1/******************************************************************** 2 * COPYRIGHT: 3 * Copyright (c) 1997-2012, International Business Machines Corporation and 4 * others. All Rights Reserved. 5 ********************************************************************/ 6/******************************************************************************** 7* 8* File CBIAPTS.C 9* 10* Modification History: 11* Name Description 12* Madhu Katragadda Creation 13*********************************************************************************/ 14/*C API TEST FOR BREAKITERATOR */ 15/** 16* This is an API test. It doesn't test very many cases, and doesn't 17* try to test the full functionality. It just calls each function in the class and 18* verifies that it works on a basic level. 19**/ 20 21#include "unicode/utypes.h" 22 23#if !UCONFIG_NO_BREAK_ITERATION 24 25#include <stdlib.h> 26#include <string.h> 27#include "unicode/uloc.h" 28#include "unicode/ubrk.h" 29#include "unicode/ustring.h" 30#include "unicode/ucnv.h" 31#include "unicode/utext.h" 32#include "cintltst.h" 33#include "cbiapts.h" 34 35#define TEST_ASSERT_SUCCESS(status) {if (U_FAILURE(status)) { \ 36log_data_err("Failure at file %s, line %d, error = %s (Are you missing data?)\n", __FILE__, __LINE__, u_errorName(status));}} 37 38#define TEST_ASSERT(expr) {if ((expr)==FALSE) { \ 39log_data_err("Test Failure at file %s, line %d (Are you missing data?)\n", __FILE__, __LINE__);}} 40 41#if !UCONFIG_NO_FILE_IO 42static void TestBreakIteratorSafeClone(void); 43#endif 44static void TestBreakIteratorRules(void); 45static void TestBreakIteratorRuleError(void); 46static void TestBreakIteratorStatusVec(void); 47static void TestBreakIteratorUText(void); 48static void TestBreakIteratorTailoring(void); 49static void TestBreakIteratorRefresh(void); 50 51void addBrkIterAPITest(TestNode** root); 52 53void addBrkIterAPITest(TestNode** root) 54{ 55#if !UCONFIG_NO_FILE_IO 56 addTest(root, &TestBreakIteratorCAPI, "tstxtbd/cbiapts/TestBreakIteratorCAPI"); 57 addTest(root, &TestBreakIteratorSafeClone, "tstxtbd/cbiapts/TestBreakIteratorSafeClone"); 58 addTest(root, &TestBreakIteratorUText, "tstxtbd/cbiapts/TestBreakIteratorUText"); 59#endif 60 addTest(root, &TestBreakIteratorRules, "tstxtbd/cbiapts/TestBreakIteratorRules"); 61 addTest(root, &TestBreakIteratorRuleError, "tstxtbd/cbiapts/TestBreakIteratorRuleError"); 62 addTest(root, &TestBreakIteratorStatusVec, "tstxtbd/cbiapts/TestBreakIteratorStatusVec"); 63 addTest(root, &TestBreakIteratorTailoring, "tstxtbd/cbiapts/TestBreakIteratorTailoring"); 64 addTest(root, &TestBreakIteratorRefresh, "tstxtbd/cbiapts/TestBreakIteratorRefresh"); 65} 66 67#define CLONETEST_ITERATOR_COUNT 2 68 69/* 70 * Utility function for converting char * to UChar * strings, to 71 * simplify the test code. Converted strings are put in heap allocated 72 * storage. A hook (probably a local in the caller's code) allows all 73 * strings converted with that hook to be freed with a single call. 74 */ 75typedef struct StringStruct { 76 struct StringStruct *link; 77 UChar str[1]; 78 } StringStruct; 79 80 81static UChar* toUChar(const char *src, void **freeHook) { 82 /* Structure of the memory that we allocate on the heap */ 83 84 int32_t numUChars; 85 int32_t destSize; 86 UChar stackBuf[2000 + sizeof(void *)/sizeof(UChar)]; 87 StringStruct *dest; 88 UConverter *cnv; 89 90 UErrorCode status = U_ZERO_ERROR; 91 if (src == NULL) { 92 return NULL; 93 }; 94 95 cnv = ucnv_open(NULL, &status); 96 if(U_FAILURE(status) || cnv == NULL) { 97 return NULL; 98 } 99 ucnv_reset(cnv); 100 numUChars = ucnv_toUChars(cnv, 101 stackBuf, 102 2000, 103 src, -1, 104 &status); 105 106 destSize = (numUChars+1) * sizeof(UChar) + sizeof(struct StringStruct); 107 dest = (StringStruct *)malloc(destSize); 108 if (dest != NULL) { 109 if (status == U_BUFFER_OVERFLOW_ERROR || status == U_STRING_NOT_TERMINATED_WARNING) { 110 ucnv_toUChars(cnv, dest->str, numUChars+1, src, -1, &status); 111 } else if (status == U_ZERO_ERROR) { 112 u_strcpy(dest->str, stackBuf); 113 } else { 114 free(dest); 115 dest = NULL; 116 } 117 } 118 119 ucnv_reset(cnv); /* be good citizens */ 120 ucnv_close(cnv); 121 if (dest == NULL) { 122 return NULL; 123 } 124 125 dest->link = (StringStruct*)(*freeHook); 126 *freeHook = dest; 127 return dest->str; 128} 129 130static void freeToUCharStrings(void **hook) { 131 StringStruct *s = *(StringStruct **)hook; 132 while (s != NULL) { 133 StringStruct *next = s->link; 134 free(s); 135 s = next; 136 } 137} 138 139 140#if !UCONFIG_NO_FILE_IO 141static void TestBreakIteratorCAPI() 142{ 143 UErrorCode status = U_ZERO_ERROR; 144 UBreakIterator *word, *sentence, *line, *character, *b, *bogus; 145 int32_t start,pos,end,to; 146 int32_t i; 147 int32_t count = 0; 148 149 UChar text[50]; 150 151 /* Note: the adjacent "" are concatenating strings, not adding a \" to the 152 string, which is probably what whoever wrote this intended. Don't fix, 153 because it would throw off the hard coded break positions in the following 154 tests. */ 155 u_uastrcpy(text, "He's from Africa. ""Mr. Livingston, I presume?"" Yeah"); 156 157 158/*test ubrk_open()*/ 159 log_verbose("\nTesting BreakIterator open functions\n"); 160 161 /* Use french for fun */ 162 word = ubrk_open(UBRK_WORD, "en_US", text, u_strlen(text), &status); 163 if(status == U_FILE_ACCESS_ERROR) { 164 log_data_err("Check your data - it doesn't seem to be around\n"); 165 return; 166 } else if(U_FAILURE(status)){ 167 log_err_status(status, "FAIL: Error in ubrk_open() for word breakiterator: %s\n", myErrorName(status)); 168 } 169 else{ 170 log_verbose("PASS: Successfully opened word breakiterator\n"); 171 } 172 173 sentence = ubrk_open(UBRK_SENTENCE, "en_US", text, u_strlen(text), &status); 174 if(U_FAILURE(status)){ 175 log_err_status(status, "FAIL: Error in ubrk_open() for sentence breakiterator: %s\n", myErrorName(status)); 176 return; 177 } 178 else{ 179 log_verbose("PASS: Successfully opened sentence breakiterator\n"); 180 } 181 182 line = ubrk_open(UBRK_LINE, "en_US", text, u_strlen(text), &status); 183 if(U_FAILURE(status)){ 184 log_err("FAIL: Error in ubrk_open() for line breakiterator: %s\n", myErrorName(status)); 185 return; 186 } 187 else{ 188 log_verbose("PASS: Successfully opened line breakiterator\n"); 189 } 190 191 character = ubrk_open(UBRK_CHARACTER, "en_US", text, u_strlen(text), &status); 192 if(U_FAILURE(status)){ 193 log_err("FAIL: Error in ubrk_open() for character breakiterator: %s\n", myErrorName(status)); 194 return; 195 } 196 else{ 197 log_verbose("PASS: Successfully opened character breakiterator\n"); 198 } 199 /*trying to open an illegal iterator*/ 200 bogus = ubrk_open((UBreakIteratorType)5, "en_US", text, u_strlen(text), &status); 201 if(U_SUCCESS(status)){ 202 log_err("FAIL: Error in ubrk_open() for BOGUS breakiterator. Expected U_ILLEGAL_ARGUMENT_ERROR\n"); 203 } 204 if(U_FAILURE(status)){ 205 if(status != U_ILLEGAL_ARGUMENT_ERROR){ 206 log_err("FAIL: Error in ubrk_open() for BOGUS breakiterator. Expected U_ILLEGAL_ARGUMENT_ERROR\n Got %s\n", myErrorName(status)); 207 } 208 } 209 status=U_ZERO_ERROR; 210 211 212/* ======= Test ubrk_countAvialable() and ubrk_getAvialable() */ 213 214 log_verbose("\nTesting ubrk_countAvailable() and ubrk_getAvailable()\n"); 215 count=ubrk_countAvailable(); 216 /* use something sensible w/o hardcoding the count */ 217 if(count < 0){ 218 log_err("FAIL: Error in ubrk_countAvialable() returned %d\n", count); 219 } 220 else{ 221 log_verbose("PASS: ubrk_countAvialable() successful returned %d\n", count); 222 } 223 for(i=0;i<count;i++) 224 { 225 log_verbose("%s\n", ubrk_getAvailable(i)); 226 if (ubrk_getAvailable(i) == 0) 227 log_err("No locale for which breakiterator is applicable\n"); 228 else 229 log_verbose("A locale %s for which breakiterator is applicable\n",ubrk_getAvailable(i)); 230 } 231 232/*========Test ubrk_first(), ubrk_last()...... and other functions*/ 233 234 log_verbose("\nTesting the functions for word\n"); 235 start = ubrk_first(word); 236 if(start!=0) 237 log_err("error ubrk_start(word) did not return 0\n"); 238 log_verbose("first (word = %d\n", (int32_t)start); 239 pos=ubrk_next(word); 240 if(pos!=4) 241 log_err("error ubrk_next(word) did not return 4\n"); 242 log_verbose("next (word = %d\n", (int32_t)pos); 243 pos=ubrk_following(word, 4); 244 if(pos!=5) 245 log_err("error ubrl_following(word,4) did not return 6\n"); 246 log_verbose("next (word = %d\n", (int32_t)pos); 247 end=ubrk_last(word); 248 if(end!=49) 249 log_err("error ubrk_last(word) did not return 49\n"); 250 log_verbose("last (word = %d\n", (int32_t)end); 251 252 pos=ubrk_previous(word); 253 log_verbose("%d %d\n", end, pos); 254 255 pos=ubrk_previous(word); 256 log_verbose("%d \n", pos); 257 258 if (ubrk_isBoundary(word, 2) != FALSE) { 259 log_err("error ubrk_isBoundary(word, 2) did not return FALSE\n"); 260 } 261 pos=ubrk_current(word); 262 if (pos != 4) { 263 log_err("error ubrk_current() != 4 after ubrk_isBoundary(word, 2)\n"); 264 } 265 if (ubrk_isBoundary(word, 4) != TRUE) { 266 log_err("error ubrk_isBoundary(word, 4) did not return TRUE\n"); 267 } 268 269 270 271 log_verbose("\nTesting the functions for character\n"); 272 ubrk_first(character); 273 pos = ubrk_following(character, 5); 274 if(pos!=6) 275 log_err("error ubrk_following(character,5) did not return 6\n"); 276 log_verbose("Following (character,5) = %d\n", (int32_t)pos); 277 pos=ubrk_following(character, 18); 278 if(pos!=19) 279 log_err("error ubrk_following(character,18) did not return 19\n"); 280 log_verbose("Followingcharacter,18) = %d\n", (int32_t)pos); 281 pos=ubrk_preceding(character, 22); 282 if(pos!=21) 283 log_err("error ubrk_preceding(character,22) did not return 21\n"); 284 log_verbose("preceding(character,22) = %d\n", (int32_t)pos); 285 286 287 log_verbose("\nTesting the functions for line\n"); 288 pos=ubrk_first(line); 289 if(pos != 0) 290 log_err("error ubrk_first(line) returned %d, expected 0\n", (int32_t)pos); 291 pos = ubrk_next(line); 292 pos=ubrk_following(line, 18); 293 if(pos!=22) 294 log_err("error ubrk_following(line) did not return 22\n"); 295 log_verbose("following (line) = %d\n", (int32_t)pos); 296 297 298 log_verbose("\nTesting the functions for sentence\n"); 299 ubrk_first(sentence); 300 pos = ubrk_current(sentence); 301 log_verbose("Current(sentence) = %d\n", (int32_t)pos); 302 pos = ubrk_last(sentence); 303 if(pos!=49) 304 log_err("error ubrk_last for sentence did not return 49\n"); 305 log_verbose("Last (sentence) = %d\n", (int32_t)pos); 306 ubrk_first(sentence); 307 to = ubrk_following( sentence, 0 ); 308 if (to == 0) log_err("ubrk_following returned 0\n"); 309 to = ubrk_preceding( sentence, to ); 310 if (to != 0) log_err("ubrk_preceding didn't return 0\n"); 311 if (ubrk_first(sentence)!=ubrk_current(sentence)) { 312 log_err("error in ubrk_first() or ubrk_current()\n"); 313 } 314 315 316 /*---- */ 317 /*Testing ubrk_open and ubrk_close()*/ 318 log_verbose("\nTesting open and close for us locale\n"); 319 b = ubrk_open(UBRK_WORD, "fr_FR", text, u_strlen(text), &status); 320 if (U_FAILURE(status)) { 321 log_err("ubrk_open for word returned NULL: %s\n", myErrorName(status)); 322 } 323 ubrk_close(b); 324 325 /* Test setText and setUText */ 326 { 327 UChar s1[] = {0x41, 0x42, 0x20, 0}; 328 UChar s2[] = {0x41, 0x42, 0x43, 0x44, 0x45, 0}; 329 UText *ut = NULL; 330 UBreakIterator *bb; 331 int j; 332 333 log_verbose("\nTesting ubrk_setText() and ubrk_setUText()\n"); 334 status = U_ZERO_ERROR; 335 bb = ubrk_open(UBRK_WORD, "en_US", NULL, 0, &status); 336 TEST_ASSERT_SUCCESS(status); 337 ubrk_setText(bb, s1, -1, &status); 338 TEST_ASSERT_SUCCESS(status); 339 ubrk_first(bb); 340 j = ubrk_next(bb); 341 TEST_ASSERT(j == 2); 342 ut = utext_openUChars(ut, s2, -1, &status); 343 ubrk_setUText(bb, ut, &status); 344 TEST_ASSERT_SUCCESS(status); 345 j = ubrk_next(bb); 346 TEST_ASSERT(j == 5); 347 348 ubrk_close(bb); 349 utext_close(ut); 350 } 351 352 ubrk_close(word); 353 ubrk_close(sentence); 354 ubrk_close(line); 355 ubrk_close(character); 356} 357 358static void TestBreakIteratorSafeClone(void) 359{ 360 UChar text[51]; /* Keep this odd to test for 64-bit memory alignment */ 361 /* NOTE: This doesn't reliably force mis-alignment of following items. */ 362 uint8_t buffer [CLONETEST_ITERATOR_COUNT] [U_BRK_SAFECLONE_BUFFERSIZE]; 363 int32_t bufferSize = U_BRK_SAFECLONE_BUFFERSIZE; 364 365 UBreakIterator * someIterators [CLONETEST_ITERATOR_COUNT]; 366 UBreakIterator * someClonedIterators [CLONETEST_ITERATOR_COUNT]; 367 368 UBreakIterator * brk; 369 UErrorCode status = U_ZERO_ERROR; 370 int32_t start,pos; 371 int32_t i; 372 373 /*Testing ubrk_safeClone */ 374 375 /* Note: the adjacent "" are concatenating strings, not adding a \" to the 376 string, which is probably what whoever wrote this intended. Don't fix, 377 because it would throw off the hard coded break positions in the following 378 tests. */ 379 u_uastrcpy(text, "He's from Africa. ""Mr. Livingston, I presume?"" Yeah"); 380 381 /* US & Thai - rule-based & dictionary based */ 382 someIterators[0] = ubrk_open(UBRK_WORD, "en_US", text, u_strlen(text), &status); 383 if(!someIterators[0] || U_FAILURE(status)) { 384 log_data_err("Couldn't open en_US word break iterator - %s\n", u_errorName(status)); 385 return; 386 } 387 388 someIterators[1] = ubrk_open(UBRK_WORD, "th_TH", text, u_strlen(text), &status); 389 if(!someIterators[1] || U_FAILURE(status)) { 390 log_data_err("Couldn't open th_TH word break iterator - %s\n", u_errorName(status)); 391 return; 392 } 393 394 /* test each type of iterator */ 395 for (i = 0; i < CLONETEST_ITERATOR_COUNT; i++) 396 { 397 398 /* Check the various error & informational states */ 399 400 /* Null status - just returns NULL */ 401 if (0 != ubrk_safeClone(someIterators[i], buffer[i], &bufferSize, 0)) 402 { 403 log_err("FAIL: Cloned Iterator failed to deal correctly with null status\n"); 404 } 405 /* error status - should return 0 & keep error the same */ 406 status = U_MEMORY_ALLOCATION_ERROR; 407 if (0 != ubrk_safeClone(someIterators[i], buffer[i], &bufferSize, &status) || status != U_MEMORY_ALLOCATION_ERROR) 408 { 409 log_err("FAIL: Cloned Iterator failed to deal correctly with incoming error status\n"); 410 } 411 status = U_ZERO_ERROR; 412 413 /* Null buffer size pointer - just returns NULL & set error to U_ILLEGAL_ARGUMENT_ERROR*/ 414 if (0 != ubrk_safeClone(someIterators[i], buffer[i], 0, &status) || status != U_ILLEGAL_ARGUMENT_ERROR) 415 { 416 log_err("FAIL: Cloned Iterator failed to deal correctly with null bufferSize pointer\n"); 417 } 418 status = U_ZERO_ERROR; 419 420 /* buffer size pointer is 0 - fill in pbufferSize with a size */ 421 bufferSize = 0; 422 if (0 != ubrk_safeClone(someIterators[i], buffer[i], &bufferSize, &status) || U_FAILURE(status) || bufferSize <= 0) 423 { 424 log_err("FAIL: Cloned Iterator failed a sizing request ('preflighting')\n"); 425 } 426 /* Verify our define is large enough */ 427 if (U_BRK_SAFECLONE_BUFFERSIZE < bufferSize) 428 { 429 log_err("FAIL: Pre-calculated buffer size is too small - %d but needed %d\n", U_BRK_SAFECLONE_BUFFERSIZE, bufferSize); 430 } 431 /* Verify we can use this run-time calculated size */ 432 if (0 == (brk = ubrk_safeClone(someIterators[i], buffer[i], &bufferSize, &status)) || U_FAILURE(status)) 433 { 434 log_err("FAIL: Iterator can't be cloned with run-time size\n"); 435 } 436 if (brk) 437 ubrk_close(brk); 438 /* size one byte too small - should allocate & let us know */ 439 --bufferSize; 440 if (0 == (brk = ubrk_safeClone(someIterators[i], 0, &bufferSize, &status)) || status != U_SAFECLONE_ALLOCATED_WARNING) 441 { 442 log_err("FAIL: Cloned Iterator failed to deal correctly with too-small buffer size\n"); 443 } 444 if (brk) 445 ubrk_close(brk); 446 status = U_ZERO_ERROR; 447 bufferSize = U_BRK_SAFECLONE_BUFFERSIZE; 448 449 /* Null buffer pointer - return Iterator & set error to U_SAFECLONE_ALLOCATED_ERROR */ 450 if (0 == (brk = ubrk_safeClone(someIterators[i], 0, &bufferSize, &status)) || status != U_SAFECLONE_ALLOCATED_WARNING) 451 { 452 log_err("FAIL: Cloned Iterator failed to deal correctly with null buffer pointer\n"); 453 } 454 if (brk) 455 ubrk_close(brk); 456 status = U_ZERO_ERROR; 457 458 /* Mis-aligned buffer pointer. */ 459 { 460 char stackBuf[U_BRK_SAFECLONE_BUFFERSIZE+sizeof(void *)]; 461 void *p; 462 int32_t offset; 463 464 brk = ubrk_safeClone(someIterators[i], &stackBuf[1], &bufferSize, &status); 465 if (U_FAILURE(status) || brk == 0) { 466 log_err("FAIL: Cloned Iterator failed with misaligned buffer pointer\n"); 467 } 468 if (status == U_SAFECLONE_ALLOCATED_WARNING) { 469 log_err("FAIL: Cloned Iterator allocated when using a mis-aligned buffer.\n"); 470 } 471 offset = (int32_t)((char *)&p-(char*)brk); 472 if (offset < 0) { 473 offset = -offset; 474 } 475 if (offset % sizeof(void *) != 0) { 476 log_err("FAIL: Cloned Iterator failed to align correctly with misaligned buffer pointer\n"); 477 } 478 if (brk) 479 ubrk_close(brk); 480 } 481 482 483 /* Null Iterator - return NULL & set U_ILLEGAL_ARGUMENT_ERROR */ 484 if (0 != ubrk_safeClone(0, buffer[i], &bufferSize, &status) || status != U_ILLEGAL_ARGUMENT_ERROR) 485 { 486 log_err("FAIL: Cloned Iterator failed to deal correctly with null Iterator pointer\n"); 487 } 488 status = U_ZERO_ERROR; 489 490 /* Do these cloned Iterators work at all - make a first & next call */ 491 bufferSize = U_BRK_SAFECLONE_BUFFERSIZE; 492 someClonedIterators[i] = ubrk_safeClone(someIterators[i], buffer[i], &bufferSize, &status); 493 494 start = ubrk_first(someClonedIterators[i]); 495 if(start!=0) 496 log_err("error ubrk_start(clone) did not return 0\n"); 497 pos=ubrk_next(someClonedIterators[i]); 498 if(pos!=4) 499 log_err("error ubrk_next(clone) did not return 4\n"); 500 501 ubrk_close(someClonedIterators[i]); 502 ubrk_close(someIterators[i]); 503 } 504} 505#endif 506 507 508/* 509// Open a break iterator from char * rules. Take care of conversion 510// of the rules and error checking. 511*/ 512static UBreakIterator * testOpenRules(char *rules) { 513 UErrorCode status = U_ZERO_ERROR; 514 UChar *ruleSourceU = NULL; 515 void *strCleanUp = NULL; 516 UParseError parseErr; 517 UBreakIterator *bi; 518 519 ruleSourceU = toUChar(rules, &strCleanUp); 520 521 bi = ubrk_openRules(ruleSourceU, -1, /* The rules */ 522 NULL, -1, /* The text to be iterated over. */ 523 &parseErr, &status); 524 525 if (U_FAILURE(status)) { 526 log_data_err("FAIL: ubrk_openRules: ICU Error \"%s\" (Are you missing data?)\n", u_errorName(status)); 527 bi = 0; 528 }; 529 freeToUCharStrings(&strCleanUp); 530 return bi; 531 532} 533 534/* 535 * TestBreakIteratorRules - Verify that a break iterator can be created from 536 * a set of source rules. 537 */ 538static void TestBreakIteratorRules() { 539 /* Rules will keep together any run of letters not including 'a', OR 540 * keep together 'abc', but only when followed by 'def', OTHERWISE 541 * just return one char at a time. 542 */ 543 char rules[] = "abc{666}/def;\n [\\p{L} - [a]]* {2}; . {1};"; 544 /* 0123456789012345678 */ 545 char data[] = "abcdex abcdefgh-def"; /* the test data string */ 546 char breaks[] = "** ** * ** *"; /* * the expected break positions */ 547 char tags[] = "01 21 6 21 2"; /* expected tag values at break positions */ 548 int32_t tagMap[] = {0, 1, 2, 3, 4, 5, 666}; 549 550 UChar *uData; 551 void *freeHook = NULL; 552 UErrorCode status = U_ZERO_ERROR; 553 int32_t pos; 554 int i; 555 556 UBreakIterator *bi = testOpenRules(rules); 557 if (bi == NULL) {return;} 558 uData = toUChar(data, &freeHook); 559 ubrk_setText(bi, uData, -1, &status); 560 561 pos = ubrk_first(bi); 562 for (i=0; i<sizeof(breaks); i++) { 563 if (pos == i && breaks[i] != '*') { 564 log_err("FAIL: unexpected break at position %d found\n", pos); 565 break; 566 } 567 if (pos != i && breaks[i] == '*') { 568 log_err("FAIL: expected break at position %d not found.\n", i); 569 break; 570 } 571 if (pos == i) { 572 int32_t tag, expectedTag; 573 tag = ubrk_getRuleStatus(bi); 574 expectedTag = tagMap[tags[i]&0xf]; 575 if (tag != expectedTag) { 576 log_err("FAIL: incorrect tag value. Position = %d; expected tag %d, got %d", 577 pos, expectedTag, tag); 578 break; 579 } 580 pos = ubrk_next(bi); 581 } 582 } 583 584 freeToUCharStrings(&freeHook); 585 ubrk_close(bi); 586} 587 588static void TestBreakIteratorRuleError() { 589/* 590 * TestBreakIteratorRuleError - Try to create a BI from rules with syntax errors, 591 * check that the error is reported correctly. 592 */ 593 char rules[] = " # This is a rule comment on line 1\n" 594 "[:L:]; # this rule is OK.\n" 595 "abcdefg); # Error, mismatched parens\n"; 596 UChar *uRules; 597 void *freeHook = NULL; 598 UErrorCode status = U_ZERO_ERROR; 599 UParseError parseErr; 600 UBreakIterator *bi; 601 602 uRules = toUChar(rules, &freeHook); 603 bi = ubrk_openRules(uRules, -1, /* The rules */ 604 NULL, -1, /* The text to be iterated over. */ 605 &parseErr, &status); 606 if (U_SUCCESS(status)) { 607 log_err("FAIL: construction of break iterator succeeded when it should have failed.\n"); 608 ubrk_close(bi); 609 } else { 610 if (parseErr.line != 3 || parseErr.offset != 8) { 611 log_data_err("FAIL: incorrect error position reported. Got line %d, char %d, expected line 3, char 7 (Are you missing data?)\n", 612 parseErr.line, parseErr.offset); 613 } 614 } 615 freeToUCharStrings(&freeHook); 616} 617 618 619/* 620* TestsBreakIteratorStatusVals() Test the ubrk_getRuleStatusVec() funciton 621*/ 622static void TestBreakIteratorStatusVec() { 623 #define RULE_STRING_LENGTH 200 624 UChar rules[RULE_STRING_LENGTH]; 625 626 #define TEST_STRING_LENGTH 25 627 UChar testString[TEST_STRING_LENGTH]; 628 UBreakIterator *bi = NULL; 629 int32_t pos = 0; 630 int32_t vals[10]; 631 int32_t numVals; 632 UErrorCode status = U_ZERO_ERROR; 633 634 u_uastrncpy(rules, "[A-N]{100}; \n" 635 "[a-w]{200}; \n" 636 "[\\p{L}]{300}; \n" 637 "[\\p{N}]{400}; \n" 638 "[0-5]{500}; \n" 639 "!.*;\n", RULE_STRING_LENGTH); 640 u_uastrncpy(testString, "ABC", TEST_STRING_LENGTH); 641 642 643 bi = ubrk_openRules(rules, -1, testString, -1, NULL, &status); 644 TEST_ASSERT_SUCCESS(status); 645 TEST_ASSERT(bi != NULL); 646 647 /* The TEST_ASSERT above should change too... */ 648 if (bi != NULL) { 649 pos = ubrk_next(bi); 650 TEST_ASSERT(pos == 1); 651 652 memset(vals, -1, sizeof(vals)); 653 numVals = ubrk_getRuleStatusVec(bi, vals, 10, &status); 654 TEST_ASSERT_SUCCESS(status); 655 TEST_ASSERT(numVals == 2); 656 TEST_ASSERT(vals[0] == 100); 657 TEST_ASSERT(vals[1] == 300); 658 TEST_ASSERT(vals[2] == -1); 659 660 numVals = ubrk_getRuleStatusVec(bi, vals, 0, &status); 661 TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); 662 TEST_ASSERT(numVals == 2); 663 } 664 665 ubrk_close(bi); 666} 667 668 669/* 670 * static void TestBreakIteratorUText(void); 671 * 672 * Test that ubrk_setUText() is present and works for a simple case. 673 */ 674static void TestBreakIteratorUText(void) { 675 const char *UTF8Str = "\x41\xc3\x85\x5A\x20\x41\x52\x69\x6E\x67"; /* c3 85 is utf-8 for A with a ring on top */ 676 /* 0 1 2 34567890 */ 677 678 UErrorCode status = U_ZERO_ERROR; 679 UBreakIterator *bi = NULL; 680 int32_t pos = 0; 681 682 683 UText *ut = utext_openUTF8(NULL, UTF8Str, -1, &status); 684 TEST_ASSERT_SUCCESS(status); 685 686 bi = ubrk_open(UBRK_WORD, "en_US", NULL, 0, &status); 687 if (U_FAILURE(status)) { 688 log_err_status(status, "Failure at file %s, line %d, error = %s\n", __FILE__, __LINE__, u_errorName(status)); 689 return; 690 } 691 692 ubrk_setUText(bi, ut, &status); 693 if (U_FAILURE(status)) { 694 log_err("Failure at file %s, line %d, error = %s\n", __FILE__, __LINE__, u_errorName(status)); 695 return; 696 } 697 698 pos = ubrk_first(bi); 699 TEST_ASSERT(pos == 0); 700 701 pos = ubrk_next(bi); 702 TEST_ASSERT(pos == 4); 703 704 pos = ubrk_next(bi); 705 TEST_ASSERT(pos == 5); 706 707 pos = ubrk_next(bi); 708 TEST_ASSERT(pos == 10); 709 710 pos = ubrk_next(bi); 711 TEST_ASSERT(pos == UBRK_DONE); 712 ubrk_close(bi); 713 utext_close(ut); 714} 715 716/* 717 * static void TestBreakIteratorTailoring(void); 718 * 719 * Test break iterator tailorings from CLDR data. 720 */ 721 722/* Thai/Lao grapheme break tailoring */ 723static const UChar thTest[] = { 0x0020, 0x0E40, 0x0E01, 0x0020, 724 0x0E01, 0x0E30, 0x0020, 0x0E01, 0x0E33, 0x0020, 0 }; 725/*in Unicode 6.1 en should behave just like th for this*/ 726/*static const int32_t thTestOffs_enFwd[] = { 1, 3, 4, 6, 7, 9, 10 };*/ 727static const int32_t thTestOffs_thFwd[] = { 1, 2, 3, 4, 5, 6, 7, 9, 10 }; 728/*static const int32_t thTestOffs_enRev[] = { 9, 7, 6, 4, 3, 1, 0 };*/ 729static const int32_t thTestOffs_thRev[] = { 9, 7, 6, 5, 4, 3, 2, 1, 0 }; 730 731/* Hebrew line break tailoring, for cldrbug 3028 */ 732static const UChar heTest[] = { 0x0020, 0x002D, 0x0031, 0x0032, 0x0020, 733 0x0061, 0x002D, 0x006B, 0x0020, 734 0x0061, 0x0300, 0x2010, 0x006B, 0x0020, 735 0x05DE, 0x05D4, 0x002D, 0x0069, 0x0020, 736 0x05D1, 0x05BC, 0x2010, 0x0047, 0x0020, 0 }; 737/*in Unicode 6.1 en should behave just like he for this*/ 738/*static const int32_t heTestOffs_enFwd[] = { 1, 5, 7, 9, 12, 14, 17, 19, 22, 24 };*/ 739static const int32_t heTestOffs_heFwd[] = { 1, 5, 7, 9, 12, 14, 19, 24 }; 740/*static const int32_t heTestOffs_enRev[] = { 22, 19, 17, 14, 12, 9, 7, 5, 1, 0 };*/ 741static const int32_t heTestOffs_heRev[] = { 19, 14, 12, 9, 7, 5, 1, 0 }; 742 743/* Finnish line break tailoring, for cldrbug 3029 */ 744static const UChar fiTest[] = { /* 00 */ 0x0020, 0x002D, 0x0031, 0x0032, 0x0020, 745 /* 05 */ 0x0061, 0x002D, 0x006B, 0x0020, 746 /* 09 */ 0x0061, 0x0300, 0x2010, 0x006B, 0x0020, 747 /* 14 */ 0x0061, 0x0020, 0x002D, 0x006B, 0x0020, 748 /* 19 */ 0x0061, 0x0300, 0x0020, 0x2010, 0x006B, 0x0020, 0 }; 749static const int32_t fiTestOffs_enFwd[] = { 1, 5, 7, 9, 12, 14, 16, 17, 19, 22, 23, 25 }; 750static const int32_t fiTestOffs_fiFwd[] = { 1, 5, 7, 9, 12, 14, 16, 19, 22, 25 }; 751static const int32_t fiTestOffs_enRev[] = { 23, 22, 19, 17, 16, 14, 12, 9, 7, 5, 1, 0 }; 752static const int32_t fiTestOffs_fiRev[] = { 22, 19, 16, 14, 12, 9, 7, 5, 1, 0 }; 753 754/* Khmer dictionary-based work break, for ICU ticket #8329 */ 755static const UChar kmTest[] = { /* 00 */ 0x179F, 0x17BC, 0x1798, 0x1785, 0x17C6, 0x178E, 0x17B6, 0x1799, 0x1796, 0x17C1, 756 /* 10 */ 0x179B, 0x1794, 0x1793, 0x17D2, 0x178F, 0x17B7, 0x1785, 0x178A, 0x17BE, 0x1798, 757 /* 20 */ 0x17D2, 0x1794, 0x17B8, 0x17A2, 0x1792, 0x17B7, 0x179F, 0x17D2, 0x178B, 0x17B6, 758 /* 30 */ 0x1793, 0x17A2, 0x179A, 0x1796, 0x17D2, 0x179A, 0x17C7, 0x1782, 0x17BB, 0x178E, 759 /* 40 */ 0x178A, 0x179B, 0x17CB, 0x1796, 0x17D2, 0x179A, 0x17C7, 0x17A2, 0x1784, 0x17D2, 760 /* 50 */ 0x1782, 0 }; 761static const int32_t kmTestOffs_kmFwd[] = { 3, /*8,*/ 11, 17, 23, 31, /*33,*/ 40, 43, 51 }; /* TODO: Investigate failure to break at offset 8 */ 762static const int32_t kmTestOffs_kmRev[] = { 43, 40, /*33,*/ 31, 23, 17, 11, /*8,*/ 3, 0 }; 763 764typedef struct { 765 const char * locale; 766 UBreakIteratorType type; 767 const UChar * test; 768 const int32_t * offsFwd; 769 const int32_t * offsRev; 770 int32_t numOffsets; 771} RBBITailoringTest; 772 773static const RBBITailoringTest tailoringTests[] = { 774 { "en", UBRK_CHARACTER, thTest, thTestOffs_thFwd, thTestOffs_thRev, sizeof(thTestOffs_thFwd)/sizeof(thTestOffs_thFwd[0]) }, 775 { "en_US_POSIX", UBRK_CHARACTER, thTest, thTestOffs_thFwd, thTestOffs_thRev, sizeof(thTestOffs_thFwd)/sizeof(thTestOffs_thFwd[0]) }, 776 { "en", UBRK_LINE, heTest, heTestOffs_heFwd, heTestOffs_heRev, sizeof(heTestOffs_heFwd)/sizeof(heTestOffs_heFwd[0]) }, 777 { "he", UBRK_LINE, heTest, heTestOffs_heFwd, heTestOffs_heRev, sizeof(heTestOffs_heFwd)/sizeof(heTestOffs_heFwd[0]) }, 778 { "en", UBRK_LINE, fiTest, fiTestOffs_enFwd, fiTestOffs_enRev, sizeof(fiTestOffs_enFwd)/sizeof(fiTestOffs_enFwd[0]) }, 779 { "fi", UBRK_LINE, fiTest, fiTestOffs_fiFwd, fiTestOffs_fiRev, sizeof(fiTestOffs_fiFwd)/sizeof(fiTestOffs_fiFwd[0]) }, 780 { "km", UBRK_WORD, kmTest, kmTestOffs_kmFwd, kmTestOffs_kmRev, sizeof(kmTestOffs_kmFwd)/sizeof(kmTestOffs_kmFwd[0]) }, 781 { NULL, 0, NULL, NULL, NULL, 0 }, 782}; 783 784static void TestBreakIteratorTailoring(void) { 785 const RBBITailoringTest * testPtr; 786 for (testPtr = tailoringTests; testPtr->locale != NULL; ++testPtr) { 787 UErrorCode status = U_ZERO_ERROR; 788 UBreakIterator* ubrkiter = ubrk_open(testPtr->type, testPtr->locale, testPtr->test, -1, &status); 789 if ( U_SUCCESS(status) ) { 790 int32_t offset, offsindx; 791 UBool foundError; 792 793 foundError = FALSE; 794 for (offsindx = 0; (offset = ubrk_next(ubrkiter)) != UBRK_DONE; ++offsindx) { 795 if (!foundError && offsindx >= testPtr->numOffsets) { 796 log_err("FAIL: locale %s, break type %d, ubrk_next expected UBRK_DONE, got %d\n", 797 testPtr->locale, testPtr->type, offset); 798 foundError = TRUE; 799 } else if (!foundError && offset != testPtr->offsFwd[offsindx]) { 800 log_err("FAIL: locale %s, break type %d, ubrk_next expected %d, got %d\n", 801 testPtr->locale, testPtr->type, testPtr->offsFwd[offsindx], offset); 802 foundError = TRUE; 803 } 804 } 805 if (!foundError && offsindx < testPtr->numOffsets) { 806 log_err("FAIL: locale %s, break type %d, ubrk_next expected %d, got UBRK_DONE\n", 807 testPtr->locale, testPtr->type, testPtr->offsFwd[offsindx]); 808 } 809 810 foundError = FALSE; 811 for (offsindx = 0; (offset = ubrk_previous(ubrkiter)) != UBRK_DONE; ++offsindx) { 812 if (!foundError && offsindx >= testPtr->numOffsets) { 813 log_err("FAIL: locale %s, break type %d, ubrk_previous expected UBRK_DONE, got %d\n", 814 testPtr->locale, testPtr->type, offset); 815 foundError = TRUE; 816 } else if (!foundError && offset != testPtr->offsRev[offsindx]) { 817 log_err("FAIL: locale %s, break type %d, ubrk_previous expected %d, got %d\n", 818 testPtr->locale, testPtr->type, testPtr->offsRev[offsindx], offset); 819 foundError = TRUE; 820 } 821 } 822 if (!foundError && offsindx < testPtr->numOffsets) { 823 log_err("FAIL: locale %s, break type %d, ubrk_previous expected %d, got UBRK_DONE\n", 824 testPtr->locale, testPtr->type, testPtr->offsRev[offsindx]); 825 } 826 827 ubrk_close(ubrkiter); 828 } else { 829 log_err_status(status, "FAIL: locale %s, break type %d, ubrk_open status: %s\n", testPtr->locale, testPtr->type, u_errorName(status)); 830 } 831 } 832} 833 834 835static void TestBreakIteratorRefresh(void) { 836 /* 837 * RefreshInput changes out the input of a Break Iterator without 838 * changing anything else in the iterator's state. Used with Java JNI, 839 * when Java moves the underlying string storage. This test 840 * runs a ubrk_next() repeatedly, moving the text in the middle of the sequence. 841 * The right set of boundaries should still be found. 842 */ 843 UChar testStr[] = {0x20, 0x41, 0x20, 0x42, 0x20, 0x43, 0x20, 0x44, 0x0}; /* = " A B C D" */ 844 UChar movedStr[] = {0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0}; 845 UErrorCode status = U_ZERO_ERROR; 846 UBreakIterator *bi; 847 UText ut1 = UTEXT_INITIALIZER; 848 UText ut2 = UTEXT_INITIALIZER; 849 850 bi = ubrk_open(UBRK_LINE, "en_US", NULL, 0, &status); 851 TEST_ASSERT_SUCCESS(status); 852 853 utext_openUChars(&ut1, testStr, -1, &status); 854 TEST_ASSERT_SUCCESS(status); 855 ubrk_setUText(bi, &ut1, &status); 856 TEST_ASSERT_SUCCESS(status); 857 858 if (U_SUCCESS(status)) { 859 /* Line boundaries will occur before each letter in the original string */ 860 TEST_ASSERT(1 == ubrk_next(bi)); 861 TEST_ASSERT(3 == ubrk_next(bi)); 862 863 /* Move the string, kill the original string. */ 864 u_strcpy(movedStr, testStr); 865 u_memset(testStr, 0x20, u_strlen(testStr)); 866 utext_openUChars(&ut2, movedStr, -1, &status); 867 TEST_ASSERT_SUCCESS(status); 868 ubrk_refreshUText(bi, &ut2, &status); 869 TEST_ASSERT_SUCCESS(status); 870 871 /* Find the following matches, now working in the moved string. */ 872 TEST_ASSERT(5 == ubrk_next(bi)); 873 TEST_ASSERT(7 == ubrk_next(bi)); 874 TEST_ASSERT(8 == ubrk_next(bi)); 875 TEST_ASSERT(UBRK_DONE == ubrk_next(bi)); 876 TEST_ASSERT_SUCCESS(status); 877 878 utext_close(&ut1); 879 utext_close(&ut2); 880 } 881 ubrk_close(bi); 882} 883 884#endif /* #if !UCONFIG_NO_BREAK_ITERATION */ 885