1 | // © 2016 and later: Unicode, Inc. and others. |
2 | // License & terms of use: http://www.unicode.org/copyright.html |
3 | /* |
4 | ********************************************************************** |
5 | * Copyright (C) 2001-2014 IBM and others. All rights reserved. |
6 | ********************************************************************** |
7 | * Date Name Description |
8 | * 03/22/2000 helena Creation. |
9 | ********************************************************************** |
10 | */ |
11 | |
12 | #include "unicode/utypes.h" |
13 | |
14 | #if !UCONFIG_NO_COLLATION && !UCONFIG_NO_BREAK_ITERATION |
15 | |
16 | #include "unicode/stsearch.h" |
17 | #include "usrchimp.h" |
18 | #include "cmemory.h" |
19 | |
20 | U_NAMESPACE_BEGIN |
21 | |
22 | UOBJECT_DEFINE_RTTI_IMPLEMENTATION(StringSearch) |
23 | |
24 | // public constructors and destructors ----------------------------------- |
25 | |
26 | StringSearch::StringSearch(const UnicodeString &pattern, |
27 | const UnicodeString &text, |
28 | const Locale &locale, |
29 | BreakIterator *breakiter, |
30 | UErrorCode &status) : |
31 | SearchIterator(text, breakiter), |
32 | m_pattern_(pattern) |
33 | { |
34 | if (U_FAILURE(status)) { |
35 | m_strsrch_ = NULL; |
36 | return; |
37 | } |
38 | |
39 | m_strsrch_ = usearch_open(m_pattern_.getBuffer(), m_pattern_.length(), |
40 | m_text_.getBuffer(), m_text_.length(), |
41 | locale.getName(), (UBreakIterator *)breakiter, |
42 | &status); |
43 | uprv_free(m_search_); |
44 | m_search_ = NULL; |
45 | |
46 | if (U_SUCCESS(status)) { |
47 | // m_search_ has been created by the base SearchIterator class |
48 | m_search_ = m_strsrch_->search; |
49 | } |
50 | } |
51 | |
52 | StringSearch::StringSearch(const UnicodeString &pattern, |
53 | const UnicodeString &text, |
54 | RuleBasedCollator *coll, |
55 | BreakIterator *breakiter, |
56 | UErrorCode &status) : |
57 | SearchIterator(text, breakiter), |
58 | m_pattern_(pattern) |
59 | { |
60 | if (U_FAILURE(status)) { |
61 | m_strsrch_ = NULL; |
62 | return; |
63 | } |
64 | if (coll == NULL) { |
65 | status = U_ILLEGAL_ARGUMENT_ERROR; |
66 | m_strsrch_ = NULL; |
67 | return; |
68 | } |
69 | m_strsrch_ = usearch_openFromCollator(m_pattern_.getBuffer(), |
70 | m_pattern_.length(), |
71 | m_text_.getBuffer(), |
72 | m_text_.length(), coll->toUCollator(), |
73 | (UBreakIterator *)breakiter, |
74 | &status); |
75 | uprv_free(m_search_); |
76 | m_search_ = NULL; |
77 | |
78 | if (U_SUCCESS(status)) { |
79 | // m_search_ has been created by the base SearchIterator class |
80 | m_search_ = m_strsrch_->search; |
81 | } |
82 | } |
83 | |
84 | StringSearch::StringSearch(const UnicodeString &pattern, |
85 | CharacterIterator &text, |
86 | const Locale &locale, |
87 | BreakIterator *breakiter, |
88 | UErrorCode &status) : |
89 | SearchIterator(text, breakiter), |
90 | m_pattern_(pattern) |
91 | { |
92 | if (U_FAILURE(status)) { |
93 | m_strsrch_ = NULL; |
94 | return; |
95 | } |
96 | m_strsrch_ = usearch_open(m_pattern_.getBuffer(), m_pattern_.length(), |
97 | m_text_.getBuffer(), m_text_.length(), |
98 | locale.getName(), (UBreakIterator *)breakiter, |
99 | &status); |
100 | uprv_free(m_search_); |
101 | m_search_ = NULL; |
102 | |
103 | if (U_SUCCESS(status)) { |
104 | // m_search_ has been created by the base SearchIterator class |
105 | m_search_ = m_strsrch_->search; |
106 | } |
107 | } |
108 | |
109 | StringSearch::StringSearch(const UnicodeString &pattern, |
110 | CharacterIterator &text, |
111 | RuleBasedCollator *coll, |
112 | BreakIterator *breakiter, |
113 | UErrorCode &status) : |
114 | SearchIterator(text, breakiter), |
115 | m_pattern_(pattern) |
116 | { |
117 | if (U_FAILURE(status)) { |
118 | m_strsrch_ = NULL; |
119 | return; |
120 | } |
121 | if (coll == NULL) { |
122 | status = U_ILLEGAL_ARGUMENT_ERROR; |
123 | m_strsrch_ = NULL; |
124 | return; |
125 | } |
126 | m_strsrch_ = usearch_openFromCollator(m_pattern_.getBuffer(), |
127 | m_pattern_.length(), |
128 | m_text_.getBuffer(), |
129 | m_text_.length(), coll->toUCollator(), |
130 | (UBreakIterator *)breakiter, |
131 | &status); |
132 | uprv_free(m_search_); |
133 | m_search_ = NULL; |
134 | |
135 | if (U_SUCCESS(status)) { |
136 | // m_search_ has been created by the base SearchIterator class |
137 | m_search_ = m_strsrch_->search; |
138 | } |
139 | } |
140 | |
141 | StringSearch::StringSearch(const StringSearch &that) : |
142 | SearchIterator(that.m_text_, that.m_breakiterator_), |
143 | m_pattern_(that.m_pattern_) |
144 | { |
145 | UErrorCode status = U_ZERO_ERROR; |
146 | |
147 | // Free m_search_ from the superclass |
148 | uprv_free(m_search_); |
149 | m_search_ = NULL; |
150 | |
151 | if (that.m_strsrch_ == NULL) { |
152 | // This was not a good copy |
153 | m_strsrch_ = NULL; |
154 | } |
155 | else { |
156 | // Make a deep copy |
157 | m_strsrch_ = usearch_openFromCollator(m_pattern_.getBuffer(), |
158 | m_pattern_.length(), |
159 | m_text_.getBuffer(), |
160 | m_text_.length(), |
161 | that.m_strsrch_->collator, |
162 | (UBreakIterator *)that.m_breakiterator_, |
163 | &status); |
164 | if (U_SUCCESS(status)) { |
165 | // m_search_ has been created by the base SearchIterator class |
166 | m_search_ = m_strsrch_->search; |
167 | } |
168 | } |
169 | } |
170 | |
171 | StringSearch::~StringSearch() |
172 | { |
173 | if (m_strsrch_ != NULL) { |
174 | usearch_close(m_strsrch_); |
175 | m_search_ = NULL; |
176 | } |
177 | } |
178 | |
179 | StringSearch * |
180 | StringSearch::clone() const { |
181 | return new StringSearch(*this); |
182 | } |
183 | |
184 | // operator overloading --------------------------------------------- |
185 | StringSearch & StringSearch::operator=(const StringSearch &that) |
186 | { |
187 | if ((*this) != that) { |
188 | UErrorCode status = U_ZERO_ERROR; |
189 | m_text_ = that.m_text_; |
190 | m_breakiterator_ = that.m_breakiterator_; |
191 | m_pattern_ = that.m_pattern_; |
192 | // all m_search_ in the parent class is linked up with m_strsrch_ |
193 | usearch_close(m_strsrch_); |
194 | m_strsrch_ = usearch_openFromCollator(m_pattern_.getBuffer(), |
195 | m_pattern_.length(), |
196 | m_text_.getBuffer(), |
197 | m_text_.length(), |
198 | that.m_strsrch_->collator, |
199 | NULL, &status); |
200 | // Check null pointer |
201 | if (m_strsrch_ != NULL) { |
202 | m_search_ = m_strsrch_->search; |
203 | } |
204 | } |
205 | return *this; |
206 | } |
207 | |
208 | UBool StringSearch::operator==(const SearchIterator &that) const |
209 | { |
210 | if (this == &that) { |
211 | return TRUE; |
212 | } |
213 | if (SearchIterator::operator ==(that)) { |
214 | StringSearch &thatsrch = (StringSearch &)that; |
215 | return (this->m_pattern_ == thatsrch.m_pattern_ && |
216 | this->m_strsrch_->collator == thatsrch.m_strsrch_->collator); |
217 | } |
218 | return FALSE; |
219 | } |
220 | |
221 | // public get and set methods ---------------------------------------- |
222 | |
223 | void StringSearch::setOffset(int32_t position, UErrorCode &status) |
224 | { |
225 | // status checked in usearch_setOffset |
226 | usearch_setOffset(m_strsrch_, position, &status); |
227 | } |
228 | |
229 | int32_t StringSearch::getOffset(void) const |
230 | { |
231 | return usearch_getOffset(m_strsrch_); |
232 | } |
233 | |
234 | void StringSearch::setText(const UnicodeString &text, UErrorCode &status) |
235 | { |
236 | if (U_SUCCESS(status)) { |
237 | m_text_ = text; |
238 | usearch_setText(m_strsrch_, text.getBuffer(), text.length(), &status); |
239 | } |
240 | } |
241 | |
242 | void StringSearch::setText(CharacterIterator &text, UErrorCode &status) |
243 | { |
244 | if (U_SUCCESS(status)) { |
245 | text.getText(m_text_); |
246 | usearch_setText(m_strsrch_, m_text_.getBuffer(), m_text_.length(), &status); |
247 | } |
248 | } |
249 | |
250 | RuleBasedCollator * StringSearch::getCollator() const |
251 | { |
252 | // Note the const_cast. It would be cleaner if this const method returned a const collator. |
253 | return RuleBasedCollator::rbcFromUCollator(const_cast<UCollator *>(m_strsrch_->collator)); |
254 | } |
255 | |
256 | void StringSearch::setCollator(RuleBasedCollator *coll, UErrorCode &status) |
257 | { |
258 | if (U_SUCCESS(status)) { |
259 | usearch_setCollator(m_strsrch_, coll->toUCollator(), &status); |
260 | } |
261 | } |
262 | |
263 | void StringSearch::setPattern(const UnicodeString &pattern, |
264 | UErrorCode &status) |
265 | { |
266 | if (U_SUCCESS(status)) { |
267 | m_pattern_ = pattern; |
268 | usearch_setPattern(m_strsrch_, m_pattern_.getBuffer(), m_pattern_.length(), |
269 | &status); |
270 | } |
271 | } |
272 | |
273 | const UnicodeString & StringSearch::getPattern() const |
274 | { |
275 | return m_pattern_; |
276 | } |
277 | |
278 | // public methods ---------------------------------------------------- |
279 | |
280 | void StringSearch::reset() |
281 | { |
282 | usearch_reset(m_strsrch_); |
283 | } |
284 | |
285 | StringSearch * StringSearch::safeClone() const |
286 | { |
287 | UErrorCode status = U_ZERO_ERROR; |
288 | StringSearch *result = new StringSearch(m_pattern_, m_text_, |
289 | getCollator(), |
290 | m_breakiterator_, |
291 | status); |
292 | /* test for NULL */ |
293 | if (result == 0) { |
294 | status = U_MEMORY_ALLOCATION_ERROR; |
295 | return 0; |
296 | } |
297 | result->setOffset(getOffset(), status); |
298 | result->setMatchStart(m_strsrch_->search->matchedIndex); |
299 | result->setMatchLength(m_strsrch_->search->matchedLength); |
300 | if (U_FAILURE(status)) { |
301 | return NULL; |
302 | } |
303 | return result; |
304 | } |
305 | |
306 | // protected method ------------------------------------------------- |
307 | |
308 | int32_t StringSearch::handleNext(int32_t position, UErrorCode &status) |
309 | { |
310 | // values passed here are already in the pre-shift position |
311 | if (U_SUCCESS(status)) { |
312 | if (m_strsrch_->pattern.cesLength == 0) { |
313 | m_search_->matchedIndex = |
314 | m_search_->matchedIndex == USEARCH_DONE ? |
315 | getOffset() : m_search_->matchedIndex + 1; |
316 | m_search_->matchedLength = 0; |
317 | ucol_setOffset(m_strsrch_->textIter, m_search_->matchedIndex, |
318 | &status); |
319 | if (m_search_->matchedIndex == m_search_->textLength) { |
320 | m_search_->matchedIndex = USEARCH_DONE; |
321 | } |
322 | } |
323 | else { |
324 | // looking at usearch.cpp, this part is shifted out to |
325 | // StringSearch instead of SearchIterator because m_strsrch_ is |
326 | // not accessible in SearchIterator |
327 | #if 0 |
328 | if (position + m_strsrch_->pattern.defaultShiftSize |
329 | > m_search_->textLength) { |
330 | setMatchNotFound(); |
331 | return USEARCH_DONE; |
332 | } |
333 | #endif |
334 | if (m_search_->matchedLength <= 0) { |
335 | // the flipping direction issue has already been handled |
336 | // in next() |
337 | // for boundary check purposes. this will ensure that the |
338 | // next match will not preceed the current offset |
339 | // note search->matchedIndex will always be set to something |
340 | // in the code |
341 | m_search_->matchedIndex = position - 1; |
342 | } |
343 | |
344 | ucol_setOffset(m_strsrch_->textIter, position, &status); |
345 | |
346 | #if 0 |
347 | for (;;) { |
348 | if (m_search_->isCanonicalMatch) { |
349 | // can't use exact here since extra accents are allowed. |
350 | usearch_handleNextCanonical(m_strsrch_, &status); |
351 | } |
352 | else { |
353 | usearch_handleNextExact(m_strsrch_, &status); |
354 | } |
355 | if (U_FAILURE(status)) { |
356 | return USEARCH_DONE; |
357 | } |
358 | if (m_breakiterator_ == NULL |
359 | #if !UCONFIG_NO_BREAK_ITERATION |
360 | || |
361 | m_search_->matchedIndex == USEARCH_DONE || |
362 | (m_breakiterator_->isBoundary(m_search_->matchedIndex) && |
363 | m_breakiterator_->isBoundary(m_search_->matchedIndex + |
364 | m_search_->matchedLength)) |
365 | #endif |
366 | ) { |
367 | if (m_search_->matchedIndex == USEARCH_DONE) { |
368 | ucol_setOffset(m_strsrch_->textIter, |
369 | m_search_->textLength, &status); |
370 | } |
371 | else { |
372 | ucol_setOffset(m_strsrch_->textIter, |
373 | m_search_->matchedIndex, &status); |
374 | } |
375 | return m_search_->matchedIndex; |
376 | } |
377 | } |
378 | #else |
379 | // if m_strsrch_->breakIter is always the same as m_breakiterator_ |
380 | // then we don't need to check the match boundaries here because |
381 | // usearch_handleNextXXX will already have done it. |
382 | if (m_search_->isCanonicalMatch) { |
383 | // *could* actually use exact here 'cause no extra accents allowed... |
384 | usearch_handleNextCanonical(m_strsrch_, &status); |
385 | } else { |
386 | usearch_handleNextExact(m_strsrch_, &status); |
387 | } |
388 | |
389 | if (U_FAILURE(status)) { |
390 | return USEARCH_DONE; |
391 | } |
392 | |
393 | if (m_search_->matchedIndex == USEARCH_DONE) { |
394 | ucol_setOffset(m_strsrch_->textIter, m_search_->textLength, &status); |
395 | } else { |
396 | ucol_setOffset(m_strsrch_->textIter, m_search_->matchedIndex, &status); |
397 | } |
398 | |
399 | return m_search_->matchedIndex; |
400 | #endif |
401 | } |
402 | } |
403 | return USEARCH_DONE; |
404 | } |
405 | |
406 | int32_t StringSearch::handlePrev(int32_t position, UErrorCode &status) |
407 | { |
408 | // values passed here are already in the pre-shift position |
409 | if (U_SUCCESS(status)) { |
410 | if (m_strsrch_->pattern.cesLength == 0) { |
411 | m_search_->matchedIndex = |
412 | (m_search_->matchedIndex == USEARCH_DONE ? getOffset() : |
413 | m_search_->matchedIndex); |
414 | if (m_search_->matchedIndex == 0) { |
415 | setMatchNotFound(); |
416 | } |
417 | else { |
418 | m_search_->matchedIndex --; |
419 | ucol_setOffset(m_strsrch_->textIter, m_search_->matchedIndex, |
420 | &status); |
421 | m_search_->matchedLength = 0; |
422 | } |
423 | } |
424 | else { |
425 | // looking at usearch.cpp, this part is shifted out to |
426 | // StringSearch instead of SearchIterator because m_strsrch_ is |
427 | // not accessible in SearchIterator |
428 | #if 0 |
429 | if (!m_search_->isOverlap && |
430 | position - m_strsrch_->pattern.defaultShiftSize < 0) { |
431 | setMatchNotFound(); |
432 | return USEARCH_DONE; |
433 | } |
434 | |
435 | for (;;) { |
436 | if (m_search_->isCanonicalMatch) { |
437 | // can't use exact here since extra accents are allowed. |
438 | usearch_handlePreviousCanonical(m_strsrch_, &status); |
439 | } |
440 | else { |
441 | usearch_handlePreviousExact(m_strsrch_, &status); |
442 | } |
443 | if (U_FAILURE(status)) { |
444 | return USEARCH_DONE; |
445 | } |
446 | if (m_breakiterator_ == NULL |
447 | #if !UCONFIG_NO_BREAK_ITERATION |
448 | || |
449 | m_search_->matchedIndex == USEARCH_DONE || |
450 | (m_breakiterator_->isBoundary(m_search_->matchedIndex) && |
451 | m_breakiterator_->isBoundary(m_search_->matchedIndex + |
452 | m_search_->matchedLength)) |
453 | #endif |
454 | ) { |
455 | return m_search_->matchedIndex; |
456 | } |
457 | } |
458 | #else |
459 | ucol_setOffset(m_strsrch_->textIter, position, &status); |
460 | |
461 | if (m_search_->isCanonicalMatch) { |
462 | // *could* use exact match here since extra accents *not* allowed! |
463 | usearch_handlePreviousCanonical(m_strsrch_, &status); |
464 | } else { |
465 | usearch_handlePreviousExact(m_strsrch_, &status); |
466 | } |
467 | |
468 | if (U_FAILURE(status)) { |
469 | return USEARCH_DONE; |
470 | } |
471 | |
472 | return m_search_->matchedIndex; |
473 | #endif |
474 | } |
475 | |
476 | return m_search_->matchedIndex; |
477 | } |
478 | return USEARCH_DONE; |
479 | } |
480 | |
481 | U_NAMESPACE_END |
482 | |
483 | #endif /* #if !UCONFIG_NO_COLLATION */ |
484 | |