const BYTE llCode = (litLength>63) ? (BYTE)ZSTD_highbit32(litLength) + LL_deltaCode : LL_Code[litLength]; - price += LL_bits[llCode] + ssPtr->log2litLengthSum - ZSTD_highbit32(ssPtr->litLengthFreq[llCode]+1); + price += LL_bits[llCode] + optPtr->log2litLengthSum - ZSTD_highbit32(optPtr->litLengthFreq[llCode]+1); } return price; } -FORCE_INLINE U32 ZSTD_getPrice(seqStore_t* seqStorePtr, U32 litLength, const BYTE* literals, U32 offset, U32 matchLength, const int ultra) +FORCE_INLINE_TEMPLATE U32 ZSTD_getPrice(optState_t* optPtr, U32 litLength, const BYTE* literals, U32 offset, U32 matchLength, const int ultra) { /* offset */ U32 price; BYTE const offCode = (BYTE)ZSTD_highbit32(offset+1); - if (seqStorePtr->staticPrices) - return ZSTD_getLiteralPrice(seqStorePtr, litLength, literals) + ZSTD_highbit32((U32)matchLength+1) + 16 + offCode; + if (optPtr->staticPrices) + return ZSTD_getLiteralPrice(optPtr, litLength, literals) + ZSTD_highbit32((U32)matchLength+1) + 16 + offCode; - price = offCode + seqStorePtr->log2offCodeSum - ZSTD_highbit32(seqStorePtr->offCodeFreq[offCode]+1); + price = offCode + optPtr->log2offCodeSum - ZSTD_highbit32(optPtr->offCodeFreq[offCode]+1); if (!ultra && offCode >= 20) price += (offCode-19)*2; /* match Length */ { const BYTE ML_deltaCode = 36; const BYTE mlCode = (matchLength>127) ? (BYTE)ZSTD_highbit32(matchLength) + ML_deltaCode : ML_Code[matchLength]; - price += ML_bits[mlCode] + seqStorePtr->log2matchLengthSum - ZSTD_highbit32(seqStorePtr->matchLengthFreq[mlCode]+1); + price += ML_bits[mlCode] + optPtr->log2matchLengthSum - ZSTD_highbit32(optPtr->matchLengthFreq[mlCode]+1); } - return price + ZSTD_getLiteralPrice(seqStorePtr, litLength, literals) + seqStorePtr->factor; + return price + ZSTD_getLiteralPrice(optPtr, litLength, literals) + optPtr->factor; } -MEM_STATIC void ZSTD_updatePrice(seqStore_t* seqStorePtr, U32 litLength, const BYTE* literals, U32 offset, U32 matchLength) +static void ZSTD_updatePrice(optState_t* optPtr, U32 litLength, const BYTE* literals, U32 offset, U32 matchLength) { U32 u; /* literals */ - seqStorePtr->litSum += litLength*ZSTD_LITFREQ_ADD; + optPtr->litSum += litLength*ZSTD_LITFREQ_ADD; for (u=0; u < litLength; u++) - seqStorePtr->litFreq[literals[u]] += ZSTD_LITFREQ_ADD; + optPtr->litFreq[literals[u]] += ZSTD_LITFREQ_ADD; /* literal Length */ { const BYTE LL_deltaCode = 19; const BYTE llCode = (litLength>63) ? (BYTE)ZSTD_highbit32(litLength) + LL_deltaCode : LL_Code[litLength]; - seqStorePtr->litLengthFreq[llCode]++; - seqStorePtr->litLengthSum++; + optPtr->litLengthFreq[llCode]++; + optPtr->litLengthSum++; } /* match offset */ { BYTE const offCode = (BYTE)ZSTD_highbit32(offset+1); - seqStorePtr->offCodeSum++; - seqStorePtr->offCodeFreq[offCode]++; + optPtr->offCodeSum++; + optPtr->offCodeFreq[offCode]++; } /* match Length */ { const BYTE ML_deltaCode = 36; const BYTE mlCode = (matchLength>127) ? (BYTE)ZSTD_highbit32(matchLength) + ML_deltaCode : ML_Code[matchLength]; - seqStorePtr->matchLengthFreq[mlCode]++; - seqStorePtr->matchLengthSum++; + optPtr->matchLengthFreq[mlCode]++; + optPtr->matchLengthSum++; } - ZSTD_setLog2Prices(seqStorePtr); + ZSTD_setLog2Prices(optPtr); } @@ -203,7 +203,7 @@ MEM_STATIC void ZSTD_updatePrice(seqStore_t* seqStorePtr, U32 litLength, const B /* function safe only for comparisons */ -MEM_STATIC U32 ZSTD_readMINMATCH(const void* memPtr, U32 length) +static U32 ZSTD_readMINMATCH(const void* memPtr, U32 length) { switch (length) { @@ -219,7 +219,7 @@ MEM_STATIC U32 ZSTD_readMINMATCH(const void* memPtr, U32 length) /* Update hashTable3 up to ip (excluded) Assumption : always within prefix (i.e. not within extDict) */ -FORCE_INLINE +static U32 ZSTD_insertAndFindFirstIndexHash3 (ZSTD_CCtx* zc, const BYTE* ip) { U32* const hashTable3 = zc->hashTable3; @@ -412,11 +412,12 @@ static U32 ZSTD_BtGetAllMatches_selectMLS_extDict ( /*-******************************* * Optimal parser *********************************/ -FORCE_INLINE +FORCE_INLINE_TEMPLATE void ZSTD_compressBlock_opt_generic(ZSTD_CCtx* ctx, const void* src, size_t srcSize, const int ultra) { seqStore_t* seqStorePtr = &(ctx->seqStore); + optState_t* optStatePtr = &(ctx->optState); const BYTE* const istart = (const BYTE*)src; const BYTE* ip = istart; const BYTE* anchor = istart; @@ -430,16 +431,16 @@ void ZSTD_compressBlock_opt_generic(ZSTD_CCtx* ctx, const U32 mls = ctx->appliedParams.cParams.searchLength; const U32 minMatch = (ctx->appliedParams.cParams.searchLength == 3) ? 3 : 4; - ZSTD_optimal_t* opt = seqStorePtr->priceTable; - ZSTD_match_t* matches = seqStorePtr->matchTable; + ZSTD_optimal_t* opt = optStatePtr->priceTable; + ZSTD_match_t* matches = optStatePtr->matchTable; const BYTE* inr; U32 offset, rep[ZSTD_REP_NUM]; /* init */ ctx->nextToUpdate3 = ctx->nextToUpdate; - ZSTD_rescaleFreqs(seqStorePtr, (const BYTE*)src, srcSize); + ZSTD_rescaleFreqs(optStatePtr, (const BYTE*)src, srcSize); ip += (ip==prefixStart); - { U32 i; for (i=0; i<ZSTD_REP_NUM; i++) rep[i]=ctx->rep[i]; } + { U32 i; for (i=0; i<ZSTD_REP_NUM; i++) rep[i]=seqStorePtr->rep[i]; } /* Match Loop */ while (ip < ilimit) { @@ -462,7 +463,7 @@ void ZSTD_compressBlock_opt_generic(ZSTD_CCtx* ctx, } best_off = i - (ip == anchor); do { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); if (mlen > last_pos || price < opt[mlen].price) SET_PRICE(mlen, mlen, i, litlen, price); /* note : macro modifies last_pos */ mlen--; @@ -487,7 +488,7 @@ void ZSTD_compressBlock_opt_generic(ZSTD_CCtx* ctx, mlen = (u>0) ? matches[u-1].len+1 : best_mlen; best_mlen = matches[u].len; while (mlen <= best_mlen) { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); if (mlen > last_pos || price < opt[mlen].price) SET_PRICE(mlen, mlen, matches[u].off, litlen, price); /* note : macro modifies last_pos */ mlen++; @@ -507,12 +508,12 @@ void ZSTD_compressBlock_opt_generic(ZSTD_CCtx* ctx, if (opt[cur-1].mlen == 1) { litlen = opt[cur-1].litlen + 1; if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr-litlen); + price = opt[cur - litlen].price + ZSTD_getLiteralPrice(optStatePtr, litlen, inr-litlen); } else - price = ZSTD_getLiteralPrice(seqStorePtr, litlen, anchor); + price = ZSTD_getLiteralPrice(optStatePtr, litlen, anchor); } else { litlen = 1; - price = opt[cur - 1].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr-1); + price = opt[cur - 1].price + ZSTD_getLiteralPrice(optStatePtr, litlen, inr-1); } if (cur > last_pos || price <= opt[cur].price) @@ -554,12 +555,12 @@ void ZSTD_compressBlock_opt_generic(ZSTD_CCtx* ctx, if (opt[cur].mlen == 1) { litlen = opt[cur].litlen; if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, inr-litlen, best_off, mlen - MINMATCH, ultra); + price = opt[cur - litlen].price + ZSTD_getPrice(optStatePtr, litlen, inr-litlen, best_off, mlen - MINMATCH, ultra); } else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); } else { litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, best_off, mlen - MINMATCH, ultra); + price = opt[cur].price + ZSTD_getPrice(optStatePtr, 0, NULL, best_off, mlen - MINMATCH, ultra); } if (cur + mlen > last_pos || price <= opt[cur + mlen].price) @@ -586,12 +587,12 @@ void ZSTD_compressBlock_opt_generic(ZSTD_CCtx* ctx, if (opt[cur].mlen == 1) { litlen = opt[cur].litlen; if (cur > litlen) - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, ip+cur-litlen, matches[u].off-1, mlen - MINMATCH, ultra); + price = opt[cur - litlen].price + ZSTD_getPrice(optStatePtr, litlen, ip+cur-litlen, matches[u].off-1, mlen - MINMATCH, ultra); else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); } else { litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, matches[u].off-1, mlen - MINMATCH, ultra); + price = opt[cur].price + ZSTD_getPrice(optStatePtr, 0, NULL, matches[u].off-1, mlen - MINMATCH, ultra); } if (cur + mlen > last_pos || (price < opt[cur + mlen].price)) @@ -645,13 +646,13 @@ _storeSequence: /* cur, last_pos, best_mlen, best_off have to be set */ if (litLength==0) offset--; } - ZSTD_updatePrice(seqStorePtr, litLength, anchor, offset, mlen-MINMATCH); + ZSTD_updatePrice(optStatePtr, litLength, anchor, offset, mlen-MINMATCH); ZSTD_storeSeq(seqStorePtr, litLength, anchor, offset, mlen-MINMATCH); anchor = ip = ip + mlen; } } /* for (cur=0; cur < last_pos; ) */ /* Save reps for next block */ - { int i; for (i=0; i<ZSTD_REP_NUM; i++) ctx->repToConfirm[i] = rep[i]; } + { int i; for (i=0; i<ZSTD_REP_NUM; i++) seqStorePtr->repToConfirm[i] = rep[i]; } /* Last Literals */ { size_t const lastLLSize = iend - anchor; @@ -661,11 +662,12 @@ _storeSequence: /* cur, last_pos, best_mlen, best_off have to be set */ } -FORCE_INLINE +FORCE_INLINE_TEMPLATE void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx* ctx, const void* src, size_t srcSize, const int ultra) { seqStore_t* seqStorePtr = &(ctx->seqStore); + optState_t* optStatePtr = &(ctx->optState); const BYTE* const istart = (const BYTE*)src; const BYTE* ip = istart; const BYTE* anchor = istart; @@ -683,16 +685,16 @@ void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx* ctx, const U32 mls = ctx->appliedParams.cParams.searchLength; const U32 minMatch = (ctx->appliedParams.cParams.searchLength == 3) ? 3 : 4; - ZSTD_optimal_t* opt = seqStorePtr->priceTable; - ZSTD_match_t* matches = seqStorePtr->matchTable; + ZSTD_optimal_t* opt = optStatePtr->priceTable; + ZSTD_match_t* matches = optStatePtr->matchTable; const BYTE* inr; /* init */ U32 offset, rep[ZSTD_REP_NUM]; - { U32 i; for (i=0; i<ZSTD_REP_NUM; i++) rep[i]=ctx->rep[i]; } + { U32 i; for (i=0; i<ZSTD_REP_NUM; i++) rep[i]=seqStorePtr->rep[i]; } ctx->nextToUpdate3 = ctx->nextToUpdate; - ZSTD_rescaleFreqs(seqStorePtr, (const BYTE*)src, srcSize); + ZSTD_rescaleFreqs(optStatePtr, (const BYTE*)src, srcSize); ip += (ip==prefixStart); /* Match Loop */ @@ -726,7 +728,7 @@ void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx* ctx, best_off = i - (ip==anchor); litlen = opt[0].litlen; do { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); if (mlen > last_pos || price < opt[mlen].price) SET_PRICE(mlen, mlen, i, litlen, price); /* note : macro modifies last_pos */ mlen--; @@ -756,7 +758,7 @@ void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx* ctx, best_mlen = matches[u].len; litlen = opt[0].litlen; while (mlen <= best_mlen) { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); if (mlen > last_pos || price < opt[mlen].price) SET_PRICE(mlen, mlen, matches[u].off, litlen, price); mlen++; @@ -773,12 +775,12 @@ void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx* ctx, if (opt[cur-1].mlen == 1) { litlen = opt[cur-1].litlen + 1; if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr-litlen); + price = opt[cur - litlen].price + ZSTD_getLiteralPrice(optStatePtr, litlen, inr-litlen); } else - price = ZSTD_getLiteralPrice(seqStorePtr, litlen, anchor); + price = ZSTD_getLiteralPrice(optStatePtr, litlen, anchor); } else { litlen = 1; - price = opt[cur - 1].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr-1); + price = opt[cur - 1].price + ZSTD_getLiteralPrice(optStatePtr, litlen, inr-1); } if (cur > last_pos || price <= opt[cur].price) @@ -826,12 +828,12 @@ void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx* ctx, if (opt[cur].mlen == 1) { litlen = opt[cur].litlen; if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, inr-litlen, best_off, mlen - MINMATCH, ultra); + price = opt[cur - litlen].price + ZSTD_getPrice(optStatePtr, litlen, inr-litlen, best_off, mlen - MINMATCH, ultra); } else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); } else { litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, best_off, mlen - MINMATCH, ultra); + price = opt[cur].price + ZSTD_getPrice(optStatePtr, 0, NULL, best_off, mlen - MINMATCH, ultra); } if (cur + mlen > last_pos || price <= opt[cur + mlen].price) @@ -858,12 +860,12 @@ void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx* ctx, if (opt[cur].mlen == 1) { litlen = opt[cur].litlen; if (cur > litlen) - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, ip+cur-litlen, matches[u].off-1, mlen - MINMATCH, ultra); + price = opt[cur - litlen].price + ZSTD_getPrice(optStatePtr, litlen, ip+cur-litlen, matches[u].off-1, mlen - MINMATCH, ultra); else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); + price = ZSTD_getPrice(optStatePtr, litlen, anchor, matches[u].off-1, mlen - MINMATCH, ultra); } else { litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, matches[u].off-1, mlen - MINMATCH, ultra); + price = opt[cur].price + ZSTD_getPrice(optStatePtr, 0, NULL, matches[u].off-1, mlen - MINMATCH, ultra); } if (cur + mlen > last_pos || (price < opt[cur + mlen].price)) @@ -918,13 +920,13 @@ _storeSequence: /* cur, last_pos, best_mlen, best_off have to be set */ if (litLength==0) offset--; } - ZSTD_updatePrice(seqStorePtr, litLength, anchor, offset, mlen-MINMATCH); + ZSTD_updatePrice(optStatePtr, litLength, anchor, offset, mlen-MINMATCH); ZSTD_storeSeq(seqStorePtr, litLength, anchor, offset, mlen-MINMATCH); anchor = ip = ip + mlen; } } /* for (cur=0; cur < last_pos; ) */ /* Save reps for next block */ - { int i; for (i=0; i<ZSTD_REP_NUM; i++) ctx->repToConfirm[i] = rep[i]; } + { int i; for (i=0; i<ZSTD_REP_NUM; i++) seqStorePtr->repToConfirm[i] = rep[i]; } /* Last Literals */ { size_t lastLLSize = iend - anchor; diff --git a/thirdparty/zstd/compress/zstdmt_compress.c b/thirdparty/zstd/compress/zstdmt_compress.c index 0cee01eacb..8564bc4392 100644 --- a/thirdparty/zstd/compress/zstdmt_compress.c +++ b/thirdparty/zstd/compress/zstdmt_compress.c @@ -1,15 +1,16 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ /* ====== Tuning parameters ====== */ -#define ZSTDMT_NBTHREADS_MAX 128 +#define ZSTDMT_NBTHREADS_MAX 256 +#define ZSTDMT_OVERLAPLOG_DEFAULT 6 /* ====== Compiler specifics ====== */ @@ -73,6 +74,7 @@ static unsigned long long GetCurrentClockTimeMicroseconds(void) /* ===== Buffer Pool ===== */ +/* a single Buffer Pool can be invoked from multiple threads in parallel */ typedef struct buffer_s { void* start; @@ -82,6 +84,8 @@ typedef struct buffer_s { static const buffer_t g_nullBuffer = { NULL, 0 }; typedef struct ZSTDMT_bufferPool_s { + pthread_mutex_t poolMutex; + size_t bufferSize; unsigned totalBuffers; unsigned nbBuffers; ZSTD_customMem cMem; @@ -90,10 +94,15 @@ typedef struct ZSTDMT_bufferPool_s { static ZSTDMT_bufferPool* ZSTDMT_createBufferPool(unsigned nbThreads, ZSTD_customMem cMem) { - unsigned const maxNbBuffers = 2*nbThreads + 2; + unsigned const maxNbBuffers = 2*nbThreads + 3; ZSTDMT_bufferPool* const bufPool = (ZSTDMT_bufferPool*)ZSTD_calloc( sizeof(ZSTDMT_bufferPool) + (maxNbBuffers-1) * sizeof(buffer_t), cMem); if (bufPool==NULL) return NULL; + if (pthread_mutex_init(&bufPool->poolMutex, NULL)) { + ZSTD_free(bufPool, cMem); + return NULL; + } + bufPool->bufferSize = 64 KB; bufPool->totalBuffers = maxNbBuffers; bufPool->nbBuffers = 0; bufPool->cMem = cMem; @@ -106,6 +115,7 @@ static void ZSTDMT_freeBufferPool(ZSTDMT_bufferPool* bufPool) if (!bufPool) return; /* compatibility with free on NULL */ for (u=0; u<bufPool->totalBuffers; u++) ZSTD_free(bufPool->bTable[u].start, bufPool->cMem); + pthread_mutex_destroy(&bufPool->poolMutex); ZSTD_free(bufPool, bufPool->cMem); } @@ -116,65 +126,85 @@ static size_t ZSTDMT_sizeof_bufferPool(ZSTDMT_bufferPool* bufPool) + (bufPool->totalBuffers - 1) * sizeof(buffer_t); unsigned u; size_t totalBufferSize = 0; + pthread_mutex_lock(&bufPool->poolMutex); for (u=0; u<bufPool->totalBuffers; u++) totalBufferSize += bufPool->bTable[u].size; + pthread_mutex_unlock(&bufPool->poolMutex); return poolSize + totalBufferSize; } -/** ZSTDMT_getBuffer() : - * assumption : invocation from main thread only ! */ -static buffer_t ZSTDMT_getBuffer(ZSTDMT_bufferPool* pool, size_t bSize) +static void ZSTDMT_setBufferSize(ZSTDMT_bufferPool* bufPool, size_t bSize) { - if (pool->nbBuffers) { /* try to use an existing buffer */ - buffer_t const buf = pool->bTable[--(pool->nbBuffers)]; + bufPool->bufferSize = bSize; +} + +/** ZSTDMT_getBuffer() : + * assumption : bufPool must be valid */ +static buffer_t ZSTDMT_getBuffer(ZSTDMT_bufferPool* bufPool) +{ + size_t const bSize = bufPool->bufferSize; + DEBUGLOG(5, "ZSTDMT_getBuffer"); + pthread_mutex_lock(&bufPool->poolMutex); + if (bufPool->nbBuffers) { /* try to use an existing buffer */ + buffer_t const buf = bufPool->bTable[--(bufPool->nbBuffers)]; size_t const availBufferSize = buf.size; - if ((availBufferSize >= bSize) & (availBufferSize <= 10*bSize)) + if ((availBufferSize >= bSize) & (availBufferSize <= 10*bSize)) { /* large enough, but not too much */ + pthread_mutex_unlock(&bufPool->poolMutex); return buf; + } /* size conditions not respected : scratch this buffer, create new one */ - ZSTD_free(buf.start, pool->cMem); + DEBUGLOG(5, "existing buffer does not meet size conditions => freeing"); + ZSTD_free(buf.start, bufPool->cMem); } + pthread_mutex_unlock(&bufPool->poolMutex); /* create new buffer */ + DEBUGLOG(5, "create a new buffer"); { buffer_t buffer; - void* const start = ZSTD_malloc(bSize, pool->cMem); - if (start==NULL) bSize = 0; + void* const start = ZSTD_malloc(bSize, bufPool->cMem); buffer.start = start; /* note : start can be NULL if malloc fails ! */ - buffer.size = bSize; + buffer.size = (start==NULL) ? 0 : bSize; return buffer; } } /* store buffer for later re-use, up to pool capacity */ -static void ZSTDMT_releaseBuffer(ZSTDMT_bufferPool* pool, buffer_t buf) -{ - if (buf.start == NULL) return; /* release on NULL */ - if (pool->nbBuffers < pool->totalBuffers) { - pool->bTable[pool->nbBuffers++] = buf; /* store for later re-use */ +static void ZSTDMT_releaseBuffer(ZSTDMT_bufferPool* bufPool, buffer_t buf) +{ + if (buf.start == NULL) return; /* compatible with release on NULL */ + DEBUGLOG(5, "ZSTDMT_releaseBuffer"); + pthread_mutex_lock(&bufPool->poolMutex); + if (bufPool->nbBuffers < bufPool->totalBuffers) { + bufPool->bTable[bufPool->nbBuffers++] = buf; /* stored for later use */ + pthread_mutex_unlock(&bufPool->poolMutex); return; } + pthread_mutex_unlock(&bufPool->poolMutex); /* Reached bufferPool capacity (should not happen) */ - ZSTD_free(buf.start, pool->cMem); + DEBUGLOG(5, "buffer pool capacity reached => freeing "); + ZSTD_free(buf.start, bufPool->cMem); } /* ===== CCtx Pool ===== */ +/* a single CCtx Pool can be invoked from multiple threads in parallel */ typedef struct { + pthread_mutex_t poolMutex; unsigned totalCCtx; unsigned availCCtx; ZSTD_customMem cMem; ZSTD_CCtx* cctx[1]; /* variable size */ } ZSTDMT_CCtxPool; -/* assumption : CCtxPool invocation only from main thread */ - /* note : all CCtx borrowed from the pool should be released back to the pool _before_ freeing the pool */ static void ZSTDMT_freeCCtxPool(ZSTDMT_CCtxPool* pool) { unsigned u; for (u=0; u<pool->totalCCtx; u++) ZSTD_freeCCtx(pool->cctx[u]); /* note : compatible with free on NULL */ + pthread_mutex_destroy(&pool->poolMutex); ZSTD_free(pool, pool->cMem); } @@ -186,6 +216,10 @@ static ZSTDMT_CCtxPool* ZSTDMT_createCCtxPool(unsigned nbThreads, ZSTDMT_CCtxPool* const cctxPool = (ZSTDMT_CCtxPool*) ZSTD_calloc( sizeof(ZSTDMT_CCtxPool) + (nbThreads-1)*sizeof(ZSTD_CCtx*), cMem); if (!cctxPool) return NULL; + if (pthread_mutex_init(&cctxPool->poolMutex, NULL)) { + ZSTD_free(cctxPool, cMem); + return NULL; + } cctxPool->cMem = cMem; cctxPool->totalCCtx = nbThreads; cctxPool->availCCtx = 1; /* at least one cctx for single-thread mode */ @@ -198,50 +232,57 @@ static ZSTDMT_CCtxPool* ZSTDMT_createCCtxPool(unsigned nbThreads, /* only works during initialization phase, not during compression */ static size_t ZSTDMT_sizeof_CCtxPool(ZSTDMT_CCtxPool* cctxPool) { - unsigned const nbThreads = cctxPool->totalCCtx; - size_t const poolSize = sizeof(*cctxPool) - + (nbThreads-1)*sizeof(ZSTD_CCtx*); - unsigned u; - size_t totalCCtxSize = 0; - for (u=0; u<nbThreads; u++) - totalCCtxSize += ZSTD_sizeof_CCtx(cctxPool->cctx[u]); - - return poolSize + totalCCtxSize; + pthread_mutex_lock(&cctxPool->poolMutex); + { unsigned const nbThreads = cctxPool->totalCCtx; + size_t const poolSize = sizeof(*cctxPool) + + (nbThreads-1)*sizeof(ZSTD_CCtx*); + unsigned u; + size_t totalCCtxSize = 0; + for (u=0; u<nbThreads; u++) { + totalCCtxSize += ZSTD_sizeof_CCtx(cctxPool->cctx[u]); + } + pthread_mutex_unlock(&cctxPool->poolMutex); + return poolSize + totalCCtxSize; + } } -static ZSTD_CCtx* ZSTDMT_getCCtx(ZSTDMT_CCtxPool* pool) +static ZSTD_CCtx* ZSTDMT_getCCtx(ZSTDMT_CCtxPool* cctxPool) { - if (pool->availCCtx) { - pool->availCCtx--; - return pool->cctx[pool->availCCtx]; - } - return ZSTD_createCCtx(); /* note : can be NULL, when creation fails ! */ + DEBUGLOG(5, "ZSTDMT_getCCtx"); + pthread_mutex_lock(&cctxPool->poolMutex); + if (cctxPool->availCCtx) { + cctxPool->availCCtx--; + { ZSTD_CCtx* const cctx = cctxPool->cctx[cctxPool->availCCtx]; + pthread_mutex_unlock(&cctxPool->poolMutex); + return cctx; + } } + pthread_mutex_unlock(&cctxPool->poolMutex); + DEBUGLOG(5, "create one more CCtx"); + return ZSTD_createCCtx_advanced(cctxPool->cMem); /* note : can be NULL, when creation fails ! */ } static void ZSTDMT_releaseCCtx(ZSTDMT_CCtxPool* pool, ZSTD_CCtx* cctx) { if (cctx==NULL) return; /* compatibility with release on NULL */ + pthread_mutex_lock(&pool->poolMutex); if (pool->availCCtx < pool->totalCCtx) pool->cctx[pool->availCCtx++] = cctx; - else + else { /* pool overflow : should not happen, since totalCCtx==nbThreads */ + DEBUGLOG(5, "CCtx pool overflow : free cctx"); ZSTD_freeCCtx(cctx); + } + pthread_mutex_unlock(&pool->poolMutex); } /* ===== Thread worker ===== */ typedef struct { - buffer_t buffer; - size_t filled; -} inBuff_t; - -typedef struct { - ZSTD_CCtx* cctx; buffer_t src; const void* srcStart; - size_t srcSize; size_t dictSize; + size_t srcSize; buffer_t dstBuff; size_t cSize; size_t dstFlushed; @@ -253,6 +294,8 @@ typedef struct { pthread_cond_t* jobCompleted_cond; ZSTD_parameters params; const ZSTD_CDict* cdict; + ZSTDMT_CCtxPool* cctxPool; + ZSTDMT_bufferPool* bufPool; unsigned long long fullFrameSize; } ZSTDMT_jobDescription; @@ -260,37 +303,56 @@ typedef struct { void ZSTDMT_compressChunk(void* jobDescription) { ZSTDMT_jobDescription* const job = (ZSTDMT_jobDescription*)jobDescription; + ZSTD_CCtx* cctx = ZSTDMT_getCCtx(job->cctxPool); const void* const src = (const char*)job->srcStart + job->dictSize; - buffer_t const dstBuff = job->dstBuff; + buffer_t dstBuff = job->dstBuff; DEBUGLOG(5, "job (first:%u) (last:%u) : dictSize %u, srcSize %u", job->firstChunk, job->lastChunk, (U32)job->dictSize, (U32)job->srcSize); + + if (cctx==NULL) { + job->cSize = ERROR(memory_allocation); + goto _endJob; + } + + if (dstBuff.start == NULL) { + dstBuff = ZSTDMT_getBuffer(job->bufPool); + if (dstBuff.start==NULL) { + job->cSize = ERROR(memory_allocation); + goto _endJob; + } + job->dstBuff = dstBuff; + } + if (job->cdict) { /* should only happen for first segment */ - size_t const initError = ZSTD_compressBegin_usingCDict_advanced(job->cctx, job->cdict, job->params.fParams, job->fullFrameSize); + size_t const initError = ZSTD_compressBegin_usingCDict_advanced(cctx, job->cdict, job->params.fParams, job->fullFrameSize); DEBUGLOG(5, "using CDict"); if (ZSTD_isError(initError)) { job->cSize = initError; goto _endJob; } } else { /* srcStart points at reloaded section */ if (!job->firstChunk) job->params.fParams.contentSizeFlag = 0; /* ensure no srcSize control */ - { size_t const dictModeError = ZSTD_setCCtxParameter(job->cctx, ZSTD_p_forceRawDict, 1); /* Force loading dictionary in "content-only" mode (no header analysis) */ - size_t const initError = ZSTD_compressBegin_advanced(job->cctx, job->srcStart, job->dictSize, job->params, job->fullFrameSize); + { size_t const dictModeError = ZSTD_setCCtxParameter(cctx, ZSTD_p_forceRawDict, 1); /* Force loading dictionary in "content-only" mode (no header analysis) */ + size_t const initError = ZSTD_compressBegin_advanced(cctx, job->srcStart, job->dictSize, job->params, job->fullFrameSize); if (ZSTD_isError(initError) || ZSTD_isError(dictModeError)) { job->cSize = initError; goto _endJob; } - ZSTD_setCCtxParameter(job->cctx, ZSTD_p_forceWindow, 1); + ZSTD_setCCtxParameter(cctx, ZSTD_p_forceWindow, 1); } } if (!job->firstChunk) { /* flush and overwrite frame header when it's not first segment */ - size_t const hSize = ZSTD_compressContinue(job->cctx, dstBuff.start, dstBuff.size, src, 0); + size_t const hSize = ZSTD_compressContinue(cctx, dstBuff.start, dstBuff.size, src, 0); if (ZSTD_isError(hSize)) { job->cSize = hSize; goto _endJob; } - ZSTD_invalidateRepCodes(job->cctx); + ZSTD_invalidateRepCodes(cctx); } DEBUGLOG(5, "Compressing : "); DEBUG_PRINTHEX(4, job->srcStart, 12); job->cSize = (job->lastChunk) ? - ZSTD_compressEnd (job->cctx, dstBuff.start, dstBuff.size, src, job->srcSize) : - ZSTD_compressContinue(job->cctx, dstBuff.start, dstBuff.size, src, job->srcSize); + ZSTD_compressEnd (cctx, dstBuff.start, dstBuff.size, src, job->srcSize) : + ZSTD_compressContinue(cctx, dstBuff.start, dstBuff.size, src, job->srcSize); DEBUGLOG(5, "compressed %u bytes into %u bytes (first:%u) (last:%u)", (unsigned)job->srcSize, (unsigned)job->cSize, job->firstChunk, job->lastChunk); DEBUGLOG(5, "dstBuff.size : %u ; => %s", (U32)dstBuff.size, ZSTD_getErrorName(job->cSize)); _endJob: + ZSTDMT_releaseCCtx(job->cctxPool, cctx); + ZSTDMT_releaseBuffer(job->bufPool, job->src); + job->src = g_nullBuffer; job->srcStart = NULL; PTHREAD_MUTEX_LOCK(job->jobCompleted_mutex); job->jobCompleted = 1; job->jobScanned = 0; @@ -303,15 +365,19 @@ _endJob: /* ===== Multi-threaded compression ===== */ /* ------------------------------------------ */ +typedef struct { + buffer_t buffer; + size_t filled; +} inBuff_t; + struct ZSTDMT_CCtx_s { POOL_ctx* factory; ZSTDMT_jobDescription* jobs; - ZSTDMT_bufferPool* buffPool; + ZSTDMT_bufferPool* bufPool; ZSTDMT_CCtxPool* cctxPool; pthread_mutex_t jobCompleted_mutex; pthread_cond_t jobCompleted_cond; size_t targetSectionSize; - size_t marginSize; size_t inBuffSize; size_t dictSize; size_t targetDictSize; @@ -324,7 +390,7 @@ struct ZSTDMT_CCtx_s { unsigned nextJobID; unsigned frameEnded; unsigned allJobsCompleted; - unsigned overlapRLog; + unsigned overlapLog; unsigned long long frameContentSize; size_t sectionSize; ZSTD_customMem cMem; @@ -347,7 +413,8 @@ ZSTDMT_CCtx* ZSTDMT_createCCtx_advanced(unsigned nbThreads, ZSTD_customMem cMem) U32 nbJobs = nbThreads + 2; DEBUGLOG(3, "ZSTDMT_createCCtx_advanced"); - if ((nbThreads < 1) | (nbThreads > ZSTDMT_NBTHREADS_MAX)) return NULL; + if (nbThreads < 1) return NULL; + nbThreads = MIN(nbThreads , ZSTDMT_NBTHREADS_MAX); if ((cMem.customAlloc!=NULL) ^ (cMem.customFree!=NULL)) /* invalid custom allocator */ return NULL; @@ -358,18 +425,24 @@ ZSTDMT_CCtx* ZSTDMT_createCCtx_advanced(unsigned nbThreads, ZSTD_customMem cMem) mtctx->nbThreads = nbThreads; mtctx->allJobsCompleted = 1; mtctx->sectionSize = 0; - mtctx->overlapRLog = 3; - mtctx->factory = POOL_create(nbThreads, 1); + mtctx->overlapLog = ZSTDMT_OVERLAPLOG_DEFAULT; + mtctx->factory = POOL_create(nbThreads, 0); mtctx->jobs = ZSTDMT_allocJobsTable(&nbJobs, cMem); mtctx->jobIDMask = nbJobs - 1; - mtctx->buffPool = ZSTDMT_createBufferPool(nbThreads, cMem); + mtctx->bufPool = ZSTDMT_createBufferPool(nbThreads, cMem); mtctx->cctxPool = ZSTDMT_createCCtxPool(nbThreads, cMem); - if (!mtctx->factory | !mtctx->jobs | !mtctx->buffPool | !mtctx->cctxPool) { + if (!mtctx->factory | !mtctx->jobs | !mtctx->bufPool | !mtctx->cctxPool) { + ZSTDMT_freeCCtx(mtctx); + return NULL; + } + if (pthread_mutex_init(&mtctx->jobCompleted_mutex, NULL)) { + ZSTDMT_freeCCtx(mtctx); + return NULL; + } + if (pthread_cond_init(&mtctx->jobCompleted_cond, NULL)) { ZSTDMT_freeCCtx(mtctx); return NULL; } - pthread_mutex_init(&mtctx->jobCompleted_mutex, NULL); /* Todo : check init function return */ - pthread_cond_init(&mtctx->jobCompleted_cond, NULL); DEBUGLOG(3, "mt_cctx created, for %u threads", nbThreads); return mtctx; } @@ -386,15 +459,13 @@ static void ZSTDMT_releaseAllJobResources(ZSTDMT_CCtx* mtctx) unsigned jobID; DEBUGLOG(3, "ZSTDMT_releaseAllJobResources"); for (jobID=0; jobID <= mtctx->jobIDMask; jobID++) { - ZSTDMT_releaseBuffer(mtctx->buffPool, mtctx->jobs[jobID].dstBuff); + ZSTDMT_releaseBuffer(mtctx->bufPool, mtctx->jobs[jobID].dstBuff); mtctx->jobs[jobID].dstBuff = g_nullBuffer; - ZSTDMT_releaseBuffer(mtctx->buffPool, mtctx->jobs[jobID].src); + ZSTDMT_releaseBuffer(mtctx->bufPool, mtctx->jobs[jobID].src); mtctx->jobs[jobID].src = g_nullBuffer; - ZSTDMT_releaseCCtx(mtctx->cctxPool, mtctx->jobs[jobID].cctx); - mtctx->jobs[jobID].cctx = NULL; } memset(mtctx->jobs, 0, (mtctx->jobIDMask+1)*sizeof(ZSTDMT_jobDescription)); - ZSTDMT_releaseBuffer(mtctx->buffPool, mtctx->inBuff.buffer); + ZSTDMT_releaseBuffer(mtctx->bufPool, mtctx->inBuff.buffer); mtctx->inBuff.buffer = g_nullBuffer; mtctx->allJobsCompleted = 1; } @@ -404,7 +475,7 @@ size_t ZSTDMT_freeCCtx(ZSTDMT_CCtx* mtctx) if (mtctx==NULL) return 0; /* compatible with free on NULL */ POOL_free(mtctx->factory); if (!mtctx->allJobsCompleted) ZSTDMT_releaseAllJobResources(mtctx); /* stop workers first */ - ZSTDMT_freeBufferPool(mtctx->buffPool); /* release job resources into pools first */ + ZSTDMT_freeBufferPool(mtctx->bufPool); /* release job resources into pools first */ ZSTD_free(mtctx->jobs, mtctx->cMem); ZSTDMT_freeCCtxPool(mtctx->cctxPool); ZSTD_freeCDict(mtctx->cdictLocal); @@ -418,11 +489,11 @@ size_t ZSTDMT_sizeof_CCtx(ZSTDMT_CCtx* mtctx) { if (mtctx == NULL) return 0; /* supports sizeof NULL */ return sizeof(*mtctx) - + POOL_sizeof(mtctx->factory) - + ZSTDMT_sizeof_bufferPool(mtctx->buffPool) - + (mtctx->jobIDMask+1) * sizeof(ZSTDMT_jobDescription) - + ZSTDMT_sizeof_CCtxPool(mtctx->cctxPool) - + ZSTD_sizeof_CDict(mtctx->cdictLocal); + + POOL_sizeof(mtctx->factory) + + ZSTDMT_sizeof_bufferPool(mtctx->bufPool) + + (mtctx->jobIDMask+1) * sizeof(ZSTDMT_jobDescription) + + ZSTDMT_sizeof_CCtxPool(mtctx->cctxPool) + + ZSTD_sizeof_CDict(mtctx->cdictLocal); } size_t ZSTDMT_setMTCtxParameter(ZSTDMT_CCtx* mtctx, ZSDTMT_parameter parameter, unsigned value) @@ -434,10 +505,10 @@ size_t ZSTDMT_setMTCtxParameter(ZSTDMT_CCtx* mtctx, ZSDTMT_parameter parameter, return 0; case ZSTDMT_p_overlapSectionLog : DEBUGLOG(5, "ZSTDMT_p_overlapSectionLog : %u", value); - mtctx->overlapRLog = (value >= 9) ? 0 : 9 - value; + mtctx->overlapLog = (value >= 9) ? 9 : value; return 0; default : - return ERROR(compressionParameter_unsupported); + return ERROR(parameter_unsupported); } } @@ -459,12 +530,13 @@ static unsigned computeNbChunks(size_t srcSize, unsigned windowLog, unsigned nbT size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, - void* dst, size_t dstCapacity, - const void* src, size_t srcSize, - const ZSTD_CDict* cdict, - ZSTD_parameters const params, - unsigned overlapRLog) + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const ZSTD_CDict* cdict, + ZSTD_parameters const params, + unsigned overlapLog) { + unsigned const overlapRLog = (overlapLog>9) ? 0 : 9-overlapLog; size_t const overlapSize = (overlapRLog>=9) ? 0 : (size_t)1 << (params.cParams.windowLog - overlapRLog); unsigned nbChunks = computeNbChunks(srcSize, params.cParams.windowLog, mtctx->nbThreads); size_t const proposedChunkSize = (srcSize + (nbChunks-1)) / nbChunks; @@ -473,6 +545,7 @@ size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, size_t remainingSrcSize = srcSize; unsigned const compressWithinDst = (dstCapacity >= ZSTD_compressBound(srcSize)) ? nbChunks : (unsigned)(dstCapacity / ZSTD_compressBound(avgChunkSize)); /* presumes avgChunkSize >= 256 KB, which should be the case */ size_t frameStartPos = 0, dstBufferPos = 0; + XXH64_state_t xxh64; DEBUGLOG(4, "nbChunks : %2u (chunkSize : %u bytes) ", nbChunks, (U32)avgChunkSize); if (nbChunks==1) { /* fallback to single-thread mode */ @@ -480,7 +553,9 @@ size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, if (cdict) return ZSTD_compress_usingCDict_advanced(cctx, dst, dstCapacity, src, srcSize, cdict, params.fParams); return ZSTD_compress_advanced(cctx, dst, dstCapacity, src, srcSize, NULL, 0, params); } - assert(avgChunkSize >= 256 KB); /* condition for ZSTD_compressBound(A) + ZSTD_compressBound(B) <= ZSTD_compressBound(A+B), which is useful to avoid allocating extra buffers */ + assert(avgChunkSize >= 256 KB); /* condition for ZSTD_compressBound(A) + ZSTD_compressBound(B) <= ZSTD_compressBound(A+B), which is required for compressWithinDst */ + ZSTDMT_setBufferSize(mtctx->bufPool, ZSTD_compressBound(avgChunkSize) ); + XXH64_reset(&xxh64, 0); if (nbChunks > mtctx->jobIDMask+1) { /* enlarge job table */ U32 nbJobs = nbChunks; @@ -496,17 +571,10 @@ size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, size_t const chunkSize = MIN(remainingSrcSize, avgChunkSize); size_t const dstBufferCapacity = ZSTD_compressBound(chunkSize); buffer_t const dstAsBuffer = { (char*)dst + dstBufferPos, dstBufferCapacity }; - buffer_t const dstBuffer = u < compressWithinDst ? dstAsBuffer : ZSTDMT_getBuffer(mtctx->buffPool, dstBufferCapacity); - ZSTD_CCtx* const cctx = ZSTDMT_getCCtx(mtctx->cctxPool); + buffer_t const dstBuffer = u < compressWithinDst ? dstAsBuffer : g_nullBuffer; size_t dictSize = u ? overlapSize : 0; - if ((cctx==NULL) || (dstBuffer.start==NULL)) { - mtctx->jobs[u].cSize = ERROR(memory_allocation); /* job result */ - mtctx->jobs[u].jobCompleted = 1; - nbChunks = u+1; /* only wait and free u jobs, instead of initially expected nbChunks ones */ - break; /* let's wait for previous jobs to complete, but don't start new ones */ - } - + mtctx->jobs[u].src = g_nullBuffer; mtctx->jobs[u].srcStart = srcStart + frameStartPos - dictSize; mtctx->jobs[u].dictSize = dictSize; mtctx->jobs[u].srcSize = chunkSize; @@ -516,13 +584,18 @@ size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, /* do not calculate checksum within sections, but write it in header for first section */ if (u!=0) mtctx->jobs[u].params.fParams.checksumFlag = 0; mtctx->jobs[u].dstBuff = dstBuffer; - mtctx->jobs[u].cctx = cctx; + mtctx->jobs[u].cctxPool = mtctx->cctxPool; + mtctx->jobs[u].bufPool = mtctx->bufPool; mtctx->jobs[u].firstChunk = (u==0); mtctx->jobs[u].lastChunk = (u==nbChunks-1); mtctx->jobs[u].jobCompleted = 0; mtctx->jobs[u].jobCompleted_mutex = &mtctx->jobCompleted_mutex; mtctx->jobs[u].jobCompleted_cond = &mtctx->jobCompleted_cond; + if (params.fParams.checksumFlag) { + XXH64_update(&xxh64, srcStart + frameStartPos, chunkSize); + } + DEBUGLOG(5, "posting job %u (%u bytes)", u, (U32)chunkSize); DEBUG_PRINTHEX(6, mtctx->jobs[u].srcStart, 12); POOL_add(mtctx->factory, ZSTDMT_compressChunk, &mtctx->jobs[u]); @@ -533,8 +606,8 @@ size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, } } /* collect result */ - { unsigned chunkID; - size_t error = 0, dstPos = 0; + { size_t error = 0, dstPos = 0; + unsigned chunkID; for (chunkID=0; chunkID<nbChunks; chunkID++) { DEBUGLOG(5, "waiting for chunk %u ", chunkID); PTHREAD_MUTEX_LOCK(&mtctx->jobCompleted_mutex); @@ -545,8 +618,6 @@ size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, pthread_mutex_unlock(&mtctx->jobCompleted_mutex); DEBUGLOG(5, "ready to write chunk %u ", chunkID); - ZSTDMT_releaseCCtx(mtctx->cctxPool, mtctx->jobs[chunkID].cctx); - mtctx->jobs[chunkID].cctx = NULL; mtctx->jobs[chunkID].srcStart = NULL; { size_t const cSize = mtctx->jobs[chunkID].cSize; if (ZSTD_isError(cSize)) error = cSize; @@ -556,13 +627,25 @@ size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, memmove((char*)dst + dstPos, mtctx->jobs[chunkID].dstBuff.start, cSize); /* may overlap when chunk compressed within dst */ if (chunkID >= compressWithinDst) { /* chunk compressed into its own buffer, which must be released */ DEBUGLOG(5, "releasing buffer %u>=%u", chunkID, compressWithinDst); - ZSTDMT_releaseBuffer(mtctx->buffPool, mtctx->jobs[chunkID].dstBuff); + ZSTDMT_releaseBuffer(mtctx->bufPool, mtctx->jobs[chunkID].dstBuff); } mtctx->jobs[chunkID].dstBuff = g_nullBuffer; } dstPos += cSize ; } - } + } /* for (chunkID=0; chunkID<nbChunks; chunkID++) */ + + DEBUGLOG(4, "checksumFlag : %u ", params.fParams.checksumFlag); + if (params.fParams.checksumFlag) { + U32 const checksum = (U32)XXH64_digest(&xxh64); + if (dstPos + 4 > dstCapacity) { + error = ERROR(dstSize_tooSmall); + } else { + DEBUGLOG(4, "writing checksum : %08X \n", checksum); + MEM_writeLE32((char*)dst + dstPos, checksum); + dstPos += 4; + } } + if (!error) DEBUGLOG(4, "compressed size : %u ", (U32)dstPos); return error ? error : dstPos; } @@ -574,10 +657,10 @@ size_t ZSTDMT_compressCCtx(ZSTDMT_CCtx* mtctx, const void* src, size_t srcSize, int compressionLevel) { - U32 const overlapRLog = (compressionLevel >= ZSTD_maxCLevel()) ? 0 : 3; + U32 const overlapLog = (compressionLevel >= ZSTD_maxCLevel()) ? 9 : ZSTDMT_OVERLAPLOG_DEFAULT; ZSTD_parameters params = ZSTD_getParams(compressionLevel, srcSize, 0); params.fParams.contentSizeFlag = 1; - return ZSTDMT_compress_advanced(mtctx, dst, dstCapacity, src, srcSize, NULL, params, overlapRLog); + return ZSTDMT_compress_advanced(mtctx, dst, dstCapacity, src, srcSize, NULL, params, overlapLog); } @@ -615,8 +698,8 @@ size_t ZSTDMT_initCStream_internal(ZSTDMT_CCtx* zcs, if (zcs->nbThreads==1) { DEBUGLOG(4, "single thread mode"); return ZSTD_initCStream_internal(zcs->cctxPool->cctx[0], - dict, dictSize, cdict, - params, pledgedSrcSize); + dict, dictSize, cdict, + params, pledgedSrcSize); } if (zcs->allJobsCompleted == 0) { /* previous compression not correctly finished */ @@ -642,18 +725,16 @@ size_t ZSTDMT_initCStream_internal(ZSTDMT_CCtx* zcs, zcs->cdict = cdict; } - zcs->targetDictSize = (zcs->overlapRLog>=9) ? 0 : (size_t)1 << (zcs->params.cParams.windowLog - zcs->overlapRLog); - DEBUGLOG(4, "overlapRLog : %u ", zcs->overlapRLog); + zcs->targetDictSize = (zcs->overlapLog==0) ? 0 : (size_t)1 << (zcs->params.cParams.windowLog - (9 - zcs->overlapLog)); + DEBUGLOG(4, "overlapLog : %u ", zcs->overlapLog); DEBUGLOG(4, "overlap Size : %u KB", (U32)(zcs->targetDictSize>>10)); zcs->targetSectionSize = zcs->sectionSize ? zcs->sectionSize : (size_t)1 << (zcs->params.cParams.windowLog + 2); zcs->targetSectionSize = MAX(ZSTDMT_SECTION_SIZE_MIN, zcs->targetSectionSize); zcs->targetSectionSize = MAX(zcs->targetDictSize, zcs->targetSectionSize); DEBUGLOG(4, "Section Size : %u KB", (U32)(zcs->targetSectionSize>>10)); - zcs->marginSize = zcs->targetSectionSize >> 2; - zcs->inBuffSize = zcs->targetDictSize + zcs->targetSectionSize + zcs->marginSize; - zcs->inBuff.buffer = ZSTDMT_getBuffer(zcs->buffPool, zcs->inBuffSize); - if (zcs->inBuff.buffer.start == NULL) return ERROR(memory_allocation); - zcs->inBuff.filled = 0; + zcs->inBuffSize = zcs->targetDictSize + zcs->targetSectionSize; + ZSTDMT_setBufferSize(zcs->bufPool, MAX(zcs->inBuffSize, ZSTD_compressBound(zcs->targetSectionSize)) ); + zcs->inBuff.buffer = g_nullBuffer; zcs->dictSize = 0; zcs->doneJobID = 0; zcs->nextJobID = 0; @@ -664,8 +745,9 @@ size_t ZSTDMT_initCStream_internal(ZSTDMT_CCtx* zcs, } size_t ZSTDMT_initCStream_advanced(ZSTDMT_CCtx* mtctx, - const void* dict, size_t dictSize, - ZSTD_parameters params, unsigned long long pledgedSrcSize) + const void* dict, size_t dictSize, + ZSTD_parameters params, + unsigned long long pledgedSrcSize) { DEBUGLOG(5, "ZSTDMT_initCStream_advanced"); return ZSTDMT_initCStream_internal(mtctx, dict, dictSize, NULL, params, pledgedSrcSize); @@ -701,19 +783,8 @@ size_t ZSTDMT_initCStream(ZSTDMT_CCtx* zcs, int compressionLevel) { static size_t ZSTDMT_createCompressionJob(ZSTDMT_CCtx* zcs, size_t srcSize, unsigned endFrame) { - size_t const dstBufferCapacity = ZSTD_compressBound(srcSize); - buffer_t const dstBuffer = ZSTDMT_getBuffer(zcs->buffPool, dstBufferCapacity); - ZSTD_CCtx* const cctx = ZSTDMT_getCCtx(zcs->cctxPool); unsigned const jobID = zcs->nextJobID & zcs->jobIDMask; - if ((cctx==NULL) || (dstBuffer.start==NULL)) { - zcs->jobs[jobID].jobCompleted = 1; - zcs->nextJobID++; - ZSTDMT_waitForAllJobsCompleted(zcs); - ZSTDMT_releaseAllJobResources(zcs); - return ERROR(memory_allocation); - } - DEBUGLOG(4, "preparing job %u to compress %u bytes with %u preload ", zcs->nextJobID, (U32)srcSize, (U32)zcs->dictSize); zcs->jobs[jobID].src = zcs->inBuff.buffer; @@ -726,8 +797,9 @@ static size_t ZSTDMT_createCompressionJob(ZSTDMT_CCtx* zcs, size_t srcSize, unsi if (zcs->nextJobID) zcs->jobs[jobID].params.fParams.checksumFlag = 0; zcs->jobs[jobID].cdict = zcs->nextJobID==0 ? zcs->cdict : NULL; zcs->jobs[jobID].fullFrameSize = zcs->frameContentSize; - zcs->jobs[jobID].dstBuff = dstBuffer; - zcs->jobs[jobID].cctx = cctx; + zcs->jobs[jobID].dstBuff = g_nullBuffer; + zcs->jobs[jobID].cctxPool = zcs->cctxPool; + zcs->jobs[jobID].bufPool = zcs->bufPool; zcs->jobs[jobID].firstChunk = (zcs->nextJobID==0); zcs->jobs[jobID].lastChunk = endFrame; zcs->jobs[jobID].jobCompleted = 0; @@ -735,11 +807,13 @@ static size_t ZSTDMT_createCompressionJob(ZSTDMT_CCtx* zcs, size_t srcSize, unsi zcs->jobs[jobID].jobCompleted_mutex = &zcs->jobCompleted_mutex; zcs->jobs[jobID].jobCompleted_cond = &zcs->jobCompleted_cond; + if (zcs->params.fParams.checksumFlag) + XXH64_update(&zcs->xxhState, (const char*)zcs->inBuff.buffer.start + zcs->dictSize, srcSize); + /* get a new buffer for next input */ if (!endFrame) { size_t const newDictSize = MIN(srcSize + zcs->dictSize, zcs->targetDictSize); - DEBUGLOG(5, "ZSTDMT_createCompressionJob::endFrame = %u", endFrame); - zcs->inBuff.buffer = ZSTDMT_getBuffer(zcs->buffPool, zcs->inBuffSize); + zcs->inBuff.buffer = ZSTDMT_getBuffer(zcs->bufPool); if (zcs->inBuff.buffer.start == NULL) { /* not enough memory to allocate next input buffer */ zcs->jobs[jobID].jobCompleted = 1; zcs->nextJobID++; @@ -747,26 +821,20 @@ static size_t ZSTDMT_createCompressionJob(ZSTDMT_CCtx* zcs, size_t srcSize, unsi ZSTDMT_releaseAllJobResources(zcs); return ERROR(memory_allocation); } - DEBUGLOG(5, "inBuff currently filled to %u", (U32)zcs->inBuff.filled); zcs->inBuff.filled -= srcSize + zcs->dictSize - newDictSize; - DEBUGLOG(5, "new job : inBuff filled to %u, with %u dict and %u src", - (U32)zcs->inBuff.filled, (U32)newDictSize, - (U32)(zcs->inBuff.filled - newDictSize)); memmove(zcs->inBuff.buffer.start, (const char*)zcs->jobs[jobID].srcStart + zcs->dictSize + srcSize - newDictSize, zcs->inBuff.filled); - DEBUGLOG(5, "new inBuff pre-filled"); zcs->dictSize = newDictSize; } else { /* if (endFrame==1) */ - DEBUGLOG(5, "ZSTDMT_createCompressionJob::endFrame = %u", endFrame); zcs->inBuff.buffer = g_nullBuffer; zcs->inBuff.filled = 0; zcs->dictSize = 0; zcs->frameEnded = 1; - if (zcs->nextJobID == 0) + if (zcs->nextJobID == 0) { /* single chunk exception : checksum is calculated directly within worker thread */ zcs->params.fParams.checksumFlag = 0; - } + } } DEBUGLOG(4, "posting job %u : %u bytes (end:%u) (note : doneJob = %u=>%u)", zcs->nextJobID, @@ -804,11 +872,8 @@ static size_t ZSTDMT_flushNextJob(ZSTDMT_CCtx* zcs, ZSTD_outBuffer* output, unsi ZSTDMT_releaseAllJobResources(zcs); return job.cSize; } - ZSTDMT_releaseCCtx(zcs->cctxPool, job.cctx); - zcs->jobs[wJobID].cctx = NULL; DEBUGLOG(5, "zcs->params.fParams.checksumFlag : %u ", zcs->params.fParams.checksumFlag); if (zcs->params.fParams.checksumFlag) { - XXH64_update(&zcs->xxhState, (const char*)job.srcStart + job.dictSize, job.srcSize); if (zcs->frameEnded && (zcs->doneJobID+1 == zcs->nextJobID)) { /* write checksum at end of last section */ U32 const checksum = (U32)XXH64_digest(&zcs->xxhState); DEBUGLOG(5, "writing checksum : %08X \n", checksum); @@ -816,9 +881,6 @@ static size_t ZSTDMT_flushNextJob(ZSTDMT_CCtx* zcs, ZSTD_outBuffer* output, unsi job.cSize += 4; zcs->jobs[wJobID].cSize += 4; } } - ZSTDMT_releaseBuffer(zcs->buffPool, job.src); - zcs->jobs[wJobID].srcStart = NULL; - zcs->jobs[wJobID].src = g_nullBuffer; zcs->jobs[wJobID].jobScanned = 1; } { size_t const toWrite = MIN(job.cSize - job.dstFlushed, output->size - output->pos); @@ -828,7 +890,7 @@ static size_t ZSTDMT_flushNextJob(ZSTDMT_CCtx* zcs, ZSTD_outBuffer* output, unsi job.dstFlushed += toWrite; } if (job.dstFlushed == job.cSize) { /* output buffer fully flushed => move to next one */ - ZSTDMT_releaseBuffer(zcs->buffPool, job.dstBuff); + ZSTDMT_releaseBuffer(zcs->bufPool, job.dstBuff); zcs->jobs[wJobID].dstBuff = g_nullBuffer; zcs->jobs[wJobID].jobCompleted = 0; zcs->doneJobID++; @@ -852,18 +914,18 @@ size_t ZSTDMT_compressStream_generic(ZSTDMT_CCtx* mtctx, ZSTD_inBuffer* input, ZSTD_EndDirective endOp) { - size_t const newJobThreshold = mtctx->dictSize + mtctx->targetSectionSize + mtctx->marginSize; + size_t const newJobThreshold = mtctx->dictSize + mtctx->targetSectionSize; assert(output->pos <= output->size); assert(input->pos <= input->size); if ((mtctx->frameEnded) && (endOp==ZSTD_e_continue)) { /* current frame being ended. Only flush/end are allowed. Or start new frame with init */ return ERROR(stage_wrong); } - if (mtctx->nbThreads==1) { + if (mtctx->nbThreads==1) { /* delegate to single-thread (synchronous) */ return ZSTD_compressStream_generic(mtctx->cctxPool->cctx[0], output, input, endOp); } - /* single-pass shortcut (note : this is blocking-mode) */ + /* single-pass shortcut (note : this is synchronous-mode) */ if ( (mtctx->nextJobID==0) /* just started */ && (mtctx->inBuff.filled==0) /* nothing buffered */ && (endOp==ZSTD_e_end) /* end order */ @@ -871,24 +933,29 @@ size_t ZSTDMT_compressStream_generic(ZSTDMT_CCtx* mtctx, size_t const cSize = ZSTDMT_compress_advanced(mtctx, (char*)output->dst + output->pos, output->size - output->pos, (const char*)input->src + input->pos, input->size - input->pos, - mtctx->cdict, mtctx->params, mtctx->overlapRLog); + mtctx->cdict, mtctx->params, mtctx->overlapLog); if (ZSTD_isError(cSize)) return cSize; input->pos = input->size; output->pos += cSize; - ZSTDMT_releaseBuffer(mtctx->buffPool, mtctx->inBuff.buffer); /* was allocated in initStream */ + ZSTDMT_releaseBuffer(mtctx->bufPool, mtctx->inBuff.buffer); /* was allocated in initStream */ mtctx->allJobsCompleted = 1; mtctx->frameEnded = 1; return 0; } /* fill input buffer */ - if ((input->src) && (mtctx->inBuff.buffer.start)) { /* support NULL input */ - size_t const toLoad = MIN(input->size - input->pos, mtctx->inBuffSize - mtctx->inBuff.filled); - DEBUGLOG(2, "inBuff:%08X; inBuffSize=%u; ToCopy=%u", (U32)(size_t)mtctx->inBuff.buffer.start, (U32)mtctx->inBuffSize, (U32)toLoad); - memcpy((char*)mtctx->inBuff.buffer.start + mtctx->inBuff.filled, (const char*)input->src + input->pos, toLoad); - input->pos += toLoad; - mtctx->inBuff.filled += toLoad; - } + if (input->size > input->pos) { /* support NULL input */ + if (mtctx->inBuff.buffer.start == NULL) { + mtctx->inBuff.buffer = ZSTDMT_getBuffer(mtctx->bufPool); + if (mtctx->inBuff.buffer.start == NULL) return ERROR(memory_allocation); + mtctx->inBuff.filled = 0; + } + { size_t const toLoad = MIN(input->size - input->pos, mtctx->inBuffSize - mtctx->inBuff.filled); + DEBUGLOG(5, "inBuff:%08X; inBuffSize=%u; ToCopy=%u", (U32)(size_t)mtctx->inBuff.buffer.start, (U32)mtctx->inBuffSize, (U32)toLoad); + memcpy((char*)mtctx->inBuff.buffer.start + mtctx->inBuff.filled, (const char*)input->src + input->pos, toLoad); + input->pos += toLoad; + mtctx->inBuff.filled += toLoad; + } } if ( (mtctx->inBuff.filled >= newJobThreshold) /* filled enough : let's compress */ && (mtctx->nextJobID <= mtctx->doneJobID + mtctx->jobIDMask) ) { /* avoid overwriting job round buffer */ diff --git a/thirdparty/zstd/compress/zstdmt_compress.h b/thirdparty/zstd/compress/zstdmt_compress.h index fad63b6d86..0f0fc2b03f 100644 --- a/thirdparty/zstd/compress/zstdmt_compress.h +++ b/thirdparty/zstd/compress/zstdmt_compress.h @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ #ifndef ZSTDMT_COMPRESS_H @@ -15,10 +15,11 @@ #endif -/* Note : All prototypes defined in this file are labelled experimental. - * No guarantee of API continuity is provided on any of them. - * In fact, the expectation is that these prototypes will be replaced - * by ZSTD_compress_generic() API in the near future */ +/* Note : This is an internal API. + * Some methods are still exposed (ZSTDLIB_API), + * because it used to be the only way to invoke MT compression. + * Now, it's recommended to use ZSTD_compress_generic() instead. + * These methods will stop being exposed in a future version */ /* === Dependencies === */ #include <stddef.h> /* size_t */ @@ -67,7 +68,7 @@ ZSTDLIB_API size_t ZSTDMT_compress_advanced(ZSTDMT_CCtx* mtctx, const void* src, size_t srcSize, const ZSTD_CDict* cdict, ZSTD_parameters const params, - unsigned overlapRLog); + unsigned overlapLog); ZSTDLIB_API size_t ZSTDMT_initCStream_advanced(ZSTDMT_CCtx* mtctx, const void* dict, size_t dictSize, /* dict can be released after init, a local copy is preserved within zcs */ diff --git a/thirdparty/zstd/decompress/huf_decompress.c b/thirdparty/zstd/decompress/huf_decompress.c index 2a1b70ea5e..79ded96bf6 100644 --- a/thirdparty/zstd/decompress/huf_decompress.c +++ b/thirdparty/zstd/decompress/huf_decompress.c @@ -33,37 +33,21 @@ ****************************************************************** */ /* ************************************************************** -* Compiler specifics -****************************************************************/ -#ifdef _MSC_VER /* Visual Studio */ -# define FORCE_INLINE static __forceinline -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ -#else -# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ -# ifdef __GNUC__ -# define FORCE_INLINE static inline __attribute__((always_inline)) -# else -# define FORCE_INLINE static inline -# endif -# else -# define FORCE_INLINE static -# endif /* __STDC_VERSION__ */ -#endif - - -/* ************************************************************** * Dependencies ****************************************************************/ #include <string.h> /* memcpy, memset */ #include "bitstream.h" /* BIT_* */ +#include "compiler.h" #include "fse.h" /* header compression */ #define HUF_STATIC_LINKING_ONLY #include "huf.h" +#include "error_private.h" /* ************************************************************** * Error Management ****************************************************************/ +#define HUF_isError ERR_isError #define HUF_STATIC_ASSERT(c) { enum { HUF_static_assert = 1/(int)(!!(c)) }; } /* use only *after* variable declarations */ @@ -180,7 +164,7 @@ static BYTE HUF_decodeSymbolX2(BIT_DStream_t* Dstream, const HUF_DEltX2* dt, con if (MEM_64bits()) \ HUF_DECODE_SYMBOLX2_0(ptr, DStreamPtr) -FORCE_INLINE size_t HUF_decodeStreamX2(BYTE* p, BIT_DStream_t* const bitDPtr, BYTE* const pEnd, const HUF_DEltX2* const dt, const U32 dtLog) +HINT_INLINE size_t HUF_decodeStreamX2(BYTE* p, BIT_DStream_t* const bitDPtr, BYTE* const pEnd, const HUF_DEltX2* const dt, const U32 dtLog) { BYTE* const pStart = p; @@ -639,7 +623,7 @@ static U32 HUF_decodeLastSymbolX4(void* op, BIT_DStream_t* DStream, const HUF_DE if (MEM_64bits()) \ ptr += HUF_decodeSymbolX4(ptr, DStreamPtr, dt, dtLog) -FORCE_INLINE size_t HUF_decodeStreamX4(BYTE* p, BIT_DStream_t* bitDPtr, BYTE* const pEnd, const HUF_DEltX4* const dt, const U32 dtLog) +HINT_INLINE size_t HUF_decodeStreamX4(BYTE* p, BIT_DStream_t* bitDPtr, BYTE* const pEnd, const HUF_DEltX4* const dt, const U32 dtLog) { BYTE* const pStart = p; @@ -917,11 +901,11 @@ static const algo_time_t algoTime[16 /* Quantization */][3 /* single, double, qu * Tells which decoder is likely to decode faster, * based on a set of pre-determined metrics. * @return : 0==HUF_decompress4X2, 1==HUF_decompress4X4 . -* Assumption : 0 < cSrcSize < dstSize <= 128 KB */ +* Assumption : 0 < cSrcSize, dstSize <= 128 KB */ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize) { /* decoder timing evaluation */ - U32 const Q = (U32)(cSrcSize * 16 / dstSize); /* Q < 16 since dstSize > cSrcSize */ + U32 const Q = cSrcSize >= dstSize ? 15 : (U32)(cSrcSize * 16 / dstSize); /* Q < 16 */ U32 const D256 = (U32)(dstSize >> 8); U32 const DTime0 = algoTime[Q][0].tableTime + (algoTime[Q][0].decode256Time * D256); U32 DTime1 = algoTime[Q][1].tableTime + (algoTime[Q][1].decode256Time * D256); @@ -977,7 +961,7 @@ size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, { /* validation checks */ if (dstSize == 0) return ERROR(dstSize_tooSmall); - if ((cSrcSize >= dstSize) || (cSrcSize <= 1)) return ERROR(corruption_detected); /* invalid */ + if (cSrcSize == 0) return ERROR(corruption_detected); { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); return algoNb ? HUF_decompress4X4_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize): diff --git a/thirdparty/zstd/decompress/zstd_decompress.c b/thirdparty/zstd/decompress/zstd_decompress.c index 465bcc20b5..d2bc545e52 100644 --- a/thirdparty/zstd/decompress/zstd_decompress.c +++ b/thirdparty/zstd/decompress/zstd_decompress.c @@ -1,10 +1,10 @@ -/** +/* * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ @@ -53,15 +53,6 @@ # include "zstd_legacy.h" #endif -#if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_I86)) /* _mm_prefetch() is not defined outside of x86/x64 */ -# include <mmintrin.h> /* https://msdn.microsoft.com/fr-fr/library/84szxsww(v=vs.90).aspx */ -# define ZSTD_PREFETCH(ptr) _mm_prefetch((const char*)ptr, _MM_HINT_T0) -#elif defined(__GNUC__) -# define ZSTD_PREFETCH(ptr) __builtin_prefetch(ptr, 0, 0) -#else -# define ZSTD_PREFETCH(ptr) /* disabled */ -#endif - /*-************************************* * Errors @@ -95,7 +86,7 @@ typedef struct { HUF_DTable hufTable[HUF_DTABLE_SIZE(HufLog)]; /* can accommodate HUF_decompress4X */ U32 workspace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32]; U32 rep[ZSTD_REP_NUM]; -} ZSTD_entropyTables_t; +} ZSTD_entropyDTables_t; struct ZSTD_DCtx_s { @@ -103,7 +94,7 @@ struct ZSTD_DCtx_s const FSE_DTable* MLTptr; const FSE_DTable* OFTptr; const HUF_DTable* HUFptr; - ZSTD_entropyTables_t entropy; + ZSTD_entropyDTables_t entropy; const void* previousDstEnd; /* detect continuity */ const void* base; /* start of current segment */ const void* vBase; /* virtual start of previous segment if it was just before current one */ @@ -304,15 +295,18 @@ size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t src return ZSTD_skippableHeaderSize; /* magic number + frame length */ memset(zfhPtr, 0, sizeof(*zfhPtr)); zfhPtr->frameContentSize = MEM_readLE32((const char *)src + 4); - zfhPtr->windowSize = 0; /* windowSize==0 means a frame is skippable */ + zfhPtr->frameType = ZSTD_skippableFrame; + zfhPtr->windowSize = 0; return 0; } return ERROR(prefix_unknown); } /* ensure there is enough `srcSize` to fully read/decode frame header */ - { size_t const fhsize = ZSTD_frameHeaderSize(src, srcSize); - if (srcSize < fhsize) return fhsize; } + { size_t const fhsize = ZSTD_frameHeaderSize(src, srcSize); + if (srcSize < fhsize) return fhsize; + zfhPtr->headerSize = (U32)fhsize; + } { BYTE const fhdByte = ip[4]; size_t pos = 5; @@ -320,24 +314,23 @@ size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t src U32 const checksumFlag = (fhdByte>>2)&1; U32 const singleSegment = (fhdByte>>5)&1; U32 const fcsID = fhdByte>>6; - U32 const windowSizeMax = 1U << ZSTD_WINDOWLOG_MAX; - U32 windowSize = 0; + U64 windowSize = 0; U32 dictID = 0; - U64 frameContentSize = 0; + U64 frameContentSize = ZSTD_CONTENTSIZE_UNKNOWN; if ((fhdByte & 0x08) != 0) - return ERROR(frameParameter_unsupported); /* reserved bits, must be zero */ + return ERROR(frameParameter_unsupported); /* reserved bits, must be zero */ + if (!singleSegment) { BYTE const wlByte = ip[pos++]; U32 const windowLog = (wlByte >> 3) + ZSTD_WINDOWLOG_ABSOLUTEMIN; if (windowLog > ZSTD_WINDOWLOG_MAX) return ERROR(frameParameter_windowTooLarge); - windowSize = (1U << windowLog); + windowSize = (1ULL << windowLog); windowSize += (windowSize >> 3) * (wlByte&7); } - switch(dictIDSizeCode) { - default: /* impossible */ + default: assert(0); /* impossible */ case 0 : break; case 1 : dictID = ip[pos]; pos++; break; case 2 : dictID = MEM_readLE16(ip+pos); pos+=2; break; @@ -345,14 +338,15 @@ size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t src } switch(fcsID) { - default: /* impossible */ + default: assert(0); /* impossible */ case 0 : if (singleSegment) frameContentSize = ip[pos]; break; case 1 : frameContentSize = MEM_readLE16(ip+pos)+256; break; case 2 : frameContentSize = MEM_readLE32(ip+pos); break; case 3 : frameContentSize = MEM_readLE64(ip+pos); break; } - if (!windowSize) windowSize = (U32)frameContentSize; - if (windowSize > windowSizeMax) return ERROR(frameParameter_windowTooLarge); + if (singleSegment) windowSize = frameContentSize; + + zfhPtr->frameType = ZSTD_frame; zfhPtr->frameContentSize = frameContentSize; zfhPtr->windowSize = windowSize; zfhPtr->dictID = dictID; @@ -362,10 +356,10 @@ size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t src } /** ZSTD_getFrameContentSize() : -* compatible with legacy mode -* @return : decompressed size of the single frame pointed to be `src` if known, otherwise -* - ZSTD_CONTENTSIZE_UNKNOWN if the size cannot be determined -* - ZSTD_CONTENTSIZE_ERROR if an error occurred (e.g. invalid magic number, srcSize too small) */ + * compatible with legacy mode + * @return : decompressed size of the single frame pointed to be `src` if known, otherwise + * - ZSTD_CONTENTSIZE_UNKNOWN if the size cannot be determined + * - ZSTD_CONTENTSIZE_ERROR if an error occurred (e.g. invalid magic number, srcSize too small) */ unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize) { #if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1) @@ -374,17 +368,14 @@ unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize) return ret == 0 ? ZSTD_CONTENTSIZE_UNKNOWN : ret; } #endif - { ZSTD_frameHeader fParams; - if (ZSTD_getFrameHeader(&fParams, src, srcSize) != 0) return ZSTD_CONTENTSIZE_ERROR; - if (fParams.windowSize == 0) { - /* Either skippable or empty frame, size == 0 either way */ + { ZSTD_frameHeader zfh; + if (ZSTD_getFrameHeader(&zfh, src, srcSize) != 0) + return ZSTD_CONTENTSIZE_ERROR; + if (zfh.frameType == ZSTD_skippableFrame) { return 0; - } else if (fParams.frameContentSize != 0) { - return fParams.frameContentSize; } else { - return ZSTD_CONTENTSIZE_UNKNOWN; - } - } + return zfh.frameContentSize; + } } } /** ZSTD_findDecompressedSize() : @@ -442,7 +433,8 @@ unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize) * compatible with legacy mode * @return : decompressed size if known, 0 otherwise note : 0 can mean any of the following : - - decompressed size is not present within frame header + - frame content is empty + - decompressed size field is not present in frame header - frame header unknown / not supported - frame header not complete (`srcSize` too small) */ unsigned long long ZSTD_getDecompressedSize(const void* src, size_t srcSize) @@ -460,7 +452,8 @@ static size_t ZSTD_decodeFrameHeader(ZSTD_DCtx* dctx, const void* src, size_t he size_t const result = ZSTD_getFrameHeader(&(dctx->fParams), src, headerSize); if (ZSTD_isError(result)) return result; /* invalid header */ if (result>0) return ERROR(srcSize_wrong); /* headerSize too small */ - if (dctx->fParams.dictID && (dctx->dictID != dctx->fParams.dictID)) return ERROR(dictionary_wrong); + if (dctx->fParams.dictID && (dctx->dictID != dctx->fParams.dictID)) + return ERROR(dictionary_wrong); if (dctx->fParams.checksumFlag) XXH64_reset(&dctx->xxhState, 0); return 0; } @@ -951,7 +944,7 @@ static seq_t ZSTD_decodeSequence(seqState_t* seqState) } -FORCE_INLINE +HINT_INLINE size_t ZSTD_execSequence(BYTE* op, BYTE* const oend, seq_t sequence, const BYTE** litPtr, const BYTE* const litLimit, @@ -1095,7 +1088,7 @@ static size_t ZSTD_decompressSequences( } -FORCE_INLINE seq_t ZSTD_decodeSequenceLong_generic(seqState_t* seqState, int const longOffsets) +FORCE_INLINE_TEMPLATE seq_t ZSTD_decodeSequenceLong_generic(seqState_t* seqState, int const longOffsets) { seq_t seq; @@ -1195,7 +1188,7 @@ static seq_t ZSTD_decodeSequenceLong(seqState_t* seqState, unsigned const window } } -FORCE_INLINE +HINT_INLINE size_t ZSTD_execSequenceLong(BYTE* op, BYTE* const oend, seq_t sequence, const BYTE** litPtr, const BYTE* const litLimit, @@ -1331,7 +1324,7 @@ static size_t ZSTD_decompressSequencesLong( seq_t const sequence = ZSTD_decodeSequenceLong(&seqState, windowSize32); size_t const oneSeqSize = ZSTD_execSequenceLong(op, oend, sequences[(seqNb-ADVANCED_SEQS) & STOSEQ_MASK], &litPtr, litEnd, base, vBase, dictEnd); if (ZSTD_isError(oneSeqSize)) return oneSeqSize; - ZSTD_PREFETCH(sequence.match); + PREFETCH(sequence.match); sequences[seqNb&STOSEQ_MASK] = sequence; op += oneSeqSize; } @@ -1433,28 +1426,26 @@ size_t ZSTD_generateNxBytes(void* dst, size_t dstCapacity, BYTE byte, size_t len size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize) { #if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1) - if (ZSTD_isLegacy(src, srcSize)) return ZSTD_findFrameCompressedSizeLegacy(src, srcSize); + if (ZSTD_isLegacy(src, srcSize)) + return ZSTD_findFrameCompressedSizeLegacy(src, srcSize); #endif - if (srcSize >= ZSTD_skippableHeaderSize && - (MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) { + if ( (srcSize >= ZSTD_skippableHeaderSize) + && (MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START ) { return ZSTD_skippableHeaderSize + MEM_readLE32((const BYTE*)src + 4); } else { const BYTE* ip = (const BYTE*)src; const BYTE* const ipstart = ip; size_t remainingSize = srcSize; - ZSTD_frameHeader fParams; + ZSTD_frameHeader zfh; - size_t const headerSize = ZSTD_frameHeaderSize(ip, remainingSize); - if (ZSTD_isError(headerSize)) return headerSize; - - /* Frame Header */ - { size_t const ret = ZSTD_getFrameHeader(&fParams, ip, remainingSize); + /* Extract Frame Header */ + { size_t const ret = ZSTD_getFrameHeader(&zfh, src, srcSize); if (ZSTD_isError(ret)) return ret; if (ret > 0) return ERROR(srcSize_wrong); } - ip += headerSize; - remainingSize -= headerSize; + ip += zfh.headerSize; + remainingSize -= zfh.headerSize; /* Loop on each block */ while (1) { @@ -1462,7 +1453,8 @@ size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize) size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSize, &blockProperties); if (ZSTD_isError(cBlockSize)) return cBlockSize; - if (ZSTD_blockHeaderSize + cBlockSize > remainingSize) return ERROR(srcSize_wrong); + if (ZSTD_blockHeaderSize + cBlockSize > remainingSize) + return ERROR(srcSize_wrong); ip += ZSTD_blockHeaderSize + cBlockSize; remainingSize -= ZSTD_blockHeaderSize + cBlockSize; @@ -1470,7 +1462,7 @@ size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize) if (blockProperties.lastBlock) break; } - if (fParams.checksumFlag) { /* Frame content checksum */ + if (zfh.checksumFlag) { /* Final frame content checksum */ if (remainingSize < 4) return ERROR(srcSize_wrong); ip += 4; remainingSize -= 4; @@ -1483,8 +1475,8 @@ size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize) /*! ZSTD_decompressFrame() : * @dctx must be properly initialized */ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx, - void* dst, size_t dstCapacity, - const void** srcPtr, size_t *srcSizePtr) + void* dst, size_t dstCapacity, + const void** srcPtr, size_t *srcSizePtr) { const BYTE* ip = (const BYTE*)(*srcPtr); BYTE* const ostart = (BYTE* const)dst; @@ -1493,13 +1485,15 @@ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx, size_t remainingSize = *srcSizePtr; /* check */ - if (remainingSize < ZSTD_frameHeaderSize_min+ZSTD_blockHeaderSize) return ERROR(srcSize_wrong); + if (remainingSize < ZSTD_frameHeaderSize_min+ZSTD_blockHeaderSize) + return ERROR(srcSize_wrong); /* Frame Header */ { size_t const frameHeaderSize = ZSTD_frameHeaderSize(ip, ZSTD_frameHeaderSize_prefix); if (ZSTD_isError(frameHeaderSize)) return frameHeaderSize; - if (remainingSize < frameHeaderSize+ZSTD_blockHeaderSize) return ERROR(srcSize_wrong); - CHECK_F(ZSTD_decodeFrameHeader(dctx, ip, frameHeaderSize)); + if (remainingSize < frameHeaderSize+ZSTD_blockHeaderSize) + return ERROR(srcSize_wrong); + CHECK_F( ZSTD_decodeFrameHeader(dctx, ip, frameHeaderSize) ); ip += frameHeaderSize; remainingSize -= frameHeaderSize; } @@ -1531,14 +1525,15 @@ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx, } if (ZSTD_isError(decodedSize)) return decodedSize; - if (dctx->fParams.checksumFlag) XXH64_update(&dctx->xxhState, op, decodedSize); + if (dctx->fParams.checksumFlag) + XXH64_update(&dctx->xxhState, op, decodedSize); op += decodedSize; ip += cBlockSize; remainingSize -= cBlockSize; if (blockProperties.lastBlock) break; } - if (dctx->fParams.checksumFlag) { /* Frame content checksum verification */ + if (dctx->fParams.checksumFlag) { /* Frame content checksum verification */ U32 const checkCalc = (U32)XXH64_digest(&dctx->xxhState); U32 checkRead; if (remainingSize<4) return ERROR(checksum_wrong); @@ -1560,17 +1555,13 @@ static size_t ZSTD_DDictDictSize(const ZSTD_DDict* ddict); static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize, - const void *dict, size_t dictSize, + const void* dict, size_t dictSize, const ZSTD_DDict* ddict) { void* const dststart = dst; + assert(dict==NULL || ddict==NULL); /* either dict or ddict set, not both */ if (ddict) { - if (dict) { - /* programmer error, these two cases should be mutually exclusive */ - return ERROR(GENERIC); - } - dict = ZSTD_DDictDictContent(ddict); dictSize = ZSTD_DDictDictSize(ddict); } @@ -1583,7 +1574,7 @@ static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx, size_t decodedSize; size_t const frameSize = ZSTD_findFrameCompressedSizeLegacy(src, srcSize); if (ZSTD_isError(frameSize)) return frameSize; - /* legacy support is incompatible with static dctx */ + /* legacy support is not compatible with static dctx */ if (dctx->staticSize) return ERROR(memory_allocation); decodedSize = ZSTD_decompressLegacy(dst, dstCapacity, src, frameSize, dict, dictSize); @@ -1606,16 +1597,13 @@ static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx, return ERROR(srcSize_wrong); skippableSize = MEM_readLE32((const BYTE *)src + 4) + ZSTD_skippableHeaderSize; - if (srcSize < skippableSize) { - return ERROR(srcSize_wrong); - } + if (srcSize < skippableSize) return ERROR(srcSize_wrong); src = (const BYTE *)src + skippableSize; srcSize -= skippableSize; continue; - } else { - return ERROR(prefix_unknown); } + return ERROR(prefix_unknown); } if (ddict) { @@ -1631,12 +1619,11 @@ static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx, { const size_t res = ZSTD_decompressFrame(dctx, dst, dstCapacity, &src, &srcSize); if (ZSTD_isError(res)) return res; - /* don't need to bounds check this, ZSTD_decompressFrame will have - * already */ + /* no need to bound check, ZSTD_decompressFrame already has */ dst = (BYTE*)dst + res; dstCapacity -= res; } - } + } /* while (srcSize >= ZSTD_frameHeaderSize_prefix) */ if (srcSize) return ERROR(srcSize_wrong); /* input not entirely consumed */ @@ -1735,7 +1722,7 @@ size_t ZSTD_decompressContinue(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, c return 0; } dctx->expected = 0; /* not necessary to copy more */ - + /* fall-through */ case ZSTDds_decodeFrameHeader: assert(src != NULL); memcpy(dctx->headerBuffer + ZSTD_frameHeaderSize_prefix, src, dctx->expected); @@ -1846,7 +1833,7 @@ static size_t ZSTD_refDictContent(ZSTD_DCtx* dctx, const void* dict, size_t dict /* ZSTD_loadEntropy() : * dict : must point at beginning of a valid zstd dictionary * @return : size of entropy tables read */ -static size_t ZSTD_loadEntropy(ZSTD_entropyTables_t* entropy, const void* const dict, size_t const dictSize) +static size_t ZSTD_loadEntropy(ZSTD_entropyDTables_t* entropy, const void* const dict, size_t const dictSize) { const BYTE* dictPtr = (const BYTE*)dict; const BYTE* const dictEnd = dictPtr + dictSize; @@ -1924,8 +1911,9 @@ static size_t ZSTD_decompress_insertDictionary(ZSTD_DCtx* dctx, const void* dict size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize) { - CHECK_F(ZSTD_decompressBegin(dctx)); - if (dict && dictSize) CHECK_E(ZSTD_decompress_insertDictionary(dctx, dict, dictSize), dictionary_corrupted); + CHECK_F( ZSTD_decompressBegin(dctx) ); + if (dict && dictSize) + CHECK_E(ZSTD_decompress_insertDictionary(dctx, dict, dictSize), dictionary_corrupted); return 0; } @@ -1936,7 +1924,7 @@ struct ZSTD_DDict_s { void* dictBuffer; const void* dictContent; size_t dictSize; - ZSTD_entropyTables_t entropy; + ZSTD_entropyDTables_t entropy; U32 dictID; U32 entropyPresent; ZSTD_customMem cMem; @@ -1954,7 +1942,7 @@ static size_t ZSTD_DDictDictSize(const ZSTD_DDict* ddict) size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dstDCtx, const ZSTD_DDict* ddict) { - CHECK_F(ZSTD_decompressBegin(dstDCtx)); + CHECK_F( ZSTD_decompressBegin(dstDCtx) ); if (ddict) { /* support begin on NULL */ dstDCtx->dictID = ddict->dictID; dstDCtx->base = ddict->dictContent; @@ -2135,7 +2123,7 @@ unsigned ZSTD_getDictID_fromDDict(const ZSTD_DDict* ddict) * ZSTD_getFrameHeader(), which will provide a more precise error code. */ unsigned ZSTD_getDictID_fromFrame(const void* src, size_t srcSize) { - ZSTD_frameHeader zfp = { 0 , 0 , 0 , 0 }; + ZSTD_frameHeader zfp = { 0, 0, ZSTD_frame, 0, 0, 0 }; size_t const hError = ZSTD_getFrameHeader(&zfp, src, srcSize); if (ZSTD_isError(hError)) return 0; return zfp.dictID; @@ -2230,7 +2218,7 @@ size_t ZSTD_setDStreamParameter(ZSTD_DStream* zds, { switch(paramType) { - default : return ERROR(parameter_unknown); + default : return ERROR(parameter_unsupported); case DStream_p_maxWindowSize : zds->maxWindowSize = paramValue ? paramValue : (U32)(-1); break; } return 0; @@ -2247,16 +2235,19 @@ size_t ZSTD_estimateDStreamSize(size_t windowSize) size_t const blockSize = MIN(windowSize, ZSTD_BLOCKSIZE_MAX); size_t const inBuffSize = blockSize; /* no block can be larger */ size_t const outBuffSize = windowSize + blockSize + (WILDCOPY_OVERLENGTH * 2); - return sizeof(ZSTD_DStream) + ZSTD_estimateDCtxSize() + inBuffSize + outBuffSize; + return ZSTD_estimateDCtxSize() + inBuffSize + outBuffSize; } ZSTDLIB_API size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize) { - ZSTD_frameHeader fh; - size_t const err = ZSTD_getFrameHeader(&fh, src, srcSize); + U32 const windowSizeMax = 1U << ZSTD_WINDOWLOG_MAX; + ZSTD_frameHeader zfh; + size_t const err = ZSTD_getFrameHeader(&zfh, src, srcSize); if (ZSTD_isError(err)) return err; if (err>0) return ERROR(srcSize_wrong); - return ZSTD_estimateDStreamSize(fh.windowSize); + if (zfh.windowSize > windowSizeMax) + return ERROR(frameParameter_windowTooLarge); + return ZSTD_estimateDStreamSize((size_t)zfh.windowSize); } @@ -2307,16 +2298,14 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB size_t const dictSize = zds->ddict ? zds->ddict->dictSize : 0; /* legacy support is incompatible with static dctx */ if (zds->staticSize) return ERROR(memory_allocation); - CHECK_F(ZSTD_initLegacyStream(&zds->legacyContext, zds->previousLegacyVersion, legacyVersion, - dict, dictSize)); + CHECK_F(ZSTD_initLegacyStream(&zds->legacyContext, + zds->previousLegacyVersion, legacyVersion, + dict, dictSize)); zds->legacyVersion = zds->previousLegacyVersion = legacyVersion; - return ZSTD_decompressLegacyStream(zds->legacyContext, zds->legacyVersion, output, input); - } else { - return hSize; /* error */ + return ZSTD_decompressLegacyStream(zds->legacyContext, legacyVersion, output, input); } -#else - return hSize; #endif + return hSize; /* error */ } if (hSize != 0) { /* need more input */ size_t const toLoad = hSize - zds->lhSize; /* if hSize!=0, hSize > zds->lhSize */ @@ -2367,8 +2356,8 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB if (zds->fParams.windowSize > zds->maxWindowSize) return ERROR(frameParameter_windowTooLarge); /* Adapt buffer sizes to frame header instructions */ - { size_t const blockSize = MIN(zds->fParams.windowSize, ZSTD_BLOCKSIZE_MAX); - size_t const neededOutSize = zds->fParams.windowSize + blockSize + WILDCOPY_OVERLENGTH * 2; + { size_t const blockSize = (size_t)(MIN(zds->fParams.windowSize, ZSTD_BLOCKSIZE_MAX)); + size_t const neededOutSize = (size_t)(zds->fParams.windowSize + blockSize + WILDCOPY_OVERLENGTH * 2); zds->blockSize = blockSize; if ((zds->inBuffSize < blockSize) || (zds->outBuffSize < neededOutSize)) { size_t const bufferSize = blockSize + neededOutSize; @@ -2393,7 +2382,7 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB zds->outBuffSize = neededOutSize; } } zds->streamStage = zdss_read; - /* pass-through */ + /* fall-through */ case zdss_read: DEBUGLOG(5, "stage zdss_read"); @@ -2418,8 +2407,7 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB } } if (ip==iend) { someMoreWork = 0; break; } /* no more input */ zds->streamStage = zdss_load; - /* pass-through */ - + /* fall-through */ case zdss_load: { size_t const neededInSize = ZSTD_nextSrcSizeToDecompress(zds); size_t const toLoad = neededInSize - zds->inPos; /* should always be <= remaining space within inBuff */ @@ -2441,8 +2429,7 @@ size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inB zds->outEnd = zds->outStart + decodedSize; } } zds->streamStage = zdss_flush; - /* pass-through */ - + /* fall-through */ case zdss_flush: { size_t const toFlushSize = zds->outEnd - zds->outStart; size_t const flushedSize = ZSTD_limitCopy(op, oend-op, zds->outBuff + zds->outStart, toFlushSize); diff --git a/thirdparty/zstd/zstd.h b/thirdparty/zstd/zstd.h index 58e9a5606d..13b4563fd6 100644 --- a/thirdparty/zstd/zstd.h +++ b/thirdparty/zstd/zstd.h @@ -2,11 +2,10 @@ * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of this source tree. An additional grant - * of patent rights can be found in the PATENTS file in the same directory. + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). */ - #if defined (__cplusplus) extern "C" { #endif @@ -59,7 +58,7 @@ extern "C" { /*------ Version ------*/ #define ZSTD_VERSION_MAJOR 1 #define ZSTD_VERSION_MINOR 3 -#define ZSTD_VERSION_RELEASE 0 +#define ZSTD_VERSION_RELEASE 1 #define ZSTD_VERSION_NUMBER (ZSTD_VERSION_MAJOR *100*100 + ZSTD_VERSION_MINOR *100 + ZSTD_VERSION_RELEASE) ZSTDLIB_API unsigned ZSTD_versionNumber(void); /**< useful to check dll version */ @@ -425,13 +424,6 @@ typedef struct { ZSTD_frameParameters fParams; } ZSTD_parameters; -typedef struct { - unsigned long long frameContentSize; - size_t windowSize; - unsigned dictID; - unsigned checksumFlag; -} ZSTD_frameHeader; - /*= Custom memory allocation functions */ typedef void* (*ZSTD_allocFunction) (void* opaque, size_t size); typedef void (*ZSTD_freeFunction) (void* opaque, void* address); @@ -809,7 +801,6 @@ ZSTDLIB_API size_t ZSTD_compressContinue(ZSTD_CCtx* cctx, void* dst, size_t dstC ZSTDLIB_API size_t ZSTD_compressEnd(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); - /*- Buffer-less streaming decompression (synchronous mode) @@ -874,6 +865,15 @@ ZSTDLIB_API size_t ZSTD_compressEnd(ZSTD_CCtx* cctx, void* dst, size_t dstCapaci */ /*===== Buffer-less streaming decompression functions =====*/ +typedef enum { ZSTD_frame, ZSTD_skippableFrame } ZSTD_frameType_e; +typedef struct { + unsigned long long frameContentSize; /* ZSTD_CONTENTSIZE_UNKNOWN means this field is not available. 0 means "empty" */ + unsigned long long windowSize; /* can be very large, up to <= frameContentSize */ + ZSTD_frameType_e frameType; /* if == ZSTD_skippableFrame, frameContentSize is the size of skippable content */ + unsigned headerSize; + unsigned dictID; + unsigned checksumFlag; +} ZSTD_frameHeader; ZSTDLIB_API size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize); /**< doesn't consume input */ ZSTDLIB_API size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx); ZSTDLIB_API size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); @@ -953,7 +953,9 @@ typedef enum { * Special: value 0 means "do not change strategy". */ /* frame parameters */ - ZSTD_p_contentSizeFlag=200, /* Content size is written into frame header _whenever known_ (default:1) */ + ZSTD_p_contentSizeFlag=200, /* Content size is written into frame header _whenever known_ (default:1) + * note that content size must be known at the beginning, + * it is sent using ZSTD_CCtx_setPledgedSrcSize() */ ZSTD_p_checksumFlag, /* A 32-bits checksum of content is written at end of frame (default:0) */ ZSTD_p_dictIDFlag, /* When applicable, dictID of dictionary is provided in frame header (default:1) */ |