lib/Support/ConvertUTF.c

14  * This source code is provided as is by Unicode, Inc. No claims are
39         source sequences, enhanced error detection, added casts
121     const UTF32* source = *sourceStart;
123     while (source < sourceEnd) {
128         ch = *source++;
133                     --source; /* return to the illegal value itself */
151                 --source; /* Back up source pointer! */
159     *sourceStart = source;
170     const UTF16* source = *sourceStart;
173     while (source < sourceEnd) {
174         const UTF16* oldSource = source; /*  In case we have to back up because of target overflow. */
175         ch = *source++;
178             /* If the 16 bits following the high surrogate are in the source buffer... */
179             if (source < sourceEnd) {
180                 ch2 = *source;
185                     ++source;
187                     --source; /* return to the illegal value itself */
192                 --source; /* return to the high surrogate */
199                 --source; /* return to the illegal value itself */
205             source = oldSource; /* Back up source pointer! */
210     *sourceStart = source;
224     const UTF16* source = *sourceStart;
226     while (source < sourceEnd) {
231         const UTF16* oldSource = source; /* In case we have to back up because of target overflow. */
232         ch = *source++;
235             /* If the 16 bits following the high surrogate are in the source buffer... */
236             if (source < sourceEnd) {
237                 UTF32 ch2 = *source;
242                     ++source;
244                     --source; /* return to the illegal value itself */
249                 --source; /* return to the high surrogate */
256                 --source; /* return to the illegal value itself */
272             source = oldSource; /* Back up source pointer! */
283     *sourceStart = source;
294     const UTF32* source = *sourceStart;
296     while (source < sourceEnd) {
301         ch = *source++;
305                 --source; /* return to the illegal value itself */
325             --source; /* Back up source pointer! */
336     *sourceStart = source;
347  *  length = trailingBytesForUTF8[*source]+1;
354 static Boolean isLegalUTF8(const UTF8 *source, int length) {
356     const UTF8 *srcptr = source+length;
364         switch (*source) {
373     case 1: if (*source >= 0x80 && *source < 0xC2) return false;
375     if (*source > 0xF4) return false;
385 Boolean isLegalUTF8Sequence(const UTF8 *source, const UTF8 *sourceEnd) {
386     int length = trailingBytesForUTF8[*source]+1;
387     if (length > sourceEnd - source) {
390     return isLegalUTF8(source, length);
409 Boolean isLegalUTF8String(const UTF8 **source, const UTF8 *sourceEnd) {
410     while (*source != sourceEnd) {
411         int length = trailingBytesForUTF8[**source] + 1;
412         if (length > sourceEnd - *source || !isLegalUTF8(*source, length))
414         *source += length;
425     const UTF8* source = *sourceStart;
427     while (source < sourceEnd) {
429         unsigned short extraBytesToRead = trailingBytesForUTF8[*source];
430         if (extraBytesToRead >= sourceEnd - source) {
434         if (!isLegalUTF8(source, extraBytesToRead+1)) {
442             case 5: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */
443             case 4: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */
444             case 3: ch += *source++; ch <<= 6;
445             case 2: ch += *source++; ch <<= 6;
446             case 1: ch += *source++; ch <<= 6;
447             case 0: ch += *source++;
452             source -= (extraBytesToRead+1); /* Back up source pointer! */
459                     source -= (extraBytesToRead+1); /* return to the illegal value itself */
471                 source -= (extraBytesToRead+1); /* return to the start */
479                 source -= (extraBytesToRead+1); /* Back up source pointer! */
487     *sourceStart = source;
498     const UTF8* source = *sourceStart;
500     while (source < sourceEnd) {
502         unsigned short extraBytesToRead = trailingBytesForUTF8[*source];
503         if (extraBytesToRead >= sourceEnd - source) {
507         if (!isLegalUTF8(source, extraBytesToRead+1)) {
515             case 5: ch += *source++; ch <<= 6;
516             case 4: ch += *source++; ch <<= 6;
517             case 3: ch += *source++; ch <<= 6;
518             case 2: ch += *source++; ch <<= 6;
519             case 1: ch += *source++; ch <<= 6;
520             case 0: ch += *source++;
525             source -= (extraBytesToRead+1); /* Back up the source pointer! */
535                     source -= (extraBytesToRead+1); /* return to the illegal value itself */
549     *sourceStart = source;
563                 ch += *source++;