summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorRĂ©mi Verschelde <remi@verschelde.fr>2021-03-01 12:42:09 +0100
committerGitHub <noreply@github.com>2021-03-01 12:42:09 +0100
commitb98b2744a60b880facc270a4f0135a48a7082413 (patch)
tree1ad6694d6e5489b1f1ad0750debf3e2434df13c2
parentc2799893f4198aaf814f7e7e3ce6b087136e42e4 (diff)
parent76b653528dc393f7523187d5afe9e84867b7c4c7 (diff)
Merge pull request #46551 from bruvzg/json_utf16
JSON parser: add UTF-16 surrogate pairs support.
-rw-r--r--core/io/json.cpp46
1 files changed, 46 insertions, 0 deletions
diff --git a/core/io/json.cpp b/core/io/json.cpp
index bc4527869b..0d9117fdda 100644
--- a/core/io/json.cpp
+++ b/core/io/json.cpp
@@ -234,6 +234,52 @@ Error JSON::_get_token(const char32_t *p_str, int &index, int p_len, Token &r_to
}
index += 4; //will add at the end anyway
+ if ((res & 0xfffffc00) == 0xd800) {
+ if (p_str[index + 1] != '\\' || p_str[index + 2] != 'u') {
+ r_err_str = "Invalid UTF-16 sequence in string, unpaired lead surrogate";
+ return ERR_PARSE_ERROR;
+ }
+ index += 2;
+ char32_t trail = 0;
+ for (int j = 0; j < 4; j++) {
+ char32_t c = p_str[index + j + 1];
+ if (c == 0) {
+ r_err_str = "Unterminated String";
+ return ERR_PARSE_ERROR;
+ }
+ if (!((c >= '0' && c <= '9') || (c >= 'a' && c <= 'f') || (c >= 'A' && c <= 'F'))) {
+ r_err_str = "Malformed hex constant in string";
+ return ERR_PARSE_ERROR;
+ }
+ char32_t v;
+ if (c >= '0' && c <= '9') {
+ v = c - '0';
+ } else if (c >= 'a' && c <= 'f') {
+ v = c - 'a';
+ v += 10;
+ } else if (c >= 'A' && c <= 'F') {
+ v = c - 'A';
+ v += 10;
+ } else {
+ ERR_PRINT("Bug parsing hex constant.");
+ v = 0;
+ }
+
+ trail <<= 4;
+ trail |= v;
+ }
+ if ((trail & 0xfffffc00) == 0xdc00) {
+ res = (res << 10UL) + trail - ((0xd800 << 10UL) + 0xdc00 - 0x10000);
+ index += 4; //will add at the end anyway
+ } else {
+ r_err_str = "Invalid UTF-16 sequence in string, unpaired lead surrogate";
+ return ERR_PARSE_ERROR;
+ }
+ } else if ((res & 0xfffffc00) == 0xdc00) {
+ r_err_str = "Invalid UTF-16 sequence in string, unpaired trail surrogate";
+ return ERR_PARSE_ERROR;
+ }
+
} break;
default: {
res = next;