1 | // © 2016 and later: Unicode, Inc. and others. |
2 | // License & terms of use: http://www.unicode.org/copyright.html |
3 | /* |
4 | ******************************************************************************* |
5 | * Copyright (C) 2004-2015, International Business Machines |
6 | * Corporation and others. All Rights Reserved. |
7 | ******************************************************************************* |
8 | * file name: uregex.cpp |
9 | */ |
10 | |
11 | #include "unicode/utypes.h" |
12 | |
13 | #if !UCONFIG_NO_REGULAR_EXPRESSIONS |
14 | |
15 | #include "unicode/regex.h" |
16 | #include "unicode/uregex.h" |
17 | #include "unicode/unistr.h" |
18 | #include "unicode/ustring.h" |
19 | #include "unicode/uchar.h" |
20 | #include "unicode/uobject.h" |
21 | #include "unicode/utf16.h" |
22 | #include "cmemory.h" |
23 | #include "uassert.h" |
24 | #include "uhash.h" |
25 | #include "umutex.h" |
26 | #include "uvectr32.h" |
27 | |
28 | #include "regextxt.h" |
29 | |
30 | U_NAMESPACE_BEGIN |
31 | |
32 | #define REMAINING_CAPACITY(idx,len) ((((len)-(idx))>0)?((len)-(idx)):0) |
33 | |
34 | struct RegularExpression: public UMemory { |
35 | public: |
36 | RegularExpression(); |
37 | ~RegularExpression(); |
38 | int32_t fMagic; |
39 | RegexPattern *fPat; |
40 | u_atomic_int32_t *fPatRefCount; |
41 | UChar *fPatString; |
42 | int32_t fPatStringLen; |
43 | RegexMatcher *fMatcher; |
44 | const UChar *fText; // Text from setText() |
45 | int32_t fTextLength; // Length provided by user with setText(), which |
46 | // may be -1. |
47 | UBool fOwnsText; |
48 | }; |
49 | |
50 | static const int32_t REXP_MAGIC = 0x72657870; // "rexp" in ASCII |
51 | |
52 | RegularExpression::RegularExpression() { |
53 | fMagic = REXP_MAGIC; |
54 | fPat = NULL; |
55 | fPatRefCount = NULL; |
56 | fPatString = NULL; |
57 | fPatStringLen = 0; |
58 | fMatcher = NULL; |
59 | fText = NULL; |
60 | fTextLength = 0; |
61 | fOwnsText = FALSE; |
62 | } |
63 | |
64 | RegularExpression::~RegularExpression() { |
65 | delete fMatcher; |
66 | fMatcher = NULL; |
67 | if (fPatRefCount!=NULL && umtx_atomic_dec(fPatRefCount)==0) { |
68 | delete fPat; |
69 | uprv_free(fPatString); |
70 | uprv_free((void *)fPatRefCount); |
71 | } |
72 | if (fOwnsText && fText!=NULL) { |
73 | uprv_free((void *)fText); |
74 | } |
75 | fMagic = 0; |
76 | } |
77 | |
78 | U_NAMESPACE_END |
79 | |
80 | U_NAMESPACE_USE |
81 | |
82 | //---------------------------------------------------------------------------------------- |
83 | // |
84 | // validateRE Do boilerplate style checks on API function parameters. |
85 | // Return TRUE if they look OK. |
86 | //---------------------------------------------------------------------------------------- |
87 | static UBool validateRE(const RegularExpression *re, UBool requiresText, UErrorCode *status) { |
88 | if (U_FAILURE(*status)) { |
89 | return FALSE; |
90 | } |
91 | if (re == NULL || re->fMagic != REXP_MAGIC) { |
92 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
93 | return FALSE; |
94 | } |
95 | // !!! Not sure how to update this with the new UText backing, which is stored in re->fMatcher anyway |
96 | if (requiresText && re->fText == NULL && !re->fOwnsText) { |
97 | *status = U_REGEX_INVALID_STATE; |
98 | return FALSE; |
99 | } |
100 | return TRUE; |
101 | } |
102 | |
103 | //---------------------------------------------------------------------------------------- |
104 | // |
105 | // uregex_open |
106 | // |
107 | //---------------------------------------------------------------------------------------- |
108 | U_CAPI URegularExpression * U_EXPORT2 |
109 | uregex_open( const UChar *pattern, |
110 | int32_t patternLength, |
111 | uint32_t flags, |
112 | UParseError *pe, |
113 | UErrorCode *status) { |
114 | |
115 | if (U_FAILURE(*status)) { |
116 | return NULL; |
117 | } |
118 | if (pattern == NULL || patternLength < -1 || patternLength == 0) { |
119 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
120 | return NULL; |
121 | } |
122 | int32_t actualPatLen = patternLength; |
123 | if (actualPatLen == -1) { |
124 | actualPatLen = u_strlen(pattern); |
125 | } |
126 | |
127 | RegularExpression *re = new RegularExpression; |
128 | u_atomic_int32_t *refC = (u_atomic_int32_t *)uprv_malloc(sizeof(int32_t)); |
129 | UChar *patBuf = (UChar *)uprv_malloc(sizeof(UChar)*(actualPatLen+1)); |
130 | if (re == NULL || refC == NULL || patBuf == NULL) { |
131 | *status = U_MEMORY_ALLOCATION_ERROR; |
132 | delete re; |
133 | uprv_free((void *)refC); |
134 | uprv_free(patBuf); |
135 | return NULL; |
136 | } |
137 | re->fPatRefCount = refC; |
138 | *re->fPatRefCount = 1; |
139 | |
140 | // |
141 | // Make a copy of the pattern string, so we can return it later if asked. |
142 | // For compiling the pattern, we will use a UText wrapper around |
143 | // this local copy, to avoid making even more copies. |
144 | // |
145 | re->fPatString = patBuf; |
146 | re->fPatStringLen = patternLength; |
147 | u_memcpy(patBuf, pattern, actualPatLen); |
148 | patBuf[actualPatLen] = 0; |
149 | |
150 | UText patText = UTEXT_INITIALIZER; |
151 | utext_openUChars(&patText, patBuf, patternLength, status); |
152 | |
153 | // |
154 | // Compile the pattern |
155 | // |
156 | if (pe != NULL) { |
157 | re->fPat = RegexPattern::compile(&patText, flags, *pe, *status); |
158 | } else { |
159 | re->fPat = RegexPattern::compile(&patText, flags, *status); |
160 | } |
161 | utext_close(&patText); |
162 | |
163 | if (U_FAILURE(*status)) { |
164 | goto ErrorExit; |
165 | } |
166 | |
167 | // |
168 | // Create the matcher object |
169 | // |
170 | re->fMatcher = re->fPat->matcher(*status); |
171 | if (U_SUCCESS(*status)) { |
172 | return (URegularExpression*)re; |
173 | } |
174 | |
175 | ErrorExit: |
176 | delete re; |
177 | return NULL; |
178 | |
179 | } |
180 | |
181 | //---------------------------------------------------------------------------------------- |
182 | // |
183 | // uregex_openUText |
184 | // |
185 | //---------------------------------------------------------------------------------------- |
186 | U_CAPI URegularExpression * U_EXPORT2 |
187 | uregex_openUText(UText *pattern, |
188 | uint32_t flags, |
189 | UParseError *pe, |
190 | UErrorCode *status) { |
191 | |
192 | if (U_FAILURE(*status)) { |
193 | return NULL; |
194 | } |
195 | if (pattern == NULL) { |
196 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
197 | return NULL; |
198 | } |
199 | |
200 | int64_t patternNativeLength = utext_nativeLength(pattern); |
201 | |
202 | if (patternNativeLength == 0) { |
203 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
204 | return NULL; |
205 | } |
206 | |
207 | RegularExpression *re = new RegularExpression; |
208 | |
209 | UErrorCode lengthStatus = U_ZERO_ERROR; |
210 | int32_t pattern16Length = utext_extract(pattern, 0, patternNativeLength, NULL, 0, &lengthStatus); |
211 | |
212 | u_atomic_int32_t *refC = (u_atomic_int32_t *)uprv_malloc(sizeof(int32_t)); |
213 | UChar *patBuf = (UChar *)uprv_malloc(sizeof(UChar)*(pattern16Length+1)); |
214 | if (re == NULL || refC == NULL || patBuf == NULL) { |
215 | *status = U_MEMORY_ALLOCATION_ERROR; |
216 | delete re; |
217 | uprv_free((void *)refC); |
218 | uprv_free(patBuf); |
219 | return NULL; |
220 | } |
221 | re->fPatRefCount = refC; |
222 | *re->fPatRefCount = 1; |
223 | |
224 | // |
225 | // Make a copy of the pattern string, so we can return it later if asked. |
226 | // For compiling the pattern, we will use a read-only UText wrapper |
227 | // around this local copy, to avoid making even more copies. |
228 | // |
229 | re->fPatString = patBuf; |
230 | re->fPatStringLen = pattern16Length; |
231 | utext_extract(pattern, 0, patternNativeLength, patBuf, pattern16Length+1, status); |
232 | |
233 | UText patText = UTEXT_INITIALIZER; |
234 | utext_openUChars(&patText, patBuf, pattern16Length, status); |
235 | |
236 | // |
237 | // Compile the pattern |
238 | // |
239 | if (pe != NULL) { |
240 | re->fPat = RegexPattern::compile(&patText, flags, *pe, *status); |
241 | } else { |
242 | re->fPat = RegexPattern::compile(&patText, flags, *status); |
243 | } |
244 | utext_close(&patText); |
245 | |
246 | if (U_FAILURE(*status)) { |
247 | goto ErrorExit; |
248 | } |
249 | |
250 | // |
251 | // Create the matcher object |
252 | // |
253 | re->fMatcher = re->fPat->matcher(*status); |
254 | if (U_SUCCESS(*status)) { |
255 | return (URegularExpression*)re; |
256 | } |
257 | |
258 | ErrorExit: |
259 | delete re; |
260 | return NULL; |
261 | |
262 | } |
263 | |
264 | //---------------------------------------------------------------------------------------- |
265 | // |
266 | // uregex_close |
267 | // |
268 | //---------------------------------------------------------------------------------------- |
269 | U_CAPI void U_EXPORT2 |
270 | uregex_close(URegularExpression *re2) { |
271 | RegularExpression *re = (RegularExpression*)re2; |
272 | UErrorCode status = U_ZERO_ERROR; |
273 | if (validateRE(re, FALSE, &status) == FALSE) { |
274 | return; |
275 | } |
276 | delete re; |
277 | } |
278 | |
279 | |
280 | //---------------------------------------------------------------------------------------- |
281 | // |
282 | // uregex_clone |
283 | // |
284 | //---------------------------------------------------------------------------------------- |
285 | U_CAPI URegularExpression * U_EXPORT2 |
286 | uregex_clone(const URegularExpression *source2, UErrorCode *status) { |
287 | RegularExpression *source = (RegularExpression*)source2; |
288 | if (validateRE(source, FALSE, status) == FALSE) { |
289 | return NULL; |
290 | } |
291 | |
292 | RegularExpression *clone = new RegularExpression; |
293 | if (clone == NULL) { |
294 | *status = U_MEMORY_ALLOCATION_ERROR; |
295 | return NULL; |
296 | } |
297 | |
298 | clone->fMatcher = source->fPat->matcher(*status); |
299 | if (U_FAILURE(*status)) { |
300 | delete clone; |
301 | return NULL; |
302 | } |
303 | |
304 | clone->fPat = source->fPat; |
305 | clone->fPatRefCount = source->fPatRefCount; |
306 | clone->fPatString = source->fPatString; |
307 | clone->fPatStringLen = source->fPatStringLen; |
308 | umtx_atomic_inc(source->fPatRefCount); |
309 | // Note: fText is not cloned. |
310 | |
311 | return (URegularExpression*)clone; |
312 | } |
313 | |
314 | |
315 | |
316 | |
317 | //------------------------------------------------------------------------------ |
318 | // |
319 | // uregex_pattern |
320 | // |
321 | //------------------------------------------------------------------------------ |
322 | U_CAPI const UChar * U_EXPORT2 |
323 | uregex_pattern(const URegularExpression *regexp2, |
324 | int32_t *patLength, |
325 | UErrorCode *status) { |
326 | RegularExpression *regexp = (RegularExpression*)regexp2; |
327 | |
328 | if (validateRE(regexp, FALSE, status) == FALSE) { |
329 | return NULL; |
330 | } |
331 | if (patLength != NULL) { |
332 | *patLength = regexp->fPatStringLen; |
333 | } |
334 | return regexp->fPatString; |
335 | } |
336 | |
337 | |
338 | //------------------------------------------------------------------------------ |
339 | // |
340 | // uregex_patternUText |
341 | // |
342 | //------------------------------------------------------------------------------ |
343 | U_CAPI UText * U_EXPORT2 |
344 | uregex_patternUText(const URegularExpression *regexp2, |
345 | UErrorCode *status) { |
346 | RegularExpression *regexp = (RegularExpression*)regexp2; |
347 | return regexp->fPat->patternText(*status); |
348 | } |
349 | |
350 | |
351 | //------------------------------------------------------------------------------ |
352 | // |
353 | // uregex_flags |
354 | // |
355 | //------------------------------------------------------------------------------ |
356 | U_CAPI int32_t U_EXPORT2 |
357 | uregex_flags(const URegularExpression *regexp2, UErrorCode *status) { |
358 | RegularExpression *regexp = (RegularExpression*)regexp2; |
359 | if (validateRE(regexp, FALSE, status) == FALSE) { |
360 | return 0; |
361 | } |
362 | int32_t flags = regexp->fPat->flags(); |
363 | return flags; |
364 | } |
365 | |
366 | |
367 | //------------------------------------------------------------------------------ |
368 | // |
369 | // uregex_setText |
370 | // |
371 | //------------------------------------------------------------------------------ |
372 | U_CAPI void U_EXPORT2 |
373 | uregex_setText(URegularExpression *regexp2, |
374 | const UChar *text, |
375 | int32_t textLength, |
376 | UErrorCode *status) { |
377 | RegularExpression *regexp = (RegularExpression*)regexp2; |
378 | if (validateRE(regexp, FALSE, status) == FALSE) { |
379 | return; |
380 | } |
381 | if (text == NULL || textLength < -1) { |
382 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
383 | return; |
384 | } |
385 | |
386 | if (regexp->fOwnsText && regexp->fText != NULL) { |
387 | uprv_free((void *)regexp->fText); |
388 | } |
389 | |
390 | regexp->fText = text; |
391 | regexp->fTextLength = textLength; |
392 | regexp->fOwnsText = FALSE; |
393 | |
394 | UText input = UTEXT_INITIALIZER; |
395 | utext_openUChars(&input, text, textLength, status); |
396 | regexp->fMatcher->reset(&input); |
397 | utext_close(&input); // reset() made a shallow clone, so we don't need this copy |
398 | } |
399 | |
400 | |
401 | //------------------------------------------------------------------------------ |
402 | // |
403 | // uregex_setUText |
404 | // |
405 | //------------------------------------------------------------------------------ |
406 | U_CAPI void U_EXPORT2 |
407 | uregex_setUText(URegularExpression *regexp2, |
408 | UText *text, |
409 | UErrorCode *status) { |
410 | RegularExpression *regexp = (RegularExpression*)regexp2; |
411 | if (validateRE(regexp, FALSE, status) == FALSE) { |
412 | return; |
413 | } |
414 | if (text == NULL) { |
415 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
416 | return; |
417 | } |
418 | |
419 | if (regexp->fOwnsText && regexp->fText != NULL) { |
420 | uprv_free((void *)regexp->fText); |
421 | } |
422 | |
423 | regexp->fText = NULL; // only fill it in on request |
424 | regexp->fTextLength = -1; |
425 | regexp->fOwnsText = TRUE; |
426 | regexp->fMatcher->reset(text); |
427 | } |
428 | |
429 | |
430 | |
431 | //------------------------------------------------------------------------------ |
432 | // |
433 | // uregex_getText |
434 | // |
435 | //------------------------------------------------------------------------------ |
436 | U_CAPI const UChar * U_EXPORT2 |
437 | uregex_getText(URegularExpression *regexp2, |
438 | int32_t *textLength, |
439 | UErrorCode *status) { |
440 | RegularExpression *regexp = (RegularExpression*)regexp2; |
441 | if (validateRE(regexp, FALSE, status) == FALSE) { |
442 | return NULL; |
443 | } |
444 | |
445 | if (regexp->fText == NULL) { |
446 | // need to fill in the text |
447 | UText *inputText = regexp->fMatcher->inputText(); |
448 | int64_t inputNativeLength = utext_nativeLength(inputText); |
449 | if (UTEXT_FULL_TEXT_IN_CHUNK(inputText, inputNativeLength)) { |
450 | regexp->fText = inputText->chunkContents; |
451 | regexp->fTextLength = (int32_t)inputNativeLength; |
452 | regexp->fOwnsText = FALSE; // because the UText owns it |
453 | } else { |
454 | UErrorCode lengthStatus = U_ZERO_ERROR; |
455 | regexp->fTextLength = utext_extract(inputText, 0, inputNativeLength, NULL, 0, &lengthStatus); // buffer overflow error |
456 | UChar *inputChars = (UChar *)uprv_malloc(sizeof(UChar)*(regexp->fTextLength+1)); |
457 | |
458 | utext_extract(inputText, 0, inputNativeLength, inputChars, regexp->fTextLength+1, status); |
459 | regexp->fText = inputChars; |
460 | regexp->fOwnsText = TRUE; // should already be set but just in case |
461 | } |
462 | } |
463 | |
464 | if (textLength != NULL) { |
465 | *textLength = regexp->fTextLength; |
466 | } |
467 | return regexp->fText; |
468 | } |
469 | |
470 | |
471 | //------------------------------------------------------------------------------ |
472 | // |
473 | // uregex_getUText |
474 | // |
475 | //------------------------------------------------------------------------------ |
476 | U_CAPI UText * U_EXPORT2 |
477 | uregex_getUText(URegularExpression *regexp2, |
478 | UText *dest, |
479 | UErrorCode *status) { |
480 | RegularExpression *regexp = (RegularExpression*)regexp2; |
481 | if (validateRE(regexp, FALSE, status) == FALSE) { |
482 | return dest; |
483 | } |
484 | return regexp->fMatcher->getInput(dest, *status); |
485 | } |
486 | |
487 | |
488 | //------------------------------------------------------------------------------ |
489 | // |
490 | // uregex_refreshUText |
491 | // |
492 | //------------------------------------------------------------------------------ |
493 | U_CAPI void U_EXPORT2 |
494 | uregex_refreshUText(URegularExpression *regexp2, |
495 | UText *text, |
496 | UErrorCode *status) { |
497 | RegularExpression *regexp = (RegularExpression*)regexp2; |
498 | if (validateRE(regexp, FALSE, status) == FALSE) { |
499 | return; |
500 | } |
501 | regexp->fMatcher->refreshInputText(text, *status); |
502 | } |
503 | |
504 | |
505 | //------------------------------------------------------------------------------ |
506 | // |
507 | // uregex_matches |
508 | // |
509 | //------------------------------------------------------------------------------ |
510 | U_CAPI UBool U_EXPORT2 |
511 | uregex_matches(URegularExpression *regexp2, |
512 | int32_t startIndex, |
513 | UErrorCode *status) { |
514 | return uregex_matches64( regexp2, (int64_t)startIndex, status); |
515 | } |
516 | |
517 | U_CAPI UBool U_EXPORT2 |
518 | uregex_matches64(URegularExpression *regexp2, |
519 | int64_t startIndex, |
520 | UErrorCode *status) { |
521 | RegularExpression *regexp = (RegularExpression*)regexp2; |
522 | UBool result = FALSE; |
523 | if (validateRE(regexp, TRUE, status) == FALSE) { |
524 | return result; |
525 | } |
526 | if (startIndex == -1) { |
527 | result = regexp->fMatcher->matches(*status); |
528 | } else { |
529 | result = regexp->fMatcher->matches(startIndex, *status); |
530 | } |
531 | return result; |
532 | } |
533 | |
534 | |
535 | //------------------------------------------------------------------------------ |
536 | // |
537 | // uregex_lookingAt |
538 | // |
539 | //------------------------------------------------------------------------------ |
540 | U_CAPI UBool U_EXPORT2 |
541 | uregex_lookingAt(URegularExpression *regexp2, |
542 | int32_t startIndex, |
543 | UErrorCode *status) { |
544 | return uregex_lookingAt64( regexp2, (int64_t)startIndex, status); |
545 | } |
546 | |
547 | U_CAPI UBool U_EXPORT2 |
548 | uregex_lookingAt64(URegularExpression *regexp2, |
549 | int64_t startIndex, |
550 | UErrorCode *status) { |
551 | RegularExpression *regexp = (RegularExpression*)regexp2; |
552 | UBool result = FALSE; |
553 | if (validateRE(regexp, TRUE, status) == FALSE) { |
554 | return result; |
555 | } |
556 | if (startIndex == -1) { |
557 | result = regexp->fMatcher->lookingAt(*status); |
558 | } else { |
559 | result = regexp->fMatcher->lookingAt(startIndex, *status); |
560 | } |
561 | return result; |
562 | } |
563 | |
564 | |
565 | |
566 | //------------------------------------------------------------------------------ |
567 | // |
568 | // uregex_find |
569 | // |
570 | //------------------------------------------------------------------------------ |
571 | U_CAPI UBool U_EXPORT2 |
572 | uregex_find(URegularExpression *regexp2, |
573 | int32_t startIndex, |
574 | UErrorCode *status) { |
575 | return uregex_find64( regexp2, (int64_t)startIndex, status); |
576 | } |
577 | |
578 | U_CAPI UBool U_EXPORT2 |
579 | uregex_find64(URegularExpression *regexp2, |
580 | int64_t startIndex, |
581 | UErrorCode *status) { |
582 | RegularExpression *regexp = (RegularExpression*)regexp2; |
583 | UBool result = FALSE; |
584 | if (validateRE(regexp, TRUE, status) == FALSE) { |
585 | return result; |
586 | } |
587 | if (startIndex == -1) { |
588 | regexp->fMatcher->resetPreserveRegion(); |
589 | result = regexp->fMatcher->find(*status); |
590 | } else { |
591 | result = regexp->fMatcher->find(startIndex, *status); |
592 | } |
593 | return result; |
594 | } |
595 | |
596 | |
597 | //------------------------------------------------------------------------------ |
598 | // |
599 | // uregex_findNext |
600 | // |
601 | //------------------------------------------------------------------------------ |
602 | U_CAPI UBool U_EXPORT2 |
603 | uregex_findNext(URegularExpression *regexp2, |
604 | UErrorCode *status) { |
605 | RegularExpression *regexp = (RegularExpression*)regexp2; |
606 | if (validateRE(regexp, TRUE, status) == FALSE) { |
607 | return FALSE; |
608 | } |
609 | UBool result = regexp->fMatcher->find(*status); |
610 | return result; |
611 | } |
612 | |
613 | //------------------------------------------------------------------------------ |
614 | // |
615 | // uregex_groupCount |
616 | // |
617 | //------------------------------------------------------------------------------ |
618 | U_CAPI int32_t U_EXPORT2 |
619 | uregex_groupCount(URegularExpression *regexp2, |
620 | UErrorCode *status) { |
621 | RegularExpression *regexp = (RegularExpression*)regexp2; |
622 | if (validateRE(regexp, FALSE, status) == FALSE) { |
623 | return 0; |
624 | } |
625 | int32_t result = regexp->fMatcher->groupCount(); |
626 | return result; |
627 | } |
628 | |
629 | |
630 | //------------------------------------------------------------------------------ |
631 | // |
632 | // uregex_groupNumberFromName |
633 | // |
634 | //------------------------------------------------------------------------------ |
635 | int32_t |
636 | uregex_groupNumberFromName(URegularExpression *regexp2, |
637 | const UChar *groupName, |
638 | int32_t nameLength, |
639 | UErrorCode *status) { |
640 | RegularExpression *regexp = (RegularExpression*)regexp2; |
641 | if (validateRE(regexp, FALSE, status) == FALSE) { |
642 | return 0; |
643 | } |
644 | int32_t result = regexp->fPat->groupNumberFromName(UnicodeString(groupName, nameLength), *status); |
645 | return result; |
646 | } |
647 | |
648 | int32_t |
649 | uregex_groupNumberFromCName(URegularExpression *regexp2, |
650 | const char *groupName, |
651 | int32_t nameLength, |
652 | UErrorCode *status) { |
653 | RegularExpression *regexp = (RegularExpression*)regexp2; |
654 | if (validateRE(regexp, FALSE, status) == FALSE) { |
655 | return 0; |
656 | } |
657 | return regexp->fPat->groupNumberFromName(groupName, nameLength, *status); |
658 | } |
659 | |
660 | //------------------------------------------------------------------------------ |
661 | // |
662 | // uregex_group |
663 | // |
664 | //------------------------------------------------------------------------------ |
665 | U_CAPI int32_t U_EXPORT2 |
666 | uregex_group(URegularExpression *regexp2, |
667 | int32_t groupNum, |
668 | UChar *dest, |
669 | int32_t destCapacity, |
670 | UErrorCode *status) { |
671 | RegularExpression *regexp = (RegularExpression*)regexp2; |
672 | if (validateRE(regexp, TRUE, status) == FALSE) { |
673 | return 0; |
674 | } |
675 | if (destCapacity < 0 || (destCapacity > 0 && dest == NULL)) { |
676 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
677 | return 0; |
678 | } |
679 | |
680 | if (destCapacity == 0 || regexp->fText != NULL) { |
681 | // If preflighting or if we already have the text as UChars, |
682 | // this is a little cheaper than extracting from the UText |
683 | |
684 | // |
685 | // Pick up the range of characters from the matcher |
686 | // |
687 | int32_t startIx = regexp->fMatcher->start(groupNum, *status); |
688 | int32_t endIx = regexp->fMatcher->end (groupNum, *status); |
689 | if (U_FAILURE(*status)) { |
690 | return 0; |
691 | } |
692 | |
693 | // |
694 | // Trim length based on buffer capacity |
695 | // |
696 | int32_t fullLength = endIx - startIx; |
697 | int32_t copyLength = fullLength; |
698 | if (copyLength < destCapacity) { |
699 | dest[copyLength] = 0; |
700 | } else if (copyLength == destCapacity) { |
701 | *status = U_STRING_NOT_TERMINATED_WARNING; |
702 | } else { |
703 | copyLength = destCapacity; |
704 | *status = U_BUFFER_OVERFLOW_ERROR; |
705 | } |
706 | |
707 | // |
708 | // Copy capture group to user's buffer |
709 | // |
710 | if (copyLength > 0) { |
711 | u_memcpy(dest, ®exp->fText[startIx], copyLength); |
712 | } |
713 | return fullLength; |
714 | } else { |
715 | int64_t start = regexp->fMatcher->start64(groupNum, *status); |
716 | int64_t limit = regexp->fMatcher->end64(groupNum, *status); |
717 | if (U_FAILURE(*status)) { |
718 | return 0; |
719 | } |
720 | // Note edge cases: |
721 | // Group didn't match: start == end == -1. UText trims to 0, UText gives zero length result. |
722 | // Zero Length Match: start == end. |
723 | int32_t length = utext_extract(regexp->fMatcher->inputText(), start, limit, dest, destCapacity, status); |
724 | return length; |
725 | } |
726 | |
727 | } |
728 | |
729 | |
730 | //------------------------------------------------------------------------------ |
731 | // |
732 | // uregex_groupUText |
733 | // |
734 | //------------------------------------------------------------------------------ |
735 | U_CAPI UText * U_EXPORT2 |
736 | uregex_groupUText(URegularExpression *regexp2, |
737 | int32_t groupNum, |
738 | UText *dest, |
739 | int64_t *groupLength, |
740 | UErrorCode *status) { |
741 | RegularExpression *regexp = (RegularExpression*)regexp2; |
742 | if (validateRE(regexp, TRUE, status) == FALSE) { |
743 | UErrorCode emptyTextStatus = U_ZERO_ERROR; |
744 | return (dest ? dest : utext_openUChars(NULL, NULL, 0, &emptyTextStatus)); |
745 | } |
746 | |
747 | return regexp->fMatcher->group(groupNum, dest, *groupLength, *status); |
748 | } |
749 | |
750 | //------------------------------------------------------------------------------ |
751 | // |
752 | // uregex_start |
753 | // |
754 | //------------------------------------------------------------------------------ |
755 | U_CAPI int32_t U_EXPORT2 |
756 | uregex_start(URegularExpression *regexp2, |
757 | int32_t groupNum, |
758 | UErrorCode *status) { |
759 | return (int32_t)uregex_start64( regexp2, groupNum, status); |
760 | } |
761 | |
762 | U_CAPI int64_t U_EXPORT2 |
763 | uregex_start64(URegularExpression *regexp2, |
764 | int32_t groupNum, |
765 | UErrorCode *status) { |
766 | RegularExpression *regexp = (RegularExpression*)regexp2; |
767 | if (validateRE(regexp, TRUE, status) == FALSE) { |
768 | return 0; |
769 | } |
770 | int64_t result = regexp->fMatcher->start64(groupNum, *status); |
771 | return result; |
772 | } |
773 | |
774 | //------------------------------------------------------------------------------ |
775 | // |
776 | // uregex_end |
777 | // |
778 | //------------------------------------------------------------------------------ |
779 | U_CAPI int32_t U_EXPORT2 |
780 | uregex_end(URegularExpression *regexp2, |
781 | int32_t groupNum, |
782 | UErrorCode *status) { |
783 | return (int32_t)uregex_end64( regexp2, groupNum, status); |
784 | } |
785 | |
786 | U_CAPI int64_t U_EXPORT2 |
787 | uregex_end64(URegularExpression *regexp2, |
788 | int32_t groupNum, |
789 | UErrorCode *status) { |
790 | RegularExpression *regexp = (RegularExpression*)regexp2; |
791 | if (validateRE(regexp, TRUE, status) == FALSE) { |
792 | return 0; |
793 | } |
794 | int64_t result = regexp->fMatcher->end64(groupNum, *status); |
795 | return result; |
796 | } |
797 | |
798 | //------------------------------------------------------------------------------ |
799 | // |
800 | // uregex_reset |
801 | // |
802 | //------------------------------------------------------------------------------ |
803 | U_CAPI void U_EXPORT2 |
804 | uregex_reset(URegularExpression *regexp2, |
805 | int32_t index, |
806 | UErrorCode *status) { |
807 | uregex_reset64( regexp2, (int64_t)index, status); |
808 | } |
809 | |
810 | U_CAPI void U_EXPORT2 |
811 | uregex_reset64(URegularExpression *regexp2, |
812 | int64_t index, |
813 | UErrorCode *status) { |
814 | RegularExpression *regexp = (RegularExpression*)regexp2; |
815 | if (validateRE(regexp, TRUE, status) == FALSE) { |
816 | return; |
817 | } |
818 | regexp->fMatcher->reset(index, *status); |
819 | } |
820 | |
821 | |
822 | //------------------------------------------------------------------------------ |
823 | // |
824 | // uregex_setRegion |
825 | // |
826 | //------------------------------------------------------------------------------ |
827 | U_CAPI void U_EXPORT2 |
828 | uregex_setRegion(URegularExpression *regexp2, |
829 | int32_t regionStart, |
830 | int32_t regionLimit, |
831 | UErrorCode *status) { |
832 | uregex_setRegion64( regexp2, (int64_t)regionStart, (int64_t)regionLimit, status); |
833 | } |
834 | |
835 | U_CAPI void U_EXPORT2 |
836 | uregex_setRegion64(URegularExpression *regexp2, |
837 | int64_t regionStart, |
838 | int64_t regionLimit, |
839 | UErrorCode *status) { |
840 | RegularExpression *regexp = (RegularExpression*)regexp2; |
841 | if (validateRE(regexp, TRUE, status) == FALSE) { |
842 | return; |
843 | } |
844 | regexp->fMatcher->region(regionStart, regionLimit, *status); |
845 | } |
846 | |
847 | |
848 | //------------------------------------------------------------------------------ |
849 | // |
850 | // uregex_setRegionAndStart |
851 | // |
852 | //------------------------------------------------------------------------------ |
853 | U_CAPI void U_EXPORT2 |
854 | uregex_setRegionAndStart(URegularExpression *regexp2, |
855 | int64_t regionStart, |
856 | int64_t regionLimit, |
857 | int64_t startIndex, |
858 | UErrorCode *status) { |
859 | RegularExpression *regexp = (RegularExpression*)regexp2; |
860 | if (validateRE(regexp, TRUE, status) == FALSE) { |
861 | return; |
862 | } |
863 | regexp->fMatcher->region(regionStart, regionLimit, startIndex, *status); |
864 | } |
865 | |
866 | //------------------------------------------------------------------------------ |
867 | // |
868 | // uregex_regionStart |
869 | // |
870 | //------------------------------------------------------------------------------ |
871 | U_CAPI int32_t U_EXPORT2 |
872 | uregex_regionStart(const URegularExpression *regexp2, |
873 | UErrorCode *status) { |
874 | return (int32_t)uregex_regionStart64(regexp2, status); |
875 | } |
876 | |
877 | U_CAPI int64_t U_EXPORT2 |
878 | uregex_regionStart64(const URegularExpression *regexp2, |
879 | UErrorCode *status) { |
880 | RegularExpression *regexp = (RegularExpression*)regexp2; |
881 | if (validateRE(regexp, TRUE, status) == FALSE) { |
882 | return 0; |
883 | } |
884 | return regexp->fMatcher->regionStart(); |
885 | } |
886 | |
887 | |
888 | //------------------------------------------------------------------------------ |
889 | // |
890 | // uregex_regionEnd |
891 | // |
892 | //------------------------------------------------------------------------------ |
893 | U_CAPI int32_t U_EXPORT2 |
894 | uregex_regionEnd(const URegularExpression *regexp2, |
895 | UErrorCode *status) { |
896 | return (int32_t)uregex_regionEnd64(regexp2, status); |
897 | } |
898 | |
899 | U_CAPI int64_t U_EXPORT2 |
900 | uregex_regionEnd64(const URegularExpression *regexp2, |
901 | UErrorCode *status) { |
902 | RegularExpression *regexp = (RegularExpression*)regexp2; |
903 | if (validateRE(regexp, TRUE, status) == FALSE) { |
904 | return 0; |
905 | } |
906 | return regexp->fMatcher->regionEnd(); |
907 | } |
908 | |
909 | |
910 | //------------------------------------------------------------------------------ |
911 | // |
912 | // uregex_hasTransparentBounds |
913 | // |
914 | //------------------------------------------------------------------------------ |
915 | U_CAPI UBool U_EXPORT2 |
916 | uregex_hasTransparentBounds(const URegularExpression *regexp2, |
917 | UErrorCode *status) { |
918 | RegularExpression *regexp = (RegularExpression*)regexp2; |
919 | if (validateRE(regexp, FALSE, status) == FALSE) { |
920 | return FALSE; |
921 | } |
922 | return regexp->fMatcher->hasTransparentBounds(); |
923 | } |
924 | |
925 | |
926 | //------------------------------------------------------------------------------ |
927 | // |
928 | // uregex_useTransparentBounds |
929 | // |
930 | //------------------------------------------------------------------------------ |
931 | U_CAPI void U_EXPORT2 |
932 | uregex_useTransparentBounds(URegularExpression *regexp2, |
933 | UBool b, |
934 | UErrorCode *status) { |
935 | RegularExpression *regexp = (RegularExpression*)regexp2; |
936 | if (validateRE(regexp, FALSE, status) == FALSE) { |
937 | return; |
938 | } |
939 | regexp->fMatcher->useTransparentBounds(b); |
940 | } |
941 | |
942 | |
943 | //------------------------------------------------------------------------------ |
944 | // |
945 | // uregex_hasAnchoringBounds |
946 | // |
947 | //------------------------------------------------------------------------------ |
948 | U_CAPI UBool U_EXPORT2 |
949 | uregex_hasAnchoringBounds(const URegularExpression *regexp2, |
950 | UErrorCode *status) { |
951 | RegularExpression *regexp = (RegularExpression*)regexp2; |
952 | if (validateRE(regexp, FALSE, status) == FALSE) { |
953 | return FALSE; |
954 | } |
955 | return regexp->fMatcher->hasAnchoringBounds(); |
956 | } |
957 | |
958 | |
959 | //------------------------------------------------------------------------------ |
960 | // |
961 | // uregex_useAnchoringBounds |
962 | // |
963 | //------------------------------------------------------------------------------ |
964 | U_CAPI void U_EXPORT2 |
965 | uregex_useAnchoringBounds(URegularExpression *regexp2, |
966 | UBool b, |
967 | UErrorCode *status) { |
968 | RegularExpression *regexp = (RegularExpression*)regexp2; |
969 | if (validateRE(regexp, FALSE, status) == FALSE) { |
970 | return; |
971 | } |
972 | regexp->fMatcher->useAnchoringBounds(b); |
973 | } |
974 | |
975 | |
976 | //------------------------------------------------------------------------------ |
977 | // |
978 | // uregex_hitEnd |
979 | // |
980 | //------------------------------------------------------------------------------ |
981 | U_CAPI UBool U_EXPORT2 |
982 | uregex_hitEnd(const URegularExpression *regexp2, |
983 | UErrorCode *status) { |
984 | RegularExpression *regexp = (RegularExpression*)regexp2; |
985 | if (validateRE(regexp, TRUE, status) == FALSE) { |
986 | return FALSE; |
987 | } |
988 | return regexp->fMatcher->hitEnd(); |
989 | } |
990 | |
991 | |
992 | //------------------------------------------------------------------------------ |
993 | // |
994 | // uregex_requireEnd |
995 | // |
996 | //------------------------------------------------------------------------------ |
997 | U_CAPI UBool U_EXPORT2 |
998 | uregex_requireEnd(const URegularExpression *regexp2, |
999 | UErrorCode *status) { |
1000 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1001 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1002 | return FALSE; |
1003 | } |
1004 | return regexp->fMatcher->requireEnd(); |
1005 | } |
1006 | |
1007 | |
1008 | //------------------------------------------------------------------------------ |
1009 | // |
1010 | // uregex_setTimeLimit |
1011 | // |
1012 | //------------------------------------------------------------------------------ |
1013 | U_CAPI void U_EXPORT2 |
1014 | uregex_setTimeLimit(URegularExpression *regexp2, |
1015 | int32_t limit, |
1016 | UErrorCode *status) { |
1017 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1018 | if (validateRE(regexp, FALSE, status)) { |
1019 | regexp->fMatcher->setTimeLimit(limit, *status); |
1020 | } |
1021 | } |
1022 | |
1023 | |
1024 | |
1025 | //------------------------------------------------------------------------------ |
1026 | // |
1027 | // uregex_getTimeLimit |
1028 | // |
1029 | //------------------------------------------------------------------------------ |
1030 | U_CAPI int32_t U_EXPORT2 |
1031 | uregex_getTimeLimit(const URegularExpression *regexp2, |
1032 | UErrorCode *status) { |
1033 | int32_t retVal = 0; |
1034 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1035 | if (validateRE(regexp, FALSE, status)) { |
1036 | retVal = regexp->fMatcher->getTimeLimit(); |
1037 | } |
1038 | return retVal; |
1039 | } |
1040 | |
1041 | |
1042 | |
1043 | //------------------------------------------------------------------------------ |
1044 | // |
1045 | // uregex_setStackLimit |
1046 | // |
1047 | //------------------------------------------------------------------------------ |
1048 | U_CAPI void U_EXPORT2 |
1049 | uregex_setStackLimit(URegularExpression *regexp2, |
1050 | int32_t limit, |
1051 | UErrorCode *status) { |
1052 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1053 | if (validateRE(regexp, FALSE, status)) { |
1054 | regexp->fMatcher->setStackLimit(limit, *status); |
1055 | } |
1056 | } |
1057 | |
1058 | |
1059 | |
1060 | //------------------------------------------------------------------------------ |
1061 | // |
1062 | // uregex_getStackLimit |
1063 | // |
1064 | //------------------------------------------------------------------------------ |
1065 | U_CAPI int32_t U_EXPORT2 |
1066 | uregex_getStackLimit(const URegularExpression *regexp2, |
1067 | UErrorCode *status) { |
1068 | int32_t retVal = 0; |
1069 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1070 | if (validateRE(regexp, FALSE, status)) { |
1071 | retVal = regexp->fMatcher->getStackLimit(); |
1072 | } |
1073 | return retVal; |
1074 | } |
1075 | |
1076 | |
1077 | //------------------------------------------------------------------------------ |
1078 | // |
1079 | // uregex_setMatchCallback |
1080 | // |
1081 | //------------------------------------------------------------------------------ |
1082 | U_CAPI void U_EXPORT2 |
1083 | uregex_setMatchCallback(URegularExpression *regexp2, |
1084 | URegexMatchCallback *callback, |
1085 | const void *context, |
1086 | UErrorCode *status) { |
1087 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1088 | if (validateRE(regexp, FALSE, status)) { |
1089 | regexp->fMatcher->setMatchCallback(callback, context, *status); |
1090 | } |
1091 | } |
1092 | |
1093 | |
1094 | //------------------------------------------------------------------------------ |
1095 | // |
1096 | // uregex_getMatchCallback |
1097 | // |
1098 | //------------------------------------------------------------------------------ |
1099 | U_CAPI void U_EXPORT2 |
1100 | uregex_getMatchCallback(const URegularExpression *regexp2, |
1101 | URegexMatchCallback **callback, |
1102 | const void **context, |
1103 | UErrorCode *status) { |
1104 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1105 | if (validateRE(regexp, FALSE, status)) { |
1106 | regexp->fMatcher->getMatchCallback(*callback, *context, *status); |
1107 | } |
1108 | } |
1109 | |
1110 | |
1111 | //------------------------------------------------------------------------------ |
1112 | // |
1113 | // uregex_setMatchProgressCallback |
1114 | // |
1115 | //------------------------------------------------------------------------------ |
1116 | U_CAPI void U_EXPORT2 |
1117 | uregex_setFindProgressCallback(URegularExpression *regexp2, |
1118 | URegexFindProgressCallback *callback, |
1119 | const void *context, |
1120 | UErrorCode *status) { |
1121 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1122 | if (validateRE(regexp, FALSE, status)) { |
1123 | regexp->fMatcher->setFindProgressCallback(callback, context, *status); |
1124 | } |
1125 | } |
1126 | |
1127 | |
1128 | //------------------------------------------------------------------------------ |
1129 | // |
1130 | // uregex_getMatchCallback |
1131 | // |
1132 | //------------------------------------------------------------------------------ |
1133 | U_CAPI void U_EXPORT2 |
1134 | uregex_getFindProgressCallback(const URegularExpression *regexp2, |
1135 | URegexFindProgressCallback **callback, |
1136 | const void **context, |
1137 | UErrorCode *status) { |
1138 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1139 | if (validateRE(regexp, FALSE, status)) { |
1140 | regexp->fMatcher->getFindProgressCallback(*callback, *context, *status); |
1141 | } |
1142 | } |
1143 | |
1144 | |
1145 | //------------------------------------------------------------------------------ |
1146 | // |
1147 | // uregex_replaceAll |
1148 | // |
1149 | //------------------------------------------------------------------------------ |
1150 | U_CAPI int32_t U_EXPORT2 |
1151 | uregex_replaceAll(URegularExpression *regexp2, |
1152 | const UChar *replacementText, |
1153 | int32_t replacementLength, |
1154 | UChar *destBuf, |
1155 | int32_t destCapacity, |
1156 | UErrorCode *status) { |
1157 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1158 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1159 | return 0; |
1160 | } |
1161 | if (replacementText == NULL || replacementLength < -1 || |
1162 | (destBuf == NULL && destCapacity > 0) || |
1163 | destCapacity < 0) { |
1164 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
1165 | return 0; |
1166 | } |
1167 | |
1168 | int32_t len = 0; |
1169 | |
1170 | uregex_reset(regexp2, 0, status); |
1171 | |
1172 | // Note: Seperate error code variables for findNext() and appendReplacement() |
1173 | // are used so that destination buffer overflow errors |
1174 | // in appendReplacement won't stop findNext() from working. |
1175 | // appendReplacement() and appendTail() special case incoming buffer |
1176 | // overflow errors, continuing to return the correct length. |
1177 | UErrorCode findStatus = *status; |
1178 | while (uregex_findNext(regexp2, &findStatus)) { |
1179 | len += uregex_appendReplacement(regexp2, replacementText, replacementLength, |
1180 | &destBuf, &destCapacity, status); |
1181 | } |
1182 | len += uregex_appendTail(regexp2, &destBuf, &destCapacity, status); |
1183 | |
1184 | if (U_FAILURE(findStatus)) { |
1185 | // If anything went wrong with the findNext(), make that error trump |
1186 | // whatever may have happened with the append() operations. |
1187 | // Errors in findNext() are not expected. |
1188 | *status = findStatus; |
1189 | } |
1190 | |
1191 | return len; |
1192 | } |
1193 | |
1194 | |
1195 | //------------------------------------------------------------------------------ |
1196 | // |
1197 | // uregex_replaceAllUText |
1198 | // |
1199 | //------------------------------------------------------------------------------ |
1200 | U_CAPI UText * U_EXPORT2 |
1201 | uregex_replaceAllUText(URegularExpression *regexp2, |
1202 | UText *replacementText, |
1203 | UText *dest, |
1204 | UErrorCode *status) { |
1205 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1206 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1207 | return 0; |
1208 | } |
1209 | if (replacementText == NULL) { |
1210 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
1211 | return 0; |
1212 | } |
1213 | |
1214 | dest = regexp->fMatcher->replaceAll(replacementText, dest, *status); |
1215 | return dest; |
1216 | } |
1217 | |
1218 | |
1219 | //------------------------------------------------------------------------------ |
1220 | // |
1221 | // uregex_replaceFirst |
1222 | // |
1223 | //------------------------------------------------------------------------------ |
1224 | U_CAPI int32_t U_EXPORT2 |
1225 | uregex_replaceFirst(URegularExpression *regexp2, |
1226 | const UChar *replacementText, |
1227 | int32_t replacementLength, |
1228 | UChar *destBuf, |
1229 | int32_t destCapacity, |
1230 | UErrorCode *status) { |
1231 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1232 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1233 | return 0; |
1234 | } |
1235 | if (replacementText == NULL || replacementLength < -1 || |
1236 | (destBuf == NULL && destCapacity > 0) || |
1237 | destCapacity < 0) { |
1238 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
1239 | return 0; |
1240 | } |
1241 | |
1242 | int32_t len = 0; |
1243 | UBool findSucceeded; |
1244 | uregex_reset(regexp2, 0, status); |
1245 | findSucceeded = uregex_find(regexp2, 0, status); |
1246 | if (findSucceeded) { |
1247 | len = uregex_appendReplacement(regexp2, replacementText, replacementLength, |
1248 | &destBuf, &destCapacity, status); |
1249 | } |
1250 | len += uregex_appendTail(regexp2, &destBuf, &destCapacity, status); |
1251 | |
1252 | return len; |
1253 | } |
1254 | |
1255 | |
1256 | //------------------------------------------------------------------------------ |
1257 | // |
1258 | // uregex_replaceFirstUText |
1259 | // |
1260 | //------------------------------------------------------------------------------ |
1261 | U_CAPI UText * U_EXPORT2 |
1262 | uregex_replaceFirstUText(URegularExpression *regexp2, |
1263 | UText *replacementText, |
1264 | UText *dest, |
1265 | UErrorCode *status) { |
1266 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1267 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1268 | return 0; |
1269 | } |
1270 | if (replacementText == NULL) { |
1271 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
1272 | return 0; |
1273 | } |
1274 | |
1275 | dest = regexp->fMatcher->replaceFirst(replacementText, dest, *status); |
1276 | return dest; |
1277 | } |
1278 | |
1279 | |
1280 | //------------------------------------------------------------------------------ |
1281 | // |
1282 | // uregex_appendReplacement |
1283 | // |
1284 | //------------------------------------------------------------------------------ |
1285 | |
1286 | U_NAMESPACE_BEGIN |
1287 | // |
1288 | // Dummy class, because these functions need to be friends of class RegexMatcher, |
1289 | // and stand-alone C functions don't work as friends |
1290 | // |
1291 | class RegexCImpl { |
1292 | public: |
1293 | inline static int32_t appendReplacement(RegularExpression *regexp, |
1294 | const UChar *replacementText, |
1295 | int32_t replacementLength, |
1296 | UChar **destBuf, |
1297 | int32_t *destCapacity, |
1298 | UErrorCode *status); |
1299 | |
1300 | inline static int32_t appendTail(RegularExpression *regexp, |
1301 | UChar **destBuf, |
1302 | int32_t *destCapacity, |
1303 | UErrorCode *status); |
1304 | |
1305 | inline static int32_t split(RegularExpression *regexp, |
1306 | UChar *destBuf, |
1307 | int32_t destCapacity, |
1308 | int32_t *requiredCapacity, |
1309 | UChar *destFields[], |
1310 | int32_t destFieldsCapacity, |
1311 | UErrorCode *status); |
1312 | }; |
1313 | |
1314 | U_NAMESPACE_END |
1315 | |
1316 | |
1317 | |
1318 | static const UChar BACKSLASH = 0x5c; |
1319 | static const UChar DOLLARSIGN = 0x24; |
1320 | static const UChar LEFTBRACKET = 0x7b; |
1321 | static const UChar RIGHTBRACKET = 0x7d; |
1322 | |
1323 | // |
1324 | // Move a character to an output buffer, with bounds checking on the index. |
1325 | // Index advances even if capacity is exceeded, for preflight size computations. |
1326 | // This little sequence is used a LOT. |
1327 | // |
1328 | static inline void appendToBuf(UChar c, int32_t *idx, UChar *buf, int32_t bufCapacity) { |
1329 | if (*idx < bufCapacity) { |
1330 | buf[*idx] = c; |
1331 | } |
1332 | (*idx)++; |
1333 | } |
1334 | |
1335 | |
1336 | // |
1337 | // appendReplacement, the actual implementation. |
1338 | // |
1339 | int32_t RegexCImpl::appendReplacement(RegularExpression *regexp, |
1340 | const UChar *replacementText, |
1341 | int32_t replacementLength, |
1342 | UChar **destBuf, |
1343 | int32_t *destCapacity, |
1344 | UErrorCode *status) { |
1345 | |
1346 | // If we come in with a buffer overflow error, don't suppress the operation. |
1347 | // A series of appendReplacements, appendTail need to correctly preflight |
1348 | // the buffer size when an overflow happens somewhere in the middle. |
1349 | UBool pendingBufferOverflow = FALSE; |
1350 | if (*status == U_BUFFER_OVERFLOW_ERROR && destCapacity != NULL && *destCapacity == 0) { |
1351 | pendingBufferOverflow = TRUE; |
1352 | *status = U_ZERO_ERROR; |
1353 | } |
1354 | |
1355 | // |
1356 | // Validate all paramters |
1357 | // |
1358 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1359 | return 0; |
1360 | } |
1361 | if (replacementText == NULL || replacementLength < -1 || |
1362 | destCapacity == NULL || destBuf == NULL || |
1363 | (*destBuf == NULL && *destCapacity > 0) || |
1364 | *destCapacity < 0) { |
1365 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
1366 | return 0; |
1367 | } |
1368 | |
1369 | RegexMatcher *m = regexp->fMatcher; |
1370 | if (m->fMatch == FALSE) { |
1371 | *status = U_REGEX_INVALID_STATE; |
1372 | return 0; |
1373 | } |
1374 | |
1375 | UChar *dest = *destBuf; |
1376 | int32_t capacity = *destCapacity; |
1377 | int32_t destIdx = 0; |
1378 | int32_t i; |
1379 | |
1380 | // If it wasn't supplied by the caller, get the length of the replacement text. |
1381 | // TODO: slightly smarter logic in the copy loop could watch for the NUL on |
1382 | // the fly and avoid this step. |
1383 | if (replacementLength == -1) { |
1384 | replacementLength = u_strlen(replacementText); |
1385 | } |
1386 | |
1387 | // Copy input string from the end of previous match to start of current match |
1388 | if (regexp->fText != NULL) { |
1389 | int32_t matchStart; |
1390 | int32_t lastMatchEnd; |
1391 | if (UTEXT_USES_U16(m->fInputText)) { |
1392 | lastMatchEnd = (int32_t)m->fLastMatchEnd; |
1393 | matchStart = (int32_t)m->fMatchStart; |
1394 | } else { |
1395 | // !!!: Would like a better way to do this! |
1396 | UErrorCode tempStatus = U_ZERO_ERROR; |
1397 | lastMatchEnd = utext_extract(m->fInputText, 0, m->fLastMatchEnd, NULL, 0, &tempStatus); |
1398 | tempStatus = U_ZERO_ERROR; |
1399 | matchStart = lastMatchEnd + utext_extract(m->fInputText, m->fLastMatchEnd, m->fMatchStart, NULL, 0, &tempStatus); |
1400 | } |
1401 | for (i=lastMatchEnd; i<matchStart; i++) { |
1402 | appendToBuf(regexp->fText[i], &destIdx, dest, capacity); |
1403 | } |
1404 | } else { |
1405 | UErrorCode possibleOverflowError = U_ZERO_ERROR; // ignore |
1406 | destIdx += utext_extract(m->fInputText, m->fLastMatchEnd, m->fMatchStart, |
1407 | dest==NULL?NULL:&dest[destIdx], REMAINING_CAPACITY(destIdx, capacity), |
1408 | &possibleOverflowError); |
1409 | } |
1410 | U_ASSERT(destIdx >= 0); |
1411 | |
1412 | // scan the replacement text, looking for substitutions ($n) and \escapes. |
1413 | int32_t replIdx = 0; |
1414 | while (replIdx < replacementLength && U_SUCCESS(*status)) { |
1415 | UChar c = replacementText[replIdx]; |
1416 | replIdx++; |
1417 | if (c != DOLLARSIGN && c != BACKSLASH) { |
1418 | // Common case, no substitution, no escaping, |
1419 | // just copy the char to the dest buf. |
1420 | appendToBuf(c, &destIdx, dest, capacity); |
1421 | continue; |
1422 | } |
1423 | |
1424 | if (c == BACKSLASH) { |
1425 | // Backslash Escape. Copy the following char out without further checks. |
1426 | // Note: Surrogate pairs don't need any special handling |
1427 | // The second half wont be a '$' or a '\', and |
1428 | // will move to the dest normally on the next |
1429 | // loop iteration. |
1430 | if (replIdx >= replacementLength) { |
1431 | break; |
1432 | } |
1433 | c = replacementText[replIdx]; |
1434 | |
1435 | if (c==0x55/*U*/ || c==0x75/*u*/) { |
1436 | // We have a \udddd or \Udddddddd escape sequence. |
1437 | UChar32 escapedChar = |
1438 | u_unescapeAt(uregex_ucstr_unescape_charAt, |
1439 | &replIdx, // Index is updated by unescapeAt |
1440 | replacementLength, // Length of replacement text |
1441 | (void *)replacementText); |
1442 | |
1443 | if (escapedChar != (UChar32)0xFFFFFFFF) { |
1444 | if (escapedChar <= 0xffff) { |
1445 | appendToBuf((UChar)escapedChar, &destIdx, dest, capacity); |
1446 | } else { |
1447 | appendToBuf(U16_LEAD(escapedChar), &destIdx, dest, capacity); |
1448 | appendToBuf(U16_TRAIL(escapedChar), &destIdx, dest, capacity); |
1449 | } |
1450 | continue; |
1451 | } |
1452 | // Note: if the \u escape was invalid, just fall through and |
1453 | // treat it as a plain \<anything> escape. |
1454 | } |
1455 | |
1456 | // Plain backslash escape. Just put out the escaped character. |
1457 | appendToBuf(c, &destIdx, dest, capacity); |
1458 | |
1459 | replIdx++; |
1460 | continue; |
1461 | } |
1462 | |
1463 | // We've got a $. Pick up the following capture group name or number. |
1464 | // For numbers, consume only digits that produce a valid capture group for the pattern. |
1465 | |
1466 | int32_t groupNum = 0; |
1467 | U_ASSERT(c == DOLLARSIGN); |
1468 | UChar32 c32 = -1; |
1469 | if (replIdx < replacementLength) { |
1470 | U16_GET(replacementText, 0, replIdx, replacementLength, c32); |
1471 | } |
1472 | if (u_isdigit(c32)) { |
1473 | int32_t numDigits = 0; |
1474 | int32_t numCaptureGroups = m->fPattern->fGroupMap->size(); |
1475 | for (;;) { |
1476 | if (replIdx >= replacementLength) { |
1477 | break; |
1478 | } |
1479 | U16_GET(replacementText, 0, replIdx, replacementLength, c32); |
1480 | if (u_isdigit(c32) == FALSE) { |
1481 | break; |
1482 | } |
1483 | |
1484 | int32_t digitVal = u_charDigitValue(c32); |
1485 | if (groupNum * 10 + digitVal <= numCaptureGroups) { |
1486 | groupNum = groupNum * 10 + digitVal; |
1487 | U16_FWD_1(replacementText, replIdx, replacementLength); |
1488 | numDigits++; |
1489 | } else { |
1490 | if (numDigits == 0) { |
1491 | *status = U_INDEX_OUTOFBOUNDS_ERROR; |
1492 | } |
1493 | break; |
1494 | } |
1495 | } |
1496 | } else if (c32 == LEFTBRACKET) { |
1497 | // Scan for Named Capture Group, ${name}. |
1498 | UnicodeString groupName; |
1499 | U16_FWD_1(replacementText, replIdx, replacementLength); |
1500 | while (U_SUCCESS(*status) && c32 != RIGHTBRACKET) { |
1501 | if (replIdx >= replacementLength) { |
1502 | *status = U_REGEX_INVALID_CAPTURE_GROUP_NAME; |
1503 | break; |
1504 | } |
1505 | U16_NEXT(replacementText, replIdx, replacementLength, c32); |
1506 | if ((c32 >= 0x41 && c32 <= 0x5a) || // A..Z |
1507 | (c32 >= 0x61 && c32 <= 0x7a) || // a..z |
1508 | (c32 >= 0x31 && c32 <= 0x39)) { // 0..9 |
1509 | groupName.append(c32); |
1510 | } else if (c32 == RIGHTBRACKET) { |
1511 | groupNum = regexp->fPat->fNamedCaptureMap ? |
1512 | uhash_geti(regexp->fPat->fNamedCaptureMap, &groupName) : 0; |
1513 | if (groupNum == 0) { |
1514 | // Name not defined by pattern. |
1515 | *status = U_REGEX_INVALID_CAPTURE_GROUP_NAME; |
1516 | } |
1517 | } else { |
1518 | // Character was something other than a name char or a closing '}' |
1519 | *status = U_REGEX_INVALID_CAPTURE_GROUP_NAME; |
1520 | } |
1521 | } |
1522 | } else { |
1523 | // $ not followed by {name} or digits. |
1524 | *status = U_REGEX_INVALID_CAPTURE_GROUP_NAME; |
1525 | } |
1526 | |
1527 | |
1528 | // Finally, append the capture group data to the destination. |
1529 | if (U_SUCCESS(*status)) { |
1530 | destIdx += uregex_group((URegularExpression*)regexp, groupNum, |
1531 | dest==NULL?NULL:&dest[destIdx], REMAINING_CAPACITY(destIdx, capacity), status); |
1532 | if (*status == U_BUFFER_OVERFLOW_ERROR) { |
1533 | // Ignore buffer overflow when extracting the group. We need to |
1534 | // continue on to get full size of the untruncated result. We will |
1535 | // raise our own buffer overflow error at the end. |
1536 | *status = U_ZERO_ERROR; |
1537 | } |
1538 | } |
1539 | |
1540 | if (U_FAILURE(*status)) { |
1541 | // bad group number or name. |
1542 | break; |
1543 | } |
1544 | } |
1545 | |
1546 | // |
1547 | // Nul Terminate the dest buffer if possible. |
1548 | // Set the appropriate buffer overflow or not terminated error, if needed. |
1549 | // |
1550 | if (destIdx < capacity) { |
1551 | dest[destIdx] = 0; |
1552 | } else if (U_SUCCESS(*status)) { |
1553 | if (destIdx == *destCapacity) { |
1554 | *status = U_STRING_NOT_TERMINATED_WARNING; |
1555 | } else { |
1556 | *status = U_BUFFER_OVERFLOW_ERROR; |
1557 | } |
1558 | } |
1559 | |
1560 | // |
1561 | // Return an updated dest buffer and capacity to the caller. |
1562 | // |
1563 | if (destIdx > 0 && *destCapacity > 0) { |
1564 | if (destIdx < capacity) { |
1565 | *destBuf += destIdx; |
1566 | *destCapacity -= destIdx; |
1567 | } else { |
1568 | *destBuf += capacity; |
1569 | *destCapacity = 0; |
1570 | } |
1571 | } |
1572 | |
1573 | // If we came in with a buffer overflow, make sure we go out with one also. |
1574 | // (A zero length match right at the end of the previous match could |
1575 | // make this function succeed even though a previous call had overflowed the buf) |
1576 | if (pendingBufferOverflow && U_SUCCESS(*status)) { |
1577 | *status = U_BUFFER_OVERFLOW_ERROR; |
1578 | } |
1579 | |
1580 | return destIdx; |
1581 | } |
1582 | |
1583 | // |
1584 | // appendReplacement the actual API function, |
1585 | // |
1586 | U_CAPI int32_t U_EXPORT2 |
1587 | uregex_appendReplacement(URegularExpression *regexp2, |
1588 | const UChar *replacementText, |
1589 | int32_t replacementLength, |
1590 | UChar **destBuf, |
1591 | int32_t *destCapacity, |
1592 | UErrorCode *status) { |
1593 | |
1594 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1595 | return RegexCImpl::appendReplacement( |
1596 | regexp, replacementText, replacementLength,destBuf, destCapacity, status); |
1597 | } |
1598 | |
1599 | // |
1600 | // uregex_appendReplacementUText...can just use the normal C++ method |
1601 | // |
1602 | U_CAPI void U_EXPORT2 |
1603 | uregex_appendReplacementUText(URegularExpression *regexp2, |
1604 | UText *replText, |
1605 | UText *dest, |
1606 | UErrorCode *status) { |
1607 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1608 | regexp->fMatcher->appendReplacement(dest, replText, *status); |
1609 | } |
1610 | |
1611 | |
1612 | //------------------------------------------------------------------------------ |
1613 | // |
1614 | // uregex_appendTail |
1615 | // |
1616 | //------------------------------------------------------------------------------ |
1617 | int32_t RegexCImpl::appendTail(RegularExpression *regexp, |
1618 | UChar **destBuf, |
1619 | int32_t *destCapacity, |
1620 | UErrorCode *status) |
1621 | { |
1622 | |
1623 | // If we come in with a buffer overflow error, don't suppress the operation. |
1624 | // A series of appendReplacements, appendTail need to correctly preflight |
1625 | // the buffer size when an overflow happens somewhere in the middle. |
1626 | UBool pendingBufferOverflow = FALSE; |
1627 | if (*status == U_BUFFER_OVERFLOW_ERROR && destCapacity != NULL && *destCapacity == 0) { |
1628 | pendingBufferOverflow = TRUE; |
1629 | *status = U_ZERO_ERROR; |
1630 | } |
1631 | |
1632 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1633 | return 0; |
1634 | } |
1635 | |
1636 | if (destCapacity == NULL || destBuf == NULL || |
1637 | (*destBuf == NULL && *destCapacity > 0) || |
1638 | *destCapacity < 0) |
1639 | { |
1640 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
1641 | return 0; |
1642 | } |
1643 | |
1644 | RegexMatcher *m = regexp->fMatcher; |
1645 | |
1646 | int32_t destIdx = 0; |
1647 | int32_t destCap = *destCapacity; |
1648 | UChar *dest = *destBuf; |
1649 | |
1650 | if (regexp->fText != NULL) { |
1651 | int32_t srcIdx; |
1652 | int64_t nativeIdx = (m->fMatch ? m->fMatchEnd : m->fLastMatchEnd); |
1653 | if (nativeIdx == -1) { |
1654 | srcIdx = 0; |
1655 | } else if (UTEXT_USES_U16(m->fInputText)) { |
1656 | srcIdx = (int32_t)nativeIdx; |
1657 | } else { |
1658 | UErrorCode newStatus = U_ZERO_ERROR; |
1659 | srcIdx = utext_extract(m->fInputText, 0, nativeIdx, NULL, 0, &newStatus); |
1660 | } |
1661 | |
1662 | for (;;) { |
1663 | U_ASSERT(destIdx >= 0); |
1664 | |
1665 | if (srcIdx == regexp->fTextLength) { |
1666 | break; |
1667 | } |
1668 | UChar c = regexp->fText[srcIdx]; |
1669 | if (c == 0 && regexp->fTextLength == -1) { |
1670 | regexp->fTextLength = srcIdx; |
1671 | break; |
1672 | } |
1673 | |
1674 | if (destIdx < destCap) { |
1675 | dest[destIdx] = c; |
1676 | } else { |
1677 | // We've overflowed the dest buffer. |
1678 | // If the total input string length is known, we can |
1679 | // compute the total buffer size needed without scanning through the string. |
1680 | if (regexp->fTextLength > 0) { |
1681 | destIdx += (regexp->fTextLength - srcIdx); |
1682 | break; |
1683 | } |
1684 | } |
1685 | srcIdx++; |
1686 | destIdx++; |
1687 | } |
1688 | } else { |
1689 | int64_t srcIdx; |
1690 | if (m->fMatch) { |
1691 | // The most recent call to find() succeeded. |
1692 | srcIdx = m->fMatchEnd; |
1693 | } else { |
1694 | // The last call to find() on this matcher failed(). |
1695 | // Look back to the end of the last find() that succeeded for src index. |
1696 | srcIdx = m->fLastMatchEnd; |
1697 | if (srcIdx == -1) { |
1698 | // There has been no successful match with this matcher. |
1699 | // We want to copy the whole string. |
1700 | srcIdx = 0; |
1701 | } |
1702 | } |
1703 | |
1704 | destIdx = utext_extract(m->fInputText, srcIdx, m->fInputLength, dest, destCap, status); |
1705 | } |
1706 | |
1707 | // |
1708 | // NUL terminate the output string, if possible, otherwise issue the |
1709 | // appropriate error or warning. |
1710 | // |
1711 | if (destIdx < destCap) { |
1712 | dest[destIdx] = 0; |
1713 | } else if (destIdx == destCap) { |
1714 | *status = U_STRING_NOT_TERMINATED_WARNING; |
1715 | } else { |
1716 | *status = U_BUFFER_OVERFLOW_ERROR; |
1717 | } |
1718 | |
1719 | // |
1720 | // Update the user's buffer ptr and capacity vars to reflect the |
1721 | // amount used. |
1722 | // |
1723 | if (destIdx < destCap) { |
1724 | *destBuf += destIdx; |
1725 | *destCapacity -= destIdx; |
1726 | } else if (*destBuf != NULL) { |
1727 | *destBuf += destCap; |
1728 | *destCapacity = 0; |
1729 | } |
1730 | |
1731 | if (pendingBufferOverflow && U_SUCCESS(*status)) { |
1732 | *status = U_BUFFER_OVERFLOW_ERROR; |
1733 | } |
1734 | |
1735 | return destIdx; |
1736 | } |
1737 | |
1738 | |
1739 | // |
1740 | // appendTail the actual API function |
1741 | // |
1742 | U_CAPI int32_t U_EXPORT2 |
1743 | uregex_appendTail(URegularExpression *regexp2, |
1744 | UChar **destBuf, |
1745 | int32_t *destCapacity, |
1746 | UErrorCode *status) { |
1747 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1748 | return RegexCImpl::appendTail(regexp, destBuf, destCapacity, status); |
1749 | } |
1750 | |
1751 | |
1752 | // |
1753 | // uregex_appendTailUText...can just use the normal C++ method |
1754 | // |
1755 | U_CAPI UText * U_EXPORT2 |
1756 | uregex_appendTailUText(URegularExpression *regexp2, |
1757 | UText *dest, |
1758 | UErrorCode *status) { |
1759 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1760 | return regexp->fMatcher->appendTail(dest, *status); |
1761 | } |
1762 | |
1763 | |
1764 | //------------------------------------------------------------------------------ |
1765 | // |
1766 | // copyString Internal utility to copy a string to an output buffer, |
1767 | // while managing buffer overflow and preflight size |
1768 | // computation. NUL termination is added to destination, |
1769 | // and the NUL is counted in the output size. |
1770 | // |
1771 | //------------------------------------------------------------------------------ |
1772 | #if 0 |
1773 | static void copyString(UChar *destBuffer, // Destination buffer. |
1774 | int32_t destCapacity, // Total capacity of dest buffer |
1775 | int32_t *destIndex, // Index into dest buffer. Updated on return. |
1776 | // Update not clipped to destCapacity. |
1777 | const UChar *srcPtr, // Pointer to source string |
1778 | int32_t srcLen) // Source string len. |
1779 | { |
1780 | int32_t si; |
1781 | int32_t di = *destIndex; |
1782 | UChar c; |
1783 | |
1784 | for (si=0; si<srcLen; si++) { |
1785 | c = srcPtr[si]; |
1786 | if (di < destCapacity) { |
1787 | destBuffer[di] = c; |
1788 | di++; |
1789 | } else { |
1790 | di += srcLen - si; |
1791 | break; |
1792 | } |
1793 | } |
1794 | if (di<destCapacity) { |
1795 | destBuffer[di] = 0; |
1796 | } |
1797 | di++; |
1798 | *destIndex = di; |
1799 | } |
1800 | #endif |
1801 | |
1802 | //------------------------------------------------------------------------------ |
1803 | // |
1804 | // uregex_split |
1805 | // |
1806 | //------------------------------------------------------------------------------ |
1807 | int32_t RegexCImpl::split(RegularExpression *regexp, |
1808 | UChar *destBuf, |
1809 | int32_t destCapacity, |
1810 | int32_t *requiredCapacity, |
1811 | UChar *destFields[], |
1812 | int32_t destFieldsCapacity, |
1813 | UErrorCode *status) { |
1814 | // |
1815 | // Reset for the input text |
1816 | // |
1817 | regexp->fMatcher->reset(); |
1818 | UText *inputText = regexp->fMatcher->fInputText; |
1819 | int64_t nextOutputStringStart = 0; |
1820 | int64_t inputLen = regexp->fMatcher->fInputLength; |
1821 | if (inputLen == 0) { |
1822 | return 0; |
1823 | } |
1824 | |
1825 | // |
1826 | // Loop through the input text, searching for the delimiter pattern |
1827 | // |
1828 | int32_t i; // Index of the field being processed. |
1829 | int32_t destIdx = 0; // Next available position in destBuf; |
1830 | int32_t numCaptureGroups = regexp->fMatcher->groupCount(); |
1831 | UErrorCode tStatus = U_ZERO_ERROR; // Want to ignore any buffer overflow errors so that the strings are still counted |
1832 | for (i=0; ; i++) { |
1833 | if (i>=destFieldsCapacity-1) { |
1834 | // There are one or zero output strings left. |
1835 | // Fill the last output string with whatever is left from the input, then exit the loop. |
1836 | // ( i will be == destFieldsCapacity if we filled the output array while processing |
1837 | // capture groups of the delimiter expression, in which case we will discard the |
1838 | // last capture group saved in favor of the unprocessed remainder of the |
1839 | // input string.) |
1840 | if (inputLen > nextOutputStringStart) { |
1841 | if (i != destFieldsCapacity-1) { |
1842 | // No fields are left. Recycle the last one for holding the trailing part of |
1843 | // the input string. |
1844 | i = destFieldsCapacity-1; |
1845 | destIdx = (int32_t)(destFields[i] - destFields[0]); |
1846 | } |
1847 | |
1848 | destFields[i] = &destBuf[destIdx]; |
1849 | destIdx += 1 + utext_extract(inputText, nextOutputStringStart, inputLen, |
1850 | &destBuf[destIdx], REMAINING_CAPACITY(destIdx, destCapacity), status); |
1851 | } |
1852 | break; |
1853 | } |
1854 | |
1855 | if (regexp->fMatcher->find()) { |
1856 | // We found another delimiter. Move everything from where we started looking |
1857 | // up until the start of the delimiter into the next output string. |
1858 | destFields[i] = &destBuf[destIdx]; |
1859 | |
1860 | destIdx += 1 + utext_extract(inputText, nextOutputStringStart, regexp->fMatcher->fMatchStart, |
1861 | &destBuf[destIdx], REMAINING_CAPACITY(destIdx, destCapacity), &tStatus); |
1862 | if (tStatus == U_BUFFER_OVERFLOW_ERROR) { |
1863 | tStatus = U_ZERO_ERROR; |
1864 | } else { |
1865 | *status = tStatus; |
1866 | } |
1867 | nextOutputStringStart = regexp->fMatcher->fMatchEnd; |
1868 | |
1869 | // If the delimiter pattern has capturing parentheses, the captured |
1870 | // text goes out into the next n destination strings. |
1871 | int32_t groupNum; |
1872 | for (groupNum=1; groupNum<=numCaptureGroups; groupNum++) { |
1873 | // If we've run out of output string slots, bail out. |
1874 | if (i==destFieldsCapacity-1) { |
1875 | break; |
1876 | } |
1877 | i++; |
1878 | |
1879 | // Set up to extract the capture group contents into the dest buffer. |
1880 | destFields[i] = &destBuf[destIdx]; |
1881 | tStatus = U_ZERO_ERROR; |
1882 | int32_t t = uregex_group((URegularExpression*)regexp, |
1883 | groupNum, |
1884 | destFields[i], |
1885 | REMAINING_CAPACITY(destIdx, destCapacity), |
1886 | &tStatus); |
1887 | destIdx += t + 1; // Record the space used in the output string buffer. |
1888 | // +1 for the NUL that terminates the string. |
1889 | if (tStatus == U_BUFFER_OVERFLOW_ERROR) { |
1890 | tStatus = U_ZERO_ERROR; |
1891 | } else { |
1892 | *status = tStatus; |
1893 | } |
1894 | } |
1895 | |
1896 | if (nextOutputStringStart == inputLen) { |
1897 | // The delimiter was at the end of the string. |
1898 | // Output an empty string, and then we are done. |
1899 | if (destIdx < destCapacity) { |
1900 | destBuf[destIdx] = 0; |
1901 | } |
1902 | if (i < destFieldsCapacity-1) { |
1903 | ++i; |
1904 | } |
1905 | if (destIdx < destCapacity) { |
1906 | destFields[i] = destBuf + destIdx; |
1907 | } |
1908 | ++destIdx; |
1909 | break; |
1910 | } |
1911 | |
1912 | } |
1913 | else |
1914 | { |
1915 | // We ran off the end of the input while looking for the next delimiter. |
1916 | // All the remaining text goes into the current output string. |
1917 | destFields[i] = &destBuf[destIdx]; |
1918 | destIdx += 1 + utext_extract(inputText, nextOutputStringStart, inputLen, |
1919 | &destBuf[destIdx], REMAINING_CAPACITY(destIdx, destCapacity), status); |
1920 | break; |
1921 | } |
1922 | } |
1923 | |
1924 | // Zero out any unused portion of the destFields array |
1925 | int j; |
1926 | for (j=i+1; j<destFieldsCapacity; j++) { |
1927 | destFields[j] = NULL; |
1928 | } |
1929 | |
1930 | if (requiredCapacity != NULL) { |
1931 | *requiredCapacity = destIdx; |
1932 | } |
1933 | if (destIdx > destCapacity) { |
1934 | *status = U_BUFFER_OVERFLOW_ERROR; |
1935 | } |
1936 | return i+1; |
1937 | } |
1938 | |
1939 | // |
1940 | // uregex_split The actual API function |
1941 | // |
1942 | U_CAPI int32_t U_EXPORT2 |
1943 | uregex_split(URegularExpression *regexp2, |
1944 | UChar *destBuf, |
1945 | int32_t destCapacity, |
1946 | int32_t *requiredCapacity, |
1947 | UChar *destFields[], |
1948 | int32_t destFieldsCapacity, |
1949 | UErrorCode *status) { |
1950 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1951 | if (validateRE(regexp, TRUE, status) == FALSE) { |
1952 | return 0; |
1953 | } |
1954 | if ((destBuf == NULL && destCapacity > 0) || |
1955 | destCapacity < 0 || |
1956 | destFields == NULL || |
1957 | destFieldsCapacity < 1 ) { |
1958 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
1959 | return 0; |
1960 | } |
1961 | |
1962 | return RegexCImpl::split(regexp, destBuf, destCapacity, requiredCapacity, destFields, destFieldsCapacity, status); |
1963 | } |
1964 | |
1965 | |
1966 | // |
1967 | // uregex_splitUText...can just use the normal C++ method |
1968 | // |
1969 | U_CAPI int32_t U_EXPORT2 |
1970 | uregex_splitUText(URegularExpression *regexp2, |
1971 | UText *destFields[], |
1972 | int32_t destFieldsCapacity, |
1973 | UErrorCode *status) { |
1974 | RegularExpression *regexp = (RegularExpression*)regexp2; |
1975 | return regexp->fMatcher->split(regexp->fMatcher->inputText(), destFields, destFieldsCapacity, *status); |
1976 | } |
1977 | |
1978 | |
1979 | #endif // !UCONFIG_NO_REGULAR_EXPRESSIONS |
1980 | |
1981 | |