| 1 | // © 2016 and later: Unicode, Inc. and others. |
| 2 | // License & terms of use: http://www.unicode.org/copyright.html |
| 3 | /* |
| 4 | ********************************************************************** |
| 5 | * Copyright (C) 2001-2008,2010 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/brkiter.h" |
| 17 | #include "unicode/schriter.h" |
| 18 | #include "unicode/search.h" |
| 19 | #include "usrchimp.h" |
| 20 | #include "cmemory.h" |
| 21 | |
| 22 | // public constructors and destructors ----------------------------------- |
| 23 | U_NAMESPACE_BEGIN |
| 24 | |
| 25 | SearchIterator::SearchIterator(const SearchIterator &other) |
| 26 | : UObject(other) |
| 27 | { |
| 28 | m_breakiterator_ = other.m_breakiterator_; |
| 29 | m_text_ = other.m_text_; |
| 30 | m_search_ = (USearch *)uprv_malloc(sizeof(USearch)); |
| 31 | m_search_->breakIter = other.m_search_->breakIter; |
| 32 | m_search_->isCanonicalMatch = other.m_search_->isCanonicalMatch; |
| 33 | m_search_->isOverlap = other.m_search_->isOverlap; |
| 34 | m_search_->elementComparisonType = other.m_search_->elementComparisonType; |
| 35 | m_search_->matchedIndex = other.m_search_->matchedIndex; |
| 36 | m_search_->matchedLength = other.m_search_->matchedLength; |
| 37 | m_search_->text = other.m_search_->text; |
| 38 | m_search_->textLength = other.m_search_->textLength; |
| 39 | } |
| 40 | |
| 41 | SearchIterator::~SearchIterator() |
| 42 | { |
| 43 | if (m_search_ != NULL) { |
| 44 | uprv_free(m_search_); |
| 45 | } |
| 46 | } |
| 47 | |
| 48 | // public get and set methods ---------------------------------------- |
| 49 | |
| 50 | void SearchIterator::setAttribute(USearchAttribute attribute, |
| 51 | USearchAttributeValue value, |
| 52 | UErrorCode &status) |
| 53 | { |
| 54 | if (U_SUCCESS(status)) { |
| 55 | switch (attribute) |
| 56 | { |
| 57 | case USEARCH_OVERLAP : |
| 58 | m_search_->isOverlap = (value == USEARCH_ON ? TRUE : FALSE); |
| 59 | break; |
| 60 | case USEARCH_CANONICAL_MATCH : |
| 61 | m_search_->isCanonicalMatch = (value == USEARCH_ON ? TRUE : FALSE); |
| 62 | break; |
| 63 | case USEARCH_ELEMENT_COMPARISON : |
| 64 | if (value == USEARCH_PATTERN_BASE_WEIGHT_IS_WILDCARD || value == USEARCH_ANY_BASE_WEIGHT_IS_WILDCARD) { |
| 65 | m_search_->elementComparisonType = (int16_t)value; |
| 66 | } else { |
| 67 | m_search_->elementComparisonType = 0; |
| 68 | } |
| 69 | break; |
| 70 | default: |
| 71 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 72 | } |
| 73 | } |
| 74 | if (value == USEARCH_ATTRIBUTE_VALUE_COUNT) { |
| 75 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 76 | } |
| 77 | } |
| 78 | |
| 79 | USearchAttributeValue SearchIterator::getAttribute( |
| 80 | USearchAttribute attribute) const |
| 81 | { |
| 82 | switch (attribute) { |
| 83 | case USEARCH_OVERLAP : |
| 84 | return (m_search_->isOverlap == TRUE ? USEARCH_ON : USEARCH_OFF); |
| 85 | case USEARCH_CANONICAL_MATCH : |
| 86 | return (m_search_->isCanonicalMatch == TRUE ? USEARCH_ON : |
| 87 | USEARCH_OFF); |
| 88 | case USEARCH_ELEMENT_COMPARISON : |
| 89 | { |
| 90 | int16_t value = m_search_->elementComparisonType; |
| 91 | if (value == USEARCH_PATTERN_BASE_WEIGHT_IS_WILDCARD || value == USEARCH_ANY_BASE_WEIGHT_IS_WILDCARD) { |
| 92 | return (USearchAttributeValue)value; |
| 93 | } else { |
| 94 | return USEARCH_STANDARD_ELEMENT_COMPARISON; |
| 95 | } |
| 96 | } |
| 97 | default : |
| 98 | return USEARCH_DEFAULT; |
| 99 | } |
| 100 | } |
| 101 | |
| 102 | int32_t SearchIterator::getMatchedStart() const |
| 103 | { |
| 104 | return m_search_->matchedIndex; |
| 105 | } |
| 106 | |
| 107 | int32_t SearchIterator::getMatchedLength() const |
| 108 | { |
| 109 | return m_search_->matchedLength; |
| 110 | } |
| 111 | |
| 112 | void SearchIterator::getMatchedText(UnicodeString &result) const |
| 113 | { |
| 114 | int32_t matchedindex = m_search_->matchedIndex; |
| 115 | int32_t matchedlength = m_search_->matchedLength; |
| 116 | if (matchedindex != USEARCH_DONE && matchedlength != 0) { |
| 117 | result.setTo(m_search_->text + matchedindex, matchedlength); |
| 118 | } |
| 119 | else { |
| 120 | result.remove(); |
| 121 | } |
| 122 | } |
| 123 | |
| 124 | void SearchIterator::setBreakIterator(BreakIterator *breakiter, |
| 125 | UErrorCode &status) |
| 126 | { |
| 127 | if (U_SUCCESS(status)) { |
| 128 | #if 0 |
| 129 | m_search_->breakIter = NULL; |
| 130 | // the c++ breakiterator may not make use of ubreakiterator. |
| 131 | // so we'll have to keep track of it ourselves. |
| 132 | #else |
| 133 | // Well, gee... the Constructors that take a BreakIterator |
| 134 | // all cast the BreakIterator to a UBreakIterator and |
| 135 | // pass it to the corresponding usearch_openFromXXX |
| 136 | // routine, so there's no reason not to do this. |
| 137 | // |
| 138 | // Besides, a UBreakIterator is a BreakIterator, so |
| 139 | // any subclass of BreakIterator should work fine here... |
| 140 | m_search_->breakIter = (UBreakIterator *) breakiter; |
| 141 | #endif |
| 142 | |
| 143 | m_breakiterator_ = breakiter; |
| 144 | } |
| 145 | } |
| 146 | |
| 147 | const BreakIterator * SearchIterator::getBreakIterator(void) const |
| 148 | { |
| 149 | return m_breakiterator_; |
| 150 | } |
| 151 | |
| 152 | void SearchIterator::setText(const UnicodeString &text, UErrorCode &status) |
| 153 | { |
| 154 | if (U_SUCCESS(status)) { |
| 155 | if (text.length() == 0) { |
| 156 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 157 | } |
| 158 | else { |
| 159 | m_text_ = text; |
| 160 | m_search_->text = m_text_.getBuffer(); |
| 161 | m_search_->textLength = m_text_.length(); |
| 162 | } |
| 163 | } |
| 164 | } |
| 165 | |
| 166 | void SearchIterator::setText(CharacterIterator &text, UErrorCode &status) |
| 167 | { |
| 168 | if (U_SUCCESS(status)) { |
| 169 | text.getText(m_text_); |
| 170 | setText(m_text_, status); |
| 171 | } |
| 172 | } |
| 173 | |
| 174 | const UnicodeString & SearchIterator::getText(void) const |
| 175 | { |
| 176 | return m_text_; |
| 177 | } |
| 178 | |
| 179 | // operator overloading ---------------------------------------------- |
| 180 | |
| 181 | UBool SearchIterator::operator==(const SearchIterator &that) const |
| 182 | { |
| 183 | if (this == &that) { |
| 184 | return TRUE; |
| 185 | } |
| 186 | return (m_breakiterator_ == that.m_breakiterator_ && |
| 187 | m_search_->isCanonicalMatch == that.m_search_->isCanonicalMatch && |
| 188 | m_search_->isOverlap == that.m_search_->isOverlap && |
| 189 | m_search_->elementComparisonType == that.m_search_->elementComparisonType && |
| 190 | m_search_->matchedIndex == that.m_search_->matchedIndex && |
| 191 | m_search_->matchedLength == that.m_search_->matchedLength && |
| 192 | m_search_->textLength == that.m_search_->textLength && |
| 193 | getOffset() == that.getOffset() && |
| 194 | (uprv_memcmp(m_search_->text, that.m_search_->text, |
| 195 | m_search_->textLength * sizeof(UChar)) == 0)); |
| 196 | } |
| 197 | |
| 198 | // public methods ---------------------------------------------------- |
| 199 | |
| 200 | int32_t SearchIterator::first(UErrorCode &status) |
| 201 | { |
| 202 | if (U_FAILURE(status)) { |
| 203 | return USEARCH_DONE; |
| 204 | } |
| 205 | setOffset(0, status); |
| 206 | return handleNext(0, status); |
| 207 | } |
| 208 | |
| 209 | int32_t SearchIterator::following(int32_t position, |
| 210 | UErrorCode &status) |
| 211 | { |
| 212 | if (U_FAILURE(status)) { |
| 213 | return USEARCH_DONE; |
| 214 | } |
| 215 | setOffset(position, status); |
| 216 | return handleNext(position, status); |
| 217 | } |
| 218 | |
| 219 | int32_t SearchIterator::last(UErrorCode &status) |
| 220 | { |
| 221 | if (U_FAILURE(status)) { |
| 222 | return USEARCH_DONE; |
| 223 | } |
| 224 | setOffset(m_search_->textLength, status); |
| 225 | return handlePrev(m_search_->textLength, status); |
| 226 | } |
| 227 | |
| 228 | int32_t SearchIterator::preceding(int32_t position, |
| 229 | UErrorCode &status) |
| 230 | { |
| 231 | if (U_FAILURE(status)) { |
| 232 | return USEARCH_DONE; |
| 233 | } |
| 234 | setOffset(position, status); |
| 235 | return handlePrev(position, status); |
| 236 | } |
| 237 | |
| 238 | int32_t SearchIterator::next(UErrorCode &status) |
| 239 | { |
| 240 | if (U_SUCCESS(status)) { |
| 241 | int32_t offset = getOffset(); |
| 242 | int32_t matchindex = m_search_->matchedIndex; |
| 243 | int32_t matchlength = m_search_->matchedLength; |
| 244 | m_search_->reset = FALSE; |
| 245 | if (m_search_->isForwardSearching == TRUE) { |
| 246 | int32_t textlength = m_search_->textLength; |
| 247 | if (offset == textlength || matchindex == textlength || |
| 248 | (matchindex != USEARCH_DONE && |
| 249 | matchindex + matchlength >= textlength)) { |
| 250 | // not enough characters to match |
| 251 | setMatchNotFound(); |
| 252 | return USEARCH_DONE; |
| 253 | } |
| 254 | } |
| 255 | else { |
| 256 | // switching direction. |
| 257 | // if matchedIndex == USEARCH_DONE, it means that either a |
| 258 | // setOffset has been called or that previous ran off the text |
| 259 | // string. the iterator would have been set to offset 0 if a |
| 260 | // match is not found. |
| 261 | m_search_->isForwardSearching = TRUE; |
| 262 | if (m_search_->matchedIndex != USEARCH_DONE) { |
| 263 | // there's no need to set the collation element iterator |
| 264 | // the next call to next will set the offset. |
| 265 | return matchindex; |
| 266 | } |
| 267 | } |
| 268 | |
| 269 | if (matchlength > 0) { |
| 270 | // if matchlength is 0 we are at the start of the iteration |
| 271 | if (m_search_->isOverlap) { |
| 272 | offset ++; |
| 273 | } |
| 274 | else { |
| 275 | offset += matchlength; |
| 276 | } |
| 277 | } |
| 278 | return handleNext(offset, status); |
| 279 | } |
| 280 | return USEARCH_DONE; |
| 281 | } |
| 282 | |
| 283 | int32_t SearchIterator::previous(UErrorCode &status) |
| 284 | { |
| 285 | if (U_SUCCESS(status)) { |
| 286 | int32_t offset; |
| 287 | if (m_search_->reset) { |
| 288 | offset = m_search_->textLength; |
| 289 | m_search_->isForwardSearching = FALSE; |
| 290 | m_search_->reset = FALSE; |
| 291 | setOffset(offset, status); |
| 292 | } |
| 293 | else { |
| 294 | offset = getOffset(); |
| 295 | } |
| 296 | |
| 297 | int32_t matchindex = m_search_->matchedIndex; |
| 298 | if (m_search_->isForwardSearching == TRUE) { |
| 299 | // switching direction. |
| 300 | // if matchedIndex == USEARCH_DONE, it means that either a |
| 301 | // setOffset has been called or that next ran off the text |
| 302 | // string. the iterator would have been set to offset textLength if |
| 303 | // a match is not found. |
| 304 | m_search_->isForwardSearching = FALSE; |
| 305 | if (matchindex != USEARCH_DONE) { |
| 306 | return matchindex; |
| 307 | } |
| 308 | } |
| 309 | else { |
| 310 | if (offset == 0 || matchindex == 0) { |
| 311 | // not enough characters to match |
| 312 | setMatchNotFound(); |
| 313 | return USEARCH_DONE; |
| 314 | } |
| 315 | } |
| 316 | |
| 317 | if (matchindex != USEARCH_DONE) { |
| 318 | if (m_search_->isOverlap) { |
| 319 | matchindex += m_search_->matchedLength - 2; |
| 320 | } |
| 321 | |
| 322 | return handlePrev(matchindex, status); |
| 323 | } |
| 324 | |
| 325 | return handlePrev(offset, status); |
| 326 | } |
| 327 | |
| 328 | return USEARCH_DONE; |
| 329 | } |
| 330 | |
| 331 | void SearchIterator::reset() |
| 332 | { |
| 333 | UErrorCode status = U_ZERO_ERROR; |
| 334 | setMatchNotFound(); |
| 335 | setOffset(0, status); |
| 336 | m_search_->isOverlap = FALSE; |
| 337 | m_search_->isCanonicalMatch = FALSE; |
| 338 | m_search_->elementComparisonType = 0; |
| 339 | m_search_->isForwardSearching = TRUE; |
| 340 | m_search_->reset = TRUE; |
| 341 | } |
| 342 | |
| 343 | // protected constructors and destructors ----------------------------- |
| 344 | |
| 345 | SearchIterator::SearchIterator() |
| 346 | { |
| 347 | m_search_ = (USearch *)uprv_malloc(sizeof(USearch)); |
| 348 | m_search_->breakIter = NULL; |
| 349 | m_search_->isOverlap = FALSE; |
| 350 | m_search_->isCanonicalMatch = FALSE; |
| 351 | m_search_->elementComparisonType = 0; |
| 352 | m_search_->isForwardSearching = TRUE; |
| 353 | m_search_->reset = TRUE; |
| 354 | m_search_->matchedIndex = USEARCH_DONE; |
| 355 | m_search_->matchedLength = 0; |
| 356 | m_search_->text = NULL; |
| 357 | m_search_->textLength = 0; |
| 358 | m_breakiterator_ = NULL; |
| 359 | } |
| 360 | |
| 361 | SearchIterator::SearchIterator(const UnicodeString &text, |
| 362 | BreakIterator *breakiter) : |
| 363 | m_breakiterator_(breakiter), |
| 364 | m_text_(text) |
| 365 | { |
| 366 | m_search_ = (USearch *)uprv_malloc(sizeof(USearch)); |
| 367 | m_search_->breakIter = NULL; |
| 368 | m_search_->isOverlap = FALSE; |
| 369 | m_search_->isCanonicalMatch = FALSE; |
| 370 | m_search_->elementComparisonType = 0; |
| 371 | m_search_->isForwardSearching = TRUE; |
| 372 | m_search_->reset = TRUE; |
| 373 | m_search_->matchedIndex = USEARCH_DONE; |
| 374 | m_search_->matchedLength = 0; |
| 375 | m_search_->text = m_text_.getBuffer(); |
| 376 | m_search_->textLength = text.length(); |
| 377 | } |
| 378 | |
| 379 | SearchIterator::SearchIterator(CharacterIterator &text, |
| 380 | BreakIterator *breakiter) : |
| 381 | m_breakiterator_(breakiter) |
| 382 | { |
| 383 | m_search_ = (USearch *)uprv_malloc(sizeof(USearch)); |
| 384 | m_search_->breakIter = NULL; |
| 385 | m_search_->isOverlap = FALSE; |
| 386 | m_search_->isCanonicalMatch = FALSE; |
| 387 | m_search_->elementComparisonType = 0; |
| 388 | m_search_->isForwardSearching = TRUE; |
| 389 | m_search_->reset = TRUE; |
| 390 | m_search_->matchedIndex = USEARCH_DONE; |
| 391 | m_search_->matchedLength = 0; |
| 392 | text.getText(m_text_); |
| 393 | m_search_->text = m_text_.getBuffer(); |
| 394 | m_search_->textLength = m_text_.length(); |
| 395 | m_breakiterator_ = breakiter; |
| 396 | } |
| 397 | |
| 398 | // protected methods ------------------------------------------------------ |
| 399 | |
| 400 | SearchIterator & SearchIterator::operator=(const SearchIterator &that) |
| 401 | { |
| 402 | if (this != &that) { |
| 403 | m_breakiterator_ = that.m_breakiterator_; |
| 404 | m_text_ = that.m_text_; |
| 405 | m_search_->breakIter = that.m_search_->breakIter; |
| 406 | m_search_->isCanonicalMatch = that.m_search_->isCanonicalMatch; |
| 407 | m_search_->isOverlap = that.m_search_->isOverlap; |
| 408 | m_search_->elementComparisonType = that.m_search_->elementComparisonType; |
| 409 | m_search_->matchedIndex = that.m_search_->matchedIndex; |
| 410 | m_search_->matchedLength = that.m_search_->matchedLength; |
| 411 | m_search_->text = that.m_search_->text; |
| 412 | m_search_->textLength = that.m_search_->textLength; |
| 413 | } |
| 414 | return *this; |
| 415 | } |
| 416 | |
| 417 | void SearchIterator::setMatchLength(int32_t length) |
| 418 | { |
| 419 | m_search_->matchedLength = length; |
| 420 | } |
| 421 | |
| 422 | void SearchIterator::setMatchStart(int32_t position) |
| 423 | { |
| 424 | m_search_->matchedIndex = position; |
| 425 | } |
| 426 | |
| 427 | void SearchIterator::setMatchNotFound() |
| 428 | { |
| 429 | setMatchStart(USEARCH_DONE); |
| 430 | setMatchLength(0); |
| 431 | UErrorCode status = U_ZERO_ERROR; |
| 432 | // by default no errors should be returned here since offsets are within |
| 433 | // range. |
| 434 | if (m_search_->isForwardSearching) { |
| 435 | setOffset(m_search_->textLength, status); |
| 436 | } |
| 437 | else { |
| 438 | setOffset(0, status); |
| 439 | } |
| 440 | } |
| 441 | |
| 442 | |
| 443 | U_NAMESPACE_END |
| 444 | |
| 445 | #endif /* #if !UCONFIG_NO_COLLATION */ |
| 446 | |