| 1 | // © 2016 and later: Unicode, Inc. and others. |
| 2 | // License & terms of use: http://www.unicode.org/copyright.html |
| 3 | /* |
| 4 | ******************************************************************************* |
| 5 | * Copyright (C) 1996-2015, International Business Machines |
| 6 | * Corporation and others. All Rights Reserved. |
| 7 | ******************************************************************************* |
| 8 | * file name: ucol.cpp |
| 9 | * encoding: UTF-8 |
| 10 | * tab size: 8 (not used) |
| 11 | * indentation:4 |
| 12 | * |
| 13 | * Modification history |
| 14 | * Date Name Comments |
| 15 | * 1996-1999 various members of ICU team maintained C API for collation framework |
| 16 | * 02/16/2001 synwee Added internal method getPrevSpecialCE |
| 17 | * 03/01/2001 synwee Added maxexpansion functionality. |
| 18 | * 03/16/2001 weiv Collation framework is rewritten in C and made UCA compliant |
| 19 | * 2012-2014 markus Rewritten in C++ again. |
| 20 | */ |
| 21 | |
| 22 | #include "unicode/utypes.h" |
| 23 | |
| 24 | #if !UCONFIG_NO_COLLATION |
| 25 | |
| 26 | #include "unicode/coll.h" |
| 27 | #include "unicode/tblcoll.h" |
| 28 | #include "unicode/bytestream.h" |
| 29 | #include "unicode/coleitr.h" |
| 30 | #include "unicode/ucoleitr.h" |
| 31 | #include "unicode/ustring.h" |
| 32 | #include "cmemory.h" |
| 33 | #include "collation.h" |
| 34 | #include "cstring.h" |
| 35 | #include "putilimp.h" |
| 36 | #include "uassert.h" |
| 37 | #include "utracimp.h" |
| 38 | |
| 39 | U_NAMESPACE_USE |
| 40 | |
| 41 | U_CAPI UCollator* U_EXPORT2 |
| 42 | ucol_openBinary(const uint8_t *bin, int32_t length, |
| 43 | const UCollator *base, |
| 44 | UErrorCode *status) |
| 45 | { |
| 46 | if(U_FAILURE(*status)) { return NULL; } |
| 47 | RuleBasedCollator *coll = new RuleBasedCollator( |
| 48 | bin, length, |
| 49 | RuleBasedCollator::rbcFromUCollator(base), |
| 50 | *status); |
| 51 | if(coll == NULL) { |
| 52 | *status = U_MEMORY_ALLOCATION_ERROR; |
| 53 | return NULL; |
| 54 | } |
| 55 | if(U_FAILURE(*status)) { |
| 56 | delete coll; |
| 57 | return NULL; |
| 58 | } |
| 59 | return coll->toUCollator(); |
| 60 | } |
| 61 | |
| 62 | U_CAPI int32_t U_EXPORT2 |
| 63 | ucol_cloneBinary(const UCollator *coll, |
| 64 | uint8_t *buffer, int32_t capacity, |
| 65 | UErrorCode *status) |
| 66 | { |
| 67 | if(U_FAILURE(*status)) { |
| 68 | return 0; |
| 69 | } |
| 70 | const RuleBasedCollator *rbc = RuleBasedCollator::rbcFromUCollator(coll); |
| 71 | if(rbc == NULL && coll != NULL) { |
| 72 | *status = U_UNSUPPORTED_ERROR; |
| 73 | return 0; |
| 74 | } |
| 75 | return rbc->cloneBinary(buffer, capacity, *status); |
| 76 | } |
| 77 | |
| 78 | U_CAPI UCollator* U_EXPORT2 |
| 79 | ucol_safeClone(const UCollator *coll, void * /*stackBuffer*/, int32_t * pBufferSize, UErrorCode *status) |
| 80 | { |
| 81 | if (status == NULL || U_FAILURE(*status)){ |
| 82 | return NULL; |
| 83 | } |
| 84 | if (coll == NULL) { |
| 85 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
| 86 | return NULL; |
| 87 | } |
| 88 | if (pBufferSize != NULL) { |
| 89 | int32_t inputSize = *pBufferSize; |
| 90 | *pBufferSize = 1; |
| 91 | if (inputSize == 0) { |
| 92 | return NULL; // preflighting for deprecated functionality |
| 93 | } |
| 94 | } |
| 95 | Collator *newColl = Collator::fromUCollator(coll)->clone(); |
| 96 | if (newColl == NULL) { |
| 97 | *status = U_MEMORY_ALLOCATION_ERROR; |
| 98 | return nullptr; |
| 99 | } else { |
| 100 | *status = U_SAFECLONE_ALLOCATED_WARNING; |
| 101 | } |
| 102 | return newColl->toUCollator(); |
| 103 | } |
| 104 | |
| 105 | U_CAPI void U_EXPORT2 |
| 106 | ucol_close(UCollator *coll) |
| 107 | { |
| 108 | UTRACE_ENTRY_OC(UTRACE_UCOL_CLOSE); |
| 109 | UTRACE_DATA1(UTRACE_INFO, "coll = %p" , coll); |
| 110 | if(coll != NULL) { |
| 111 | delete Collator::fromUCollator(coll); |
| 112 | } |
| 113 | UTRACE_EXIT(); |
| 114 | } |
| 115 | |
| 116 | U_CAPI int32_t U_EXPORT2 |
| 117 | ucol_mergeSortkeys(const uint8_t *src1, int32_t src1Length, |
| 118 | const uint8_t *src2, int32_t src2Length, |
| 119 | uint8_t *dest, int32_t destCapacity) { |
| 120 | /* check arguments */ |
| 121 | if( src1==NULL || src1Length<-1 || src1Length==0 || (src1Length>0 && src1[src1Length-1]!=0) || |
| 122 | src2==NULL || src2Length<-1 || src2Length==0 || (src2Length>0 && src2[src2Length-1]!=0) || |
| 123 | destCapacity<0 || (destCapacity>0 && dest==NULL) |
| 124 | ) { |
| 125 | /* error, attempt to write a zero byte and return 0 */ |
| 126 | if(dest!=NULL && destCapacity>0) { |
| 127 | *dest=0; |
| 128 | } |
| 129 | return 0; |
| 130 | } |
| 131 | |
| 132 | /* check lengths and capacity */ |
| 133 | if(src1Length<0) { |
| 134 | src1Length=(int32_t)uprv_strlen((const char *)src1)+1; |
| 135 | } |
| 136 | if(src2Length<0) { |
| 137 | src2Length=(int32_t)uprv_strlen((const char *)src2)+1; |
| 138 | } |
| 139 | |
| 140 | int32_t destLength=src1Length+src2Length; |
| 141 | if(destLength>destCapacity) { |
| 142 | /* the merged sort key does not fit into the destination */ |
| 143 | return destLength; |
| 144 | } |
| 145 | |
| 146 | /* merge the sort keys with the same number of levels */ |
| 147 | uint8_t *p=dest; |
| 148 | for(;;) { |
| 149 | /* copy level from src1 not including 00 or 01 */ |
| 150 | uint8_t b; |
| 151 | while((b=*src1)>=2) { |
| 152 | ++src1; |
| 153 | *p++=b; |
| 154 | } |
| 155 | |
| 156 | /* add a 02 merge separator */ |
| 157 | *p++=2; |
| 158 | |
| 159 | /* copy level from src2 not including 00 or 01 */ |
| 160 | while((b=*src2)>=2) { |
| 161 | ++src2; |
| 162 | *p++=b; |
| 163 | } |
| 164 | |
| 165 | /* if both sort keys have another level, then add a 01 level separator and continue */ |
| 166 | if(*src1==1 && *src2==1) { |
| 167 | ++src1; |
| 168 | ++src2; |
| 169 | *p++=1; |
| 170 | } else { |
| 171 | break; |
| 172 | } |
| 173 | } |
| 174 | |
| 175 | /* |
| 176 | * here, at least one sort key is finished now, but the other one |
| 177 | * might have some contents left from containing more levels; |
| 178 | * that contents is just appended to the result |
| 179 | */ |
| 180 | if(*src1!=0) { |
| 181 | /* src1 is not finished, therefore *src2==0, and src1 is appended */ |
| 182 | src2=src1; |
| 183 | } |
| 184 | /* append src2, "the other, unfinished sort key" */ |
| 185 | while((*p++=*src2++)!=0) {} |
| 186 | |
| 187 | /* the actual length might be less than destLength if either sort key contained illegally embedded zero bytes */ |
| 188 | return (int32_t)(p-dest); |
| 189 | } |
| 190 | |
| 191 | U_CAPI int32_t U_EXPORT2 |
| 192 | ucol_getSortKey(const UCollator *coll, |
| 193 | const UChar *source, |
| 194 | int32_t sourceLength, |
| 195 | uint8_t *result, |
| 196 | int32_t resultLength) |
| 197 | { |
| 198 | UTRACE_ENTRY(UTRACE_UCOL_GET_SORTKEY); |
| 199 | if (UTRACE_LEVEL(UTRACE_VERBOSE)) { |
| 200 | UTRACE_DATA3(UTRACE_VERBOSE, "coll=%p, source string = %vh " , coll, source, |
| 201 | ((sourceLength==-1 && source!=NULL) ? u_strlen(source) : sourceLength)); |
| 202 | } |
| 203 | |
| 204 | int32_t keySize = Collator::fromUCollator(coll)-> |
| 205 | getSortKey(source, sourceLength, result, resultLength); |
| 206 | |
| 207 | UTRACE_DATA2(UTRACE_VERBOSE, "Sort Key = %vb" , result, keySize); |
| 208 | UTRACE_EXIT_VALUE(keySize); |
| 209 | return keySize; |
| 210 | } |
| 211 | |
| 212 | U_CAPI int32_t U_EXPORT2 |
| 213 | ucol_nextSortKeyPart(const UCollator *coll, |
| 214 | UCharIterator *iter, |
| 215 | uint32_t state[2], |
| 216 | uint8_t *dest, int32_t count, |
| 217 | UErrorCode *status) |
| 218 | { |
| 219 | /* error checking */ |
| 220 | if(status==NULL || U_FAILURE(*status)) { |
| 221 | return 0; |
| 222 | } |
| 223 | UTRACE_ENTRY(UTRACE_UCOL_NEXTSORTKEYPART); |
| 224 | UTRACE_DATA6(UTRACE_VERBOSE, "coll=%p, iter=%p, state=%d %d, dest=%p, count=%d" , |
| 225 | coll, iter, state[0], state[1], dest, count); |
| 226 | |
| 227 | int32_t i = Collator::fromUCollator(coll)-> |
| 228 | internalNextSortKeyPart(iter, state, dest, count, *status); |
| 229 | |
| 230 | // Return number of meaningful sortkey bytes. |
| 231 | UTRACE_DATA4(UTRACE_VERBOSE, "dest = %vb, state=%d %d" , |
| 232 | dest,i, state[0], state[1]); |
| 233 | UTRACE_EXIT_VALUE_STATUS(i, *status); |
| 234 | return i; |
| 235 | } |
| 236 | |
| 237 | /** |
| 238 | * Produce a bound for a given sortkey and a number of levels. |
| 239 | */ |
| 240 | U_CAPI int32_t U_EXPORT2 |
| 241 | ucol_getBound(const uint8_t *source, |
| 242 | int32_t sourceLength, |
| 243 | UColBoundMode boundType, |
| 244 | uint32_t noOfLevels, |
| 245 | uint8_t *result, |
| 246 | int32_t resultLength, |
| 247 | UErrorCode *status) |
| 248 | { |
| 249 | // consistency checks |
| 250 | if(status == NULL || U_FAILURE(*status)) { |
| 251 | return 0; |
| 252 | } |
| 253 | if(source == NULL) { |
| 254 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
| 255 | return 0; |
| 256 | } |
| 257 | |
| 258 | int32_t sourceIndex = 0; |
| 259 | // Scan the string until we skip enough of the key OR reach the end of the key |
| 260 | do { |
| 261 | sourceIndex++; |
| 262 | if(source[sourceIndex] == Collation::LEVEL_SEPARATOR_BYTE) { |
| 263 | noOfLevels--; |
| 264 | } |
| 265 | } while (noOfLevels > 0 |
| 266 | && (source[sourceIndex] != 0 || sourceIndex < sourceLength)); |
| 267 | |
| 268 | if((source[sourceIndex] == 0 || sourceIndex == sourceLength) |
| 269 | && noOfLevels > 0) { |
| 270 | *status = U_SORT_KEY_TOO_SHORT_WARNING; |
| 271 | } |
| 272 | |
| 273 | |
| 274 | // READ ME: this code assumes that the values for boundType |
| 275 | // enum will not changes. They are set so that the enum value |
| 276 | // corresponds to the number of extra bytes each bound type |
| 277 | // needs. |
| 278 | if(result != NULL && resultLength >= sourceIndex+boundType) { |
| 279 | uprv_memcpy(result, source, sourceIndex); |
| 280 | switch(boundType) { |
| 281 | // Lower bound just gets terminated. No extra bytes |
| 282 | case UCOL_BOUND_LOWER: // = 0 |
| 283 | break; |
| 284 | // Upper bound needs one extra byte |
| 285 | case UCOL_BOUND_UPPER: // = 1 |
| 286 | result[sourceIndex++] = 2; |
| 287 | break; |
| 288 | // Upper long bound needs two extra bytes |
| 289 | case UCOL_BOUND_UPPER_LONG: // = 2 |
| 290 | result[sourceIndex++] = 0xFF; |
| 291 | result[sourceIndex++] = 0xFF; |
| 292 | break; |
| 293 | default: |
| 294 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
| 295 | return 0; |
| 296 | } |
| 297 | result[sourceIndex++] = 0; |
| 298 | |
| 299 | return sourceIndex; |
| 300 | } else { |
| 301 | return sourceIndex+boundType+1; |
| 302 | } |
| 303 | } |
| 304 | |
| 305 | U_CAPI void U_EXPORT2 |
| 306 | ucol_setMaxVariable(UCollator *coll, UColReorderCode group, UErrorCode *pErrorCode) { |
| 307 | if(U_FAILURE(*pErrorCode)) { return; } |
| 308 | Collator::fromUCollator(coll)->setMaxVariable(group, *pErrorCode); |
| 309 | } |
| 310 | |
| 311 | U_CAPI UColReorderCode U_EXPORT2 |
| 312 | ucol_getMaxVariable(const UCollator *coll) { |
| 313 | return Collator::fromUCollator(coll)->getMaxVariable(); |
| 314 | } |
| 315 | |
| 316 | U_CAPI uint32_t U_EXPORT2 |
| 317 | ucol_setVariableTop(UCollator *coll, const UChar *varTop, int32_t len, UErrorCode *status) { |
| 318 | if(U_FAILURE(*status) || coll == NULL) { |
| 319 | return 0; |
| 320 | } |
| 321 | return Collator::fromUCollator(coll)->setVariableTop(varTop, len, *status); |
| 322 | } |
| 323 | |
| 324 | U_CAPI uint32_t U_EXPORT2 ucol_getVariableTop(const UCollator *coll, UErrorCode *status) { |
| 325 | if(U_FAILURE(*status) || coll == NULL) { |
| 326 | return 0; |
| 327 | } |
| 328 | return Collator::fromUCollator(coll)->getVariableTop(*status); |
| 329 | } |
| 330 | |
| 331 | U_CAPI void U_EXPORT2 |
| 332 | ucol_restoreVariableTop(UCollator *coll, const uint32_t varTop, UErrorCode *status) { |
| 333 | if(U_FAILURE(*status) || coll == NULL) { |
| 334 | return; |
| 335 | } |
| 336 | Collator::fromUCollator(coll)->setVariableTop(varTop, *status); |
| 337 | } |
| 338 | |
| 339 | U_CAPI void U_EXPORT2 |
| 340 | ucol_setAttribute(UCollator *coll, UColAttribute attr, UColAttributeValue value, UErrorCode *status) { |
| 341 | if(U_FAILURE(*status) || coll == NULL) { |
| 342 | return; |
| 343 | } |
| 344 | |
| 345 | Collator::fromUCollator(coll)->setAttribute(attr, value, *status); |
| 346 | } |
| 347 | |
| 348 | U_CAPI UColAttributeValue U_EXPORT2 |
| 349 | ucol_getAttribute(const UCollator *coll, UColAttribute attr, UErrorCode *status) { |
| 350 | if(U_FAILURE(*status) || coll == NULL) { |
| 351 | return UCOL_DEFAULT; |
| 352 | } |
| 353 | |
| 354 | return Collator::fromUCollator(coll)->getAttribute(attr, *status); |
| 355 | } |
| 356 | |
| 357 | U_CAPI void U_EXPORT2 |
| 358 | ucol_setStrength( UCollator *coll, |
| 359 | UCollationStrength strength) |
| 360 | { |
| 361 | UErrorCode status = U_ZERO_ERROR; |
| 362 | ucol_setAttribute(coll, UCOL_STRENGTH, strength, &status); |
| 363 | } |
| 364 | |
| 365 | U_CAPI UCollationStrength U_EXPORT2 |
| 366 | ucol_getStrength(const UCollator *coll) |
| 367 | { |
| 368 | UErrorCode status = U_ZERO_ERROR; |
| 369 | return ucol_getAttribute(coll, UCOL_STRENGTH, &status); |
| 370 | } |
| 371 | |
| 372 | U_CAPI int32_t U_EXPORT2 |
| 373 | ucol_getReorderCodes(const UCollator *coll, |
| 374 | int32_t *dest, |
| 375 | int32_t destCapacity, |
| 376 | UErrorCode *status) { |
| 377 | if (U_FAILURE(*status)) { |
| 378 | return 0; |
| 379 | } |
| 380 | |
| 381 | return Collator::fromUCollator(coll)->getReorderCodes(dest, destCapacity, *status); |
| 382 | } |
| 383 | |
| 384 | U_CAPI void U_EXPORT2 |
| 385 | ucol_setReorderCodes(UCollator* coll, |
| 386 | const int32_t* reorderCodes, |
| 387 | int32_t reorderCodesLength, |
| 388 | UErrorCode *status) { |
| 389 | if (U_FAILURE(*status)) { |
| 390 | return; |
| 391 | } |
| 392 | |
| 393 | Collator::fromUCollator(coll)->setReorderCodes(reorderCodes, reorderCodesLength, *status); |
| 394 | } |
| 395 | |
| 396 | U_CAPI int32_t U_EXPORT2 |
| 397 | ucol_getEquivalentReorderCodes(int32_t reorderCode, |
| 398 | int32_t* dest, |
| 399 | int32_t destCapacity, |
| 400 | UErrorCode *pErrorCode) { |
| 401 | return Collator::getEquivalentReorderCodes(reorderCode, dest, destCapacity, *pErrorCode); |
| 402 | } |
| 403 | |
| 404 | U_CAPI void U_EXPORT2 |
| 405 | ucol_getVersion(const UCollator* coll, |
| 406 | UVersionInfo versionInfo) |
| 407 | { |
| 408 | Collator::fromUCollator(coll)->getVersion(versionInfo); |
| 409 | } |
| 410 | |
| 411 | U_CAPI UCollationResult U_EXPORT2 |
| 412 | ucol_strcollIter( const UCollator *coll, |
| 413 | UCharIterator *sIter, |
| 414 | UCharIterator *tIter, |
| 415 | UErrorCode *status) |
| 416 | { |
| 417 | if(!status || U_FAILURE(*status)) { |
| 418 | return UCOL_EQUAL; |
| 419 | } |
| 420 | |
| 421 | UTRACE_ENTRY(UTRACE_UCOL_STRCOLLITER); |
| 422 | UTRACE_DATA3(UTRACE_VERBOSE, "coll=%p, sIter=%p, tIter=%p" , coll, sIter, tIter); |
| 423 | |
| 424 | if(sIter == NULL || tIter == NULL || coll == NULL) { |
| 425 | *status = U_ILLEGAL_ARGUMENT_ERROR; |
| 426 | UTRACE_EXIT_VALUE_STATUS(UCOL_EQUAL, *status); |
| 427 | return UCOL_EQUAL; |
| 428 | } |
| 429 | |
| 430 | UCollationResult result = Collator::fromUCollator(coll)->compare(*sIter, *tIter, *status); |
| 431 | |
| 432 | UTRACE_EXIT_VALUE_STATUS(result, *status); |
| 433 | return result; |
| 434 | } |
| 435 | |
| 436 | |
| 437 | /* */ |
| 438 | /* ucol_strcoll Main public API string comparison function */ |
| 439 | /* */ |
| 440 | U_CAPI UCollationResult U_EXPORT2 |
| 441 | ucol_strcoll( const UCollator *coll, |
| 442 | const UChar *source, |
| 443 | int32_t sourceLength, |
| 444 | const UChar *target, |
| 445 | int32_t targetLength) |
| 446 | { |
| 447 | UTRACE_ENTRY(UTRACE_UCOL_STRCOLL); |
| 448 | if (UTRACE_LEVEL(UTRACE_VERBOSE)) { |
| 449 | UTRACE_DATA3(UTRACE_VERBOSE, "coll=%p, source=%p, target=%p" , coll, source, target); |
| 450 | UTRACE_DATA2(UTRACE_VERBOSE, "source string = %vh " , source, sourceLength); |
| 451 | UTRACE_DATA2(UTRACE_VERBOSE, "target string = %vh " , target, targetLength); |
| 452 | } |
| 453 | |
| 454 | UErrorCode status = U_ZERO_ERROR; |
| 455 | UCollationResult returnVal = Collator::fromUCollator(coll)-> |
| 456 | compare(source, sourceLength, target, targetLength, status); |
| 457 | UTRACE_EXIT_VALUE_STATUS(returnVal, status); |
| 458 | return returnVal; |
| 459 | } |
| 460 | |
| 461 | U_CAPI UCollationResult U_EXPORT2 |
| 462 | ucol_strcollUTF8( |
| 463 | const UCollator *coll, |
| 464 | const char *source, |
| 465 | int32_t sourceLength, |
| 466 | const char *target, |
| 467 | int32_t targetLength, |
| 468 | UErrorCode *status) |
| 469 | { |
| 470 | UTRACE_ENTRY(UTRACE_UCOL_STRCOLLUTF8); |
| 471 | if (UTRACE_LEVEL(UTRACE_VERBOSE)) { |
| 472 | UTRACE_DATA3(UTRACE_VERBOSE, "coll=%p, source=%p, target=%p" , coll, source, target); |
| 473 | UTRACE_DATA2(UTRACE_VERBOSE, "source string = %vb " , source, sourceLength); |
| 474 | UTRACE_DATA2(UTRACE_VERBOSE, "target string = %vb " , target, targetLength); |
| 475 | } |
| 476 | |
| 477 | if (U_FAILURE(*status)) { |
| 478 | /* do nothing */ |
| 479 | UTRACE_EXIT_VALUE_STATUS(UCOL_EQUAL, *status); |
| 480 | return UCOL_EQUAL; |
| 481 | } |
| 482 | |
| 483 | UCollationResult returnVal = Collator::fromUCollator(coll)->internalCompareUTF8( |
| 484 | source, sourceLength, target, targetLength, *status); |
| 485 | UTRACE_EXIT_VALUE_STATUS(returnVal, *status); |
| 486 | return returnVal; |
| 487 | } |
| 488 | |
| 489 | |
| 490 | /* convenience function for comparing strings */ |
| 491 | U_CAPI UBool U_EXPORT2 |
| 492 | ucol_greater( const UCollator *coll, |
| 493 | const UChar *source, |
| 494 | int32_t sourceLength, |
| 495 | const UChar *target, |
| 496 | int32_t targetLength) |
| 497 | { |
| 498 | return (ucol_strcoll(coll, source, sourceLength, target, targetLength) |
| 499 | == UCOL_GREATER); |
| 500 | } |
| 501 | |
| 502 | /* convenience function for comparing strings */ |
| 503 | U_CAPI UBool U_EXPORT2 |
| 504 | ucol_greaterOrEqual( const UCollator *coll, |
| 505 | const UChar *source, |
| 506 | int32_t sourceLength, |
| 507 | const UChar *target, |
| 508 | int32_t targetLength) |
| 509 | { |
| 510 | return (ucol_strcoll(coll, source, sourceLength, target, targetLength) |
| 511 | != UCOL_LESS); |
| 512 | } |
| 513 | |
| 514 | /* convenience function for comparing strings */ |
| 515 | U_CAPI UBool U_EXPORT2 |
| 516 | ucol_equal( const UCollator *coll, |
| 517 | const UChar *source, |
| 518 | int32_t sourceLength, |
| 519 | const UChar *target, |
| 520 | int32_t targetLength) |
| 521 | { |
| 522 | return (ucol_strcoll(coll, source, sourceLength, target, targetLength) |
| 523 | == UCOL_EQUAL); |
| 524 | } |
| 525 | |
| 526 | U_CAPI void U_EXPORT2 |
| 527 | ucol_getUCAVersion(const UCollator* coll, UVersionInfo info) { |
| 528 | const Collator *c = Collator::fromUCollator(coll); |
| 529 | if(c != NULL) { |
| 530 | UVersionInfo v; |
| 531 | c->getVersion(v); |
| 532 | // Note: This is tied to how the current implementation encodes the UCA version |
| 533 | // in the overall getVersion(). |
| 534 | // Alternatively, we could load the root collator and get at lower-level data from there. |
| 535 | // Either way, it will reflect the input collator's UCA version only |
| 536 | // if it is a known implementation. |
| 537 | // It would be cleaner to make this a virtual Collator method. |
| 538 | info[0] = v[1] >> 3; |
| 539 | info[1] = v[1] & 7; |
| 540 | info[2] = v[2] >> 6; |
| 541 | info[3] = 0; |
| 542 | } |
| 543 | } |
| 544 | |
| 545 | U_CAPI const UChar * U_EXPORT2 |
| 546 | ucol_getRules(const UCollator *coll, int32_t *length) { |
| 547 | const RuleBasedCollator *rbc = RuleBasedCollator::rbcFromUCollator(coll); |
| 548 | // OK to crash if coll==NULL: We do not want to check "this" pointers. |
| 549 | if(rbc != NULL || coll == NULL) { |
| 550 | const UnicodeString &rules = rbc->getRules(); |
| 551 | U_ASSERT(rules.getBuffer()[rules.length()] == 0); |
| 552 | *length = rules.length(); |
| 553 | return rules.getBuffer(); |
| 554 | } |
| 555 | static const UChar _NUL = 0; |
| 556 | *length = 0; |
| 557 | return &_NUL; |
| 558 | } |
| 559 | |
| 560 | U_CAPI int32_t U_EXPORT2 |
| 561 | ucol_getRulesEx(const UCollator *coll, UColRuleOption delta, UChar *buffer, int32_t bufferLen) { |
| 562 | UnicodeString rules; |
| 563 | const RuleBasedCollator *rbc = RuleBasedCollator::rbcFromUCollator(coll); |
| 564 | if(rbc != NULL || coll == NULL) { |
| 565 | rbc->getRules(delta, rules); |
| 566 | } |
| 567 | if(buffer != NULL && bufferLen > 0) { |
| 568 | UErrorCode errorCode = U_ZERO_ERROR; |
| 569 | return rules.extract(buffer, bufferLen, errorCode); |
| 570 | } else { |
| 571 | return rules.length(); |
| 572 | } |
| 573 | } |
| 574 | |
| 575 | U_CAPI const char * U_EXPORT2 |
| 576 | ucol_getLocale(const UCollator *coll, ULocDataLocaleType type, UErrorCode *status) { |
| 577 | return ucol_getLocaleByType(coll, type, status); |
| 578 | } |
| 579 | |
| 580 | U_CAPI const char * U_EXPORT2 |
| 581 | ucol_getLocaleByType(const UCollator *coll, ULocDataLocaleType type, UErrorCode *status) { |
| 582 | if(U_FAILURE(*status)) { |
| 583 | return NULL; |
| 584 | } |
| 585 | UTRACE_ENTRY(UTRACE_UCOL_GETLOCALE); |
| 586 | UTRACE_DATA1(UTRACE_INFO, "coll=%p" , coll); |
| 587 | |
| 588 | const char *result; |
| 589 | const RuleBasedCollator *rbc = RuleBasedCollator::rbcFromUCollator(coll); |
| 590 | if(rbc == NULL && coll != NULL) { |
| 591 | *status = U_UNSUPPORTED_ERROR; |
| 592 | result = NULL; |
| 593 | } else { |
| 594 | result = rbc->internalGetLocaleID(type, *status); |
| 595 | } |
| 596 | |
| 597 | UTRACE_DATA1(UTRACE_INFO, "result = %s" , result); |
| 598 | UTRACE_EXIT_STATUS(*status); |
| 599 | return result; |
| 600 | } |
| 601 | |
| 602 | U_CAPI USet * U_EXPORT2 |
| 603 | ucol_getTailoredSet(const UCollator *coll, UErrorCode *status) { |
| 604 | if(U_FAILURE(*status)) { |
| 605 | return NULL; |
| 606 | } |
| 607 | UnicodeSet *set = Collator::fromUCollator(coll)->getTailoredSet(*status); |
| 608 | if(U_FAILURE(*status)) { |
| 609 | delete set; |
| 610 | return NULL; |
| 611 | } |
| 612 | return set->toUSet(); |
| 613 | } |
| 614 | |
| 615 | U_CAPI UBool U_EXPORT2 |
| 616 | ucol_equals(const UCollator *source, const UCollator *target) { |
| 617 | return source == target || |
| 618 | (*Collator::fromUCollator(source)) == (*Collator::fromUCollator(target)); |
| 619 | } |
| 620 | |
| 621 | #endif /* #if !UCONFIG_NO_COLLATION */ |
| 622 | |