diff options
author | shane | 2013-10-17 15:02:23 -0400 |
---|---|---|
committer | shane | 2013-10-17 15:02:23 -0400 |
commit | 95d1d0cc91f36ece9da39270cd1ca0536d0cf02d (patch) | |
tree | 9d2431f1bfa3c4a562cd006bc22ef2b7280d7b66 /src | |
parent | e9995aee384bdbac6c6afc0f4548a080eb4241bc (diff) | |
download | libplist-95d1d0cc91f36ece9da39270cd1ca0536d0cf02d.tar.gz libplist-95d1d0cc91f36ece9da39270cd1ca0536d0cf02d.tar.bz2 |
UTF-16 surrogate pair fix
Handle UTF-16 surrogate pair conversion to/from UTF-8
Diffstat (limited to 'src')
-rw-r--r-- | src/bplist.c | 41 |
1 files changed, 37 insertions, 4 deletions
diff --git a/src/bplist.c b/src/bplist.c index 5db7096..d825fc8 100644 --- a/src/bplist.c +++ b/src/bplist.c @@ -286,15 +286,38 @@ static plist_t parse_string_node(char *bnode, uint64_t size) static char *plist_utf16_to_utf8(uint16_t *unistr, long len, long *items_read, long *items_written) { if (!unistr || (len <= 0)) return NULL; - char *outbuf = (char*)malloc(3*(len+1)); + char *outbuf = (char*)malloc(4*(len+1)); int p = 0; int i = 0; uint16_t wc; + uint32_t w; + int read_lead_surrogate = 0; while (i < len) { wc = unistr[i++]; - if (wc >= 0x800) { + if (wc >= 0xD800 && wc <= 0xDBFF) { + if (!read_lead_surrogate) { + read_lead_surrogate = 1; + w = 0x010000 + ((wc & 0x3FF) << 10); + } else { + // This is invalid, the next 16 bit char should be a trail surrogate. + // Handling error by skipping. + read_lead_surrogate = 0; + } + } else if (wc >= 0xDC00 && wc <= 0xDFFF) { + if (read_lead_surrogate) { + read_lead_surrogate = 0; + w = w | (wc & 0x3FF); + outbuf[p++] = (char)(0xF0 + ((w >> 18) & 0x3)); + outbuf[p++] = (char)(0x80 + ((w >> 12) & 0x3F)); + outbuf[p++] = (char)(0x80 + ((w >> 6) & 0x3F)); + outbuf[p++] = (char)(0x80 + (w & 0x3F)); + } else { + // This is invalid. A trail surrogate should always follow a lead surrogate. + // Handling error by skipping + } + } else if (wc >= 0x800) { outbuf[p++] = (char)(0xE0 + ((wc >> 12) & 0xF)); outbuf[p++] = (char)(0x80 + ((wc >> 6) & 0x3F)); outbuf[p++] = (char)(0x80 + (wc & 0x3F)); @@ -988,19 +1011,29 @@ static int is_ascii_string(char* s, int len) uint16_t *plist_utf8_to_utf16(char *unistr, long size, long *items_read, long *items_written) { - uint16_t *outbuf = (uint16_t*)malloc((size+1)*sizeof(uint16_t)); + uint16_t *outbuf = (uint16_t*)malloc(((size*2)+1)*sizeof(uint16_t)); int p = 0; int i = 0; unsigned char c0; unsigned char c1; unsigned char c2; + unsigned char c3; + + uint32_t w; while (i < size) { c0 = unistr[i]; c1 = (i < size-1) ? unistr[i+1] : 0; c2 = (i < size-2) ? unistr[i+2] : 0; - if ((c0 >= 0xE0) && (i < size-2) && (c1 >= 0x80) && (c2 >= 0x80)) { + c3 = (i < size-3) ? unistr[i+3] : 0; + if ((c0 >= 0xF0) && (i < size-3) && (c1 >= 0x80) && (c2 >= 0x80) && (c3 >= 0x80)) { + // 4 byte sequence. Need to generate UTF-16 surrogate pair + w = ((((c0 & 7) << 18) + ((c1 & 0x3F) << 12) + ((c2 & 0x3F) << 6) + (c3 & 0x3F)) & 0x0FFFFF) - 0x010000; + outbuf[p++] = 0xD800 + (w >> 10); + outbuf[p++] = 0xDC00 + (w & 0x3FF); + i+=4; + } else if ((c0 >= 0xE0) && (i < size-2) && (c1 >= 0x80) && (c2 >= 0x80)) { // 3 byte sequence outbuf[p++] = ((c2 & 0x3F) + ((c1 & 3) << 6)) + (((c1 >> 2) & 15) << 8) + ((c0 & 15) << 12); i+=3; |