1 /******************************************************************** 2 * COPYRIGHT: 3 * Copyright (c) 2005-2012, International Business Machines Corporation and 4 * others. All Rights Reserved. 5 ********************************************************************/ 6 /************************************************************************ 7 * Tests for the UText and UTextIterator text abstraction classses 8 * 9 ************************************************************************/ 10 11 #include <string.h> 12 #include <stdio.h> 13 #include <stdlib.h> 14 #include "unicode/utypes.h" 15 #include "unicode/utext.h" 16 #include "unicode/utf8.h" 17 #include "unicode/ustring.h" 18 #include "unicode/uchriter.h" 19 #include "utxttest.h" 20 21 static UBool gFailed = FALSE; 22 static int gTestNum = 0; 23 24 // Forward decl 25 UText *openFragmentedUnicodeString(UText *ut, UnicodeString *s, UErrorCode *status); 26 27 #define TEST_ASSERT(x) \ 28 { if ((x)==FALSE) {errln("Test #%d failure in file %s at line %d\n", gTestNum, __FILE__, __LINE__);\ 29 gFailed = TRUE;\ 30 }} 31 32 33 #define TEST_SUCCESS(status) \ 34 { if (U_FAILURE(status)) {errln("Test #%d failure in file %s at line %d. Error = \"%s\"\n", \ 35 gTestNum, __FILE__, __LINE__, u_errorName(status)); \ 36 gFailed = TRUE;\ 37 }} 38 39 UTextTest::UTextTest() { 40 } 41 42 UTextTest::~UTextTest() { 43 } 44 45 46 void 47 UTextTest::runIndexedTest(int32_t index, UBool exec, 48 const char* &name, char* /*par*/) { 49 switch (index) { 50 case 0: name = "TextTest"; 51 if (exec) TextTest(); break; 52 case 1: name = "ErrorTest"; 53 if (exec) ErrorTest(); break; 54 case 2: name = "FreezeTest"; 55 if (exec) FreezeTest(); break; 56 case 3: name = "Ticket5560"; 57 if (exec) Ticket5560(); break; 58 case 4: name = "Ticket6847"; 59 if (exec) Ticket6847(); break; 60 default: name = ""; break; 61 } 62 } 63 64 // 65 // Quick and dirty random number generator. 66 // (don't use library so that results are portable. 67 static uint32_t m_seed = 1; 68 static uint32_t m_rand() 69 { 70 m_seed = m_seed * 1103515245 + 12345; 71 return (uint32_t)(m_seed/65536) % 32768; 72 } 73 74 75 // 76 // TextTest() 77 // 78 // Top Level function for UText testing. 79 // Specifies the strings to be tested, with the acutal testing itself 80 // being carried out in another function, TestString(). 81 // 82 void UTextTest::TextTest() { 83 int32_t i, j; 84 85 TestString("abcd\\U00010001xyz"); 86 TestString(""); 87 88 // Supplementary chars at start or end 89 TestString("\\U00010001"); 90 TestString("abc\\U00010001"); 91 TestString("\\U00010001abc"); 92 93 // Test simple strings of lengths 1 to 60, looking for glitches at buffer boundaries 94 UnicodeString s; 95 for (i=1; i<60; i++) { 96 s.truncate(0); 97 for (j=0; j<i; j++) { 98 if (j+0x30 == 0x5c) { 99 // backslash. Needs to be escaped 100 s.append((UChar)0x5c); 101 } 102 s.append(UChar(j+0x30)); 103 } 104 TestString(s); 105 } 106 107 // Test strings with odd-aligned supplementary chars, 108 // looking for glitches at buffer boundaries 109 for (i=1; i<60; i++) { 110 s.truncate(0); 111 s.append((UChar)0x41); 112 for (j=0; j<i; j++) { 113 s.append(UChar32(j+0x11000)); 114 } 115 TestString(s); 116 } 117 118 // String of chars of randomly varying size in utf-8 representation. 119 // Exercise the mapping, and the varying sized buffer. 120 // 121 s.truncate(0); 122 UChar32 c1 = 0; 123 UChar32 c2 = 0x100; 124 UChar32 c3 = 0xa000; 125 UChar32 c4 = 0x11000; 126 for (i=0; i<1000; i++) { 127 int len8 = m_rand()%4 + 1; 128 switch (len8) { 129 case 1: 130 c1 = (c1+1)%0x80; 131 // don't put 0 into string (0 terminated strings for some tests) 132 // don't put '\', will cause unescape() to fail. 133 if (c1==0x5c || c1==0) { 134 c1++; 135 } 136 s.append(c1); 137 break; 138 case 2: 139 s.append(c2++); 140 break; 141 case 3: 142 s.append(c3++); 143 break; 144 case 4: 145 s.append(c4++); 146 break; 147 } 148 } 149 TestString(s); 150 } 151 152 153 // 154 // TestString() Run a suite of UText tests on a string. 155 // The test string is unescaped before use. 156 // 157 void UTextTest::TestString(const UnicodeString &s) { 158 int32_t i; 159 int32_t j; 160 UChar32 c; 161 int32_t cpCount = 0; 162 UErrorCode status = U_ZERO_ERROR; 163 UText *ut = NULL; 164 int32_t saLen; 165 166 UnicodeString sa = s.unescape(); 167 saLen = sa.length(); 168 169 // 170 // Build up a mapping between code points and UTF-16 code unit indexes. 171 // 172 m *cpMap = new m[sa.length() + 1]; 173 j = 0; 174 for (i=0; i<sa.length(); i=sa.moveIndex32(i, 1)) { 175 c = sa.char32At(i); 176 cpMap[j].nativeIdx = i; 177 cpMap[j].cp = c; 178 j++; 179 cpCount++; 180 } 181 cpMap[j].nativeIdx = i; // position following the last char in utf-16 string. 182 183 184 // UChar * test, null terminated 185 status = U_ZERO_ERROR; 186 UChar *buf = new UChar[saLen+1]; 187 sa.extract(buf, saLen+1, status); 188 TEST_SUCCESS(status); 189 ut = utext_openUChars(NULL, buf, -1, &status); 190 TEST_SUCCESS(status); 191 TestAccess(sa, ut, cpCount, cpMap); 192 utext_close(ut); 193 delete [] buf; 194 195 // UChar * test, with length 196 status = U_ZERO_ERROR; 197 buf = new UChar[saLen+1]; 198 sa.extract(buf, saLen+1, status); 199 TEST_SUCCESS(status); 200 ut = utext_openUChars(NULL, buf, saLen, &status); 201 TEST_SUCCESS(status); 202 TestAccess(sa, ut, cpCount, cpMap); 203 utext_close(ut); 204 delete [] buf; 205 206 207 // UnicodeString test 208 status = U_ZERO_ERROR; 209 ut = utext_openUnicodeString(NULL, &sa, &status); 210 TEST_SUCCESS(status); 211 TestAccess(sa, ut, cpCount, cpMap); 212 TestCMR(sa, ut, cpCount, cpMap, cpMap); 213 utext_close(ut); 214 215 216 // Const UnicodeString test 217 status = U_ZERO_ERROR; 218 ut = utext_openConstUnicodeString(NULL, &sa, &status); 219 TEST_SUCCESS(status); 220 TestAccess(sa, ut, cpCount, cpMap); 221 utext_close(ut); 222 223 224 // Replaceable test. (UnicodeString inherits Replaceable) 225 status = U_ZERO_ERROR; 226 ut = utext_openReplaceable(NULL, &sa, &status); 227 TEST_SUCCESS(status); 228 TestAccess(sa, ut, cpCount, cpMap); 229 TestCMR(sa, ut, cpCount, cpMap, cpMap); 230 utext_close(ut); 231 232 // Character Iterator Tests 233 status = U_ZERO_ERROR; 234 const UChar *cbuf = sa.getBuffer(); 235 CharacterIterator *ci = new UCharCharacterIterator(cbuf, saLen, status); 236 TEST_SUCCESS(status); 237 ut = utext_openCharacterIterator(NULL, ci, &status); 238 TEST_SUCCESS(status); 239 TestAccess(sa, ut, cpCount, cpMap); 240 utext_close(ut); 241 delete ci; 242 243 244 // Fragmented UnicodeString (Chunk size of one) 245 // 246 status = U_ZERO_ERROR; 247 ut = openFragmentedUnicodeString(NULL, &sa, &status); 248 TEST_SUCCESS(status); 249 TestAccess(sa, ut, cpCount, cpMap); 250 utext_close(ut); 251 252 // 253 // UTF-8 test 254 // 255 256 // Convert the test string from UnicodeString to (char *) in utf-8 format 257 int32_t u8Len = sa.extract(0, sa.length(), NULL, 0, "utf-8"); 258 char *u8String = new char[u8Len + 1]; 259 sa.extract(0, sa.length(), u8String, u8Len+1, "utf-8"); 260 261 // Build up the map of code point indices in the utf-8 string 262 m * u8Map = new m[sa.length() + 1]; 263 i = 0; // native utf-8 index 264 for (j=0; j<cpCount ; j++) { // code point number 265 u8Map[j].nativeIdx = i; 266 U8_NEXT(u8String, i, u8Len, c) 267 u8Map[j].cp = c; 268 } 269 u8Map[cpCount].nativeIdx = u8Len; // position following the last char in utf-8 string. 270 271 // Do the test itself 272 status = U_ZERO_ERROR; 273 ut = utext_openUTF8(NULL, u8String, -1, &status); 274 TEST_SUCCESS(status); 275 TestAccess(sa, ut, cpCount, u8Map); 276 utext_close(ut); 277 278 279 280 delete []cpMap; 281 delete []u8Map; 282 delete []u8String; 283 } 284 285 // TestCMR test Copy, Move and Replace operations. 286 // us UnicodeString containing the test text. 287 // ut UText containing the same test text. 288 // cpCount number of code points in the test text. 289 // nativeMap Mapping from code points to native indexes for the UText. 290 // u16Map Mapping from code points to UTF-16 indexes, for use with the UnicodeString. 291 // 292 // This function runs a whole series of opertions on each incoming UText. 293 // The UText is deep-cloned prior to each operation, so that the original UText remains unchanged. 294 // 295 void UTextTest::TestCMR(const UnicodeString &us, UText *ut, int cpCount, m *nativeMap, m *u16Map) { 296 TEST_ASSERT(utext_isWritable(ut) == TRUE); 297 298 int srcLengthType; // Loop variables for selecting the postion and length 299 int srcPosType; // of the block to operate on within the source text. 300 int destPosType; 301 302 int srcIndex = 0; // Code Point indexes of the block to operate on for 303 int srcLength = 0; // a specific test. 304 305 int destIndex = 0; // Code point index of the destination for a copy/move test. 306 307 int32_t nativeStart = 0; // Native unit indexes for a test. 308 int32_t nativeLimit = 0; 309 int32_t nativeDest = 0; 310 311 int32_t u16Start = 0; // UTF-16 indexes for a test. 312 int32_t u16Limit = 0; // used when performing the same operation in a Unicode String 313 int32_t u16Dest = 0; 314 315 // Iterate over a whole series of source index, length and a target indexes. 316 // This is done with code point indexes; these will be later translated to native 317 // indexes using the cpMap. 318 for (srcLengthType=1; srcLengthType<=3; srcLengthType++) { 319 switch (srcLengthType) { 320 case 1: srcLength = 1; break; 321 case 2: srcLength = 5; break; 322 case 3: srcLength = cpCount / 3; 323 } 324 for (srcPosType=1; srcPosType<=5; srcPosType++) { 325 switch (srcPosType) { 326 case 1: srcIndex = 0; break; 327 case 2: srcIndex = 1; break; 328 case 3: srcIndex = cpCount - srcLength; break; 329 case 4: srcIndex = cpCount - srcLength - 1; break; 330 case 5: srcIndex = cpCount / 2; break; 331 } 332 if (srcIndex < 0 || srcIndex + srcLength > cpCount) { 333 // filter out bogus test cases - 334 // those with a source range that falls of an edge of the string. 335 continue; 336 } 337 338 // 339 // Copy and move tests. 340 // iterate over a variety of destination positions. 341 // 342 for (destPosType=1; destPosType<=4; destPosType++) { 343 switch (destPosType) { 344 case 1: destIndex = 0; break; 345 case 2: destIndex = 1; break; 346 case 3: destIndex = srcIndex - 1; break; 347 case 4: destIndex = srcIndex + srcLength + 1; break; 348 case 5: destIndex = cpCount-1; break; 349 case 6: destIndex = cpCount; break; 350 } 351 if (destIndex<0 || destIndex>cpCount) { 352 // filter out bogus test cases. 353 continue; 354 } 355 356 nativeStart = nativeMap[srcIndex].nativeIdx; 357 nativeLimit = nativeMap[srcIndex+srcLength].nativeIdx; 358 nativeDest = nativeMap[destIndex].nativeIdx; 359 360 u16Start = u16Map[srcIndex].nativeIdx; 361 u16Limit = u16Map[srcIndex+srcLength].nativeIdx; 362 u16Dest = u16Map[destIndex].nativeIdx; 363 364 gFailed = FALSE; 365 TestCopyMove(us, ut, FALSE, 366 nativeStart, nativeLimit, nativeDest, 367 u16Start, u16Limit, u16Dest); 368 369 TestCopyMove(us, ut, TRUE, 370 nativeStart, nativeLimit, nativeDest, 371 u16Start, u16Limit, u16Dest); 372 373 if (gFailed) { 374 return; 375 } 376 } 377 378 // 379 // Replace tests. 380 // 381 UnicodeString fullRepString("This is an arbitrary string that will be used as replacement text"); 382 for (int32_t replStrLen=0; replStrLen<20; replStrLen++) { 383 UnicodeString repStr(fullRepString, 0, replStrLen); 384 TestReplace(us, ut, 385 nativeStart, nativeLimit, 386 u16Start, u16Limit, 387 repStr); 388 if (gFailed) { 389 return; 390 } 391 } 392 393 } 394 } 395 396 } 397 398 // 399 // TestCopyMove run a single test case for utext_copy. 400 // Test cases are created in TestCMR and dispatched here for execution. 401 // 402 void UTextTest::TestCopyMove(const UnicodeString &us, UText *ut, UBool move, 403 int32_t nativeStart, int32_t nativeLimit, int32_t nativeDest, 404 int32_t u16Start, int32_t u16Limit, int32_t u16Dest) 405 { 406 UErrorCode status = U_ZERO_ERROR; 407 UText *targetUT = NULL; 408 gTestNum++; 409 gFailed = FALSE; 410 411 // 412 // clone the UText. The test will be run in the cloned copy 413 // so that we don't alter the original. 414 // 415 targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status); 416 TEST_SUCCESS(status); 417 UnicodeString targetUS(us); // And copy the reference string. 418 419 // do the test operation first in the reference 420 targetUS.copy(u16Start, u16Limit, u16Dest); 421 if (move) { 422 // delete out the source range. 423 if (u16Limit < u16Dest) { 424 targetUS.removeBetween(u16Start, u16Limit); 425 } else { 426 int32_t amtCopied = u16Limit - u16Start; 427 targetUS.removeBetween(u16Start+amtCopied, u16Limit+amtCopied); 428 } 429 } 430 431 // Do the same operation in the UText under test 432 utext_copy(targetUT, nativeStart, nativeLimit, nativeDest, move, &status); 433 if (nativeDest > nativeStart && nativeDest < nativeLimit) { 434 TEST_ASSERT(status == U_INDEX_OUTOFBOUNDS_ERROR); 435 } else { 436 TEST_SUCCESS(status); 437 438 // Compare the results of the two parallel tests 439 int32_t usi = 0; // UnicodeString postion, utf-16 index. 440 int64_t uti = 0; // UText position, native index. 441 int32_t cpi; // char32 position (code point index) 442 UChar32 usc; // code point from Unicode String 443 UChar32 utc; // code point from UText 444 utext_setNativeIndex(targetUT, 0); 445 for (cpi=0; ; cpi++) { 446 usc = targetUS.char32At(usi); 447 utc = utext_next32(targetUT); 448 if (utc < 0) { 449 break; 450 } 451 TEST_ASSERT(uti == usi); 452 TEST_ASSERT(utc == usc); 453 usi = targetUS.moveIndex32(usi, 1); 454 uti = utext_getNativeIndex(targetUT); 455 if (gFailed) { 456 goto cleanupAndReturn; 457 } 458 } 459 int64_t expectedNativeLength = utext_nativeLength(ut); 460 if (move == FALSE) { 461 expectedNativeLength += nativeLimit - nativeStart; 462 } 463 uti = utext_getNativeIndex(targetUT); 464 TEST_ASSERT(uti == expectedNativeLength); 465 } 466 467 cleanupAndReturn: 468 utext_close(targetUT); 469 } 470 471 472 // 473 // TestReplace Test a single Replace operation. 474 // 475 void UTextTest::TestReplace( 476 const UnicodeString &us, // reference UnicodeString in which to do the replace 477 UText *ut, // UnicodeText object under test. 478 int32_t nativeStart, // Range to be replaced, in UText native units. 479 int32_t nativeLimit, 480 int32_t u16Start, // Range to be replaced, in UTF-16 units 481 int32_t u16Limit, // for use in the reference UnicodeString. 482 const UnicodeString &repStr) // The replacement string 483 { 484 UErrorCode status = U_ZERO_ERROR; 485 UText *targetUT = NULL; 486 gTestNum++; 487 gFailed = FALSE; 488 489 // 490 // clone the target UText. The test will be run in the cloned copy 491 // so that we don't alter the original. 492 // 493 targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status); 494 TEST_SUCCESS(status); 495 UnicodeString targetUS(us); // And copy the reference string. 496 497 // 498 // Do the replace operation in the Unicode String, to 499 // produce a reference result. 500 // 501 targetUS.replace(u16Start, u16Limit-u16Start, repStr); 502 503 // 504 // Do the replace on the UText under test 505 // 506 const UChar *rs = repStr.getBuffer(); 507 int32_t rsLen = repStr.length(); 508 int32_t actualDelta = utext_replace(targetUT, nativeStart, nativeLimit, rs, rsLen, &status); 509 int32_t expectedDelta = repStr.length() - (nativeLimit - nativeStart); 510 TEST_ASSERT(actualDelta == expectedDelta); 511 512 // 513 // Compare the results 514 // 515 int32_t usi = 0; // UnicodeString postion, utf-16 index. 516 int64_t uti = 0; // UText position, native index. 517 int32_t cpi; // char32 position (code point index) 518 UChar32 usc; // code point from Unicode String 519 UChar32 utc; // code point from UText 520 int64_t expectedNativeLength = 0; 521 utext_setNativeIndex(targetUT, 0); 522 for (cpi=0; ; cpi++) { 523 usc = targetUS.char32At(usi); 524 utc = utext_next32(targetUT); 525 if (utc < 0) { 526 break; 527 } 528 TEST_ASSERT(uti == usi); 529 TEST_ASSERT(utc == usc); 530 usi = targetUS.moveIndex32(usi, 1); 531 uti = utext_getNativeIndex(targetUT); 532 if (gFailed) { 533 goto cleanupAndReturn; 534 } 535 } 536 expectedNativeLength = utext_nativeLength(ut) + expectedDelta; 537 uti = utext_getNativeIndex(targetUT); 538 TEST_ASSERT(uti == expectedNativeLength); 539 540 cleanupAndReturn: 541 utext_close(targetUT); 542 } 543 544 // 545 // TestAccess Test the read only access functions on a UText, including cloning. 546 // The text is accessed in a variety of ways, and compared with 547 // the reference UnicodeString. 548 // 549 void UTextTest::TestAccess(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) { 550 // Run the standard tests on the caller-supplied UText. 551 TestAccessNoClone(us, ut, cpCount, cpMap); 552 553 // Re-run tests on a shallow clone. 554 utext_setNativeIndex(ut, 0); 555 UErrorCode status = U_ZERO_ERROR; 556 UText *shallowClone = utext_clone(NULL, ut, FALSE /*deep*/, FALSE /*readOnly*/, &status); 557 TEST_SUCCESS(status); 558 TestAccessNoClone(us, shallowClone, cpCount, cpMap); 559 560 // 561 // Rerun again on a deep clone. 562 // Note that text providers are not required to provide deep cloning, 563 // so unsupported errors are ignored. 564 // 565 status = U_ZERO_ERROR; 566 utext_setNativeIndex(shallowClone, 0); 567 UText *deepClone = utext_clone(NULL, shallowClone, TRUE, FALSE, &status); 568 utext_close(shallowClone); 569 if (status != U_UNSUPPORTED_ERROR) { 570 TEST_SUCCESS(status); 571 TestAccessNoClone(us, deepClone, cpCount, cpMap); 572 } 573 utext_close(deepClone); 574 } 575 576 577 // 578 // TestAccessNoClone() Test the read only access functions on a UText. 579 // The text is accessed in a variety of ways, and compared with 580 // the reference UnicodeString. 581 // 582 void UTextTest::TestAccessNoClone(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) { 583 UErrorCode status = U_ZERO_ERROR; 584 gTestNum++; 585 586 // 587 // Check the length from the UText 588 // 589 int64_t expectedLen = cpMap[cpCount].nativeIdx; 590 int64_t utlen = utext_nativeLength(ut); 591 TEST_ASSERT(expectedLen == utlen); 592 593 // 594 // Iterate forwards, verify that we get the correct code points 595 // at the correct native offsets. 596 // 597 int i = 0; 598 int64_t index; 599 int64_t expectedIndex = 0; 600 int64_t foundIndex = 0; 601 UChar32 expectedC; 602 UChar32 foundC; 603 int64_t len; 604 605 for (i=0; i<cpCount; i++) { 606 expectedIndex = cpMap[i].nativeIdx; 607 foundIndex = utext_getNativeIndex(ut); 608 TEST_ASSERT(expectedIndex == foundIndex); 609 expectedC = cpMap[i].cp; 610 foundC = utext_next32(ut); 611 TEST_ASSERT(expectedC == foundC); 612 foundIndex = utext_getPreviousNativeIndex(ut); 613 TEST_ASSERT(expectedIndex == foundIndex); 614 if (gFailed) { 615 return; 616 } 617 } 618 foundC = utext_next32(ut); 619 TEST_ASSERT(foundC == U_SENTINEL); 620 621 // Repeat above, using macros 622 utext_setNativeIndex(ut, 0); 623 for (i=0; i<cpCount; i++) { 624 expectedIndex = cpMap[i].nativeIdx; 625 foundIndex = UTEXT_GETNATIVEINDEX(ut); 626 TEST_ASSERT(expectedIndex == foundIndex); 627 expectedC = cpMap[i].cp; 628 foundC = UTEXT_NEXT32(ut); 629 TEST_ASSERT(expectedC == foundC); 630 if (gFailed) { 631 return; 632 } 633 } 634 foundC = UTEXT_NEXT32(ut); 635 TEST_ASSERT(foundC == U_SENTINEL); 636 637 // 638 // Forward iteration (above) should have left index at the 639 // end of the input, which should == length(). 640 // 641 len = utext_nativeLength(ut); 642 foundIndex = utext_getNativeIndex(ut); 643 TEST_ASSERT(len == foundIndex); 644 645 // 646 // Iterate backwards over entire test string 647 // 648 len = utext_getNativeIndex(ut); 649 utext_setNativeIndex(ut, len); 650 for (i=cpCount-1; i>=0; i--) { 651 expectedC = cpMap[i].cp; 652 expectedIndex = cpMap[i].nativeIdx; 653 int64_t prevIndex = utext_getPreviousNativeIndex(ut); 654 foundC = utext_previous32(ut); 655 foundIndex = utext_getNativeIndex(ut); 656 TEST_ASSERT(expectedIndex == foundIndex); 657 TEST_ASSERT(expectedC == foundC); 658 TEST_ASSERT(prevIndex == foundIndex); 659 if (gFailed) { 660 return; 661 } 662 } 663 664 // 665 // Backwards iteration, above, should have left our iterator 666 // position at zero, and continued backwards iterationshould fail. 667 // 668 foundIndex = utext_getNativeIndex(ut); 669 TEST_ASSERT(foundIndex == 0); 670 foundIndex = utext_getPreviousNativeIndex(ut); 671 TEST_ASSERT(foundIndex == 0); 672 673 674 foundC = utext_previous32(ut); 675 TEST_ASSERT(foundC == U_SENTINEL); 676 foundIndex = utext_getNativeIndex(ut); 677 TEST_ASSERT(foundIndex == 0); 678 foundIndex = utext_getPreviousNativeIndex(ut); 679 TEST_ASSERT(foundIndex == 0); 680 681 682 // And again, with the macros 683 utext_setNativeIndex(ut, len); 684 for (i=cpCount-1; i>=0; i--) { 685 expectedC = cpMap[i].cp; 686 expectedIndex = cpMap[i].nativeIdx; 687 foundC = UTEXT_PREVIOUS32(ut); 688 foundIndex = UTEXT_GETNATIVEINDEX(ut); 689 TEST_ASSERT(expectedIndex == foundIndex); 690 TEST_ASSERT(expectedC == foundC); 691 if (gFailed) { 692 return; 693 } 694 } 695 696 // 697 // Backwards iteration, above, should have left our iterator 698 // position at zero, and continued backwards iterationshould fail. 699 // 700 foundIndex = UTEXT_GETNATIVEINDEX(ut); 701 TEST_ASSERT(foundIndex == 0); 702 703 foundC = UTEXT_PREVIOUS32(ut); 704 TEST_ASSERT(foundC == U_SENTINEL); 705 foundIndex = UTEXT_GETNATIVEINDEX(ut); 706 TEST_ASSERT(foundIndex == 0); 707 if (gFailed) { 708 return; 709 } 710 711 // 712 // next32From(), prevous32From(), Iterate in a somewhat random order. 713 // 714 int cpIndex = 0; 715 for (i=0; i<cpCount; i++) { 716 cpIndex = (cpIndex + 9973) % cpCount; 717 index = cpMap[cpIndex].nativeIdx; 718 expectedC = cpMap[cpIndex].cp; 719 foundC = utext_next32From(ut, index); 720 TEST_ASSERT(expectedC == foundC); 721 if (gFailed) { 722 return; 723 } 724 } 725 726 cpIndex = 0; 727 for (i=0; i<cpCount; i++) { 728 cpIndex = (cpIndex + 9973) % cpCount; 729 index = cpMap[cpIndex+1].nativeIdx; 730 expectedC = cpMap[cpIndex].cp; 731 foundC = utext_previous32From(ut, index); 732 TEST_ASSERT(expectedC == foundC); 733 if (gFailed) { 734 return; 735 } 736 } 737 738 739 // 740 // moveIndex(int32_t delta); 741 // 742 743 // Walk through frontwards, incrementing by one 744 utext_setNativeIndex(ut, 0); 745 for (i=1; i<=cpCount; i++) { 746 utext_moveIndex32(ut, 1); 747 index = utext_getNativeIndex(ut); 748 expectedIndex = cpMap[i].nativeIdx; 749 TEST_ASSERT(expectedIndex == index); 750 index = UTEXT_GETNATIVEINDEX(ut); 751 TEST_ASSERT(expectedIndex == index); 752 } 753 754 // Walk through frontwards, incrementing by two 755 utext_setNativeIndex(ut, 0); 756 for (i=2; i<cpCount; i+=2) { 757 utext_moveIndex32(ut, 2); 758 index = utext_getNativeIndex(ut); 759 expectedIndex = cpMap[i].nativeIdx; 760 TEST_ASSERT(expectedIndex == index); 761 index = UTEXT_GETNATIVEINDEX(ut); 762 TEST_ASSERT(expectedIndex == index); 763 } 764 765 // walk through the string backwards, decrementing by one. 766 i = cpMap[cpCount].nativeIdx; 767 utext_setNativeIndex(ut, i); 768 for (i=cpCount; i>=0; i--) { 769 expectedIndex = cpMap[i].nativeIdx; 770 index = utext_getNativeIndex(ut); 771 TEST_ASSERT(expectedIndex == index); 772 index = UTEXT_GETNATIVEINDEX(ut); 773 TEST_ASSERT(expectedIndex == index); 774 utext_moveIndex32(ut, -1); 775 } 776 777 778 // walk through backwards, decrementing by three 779 i = cpMap[cpCount].nativeIdx; 780 utext_setNativeIndex(ut, i); 781 for (i=cpCount; i>=0; i-=3) { 782 expectedIndex = cpMap[i].nativeIdx; 783 index = utext_getNativeIndex(ut); 784 TEST_ASSERT(expectedIndex == index); 785 index = UTEXT_GETNATIVEINDEX(ut); 786 TEST_ASSERT(expectedIndex == index); 787 utext_moveIndex32(ut, -3); 788 } 789 790 791 // 792 // Extract 793 // 794 int bufSize = us.length() + 10; 795 UChar *buf = new UChar[bufSize]; 796 status = U_ZERO_ERROR; 797 expectedLen = us.length(); 798 len = utext_extract(ut, 0, utlen, buf, bufSize, &status); 799 TEST_SUCCESS(status); 800 TEST_ASSERT(len == expectedLen); 801 int compareResult = us.compare(buf, -1); 802 TEST_ASSERT(compareResult == 0); 803 804 status = U_ZERO_ERROR; 805 len = utext_extract(ut, 0, utlen, NULL, 0, &status); 806 if (utlen == 0) { 807 TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING); 808 } else { 809 TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); 810 } 811 TEST_ASSERT(len == expectedLen); 812 813 status = U_ZERO_ERROR; 814 u_memset(buf, 0x5555, bufSize); 815 len = utext_extract(ut, 0, utlen, buf, 1, &status); 816 if (us.length() == 0) { 817 TEST_SUCCESS(status); 818 TEST_ASSERT(buf[0] == 0); 819 } else { 820 // Buf len == 1, extracting a single 16 bit value. 821 // If the data char is supplementary, it doesn't matter whether the buffer remains unchanged, 822 // or whether the lead surrogate of the pair is extracted. 823 // It's a buffer overflow error in either case. 824 TEST_ASSERT(buf[0] == us.charAt(0) || 825 (buf[0] == 0x5555 && U_IS_SUPPLEMENTARY(us.char32At(0)))); 826 TEST_ASSERT(buf[1] == 0x5555); 827 if (us.length() == 1) { 828 TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING); 829 } else { 830 TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); 831 } 832 } 833 834 delete []buf; 835 } 836 837 // 838 // ErrorTest() Check various error and edge cases. 839 // 840 void UTextTest::ErrorTest() 841 { 842 // Close of an unitialized UText. Shouldn't blow up. 843 { 844 UText ut; 845 memset(&ut, 0, sizeof(UText)); 846 utext_close(&ut); 847 utext_close(NULL); 848 } 849 850 // Double-close of a UText. Shouldn't blow up. UText should still be usable. 851 { 852 UErrorCode status = U_ZERO_ERROR; 853 UText ut = UTEXT_INITIALIZER; 854 UnicodeString s("Hello, World"); 855 UText *ut2 = utext_openUnicodeString(&ut, &s, &status); 856 TEST_SUCCESS(status); 857 TEST_ASSERT(ut2 == &ut); 858 859 UText *ut3 = utext_close(&ut); 860 TEST_ASSERT(ut3 == &ut); 861 862 UText *ut4 = utext_close(&ut); 863 TEST_ASSERT(ut4 == &ut); 864 865 utext_openUnicodeString(&ut, &s, &status); 866 TEST_SUCCESS(status); 867 utext_close(&ut); 868 } 869 870 // Re-use of a UText, chaining through each of the types of UText 871 // (If it doesn't blow up, and doesn't leak, it's probably working fine) 872 { 873 UErrorCode status = U_ZERO_ERROR; 874 UText ut = UTEXT_INITIALIZER; 875 UText *utp; 876 UnicodeString s1("Hello, World"); 877 UChar s2[] = {(UChar)0x41, (UChar)0x42, (UChar)0}; 878 const char *s3 = "\x66\x67\x68"; 879 880 utp = utext_openUnicodeString(&ut, &s1, &status); 881 TEST_SUCCESS(status); 882 TEST_ASSERT(utp == &ut); 883 884 utp = utext_openConstUnicodeString(&ut, &s1, &status); 885 TEST_SUCCESS(status); 886 TEST_ASSERT(utp == &ut); 887 888 utp = utext_openUTF8(&ut, s3, -1, &status); 889 TEST_SUCCESS(status); 890 TEST_ASSERT(utp == &ut); 891 892 utp = utext_openUChars(&ut, s2, -1, &status); 893 TEST_SUCCESS(status); 894 TEST_ASSERT(utp == &ut); 895 896 utp = utext_close(&ut); 897 TEST_ASSERT(utp == &ut); 898 899 utp = utext_openUnicodeString(&ut, &s1, &status); 900 TEST_SUCCESS(status); 901 TEST_ASSERT(utp == &ut); 902 } 903 904 // Invalid parameters on open 905 // 906 { 907 UErrorCode status = U_ZERO_ERROR; 908 UText ut = UTEXT_INITIALIZER; 909 910 utext_openUChars(&ut, NULL, 5, &status); 911 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR); 912 913 status = U_ZERO_ERROR; 914 utext_openUChars(&ut, NULL, -1, &status); 915 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR); 916 917 status = U_ZERO_ERROR; 918 utext_openUTF8(&ut, NULL, 4, &status); 919 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR); 920 921 status = U_ZERO_ERROR; 922 utext_openUTF8(&ut, NULL, -1, &status); 923 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR); 924 } 925 926 // 927 // UTF-8 with malformed sequences. 928 // These should come through as the Unicode replacement char, \ufffd 929 // 930 { 931 UErrorCode status = U_ZERO_ERROR; 932 UText *ut = NULL; 933 const char *badUTF8 = "\x41\x81\x42\xf0\x81\x81\x43"; 934 UChar32 c; 935 936 ut = utext_openUTF8(NULL, badUTF8, -1, &status); 937 TEST_SUCCESS(status); 938 c = utext_char32At(ut, 1); 939 TEST_ASSERT(c == 0xfffd); 940 c = utext_char32At(ut, 3); 941 TEST_ASSERT(c == 0xfffd); 942 c = utext_char32At(ut, 5); 943 TEST_ASSERT(c == 0xfffd); 944 c = utext_char32At(ut, 6); 945 TEST_ASSERT(c == 0x43); 946 947 UChar buf[10]; 948 int n = utext_extract(ut, 0, 9, buf, 10, &status); 949 TEST_SUCCESS(status); 950 TEST_ASSERT(n==5); 951 TEST_ASSERT(buf[1] == 0xfffd); 952 TEST_ASSERT(buf[3] == 0xfffd); 953 TEST_ASSERT(buf[2] == 0x42); 954 utext_close(ut); 955 } 956 957 958 // 959 // isLengthExpensive - does it make the exptected transitions after 960 // getting the length of a nul terminated string? 961 // 962 { 963 UErrorCode status = U_ZERO_ERROR; 964 UnicodeString sa("Hello, this is a string"); 965 UBool isExpensive; 966 967 UChar sb[100]; 968 memset(sb, 0x20, sizeof(sb)); 969 sb[99] = 0; 970 971 UText *uta = utext_openUnicodeString(NULL, &sa, &status); 972 TEST_SUCCESS(status); 973 isExpensive = utext_isLengthExpensive(uta); 974 TEST_ASSERT(isExpensive == FALSE); 975 utext_close(uta); 976 977 UText *utb = utext_openUChars(NULL, sb, -1, &status); 978 TEST_SUCCESS(status); 979 isExpensive = utext_isLengthExpensive(utb); 980 TEST_ASSERT(isExpensive == TRUE); 981 int64_t len = utext_nativeLength(utb); 982 TEST_ASSERT(len == 99); 983 isExpensive = utext_isLengthExpensive(utb); 984 TEST_ASSERT(isExpensive == FALSE); 985 utext_close(utb); 986 } 987 988 // 989 // Index to positions not on code point boundaries. 990 // 991 { 992 const char *u8str = "\xc8\x81\xe1\x82\x83\xf1\x84\x85\x86"; 993 int32_t startMap[] = { 0, 0, 2, 2, 2, 5, 5, 5, 5, 9, 9}; 994 int32_t nextMap[] = { 2, 2, 5, 5, 5, 9, 9, 9, 9, 9, 9}; 995 int32_t prevMap[] = { 0, 0, 0, 0, 0, 2, 2, 2, 2, 5, 5}; 996 UChar32 c32Map[] = {0x201, 0x201, 0x1083, 0x1083, 0x1083, 0x044146, 0x044146, 0x044146, 0x044146, -1, -1}; 997 UChar32 pr32Map[] = { -1, -1, 0x201, 0x201, 0x201, 0x1083, 0x1083, 0x1083, 0x1083, 0x044146, 0x044146}; 998 999 // extractLen is the size, in UChars, of what will be extracted between index and index+1. 1000 // is zero when both index positions lie within the same code point. 1001 int32_t exLen[] = { 0, 1, 0, 0, 1, 0, 0, 0, 2, 0, 0}; 1002 1003 1004 UErrorCode status = U_ZERO_ERROR; 1005 UText *ut = utext_openUTF8(NULL, u8str, -1, &status); 1006 TEST_SUCCESS(status); 1007 1008 // Check setIndex 1009 int32_t i; 1010 int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t); 1011 for (i=0; i<startMapLimit; i++) { 1012 utext_setNativeIndex(ut, i); 1013 int64_t cpIndex = utext_getNativeIndex(ut); 1014 TEST_ASSERT(cpIndex == startMap[i]); 1015 cpIndex = UTEXT_GETNATIVEINDEX(ut); 1016 TEST_ASSERT(cpIndex == startMap[i]); 1017 } 1018 1019 // Check char32At 1020 for (i=0; i<startMapLimit; i++) { 1021 UChar32 c32 = utext_char32At(ut, i); 1022 TEST_ASSERT(c32 == c32Map[i]); 1023 int64_t cpIndex = utext_getNativeIndex(ut); 1024 TEST_ASSERT(cpIndex == startMap[i]); 1025 } 1026 1027 // Check utext_next32From 1028 for (i=0; i<startMapLimit; i++) { 1029 UChar32 c32 = utext_next32From(ut, i); 1030 TEST_ASSERT(c32 == c32Map[i]); 1031 int64_t cpIndex = utext_getNativeIndex(ut); 1032 TEST_ASSERT(cpIndex == nextMap[i]); 1033 } 1034 1035 // check utext_previous32From 1036 for (i=0; i<startMapLimit; i++) { 1037 gTestNum++; 1038 UChar32 c32 = utext_previous32From(ut, i); 1039 TEST_ASSERT(c32 == pr32Map[i]); 1040 int64_t cpIndex = utext_getNativeIndex(ut); 1041 TEST_ASSERT(cpIndex == prevMap[i]); 1042 } 1043 1044 // check Extract 1045 // Extract from i to i+1, which may be zero or one code points, 1046 // depending on whether the indices straddle a cp boundary. 1047 for (i=0; i<startMapLimit; i++) { 1048 UChar buf[3]; 1049 status = U_ZERO_ERROR; 1050 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status); 1051 TEST_SUCCESS(status); 1052 TEST_ASSERT(extractedLen == exLen[i]); 1053 if (extractedLen > 0) { 1054 UChar32 c32; 1055 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */ 1056 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32); 1057 TEST_ASSERT(c32 == c32Map[i]); 1058 } 1059 } 1060 1061 utext_close(ut); 1062 } 1063 1064 1065 { // Similar test, with utf16 instead of utf8 1066 // TODO: merge the common parts of these tests. 1067 1068 UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV); 1069 int32_t startMap[] ={ 0, 1, 1, 3, 4, 4, 6, 6}; 1070 int32_t nextMap[] = { 1, 3, 3, 4, 6, 6, 6, 6}; 1071 int32_t prevMap[] = { 0, 0, 0, 1, 3, 3, 4, 4}; 1072 UChar32 c32Map[] = {0x1000, 0x11000, 0x11000, 0x2000, 0x22000, 0x22000, -1, -1}; 1073 UChar32 pr32Map[] = { -1, 0x1000, 0x1000, 0x11000, 0x2000, 0x2000, 0x22000, 0x22000}; 1074 int32_t exLen[] = { 1, 0, 2, 1, 0, 2, 0, 0,}; 1075 1076 u16str = u16str.unescape(); 1077 UErrorCode status = U_ZERO_ERROR; 1078 UText *ut = utext_openUnicodeString(NULL, &u16str, &status); 1079 TEST_SUCCESS(status); 1080 1081 int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t); 1082 int i; 1083 for (i=0; i<startMapLimit; i++) { 1084 utext_setNativeIndex(ut, i); 1085 int64_t cpIndex = utext_getNativeIndex(ut); 1086 TEST_ASSERT(cpIndex == startMap[i]); 1087 } 1088 1089 // Check char32At 1090 for (i=0; i<startMapLimit; i++) { 1091 UChar32 c32 = utext_char32At(ut, i); 1092 TEST_ASSERT(c32 == c32Map[i]); 1093 int64_t cpIndex = utext_getNativeIndex(ut); 1094 TEST_ASSERT(cpIndex == startMap[i]); 1095 } 1096 1097 // Check utext_next32From 1098 for (i=0; i<startMapLimit; i++) { 1099 UChar32 c32 = utext_next32From(ut, i); 1100 TEST_ASSERT(c32 == c32Map[i]); 1101 int64_t cpIndex = utext_getNativeIndex(ut); 1102 TEST_ASSERT(cpIndex == nextMap[i]); 1103 } 1104 1105 // check utext_previous32From 1106 for (i=0; i<startMapLimit; i++) { 1107 UChar32 c32 = utext_previous32From(ut, i); 1108 TEST_ASSERT(c32 == pr32Map[i]); 1109 int64_t cpIndex = utext_getNativeIndex(ut); 1110 TEST_ASSERT(cpIndex == prevMap[i]); 1111 } 1112 1113 // check Extract 1114 // Extract from i to i+1, which may be zero or one code points, 1115 // depending on whether the indices straddle a cp boundary. 1116 for (i=0; i<startMapLimit; i++) { 1117 UChar buf[3]; 1118 status = U_ZERO_ERROR; 1119 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status); 1120 TEST_SUCCESS(status); 1121 TEST_ASSERT(extractedLen == exLen[i]); 1122 if (extractedLen > 0) { 1123 UChar32 c32; 1124 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */ 1125 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32); 1126 TEST_ASSERT(c32 == c32Map[i]); 1127 } 1128 } 1129 1130 utext_close(ut); 1131 } 1132 1133 { // Similar test, with UText over Replaceable 1134 // TODO: merge the common parts of these tests. 1135 1136 UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV); 1137 int32_t startMap[] ={ 0, 1, 1, 3, 4, 4, 6, 6}; 1138 int32_t nextMap[] = { 1, 3, 3, 4, 6, 6, 6, 6}; 1139 int32_t prevMap[] = { 0, 0, 0, 1, 3, 3, 4, 4}; 1140 UChar32 c32Map[] = {0x1000, 0x11000, 0x11000, 0x2000, 0x22000, 0x22000, -1, -1}; 1141 UChar32 pr32Map[] = { -1, 0x1000, 0x1000, 0x11000, 0x2000, 0x2000, 0x22000, 0x22000}; 1142 int32_t exLen[] = { 1, 0, 2, 1, 0, 2, 0, 0,}; 1143 1144 u16str = u16str.unescape(); 1145 UErrorCode status = U_ZERO_ERROR; 1146 UText *ut = utext_openReplaceable(NULL, &u16str, &status); 1147 TEST_SUCCESS(status); 1148 1149 int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t); 1150 int i; 1151 for (i=0; i<startMapLimit; i++) { 1152 utext_setNativeIndex(ut, i); 1153 int64_t cpIndex = utext_getNativeIndex(ut); 1154 TEST_ASSERT(cpIndex == startMap[i]); 1155 } 1156 1157 // Check char32At 1158 for (i=0; i<startMapLimit; i++) { 1159 UChar32 c32 = utext_char32At(ut, i); 1160 TEST_ASSERT(c32 == c32Map[i]); 1161 int64_t cpIndex = utext_getNativeIndex(ut); 1162 TEST_ASSERT(cpIndex == startMap[i]); 1163 } 1164 1165 // Check utext_next32From 1166 for (i=0; i<startMapLimit; i++) { 1167 UChar32 c32 = utext_next32From(ut, i); 1168 TEST_ASSERT(c32 == c32Map[i]); 1169 int64_t cpIndex = utext_getNativeIndex(ut); 1170 TEST_ASSERT(cpIndex == nextMap[i]); 1171 } 1172 1173 // check utext_previous32From 1174 for (i=0; i<startMapLimit; i++) { 1175 UChar32 c32 = utext_previous32From(ut, i); 1176 TEST_ASSERT(c32 == pr32Map[i]); 1177 int64_t cpIndex = utext_getNativeIndex(ut); 1178 TEST_ASSERT(cpIndex == prevMap[i]); 1179 } 1180 1181 // check Extract 1182 // Extract from i to i+1, which may be zero or one code points, 1183 // depending on whether the indices straddle a cp boundary. 1184 for (i=0; i<startMapLimit; i++) { 1185 UChar buf[3]; 1186 status = U_ZERO_ERROR; 1187 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status); 1188 TEST_SUCCESS(status); 1189 TEST_ASSERT(extractedLen == exLen[i]); 1190 if (extractedLen > 0) { 1191 UChar32 c32; 1192 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */ 1193 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32); 1194 TEST_ASSERT(c32 == c32Map[i]); 1195 } 1196 } 1197 1198 utext_close(ut); 1199 } 1200 } 1201 1202 1203 void UTextTest::FreezeTest() { 1204 // Check isWritable() and freeze() behavior. 1205 // 1206 1207 UnicodeString ustr("Hello, World."); 1208 const char u8str[] = {char(0x31), (char)0x32, (char)0x33, 0}; 1209 const UChar u16str[] = {(UChar)0x31, (UChar)0x32, (UChar)0x44, 0}; 1210 1211 UErrorCode status = U_ZERO_ERROR; 1212 UText *ut = NULL; 1213 UText *ut2 = NULL; 1214 1215 ut = utext_openUTF8(ut, u8str, -1, &status); 1216 TEST_SUCCESS(status); 1217 UBool writable = utext_isWritable(ut); 1218 TEST_ASSERT(writable == FALSE); 1219 utext_copy(ut, 1, 2, 0, TRUE, &status); 1220 TEST_ASSERT(status == U_NO_WRITE_PERMISSION); 1221 1222 status = U_ZERO_ERROR; 1223 ut = utext_openUChars(ut, u16str, -1, &status); 1224 TEST_SUCCESS(status); 1225 writable = utext_isWritable(ut); 1226 TEST_ASSERT(writable == FALSE); 1227 utext_copy(ut, 1, 2, 0, TRUE, &status); 1228 TEST_ASSERT(status == U_NO_WRITE_PERMISSION); 1229 1230 status = U_ZERO_ERROR; 1231 ut = utext_openUnicodeString(ut, &ustr, &status); 1232 TEST_SUCCESS(status); 1233 writable = utext_isWritable(ut); 1234 TEST_ASSERT(writable == TRUE); 1235 utext_freeze(ut); 1236 writable = utext_isWritable(ut); 1237 TEST_ASSERT(writable == FALSE); 1238 utext_copy(ut, 1, 2, 0, TRUE, &status); 1239 TEST_ASSERT(status == U_NO_WRITE_PERMISSION); 1240 1241 status = U_ZERO_ERROR; 1242 ut = utext_openUnicodeString(ut, &ustr, &status); 1243 TEST_SUCCESS(status); 1244 ut2 = utext_clone(ut2, ut, FALSE, FALSE, &status); // clone with readonly = false 1245 TEST_SUCCESS(status); 1246 writable = utext_isWritable(ut2); 1247 TEST_ASSERT(writable == TRUE); 1248 ut2 = utext_clone(ut2, ut, FALSE, TRUE, &status); // clone with readonly = true 1249 TEST_SUCCESS(status); 1250 writable = utext_isWritable(ut2); 1251 TEST_ASSERT(writable == FALSE); 1252 utext_copy(ut2, 1, 2, 0, TRUE, &status); 1253 TEST_ASSERT(status == U_NO_WRITE_PERMISSION); 1254 1255 status = U_ZERO_ERROR; 1256 ut = utext_openConstUnicodeString(ut, (const UnicodeString *)&ustr, &status); 1257 TEST_SUCCESS(status); 1258 writable = utext_isWritable(ut); 1259 TEST_ASSERT(writable == FALSE); 1260 utext_copy(ut, 1, 2, 0, TRUE, &status); 1261 TEST_ASSERT(status == U_NO_WRITE_PERMISSION); 1262 1263 // Deep Clone of a frozen UText should re-enable writing in the copy. 1264 status = U_ZERO_ERROR; 1265 ut = utext_openUnicodeString(ut, &ustr, &status); 1266 TEST_SUCCESS(status); 1267 utext_freeze(ut); 1268 ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status); // deep clone 1269 TEST_SUCCESS(status); 1270 writable = utext_isWritable(ut2); 1271 TEST_ASSERT(writable == TRUE); 1272 1273 1274 // Deep clone of a frozen UText, where the base type is intrinsically non-writable, 1275 // should NOT enable writing in the copy. 1276 status = U_ZERO_ERROR; 1277 ut = utext_openUChars(ut, u16str, -1, &status); 1278 TEST_SUCCESS(status); 1279 utext_freeze(ut); 1280 ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status); // deep clone 1281 TEST_SUCCESS(status); 1282 writable = utext_isWritable(ut2); 1283 TEST_ASSERT(writable == FALSE); 1284 1285 // cleanup 1286 utext_close(ut); 1287 utext_close(ut2); 1288 } 1289 1290 1291 // 1292 // Fragmented UText 1293 // A UText type that works with a chunk size of 1. 1294 // Intended to test for edge cases. 1295 // Input comes from a UnicodeString. 1296 // 1297 // ut.b the character. Put into both halves. 1298 // 1299 1300 U_CDECL_BEGIN 1301 static UBool U_CALLCONV 1302 fragTextAccess(UText *ut, int64_t index, UBool forward) { 1303 const UnicodeString *us = (const UnicodeString *)ut->context; 1304 UChar c; 1305 int32_t length = us->length(); 1306 if (forward && index>=0 && index<length) { 1307 c = us->charAt((int32_t)index); 1308 ut->b = c | c<<16; 1309 ut->chunkOffset = 0; 1310 ut->chunkLength = 1; 1311 ut->chunkNativeStart = index; 1312 ut->chunkNativeLimit = index+1; 1313 return true; 1314 } 1315 if (!forward && index>0 && index <=length) { 1316 c = us->charAt((int32_t)index-1); 1317 ut->b = c | c<<16; 1318 ut->chunkOffset = 1; 1319 ut->chunkLength = 1; 1320 ut->chunkNativeStart = index-1; 1321 ut->chunkNativeLimit = index; 1322 return true; 1323 } 1324 ut->b = 0; 1325 ut->chunkOffset = 0; 1326 ut->chunkLength = 0; 1327 if (index <= 0) { 1328 ut->chunkNativeStart = 0; 1329 ut->chunkNativeLimit = 0; 1330 } else { 1331 ut->chunkNativeStart = length; 1332 ut->chunkNativeLimit = length; 1333 } 1334 return false; 1335 } 1336 1337 // Function table to be used with this fragmented text provider. 1338 // Initialized in the open function. 1339 static UTextFuncs fragmentFuncs; 1340 1341 // Clone function for fragmented text provider. 1342 // Didn't really want to provide this, but it's easier to provide it than to keep it 1343 // out of the tests. 1344 // 1345 UText * 1346 cloneFragmentedUnicodeString(UText *dest, const UText *src, UBool deep, UErrorCode *status) { 1347 if (U_FAILURE(*status)) { 1348 return NULL; 1349 } 1350 if (deep) { 1351 *status = U_UNSUPPORTED_ERROR; 1352 return NULL; 1353 } 1354 dest = utext_openUnicodeString(dest, (UnicodeString *)src->context, status); 1355 utext_setNativeIndex(dest, utext_getNativeIndex(src)); 1356 return dest; 1357 } 1358 1359 U_CDECL_END 1360 1361 // Open function for the fragmented text provider. 1362 UText * 1363 openFragmentedUnicodeString(UText *ut, UnicodeString *s, UErrorCode *status) { 1364 ut = utext_openUnicodeString(ut, s, status); 1365 if (U_FAILURE(*status)) { 1366 return ut; 1367 } 1368 1369 // Copy of the function table from the stock UnicodeString UText, 1370 // and replace the entry for the access function. 1371 memcpy(&fragmentFuncs, ut->pFuncs, sizeof(fragmentFuncs)); 1372 fragmentFuncs.access = fragTextAccess; 1373 fragmentFuncs.clone = cloneFragmentedUnicodeString; 1374 ut->pFuncs = &fragmentFuncs; 1375 1376 ut->chunkContents = (UChar *)&ut->b; 1377 ut->pFuncs->access(ut, 0, TRUE); 1378 return ut; 1379 } 1380 1381 // Regression test for Ticket 5560 1382 // Clone fails to update chunkContentPointer in the cloned copy. 1383 // This is only an issue for UText types that work in a local buffer, 1384 // (UTF-8 wrapper, for example) 1385 // 1386 // The test: 1387 // 1. Create an inital UText 1388 // 2. Deep clone it. Contents should match original. 1389 // 3. Reset original to something different. 1390 // 4. Check that clone contents did not change. 1391 // 1392 void UTextTest::Ticket5560() { 1393 /* The following two strings are in UTF-8 even on EBCDIC platforms. */ 1394 static const char s1[] = {0x41,0x42,0x43,0x44,0x45,0x46,0}; /* "ABCDEF" */ 1395 static const char s2[] = {0x31,0x32,0x33,0x34,0x35,0x36,0}; /* "123456" */ 1396 UErrorCode status = U_ZERO_ERROR; 1397 1398 UText ut1 = UTEXT_INITIALIZER; 1399 UText ut2 = UTEXT_INITIALIZER; 1400 1401 utext_openUTF8(&ut1, s1, -1, &status); 1402 UChar c = utext_next32(&ut1); 1403 TEST_ASSERT(c == 0x41); // c == 'A' 1404 1405 utext_clone(&ut2, &ut1, TRUE, FALSE, &status); 1406 TEST_SUCCESS(status); 1407 c = utext_next32(&ut2); 1408 TEST_ASSERT(c == 0x42); // c == 'B' 1409 c = utext_next32(&ut1); 1410 TEST_ASSERT(c == 0x42); // c == 'B' 1411 1412 utext_openUTF8(&ut1, s2, -1, &status); 1413 c = utext_next32(&ut1); 1414 TEST_ASSERT(c == 0x31); // c == '1' 1415 c = utext_next32(&ut2); 1416 TEST_ASSERT(c == 0x43); // c == 'C' 1417 1418 utext_close(&ut1); 1419 utext_close(&ut2); 1420 } 1421 1422 1423 // Test for Ticket 6847 1424 // 1425 void UTextTest::Ticket6847() { 1426 const int STRLEN = 90; 1427 UChar s[STRLEN+1]; 1428 u_memset(s, 0x41, STRLEN); 1429 s[STRLEN] = 0; 1430 1431 UErrorCode status = U_ZERO_ERROR; 1432 UText *ut = utext_openUChars(NULL, s, -1, &status); 1433 1434 utext_setNativeIndex(ut, 0); 1435 int32_t count = 0; 1436 UChar32 c = 0; 1437 int64_t nativeIndex = UTEXT_GETNATIVEINDEX(ut); 1438 TEST_ASSERT(nativeIndex == 0); 1439 while ((c = utext_next32(ut)) != U_SENTINEL) { 1440 TEST_ASSERT(c == 0x41); 1441 TEST_ASSERT(count < STRLEN); 1442 if (count >= STRLEN) { 1443 break; 1444 } 1445 count++; 1446 nativeIndex = UTEXT_GETNATIVEINDEX(ut); 1447 TEST_ASSERT(nativeIndex == count); 1448 } 1449 TEST_ASSERT(count == STRLEN); 1450 nativeIndex = UTEXT_GETNATIVEINDEX(ut); 1451 TEST_ASSERT(nativeIndex == STRLEN); 1452 utext_close(ut); 1453 } 1454 1455
