]> git.saurik.com Git - apple/icu.git/blob - icuSources/test/intltest/utxttest.cpp
projects / apple / icu.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
ICU-531.48.tar.gz
[apple/icu.git] / icuSources / test / intltest / utxttest.cpp
1 /********************************************************************
2 * COPYRIGHT:
3 * Copyright (c) 2005-2013, 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 case 5: name = "Ticket10562";
61 if (exec) Ticket10562(); break;
62 default: name = ""; break;
63 }
64 }
65
66 //
67 // Quick and dirty random number generator.
68 // (don't use library so that results are portable.
69 static uint32_t m_seed = 1;
70 static uint32_t m_rand()
71 {
72 m_seed = m_seed * 1103515245 + 12345;
73 return (uint32_t)(m_seed/65536) % 32768;
74 }
75
76
77 //
78 // TextTest()
79 //
80 // Top Level function for UText testing.
81 // Specifies the strings to be tested, with the acutal testing itself
82 // being carried out in another function, TestString().
83 //
84 void UTextTest::TextTest() {
85 int32_t i, j;
86
87 TestString("abcd\\U00010001xyz");
88 TestString("");
89
90 // Supplementary chars at start or end
91 TestString("\\U00010001");
92 TestString("abc\\U00010001");
93 TestString("\\U00010001abc");
94
95 // Test simple strings of lengths 1 to 60, looking for glitches at buffer boundaries
96 UnicodeString s;
97 for (i=1; i<60; i++) {
98 s.truncate(0);
99 for (j=0; j<i; j++) {
100 if (j+0x30 == 0x5c) {
101 // backslash. Needs to be escaped
102 s.append((UChar)0x5c);
103 }
104 s.append(UChar(j+0x30));
105 }
106 TestString(s);
107 }
108
109 // Test strings with odd-aligned supplementary chars,
110 // looking for glitches at buffer boundaries
111 for (i=1; i<60; i++) {
112 s.truncate(0);
113 s.append((UChar)0x41);
114 for (j=0; j<i; j++) {
115 s.append(UChar32(j+0x11000));
116 }
117 TestString(s);
118 }
119
120 // String of chars of randomly varying size in utf-8 representation.
121 // Exercise the mapping, and the varying sized buffer.
122 //
123 s.truncate(0);
124 UChar32 c1 = 0;
125 UChar32 c2 = 0x100;
126 UChar32 c3 = 0xa000;
127 UChar32 c4 = 0x11000;
128 for (i=0; i<1000; i++) {
129 int len8 = m_rand()%4 + 1;
130 switch (len8) {
131 case 1:
132 c1 = (c1+1)%0x80;
133 // don't put 0 into string (0 terminated strings for some tests)
134 // don't put '\', will cause unescape() to fail.
135 if (c1==0x5c || c1==0) {
136 c1++;
137 }
138 s.append(c1);
139 break;
140 case 2:
141 s.append(c2++);
142 break;
143 case 3:
144 s.append(c3++);
145 break;
146 case 4:
147 s.append(c4++);
148 break;
149 }
150 }
151 TestString(s);
152 }
153
154
155 //
156 // TestString() Run a suite of UText tests on a string.
157 // The test string is unescaped before use.
158 //
159 void UTextTest::TestString(const UnicodeString &s) {
160 int32_t i;
161 int32_t j;
162 UChar32 c;
163 int32_t cpCount = 0;
164 UErrorCode status = U_ZERO_ERROR;
165 UText *ut = NULL;
166 int32_t saLen;
167
168 UnicodeString sa = s.unescape();
169 saLen = sa.length();
170
171 //
172 // Build up a mapping between code points and UTF-16 code unit indexes.
173 //
174 m *cpMap = new m[sa.length() + 1];
175 j = 0;
176 for (i=0; i<sa.length(); i=sa.moveIndex32(i, 1)) {
177 c = sa.char32At(i);
178 cpMap[j].nativeIdx = i;
179 cpMap[j].cp = c;
180 j++;
181 cpCount++;
182 }
183 cpMap[j].nativeIdx = i; // position following the last char in utf-16 string.
184
185
186 // UChar * test, null terminated
187 status = U_ZERO_ERROR;
188 UChar *buf = new UChar[saLen+1];
189 sa.extract(buf, saLen+1, status);
190 TEST_SUCCESS(status);
191 ut = utext_openUChars(NULL, buf, -1, &status);
192 TEST_SUCCESS(status);
193 TestAccess(sa, ut, cpCount, cpMap);
194 utext_close(ut);
195 delete [] buf;
196
197 // UChar * test, with length
198 status = U_ZERO_ERROR;
199 buf = new UChar[saLen+1];
200 sa.extract(buf, saLen+1, status);
201 TEST_SUCCESS(status);
202 ut = utext_openUChars(NULL, buf, saLen, &status);
203 TEST_SUCCESS(status);
204 TestAccess(sa, ut, cpCount, cpMap);
205 utext_close(ut);
206 delete [] buf;
207
208
209 // UnicodeString test
210 status = U_ZERO_ERROR;
211 ut = utext_openUnicodeString(NULL, &sa, &status);
212 TEST_SUCCESS(status);
213 TestAccess(sa, ut, cpCount, cpMap);
214 TestCMR(sa, ut, cpCount, cpMap, cpMap);
215 utext_close(ut);
216
217
218 // Const UnicodeString test
219 status = U_ZERO_ERROR;
220 ut = utext_openConstUnicodeString(NULL, &sa, &status);
221 TEST_SUCCESS(status);
222 TestAccess(sa, ut, cpCount, cpMap);
223 utext_close(ut);
224
225
226 // Replaceable test. (UnicodeString inherits Replaceable)
227 status = U_ZERO_ERROR;
228 ut = utext_openReplaceable(NULL, &sa, &status);
229 TEST_SUCCESS(status);
230 TestAccess(sa, ut, cpCount, cpMap);
231 TestCMR(sa, ut, cpCount, cpMap, cpMap);
232 utext_close(ut);
233
234 // Character Iterator Tests
235 status = U_ZERO_ERROR;
236 const UChar *cbuf = sa.getBuffer();
237 CharacterIterator *ci = new UCharCharacterIterator(cbuf, saLen, status);
238 TEST_SUCCESS(status);
239 ut = utext_openCharacterIterator(NULL, ci, &status);
240 TEST_SUCCESS(status);
241 TestAccess(sa, ut, cpCount, cpMap);
242 utext_close(ut);
243 delete ci;
244
245
246 // Fragmented UnicodeString (Chunk size of one)
247 //
248 status = U_ZERO_ERROR;
249 ut = openFragmentedUnicodeString(NULL, &sa, &status);
250 TEST_SUCCESS(status);
251 TestAccess(sa, ut, cpCount, cpMap);
252 utext_close(ut);
253
254 //
255 // UTF-8 test
256 //
257
258 // Convert the test string from UnicodeString to (char *) in utf-8 format
259 int32_t u8Len = sa.extract(0, sa.length(), NULL, 0, "utf-8");
260 char *u8String = new char[u8Len + 1];
261 sa.extract(0, sa.length(), u8String, u8Len+1, "utf-8");
262
263 // Build up the map of code point indices in the utf-8 string
264 m * u8Map = new m[sa.length() + 1];
265 i = 0; // native utf-8 index
266 for (j=0; j<cpCount ; j++) { // code point number
267 u8Map[j].nativeIdx = i;
268 U8_NEXT(u8String, i, u8Len, c)
269 u8Map[j].cp = c;
270 }
271 u8Map[cpCount].nativeIdx = u8Len; // position following the last char in utf-8 string.
272
273 // Do the test itself
274 status = U_ZERO_ERROR;
275 ut = utext_openUTF8(NULL, u8String, -1, &status);
276 TEST_SUCCESS(status);
277 TestAccess(sa, ut, cpCount, u8Map);
278 utext_close(ut);
279
280
281
282 delete []cpMap;
283 delete []u8Map;
284 delete []u8String;
285 }
286
287 // TestCMR test Copy, Move and Replace operations.
288 // us UnicodeString containing the test text.
289 // ut UText containing the same test text.
290 // cpCount number of code points in the test text.
291 // nativeMap Mapping from code points to native indexes for the UText.
292 // u16Map Mapping from code points to UTF-16 indexes, for use with the UnicodeString.
293 //
294 // This function runs a whole series of opertions on each incoming UText.
295 // The UText is deep-cloned prior to each operation, so that the original UText remains unchanged.
296 //
297 void UTextTest::TestCMR(const UnicodeString &us, UText *ut, int cpCount, m *nativeMap, m *u16Map) {
298 TEST_ASSERT(utext_isWritable(ut) == TRUE);
299
300 int srcLengthType; // Loop variables for selecting the postion and length
301 int srcPosType; // of the block to operate on within the source text.
302 int destPosType;
303
304 int srcIndex = 0; // Code Point indexes of the block to operate on for
305 int srcLength = 0; // a specific test.
306
307 int destIndex = 0; // Code point index of the destination for a copy/move test.
308
309 int32_t nativeStart = 0; // Native unit indexes for a test.
310 int32_t nativeLimit = 0;
311 int32_t nativeDest = 0;
312
313 int32_t u16Start = 0; // UTF-16 indexes for a test.
314 int32_t u16Limit = 0; // used when performing the same operation in a Unicode String
315 int32_t u16Dest = 0;
316
317 // Iterate over a whole series of source index, length and a target indexes.
318 // This is done with code point indexes; these will be later translated to native
319 // indexes using the cpMap.
320 for (srcLengthType=1; srcLengthType<=3; srcLengthType++) {
321 switch (srcLengthType) {
322 case 1: srcLength = 1; break;
323 case 2: srcLength = 5; break;
324 case 3: srcLength = cpCount / 3;
325 }
326 for (srcPosType=1; srcPosType<=5; srcPosType++) {
327 switch (srcPosType) {
328 case 1: srcIndex = 0; break;
329 case 2: srcIndex = 1; break;
330 case 3: srcIndex = cpCount - srcLength; break;
331 case 4: srcIndex = cpCount - srcLength - 1; break;
332 case 5: srcIndex = cpCount / 2; break;
333 }
334 if (srcIndex < 0 || srcIndex + srcLength > cpCount) {
335 // filter out bogus test cases -
336 // those with a source range that falls of an edge of the string.
337 continue;
338 }
339
340 //
341 // Copy and move tests.
342 // iterate over a variety of destination positions.
343 //
344 for (destPosType=1; destPosType<=4; destPosType++) {
345 switch (destPosType) {
346 case 1: destIndex = 0; break;
347 case 2: destIndex = 1; break;
348 case 3: destIndex = srcIndex - 1; break;
349 case 4: destIndex = srcIndex + srcLength + 1; break;
350 case 5: destIndex = cpCount-1; break;
351 case 6: destIndex = cpCount; break;
352 }
353 if (destIndex<0 || destIndex>cpCount) {
354 // filter out bogus test cases.
355 continue;
356 }
357
358 nativeStart = nativeMap[srcIndex].nativeIdx;
359 nativeLimit = nativeMap[srcIndex+srcLength].nativeIdx;
360 nativeDest = nativeMap[destIndex].nativeIdx;
361
362 u16Start = u16Map[srcIndex].nativeIdx;
363 u16Limit = u16Map[srcIndex+srcLength].nativeIdx;
364 u16Dest = u16Map[destIndex].nativeIdx;
365
366 gFailed = FALSE;
367 TestCopyMove(us, ut, FALSE,
368 nativeStart, nativeLimit, nativeDest,
369 u16Start, u16Limit, u16Dest);
370
371 TestCopyMove(us, ut, TRUE,
372 nativeStart, nativeLimit, nativeDest,
373 u16Start, u16Limit, u16Dest);
374
375 if (gFailed) {
376 return;
377 }
378 }
379
380 //
381 // Replace tests.
382 //
383 UnicodeString fullRepString("This is an arbitrary string that will be used as replacement text");
384 for (int32_t replStrLen=0; replStrLen<20; replStrLen++) {
385 UnicodeString repStr(fullRepString, 0, replStrLen);
386 TestReplace(us, ut,
387 nativeStart, nativeLimit,
388 u16Start, u16Limit,
389 repStr);
390 if (gFailed) {
391 return;
392 }
393 }
394
395 }
396 }
397
398 }
399
400 //
401 // TestCopyMove run a single test case for utext_copy.
402 // Test cases are created in TestCMR and dispatched here for execution.
403 //
404 void UTextTest::TestCopyMove(const UnicodeString &us, UText *ut, UBool move,
405 int32_t nativeStart, int32_t nativeLimit, int32_t nativeDest,
406 int32_t u16Start, int32_t u16Limit, int32_t u16Dest)
407 {
408 UErrorCode status = U_ZERO_ERROR;
409 UText *targetUT = NULL;
410 gTestNum++;
411 gFailed = FALSE;
412
413 //
414 // clone the UText. The test will be run in the cloned copy
415 // so that we don't alter the original.
416 //
417 targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status);
418 TEST_SUCCESS(status);
419 UnicodeString targetUS(us); // And copy the reference string.
420
421 // do the test operation first in the reference
422 targetUS.copy(u16Start, u16Limit, u16Dest);
423 if (move) {
424 // delete out the source range.
425 if (u16Limit < u16Dest) {
426 targetUS.removeBetween(u16Start, u16Limit);
427 } else {
428 int32_t amtCopied = u16Limit - u16Start;
429 targetUS.removeBetween(u16Start+amtCopied, u16Limit+amtCopied);
430 }
431 }
432
433 // Do the same operation in the UText under test
434 utext_copy(targetUT, nativeStart, nativeLimit, nativeDest, move, &status);
435 if (nativeDest > nativeStart && nativeDest < nativeLimit) {
436 TEST_ASSERT(status == U_INDEX_OUTOFBOUNDS_ERROR);
437 } else {
438 TEST_SUCCESS(status);
439
440 // Compare the results of the two parallel tests
441 int32_t usi = 0; // UnicodeString postion, utf-16 index.
442 int64_t uti = 0; // UText position, native index.
443 int32_t cpi; // char32 position (code point index)
444 UChar32 usc; // code point from Unicode String
445 UChar32 utc; // code point from UText
446 utext_setNativeIndex(targetUT, 0);
447 for (cpi=0; ; cpi++) {
448 usc = targetUS.char32At(usi);
449 utc = utext_next32(targetUT);
450 if (utc < 0) {
451 break;
452 }
453 TEST_ASSERT(uti == usi);
454 TEST_ASSERT(utc == usc);
455 usi = targetUS.moveIndex32(usi, 1);
456 uti = utext_getNativeIndex(targetUT);
457 if (gFailed) {
458 goto cleanupAndReturn;
459 }
460 }
461 int64_t expectedNativeLength = utext_nativeLength(ut);
462 if (move == FALSE) {
463 expectedNativeLength += nativeLimit - nativeStart;
464 }
465 uti = utext_getNativeIndex(targetUT);
466 TEST_ASSERT(uti == expectedNativeLength);
467 }
468
469 cleanupAndReturn:
470 utext_close(targetUT);
471 }
472
473
474 //
475 // TestReplace Test a single Replace operation.
476 //
477 void UTextTest::TestReplace(
478 const UnicodeString &us, // reference UnicodeString in which to do the replace
479 UText *ut, // UnicodeText object under test.
480 int32_t nativeStart, // Range to be replaced, in UText native units.
481 int32_t nativeLimit,
482 int32_t u16Start, // Range to be replaced, in UTF-16 units
483 int32_t u16Limit, // for use in the reference UnicodeString.
484 const UnicodeString &repStr) // The replacement string
485 {
486 UErrorCode status = U_ZERO_ERROR;
487 UText *targetUT = NULL;
488 gTestNum++;
489 gFailed = FALSE;
490
491 //
492 // clone the target UText. The test will be run in the cloned copy
493 // so that we don't alter the original.
494 //
495 targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status);
496 TEST_SUCCESS(status);
497 UnicodeString targetUS(us); // And copy the reference string.
498
499 //
500 // Do the replace operation in the Unicode String, to
501 // produce a reference result.
502 //
503 targetUS.replace(u16Start, u16Limit-u16Start, repStr);
504
505 //
506 // Do the replace on the UText under test
507 //
508 const UChar *rs = repStr.getBuffer();
509 int32_t rsLen = repStr.length();
510 int32_t actualDelta = utext_replace(targetUT, nativeStart, nativeLimit, rs, rsLen, &status);
511 int32_t expectedDelta = repStr.length() - (nativeLimit - nativeStart);
512 TEST_ASSERT(actualDelta == expectedDelta);
513
514 //
515 // Compare the results
516 //
517 int32_t usi = 0; // UnicodeString postion, utf-16 index.
518 int64_t uti = 0; // UText position, native index.
519 int32_t cpi; // char32 position (code point index)
520 UChar32 usc; // code point from Unicode String
521 UChar32 utc; // code point from UText
522 int64_t expectedNativeLength = 0;
523 utext_setNativeIndex(targetUT, 0);
524 for (cpi=0; ; cpi++) {
525 usc = targetUS.char32At(usi);
526 utc = utext_next32(targetUT);
527 if (utc < 0) {
528 break;
529 }
530 TEST_ASSERT(uti == usi);
531 TEST_ASSERT(utc == usc);
532 usi = targetUS.moveIndex32(usi, 1);
533 uti = utext_getNativeIndex(targetUT);
534 if (gFailed) {
535 goto cleanupAndReturn;
536 }
537 }
538 expectedNativeLength = utext_nativeLength(ut) + expectedDelta;
539 uti = utext_getNativeIndex(targetUT);
540 TEST_ASSERT(uti == expectedNativeLength);
541
542 cleanupAndReturn:
543 utext_close(targetUT);
544 }
545
546 //
547 // TestAccess Test the read only access functions on a UText, including cloning.
548 // The text is accessed in a variety of ways, and compared with
549 // the reference UnicodeString.
550 //
551 void UTextTest::TestAccess(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) {
552 // Run the standard tests on the caller-supplied UText.
553 TestAccessNoClone(us, ut, cpCount, cpMap);
554
555 // Re-run tests on a shallow clone.
556 utext_setNativeIndex(ut, 0);
557 UErrorCode status = U_ZERO_ERROR;
558 UText *shallowClone = utext_clone(NULL, ut, FALSE /*deep*/, FALSE /*readOnly*/, &status);
559 TEST_SUCCESS(status);
560 TestAccessNoClone(us, shallowClone, cpCount, cpMap);
561
562 //
563 // Rerun again on a deep clone.
564 // Note that text providers are not required to provide deep cloning,
565 // so unsupported errors are ignored.
566 //
567 status = U_ZERO_ERROR;
568 utext_setNativeIndex(shallowClone, 0);
569 UText *deepClone = utext_clone(NULL, shallowClone, TRUE, FALSE, &status);
570 utext_close(shallowClone);
571 if (status != U_UNSUPPORTED_ERROR) {
572 TEST_SUCCESS(status);
573 TestAccessNoClone(us, deepClone, cpCount, cpMap);
574 }
575 utext_close(deepClone);
576 }
577
578
579 //
580 // TestAccessNoClone() Test the read only access functions on a UText.
581 // The text is accessed in a variety of ways, and compared with
582 // the reference UnicodeString.
583 //
584 void UTextTest::TestAccessNoClone(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) {
585 UErrorCode status = U_ZERO_ERROR;
586 gTestNum++;
587
588 //
589 // Check the length from the UText
590 //
591 int64_t expectedLen = cpMap[cpCount].nativeIdx;
592 int64_t utlen = utext_nativeLength(ut);
593 TEST_ASSERT(expectedLen == utlen);
594
595 //
596 // Iterate forwards, verify that we get the correct code points
597 // at the correct native offsets.
598 //
599 int i = 0;
600 int64_t index;
601 int64_t expectedIndex = 0;
602 int64_t foundIndex = 0;
603 UChar32 expectedC;
604 UChar32 foundC;
605 int64_t len;
606
607 for (i=0; i<cpCount; i++) {
608 expectedIndex = cpMap[i].nativeIdx;
609 foundIndex = utext_getNativeIndex(ut);
610 TEST_ASSERT(expectedIndex == foundIndex);
611 expectedC = cpMap[i].cp;
612 foundC = utext_next32(ut);
613 TEST_ASSERT(expectedC == foundC);
614 foundIndex = utext_getPreviousNativeIndex(ut);
615 TEST_ASSERT(expectedIndex == foundIndex);
616 if (gFailed) {
617 return;
618 }
619 }
620 foundC = utext_next32(ut);
621 TEST_ASSERT(foundC == U_SENTINEL);
622
623 // Repeat above, using macros
624 utext_setNativeIndex(ut, 0);
625 for (i=0; i<cpCount; i++) {
626 expectedIndex = cpMap[i].nativeIdx;
627 foundIndex = UTEXT_GETNATIVEINDEX(ut);
628 TEST_ASSERT(expectedIndex == foundIndex);
629 expectedC = cpMap[i].cp;
630 foundC = UTEXT_NEXT32(ut);
631 TEST_ASSERT(expectedC == foundC);
632 if (gFailed) {
633 return;
634 }
635 }
636 foundC = UTEXT_NEXT32(ut);
637 TEST_ASSERT(foundC == U_SENTINEL);
638
639 //
640 // Forward iteration (above) should have left index at the
641 // end of the input, which should == length().
642 //
643 len = utext_nativeLength(ut);
644 foundIndex = utext_getNativeIndex(ut);
645 TEST_ASSERT(len == foundIndex);
646
647 //
648 // Iterate backwards over entire test string
649 //
650 len = utext_getNativeIndex(ut);
651 utext_setNativeIndex(ut, len);
652 for (i=cpCount-1; i>=0; i--) {
653 expectedC = cpMap[i].cp;
654 expectedIndex = cpMap[i].nativeIdx;
655 int64_t prevIndex = utext_getPreviousNativeIndex(ut);
656 foundC = utext_previous32(ut);
657 foundIndex = utext_getNativeIndex(ut);
658 TEST_ASSERT(expectedIndex == foundIndex);
659 TEST_ASSERT(expectedC == foundC);
660 TEST_ASSERT(prevIndex == foundIndex);
661 if (gFailed) {
662 return;
663 }
664 }
665
666 //
667 // Backwards iteration, above, should have left our iterator
668 // position at zero, and continued backwards iterationshould fail.
669 //
670 foundIndex = utext_getNativeIndex(ut);
671 TEST_ASSERT(foundIndex == 0);
672 foundIndex = utext_getPreviousNativeIndex(ut);
673 TEST_ASSERT(foundIndex == 0);
674
675
676 foundC = utext_previous32(ut);
677 TEST_ASSERT(foundC == U_SENTINEL);
678 foundIndex = utext_getNativeIndex(ut);
679 TEST_ASSERT(foundIndex == 0);
680 foundIndex = utext_getPreviousNativeIndex(ut);
681 TEST_ASSERT(foundIndex == 0);
682
683
684 // And again, with the macros
685 utext_setNativeIndex(ut, len);
686 for (i=cpCount-1; i>=0; i--) {
687 expectedC = cpMap[i].cp;
688 expectedIndex = cpMap[i].nativeIdx;
689 foundC = UTEXT_PREVIOUS32(ut);
690 foundIndex = UTEXT_GETNATIVEINDEX(ut);
691 TEST_ASSERT(expectedIndex == foundIndex);
692 TEST_ASSERT(expectedC == foundC);
693 if (gFailed) {
694 return;
695 }
696 }
697
698 //
699 // Backwards iteration, above, should have left our iterator
700 // position at zero, and continued backwards iterationshould fail.
701 //
702 foundIndex = UTEXT_GETNATIVEINDEX(ut);
703 TEST_ASSERT(foundIndex == 0);
704
705 foundC = UTEXT_PREVIOUS32(ut);
706 TEST_ASSERT(foundC == U_SENTINEL);
707 foundIndex = UTEXT_GETNATIVEINDEX(ut);
708 TEST_ASSERT(foundIndex == 0);
709 if (gFailed) {
710 return;
711 }
712
713 //
714 // next32From(), prevous32From(), Iterate in a somewhat random order.
715 //
716 int cpIndex = 0;
717 for (i=0; i<cpCount; i++) {
718 cpIndex = (cpIndex + 9973) % cpCount;
719 index = cpMap[cpIndex].nativeIdx;
720 expectedC = cpMap[cpIndex].cp;
721 foundC = utext_next32From(ut, index);
722 TEST_ASSERT(expectedC == foundC);
723 if (gFailed) {
724 return;
725 }
726 }
727
728 cpIndex = 0;
729 for (i=0; i<cpCount; i++) {
730 cpIndex = (cpIndex + 9973) % cpCount;
731 index = cpMap[cpIndex+1].nativeIdx;
732 expectedC = cpMap[cpIndex].cp;
733 foundC = utext_previous32From(ut, index);
734 TEST_ASSERT(expectedC == foundC);
735 if (gFailed) {
736 return;
737 }
738 }
739
740
741 //
742 // moveIndex(int32_t delta);
743 //
744
745 // Walk through frontwards, incrementing by one
746 utext_setNativeIndex(ut, 0);
747 for (i=1; i<=cpCount; i++) {
748 utext_moveIndex32(ut, 1);
749 index = utext_getNativeIndex(ut);
750 expectedIndex = cpMap[i].nativeIdx;
751 TEST_ASSERT(expectedIndex == index);
752 index = UTEXT_GETNATIVEINDEX(ut);
753 TEST_ASSERT(expectedIndex == index);
754 }
755
756 // Walk through frontwards, incrementing by two
757 utext_setNativeIndex(ut, 0);
758 for (i=2; i<cpCount; i+=2) {
759 utext_moveIndex32(ut, 2);
760 index = utext_getNativeIndex(ut);
761 expectedIndex = cpMap[i].nativeIdx;
762 TEST_ASSERT(expectedIndex == index);
763 index = UTEXT_GETNATIVEINDEX(ut);
764 TEST_ASSERT(expectedIndex == index);
765 }
766
767 // walk through the string backwards, decrementing by one.
768 i = cpMap[cpCount].nativeIdx;
769 utext_setNativeIndex(ut, i);
770 for (i=cpCount; i>=0; i--) {
771 expectedIndex = cpMap[i].nativeIdx;
772 index = utext_getNativeIndex(ut);
773 TEST_ASSERT(expectedIndex == index);
774 index = UTEXT_GETNATIVEINDEX(ut);
775 TEST_ASSERT(expectedIndex == index);
776 utext_moveIndex32(ut, -1);
777 }
778
779
780 // walk through backwards, decrementing by three
781 i = cpMap[cpCount].nativeIdx;
782 utext_setNativeIndex(ut, i);
783 for (i=cpCount; i>=0; i-=3) {
784 expectedIndex = cpMap[i].nativeIdx;
785 index = utext_getNativeIndex(ut);
786 TEST_ASSERT(expectedIndex == index);
787 index = UTEXT_GETNATIVEINDEX(ut);
788 TEST_ASSERT(expectedIndex == index);
789 utext_moveIndex32(ut, -3);
790 }
791
792
793 //
794 // Extract
795 //
796 int bufSize = us.length() + 10;
797 UChar *buf = new UChar[bufSize];
798 status = U_ZERO_ERROR;
799 expectedLen = us.length();
800 len = utext_extract(ut, 0, utlen, buf, bufSize, &status);
801 TEST_SUCCESS(status);
802 TEST_ASSERT(len == expectedLen);
803 int compareResult = us.compare(buf, -1);
804 TEST_ASSERT(compareResult == 0);
805
806 status = U_ZERO_ERROR;
807 len = utext_extract(ut, 0, utlen, NULL, 0, &status);
808 if (utlen == 0) {
809 TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING);
810 } else {
811 TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR);
812 }
813 TEST_ASSERT(len == expectedLen);
814
815 status = U_ZERO_ERROR;
816 u_memset(buf, 0x5555, bufSize);
817 len = utext_extract(ut, 0, utlen, buf, 1, &status);
818 if (us.length() == 0) {
819 TEST_SUCCESS(status);
820 TEST_ASSERT(buf[0] == 0);
821 } else {
822 // Buf len == 1, extracting a single 16 bit value.
823 // If the data char is supplementary, it doesn't matter whether the buffer remains unchanged,
824 // or whether the lead surrogate of the pair is extracted.
825 // It's a buffer overflow error in either case.
826 TEST_ASSERT(buf[0] == us.charAt(0) ||
827 (buf[0] == 0x5555 && U_IS_SUPPLEMENTARY(us.char32At(0))));
828 TEST_ASSERT(buf[1] == 0x5555);
829 if (us.length() == 1) {
830 TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING);
831 } else {
832 TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR);
833 }
834 }
835
836 delete []buf;
837 }
838
839 //
840 // ErrorTest() Check various error and edge cases.
841 //
842 void UTextTest::ErrorTest()
843 {
844 // Close of an unitialized UText. Shouldn't blow up.
845 {
846 UText ut;
847 memset(&ut, 0, sizeof(UText));
848 utext_close(&ut);
849 utext_close(NULL);
850 }
851
852 // Double-close of a UText. Shouldn't blow up. UText should still be usable.
853 {
854 UErrorCode status = U_ZERO_ERROR;
855 UText ut = UTEXT_INITIALIZER;
856 UnicodeString s("Hello, World");
857 UText *ut2 = utext_openUnicodeString(&ut, &s, &status);
858 TEST_SUCCESS(status);
859 TEST_ASSERT(ut2 == &ut);
860
861 UText *ut3 = utext_close(&ut);
862 TEST_ASSERT(ut3 == &ut);
863
864 UText *ut4 = utext_close(&ut);
865 TEST_ASSERT(ut4 == &ut);
866
867 utext_openUnicodeString(&ut, &s, &status);
868 TEST_SUCCESS(status);
869 utext_close(&ut);
870 }
871
872 // Re-use of a UText, chaining through each of the types of UText
873 // (If it doesn't blow up, and doesn't leak, it's probably working fine)
874 {
875 UErrorCode status = U_ZERO_ERROR;
876 UText ut = UTEXT_INITIALIZER;
877 UText *utp;
878 UnicodeString s1("Hello, World");
879 UChar s2[] = {(UChar)0x41, (UChar)0x42, (UChar)0};
880 const char *s3 = "\x66\x67\x68";
881
882 utp = utext_openUnicodeString(&ut, &s1, &status);
883 TEST_SUCCESS(status);
884 TEST_ASSERT(utp == &ut);
885
886 utp = utext_openConstUnicodeString(&ut, &s1, &status);
887 TEST_SUCCESS(status);
888 TEST_ASSERT(utp == &ut);
889
890 utp = utext_openUTF8(&ut, s3, -1, &status);
891 TEST_SUCCESS(status);
892 TEST_ASSERT(utp == &ut);
893
894 utp = utext_openUChars(&ut, s2, -1, &status);
895 TEST_SUCCESS(status);
896 TEST_ASSERT(utp == &ut);
897
898 utp = utext_close(&ut);
899 TEST_ASSERT(utp == &ut);
900
901 utp = utext_openUnicodeString(&ut, &s1, &status);
902 TEST_SUCCESS(status);
903 TEST_ASSERT(utp == &ut);
904 }
905
906 // Invalid parameters on open
907 //
908 {
909 UErrorCode status = U_ZERO_ERROR;
910 UText ut = UTEXT_INITIALIZER;
911
912 utext_openUChars(&ut, NULL, 5, &status);
913 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
914
915 status = U_ZERO_ERROR;
916 utext_openUChars(&ut, NULL, -1, &status);
917 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
918
919 status = U_ZERO_ERROR;
920 utext_openUTF8(&ut, NULL, 4, &status);
921 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
922
923 status = U_ZERO_ERROR;
924 utext_openUTF8(&ut, NULL, -1, &status);
925 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
926 }
927
928 //
929 // UTF-8 with malformed sequences.
930 // These should come through as the Unicode replacement char, \ufffd
931 //
932 {
933 UErrorCode status = U_ZERO_ERROR;
934 UText *ut = NULL;
935 const char *badUTF8 = "\x41\x81\x42\xf0\x81\x81\x43";
936 UChar32 c;
937
938 ut = utext_openUTF8(NULL, badUTF8, -1, &status);
939 TEST_SUCCESS(status);
940 c = utext_char32At(ut, 1);
941 TEST_ASSERT(c == 0xfffd);
942 c = utext_char32At(ut, 3);
943 TEST_ASSERT(c == 0xfffd);
944 c = utext_char32At(ut, 5);
945 TEST_ASSERT(c == 0xfffd);
946 c = utext_char32At(ut, 6);
947 TEST_ASSERT(c == 0x43);
948
949 UChar buf[10];
950 int n = utext_extract(ut, 0, 9, buf, 10, &status);
951 TEST_SUCCESS(status);
952 TEST_ASSERT(n==5);
953 TEST_ASSERT(buf[1] == 0xfffd);
954 TEST_ASSERT(buf[3] == 0xfffd);
955 TEST_ASSERT(buf[2] == 0x42);
956 utext_close(ut);
957 }
958
959
960 //
961 // isLengthExpensive - does it make the exptected transitions after
962 // getting the length of a nul terminated string?
963 //
964 {
965 UErrorCode status = U_ZERO_ERROR;
966 UnicodeString sa("Hello, this is a string");
967 UBool isExpensive;
968
969 UChar sb[100];
970 memset(sb, 0x20, sizeof(sb));
971 sb[99] = 0;
972
973 UText *uta = utext_openUnicodeString(NULL, &sa, &status);
974 TEST_SUCCESS(status);
975 isExpensive = utext_isLengthExpensive(uta);
976 TEST_ASSERT(isExpensive == FALSE);
977 utext_close(uta);
978
979 UText *utb = utext_openUChars(NULL, sb, -1, &status);
980 TEST_SUCCESS(status);
981 isExpensive = utext_isLengthExpensive(utb);
982 TEST_ASSERT(isExpensive == TRUE);
983 int64_t len = utext_nativeLength(utb);
984 TEST_ASSERT(len == 99);
985 isExpensive = utext_isLengthExpensive(utb);
986 TEST_ASSERT(isExpensive == FALSE);
987 utext_close(utb);
988 }
989
990 //
991 // Index to positions not on code point boundaries.
992 //
993 {
994 const char *u8str = "\xc8\x81\xe1\x82\x83\xf1\x84\x85\x86";
995 int32_t startMap[] = { 0, 0, 2, 2, 2, 5, 5, 5, 5, 9, 9};
996 int32_t nextMap[] = { 2, 2, 5, 5, 5, 9, 9, 9, 9, 9, 9};
997 int32_t prevMap[] = { 0, 0, 0, 0, 0, 2, 2, 2, 2, 5, 5};
998 UChar32 c32Map[] = {0x201, 0x201, 0x1083, 0x1083, 0x1083, 0x044146, 0x044146, 0x044146, 0x044146, -1, -1};
999 UChar32 pr32Map[] = { -1, -1, 0x201, 0x201, 0x201, 0x1083, 0x1083, 0x1083, 0x1083, 0x044146, 0x044146};
1000
1001 // extractLen is the size, in UChars, of what will be extracted between index and index+1.
1002 // is zero when both index positions lie within the same code point.
1003 int32_t exLen[] = { 0, 1, 0, 0, 1, 0, 0, 0, 2, 0, 0};
1004
1005
1006 UErrorCode status = U_ZERO_ERROR;
1007 UText *ut = utext_openUTF8(NULL, u8str, -1, &status);
1008 TEST_SUCCESS(status);
1009
1010 // Check setIndex
1011 int32_t i;
1012 int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t);
1013 for (i=0; i<startMapLimit; i++) {
1014 utext_setNativeIndex(ut, i);
1015 int64_t cpIndex = utext_getNativeIndex(ut);
1016 TEST_ASSERT(cpIndex == startMap[i]);
1017 cpIndex = UTEXT_GETNATIVEINDEX(ut);
1018 TEST_ASSERT(cpIndex == startMap[i]);
1019 }
1020
1021 // Check char32At
1022 for (i=0; i<startMapLimit; i++) {
1023 UChar32 c32 = utext_char32At(ut, i);
1024 TEST_ASSERT(c32 == c32Map[i]);
1025 int64_t cpIndex = utext_getNativeIndex(ut);
1026 TEST_ASSERT(cpIndex == startMap[i]);
1027 }
1028
1029 // Check utext_next32From
1030 for (i=0; i<startMapLimit; i++) {
1031 UChar32 c32 = utext_next32From(ut, i);
1032 TEST_ASSERT(c32 == c32Map[i]);
1033 int64_t cpIndex = utext_getNativeIndex(ut);
1034 TEST_ASSERT(cpIndex == nextMap[i]);
1035 }
1036
1037 // check utext_previous32From
1038 for (i=0; i<startMapLimit; i++) {
1039 gTestNum++;
1040 UChar32 c32 = utext_previous32From(ut, i);
1041 TEST_ASSERT(c32 == pr32Map[i]);
1042 int64_t cpIndex = utext_getNativeIndex(ut);
1043 TEST_ASSERT(cpIndex == prevMap[i]);
1044 }
1045
1046 // check Extract
1047 // Extract from i to i+1, which may be zero or one code points,
1048 // depending on whether the indices straddle a cp boundary.
1049 for (i=0; i<startMapLimit; i++) {
1050 UChar buf[3];
1051 status = U_ZERO_ERROR;
1052 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1053 TEST_SUCCESS(status);
1054 TEST_ASSERT(extractedLen == exLen[i]);
1055 if (extractedLen > 0) {
1056 UChar32 c32;
1057 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1058 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1059 TEST_ASSERT(c32 == c32Map[i]);
1060 }
1061 }
1062
1063 utext_close(ut);
1064 }
1065
1066
1067 { // Similar test, with utf16 instead of utf8
1068 // TODO: merge the common parts of these tests.
1069
1070 UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV);
1071 int32_t startMap[] ={ 0, 1, 1, 3, 4, 4, 6, 6};
1072 int32_t nextMap[] = { 1, 3, 3, 4, 6, 6, 6, 6};
1073 int32_t prevMap[] = { 0, 0, 0, 1, 3, 3, 4, 4};
1074 UChar32 c32Map[] = {0x1000, 0x11000, 0x11000, 0x2000, 0x22000, 0x22000, -1, -1};
1075 UChar32 pr32Map[] = { -1, 0x1000, 0x1000, 0x11000, 0x2000, 0x2000, 0x22000, 0x22000};
1076 int32_t exLen[] = { 1, 0, 2, 1, 0, 2, 0, 0,};
1077
1078 u16str = u16str.unescape();
1079 UErrorCode status = U_ZERO_ERROR;
1080 UText *ut = utext_openUnicodeString(NULL, &u16str, &status);
1081 TEST_SUCCESS(status);
1082
1083 int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t);
1084 int i;
1085 for (i=0; i<startMapLimit; i++) {
1086 utext_setNativeIndex(ut, i);
1087 int64_t cpIndex = utext_getNativeIndex(ut);
1088 TEST_ASSERT(cpIndex == startMap[i]);
1089 }
1090
1091 // Check char32At
1092 for (i=0; i<startMapLimit; i++) {
1093 UChar32 c32 = utext_char32At(ut, i);
1094 TEST_ASSERT(c32 == c32Map[i]);
1095 int64_t cpIndex = utext_getNativeIndex(ut);
1096 TEST_ASSERT(cpIndex == startMap[i]);
1097 }
1098
1099 // Check utext_next32From
1100 for (i=0; i<startMapLimit; i++) {
1101 UChar32 c32 = utext_next32From(ut, i);
1102 TEST_ASSERT(c32 == c32Map[i]);
1103 int64_t cpIndex = utext_getNativeIndex(ut);
1104 TEST_ASSERT(cpIndex == nextMap[i]);
1105 }
1106
1107 // check utext_previous32From
1108 for (i=0; i<startMapLimit; i++) {
1109 UChar32 c32 = utext_previous32From(ut, i);
1110 TEST_ASSERT(c32 == pr32Map[i]);
1111 int64_t cpIndex = utext_getNativeIndex(ut);
1112 TEST_ASSERT(cpIndex == prevMap[i]);
1113 }
1114
1115 // check Extract
1116 // Extract from i to i+1, which may be zero or one code points,
1117 // depending on whether the indices straddle a cp boundary.
1118 for (i=0; i<startMapLimit; i++) {
1119 UChar buf[3];
1120 status = U_ZERO_ERROR;
1121 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1122 TEST_SUCCESS(status);
1123 TEST_ASSERT(extractedLen == exLen[i]);
1124 if (extractedLen > 0) {
1125 UChar32 c32;
1126 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1127 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1128 TEST_ASSERT(c32 == c32Map[i]);
1129 }
1130 }
1131
1132 utext_close(ut);
1133 }
1134
1135 { // Similar test, with UText over Replaceable
1136 // TODO: merge the common parts of these tests.
1137
1138 UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV);
1139 int32_t startMap[] ={ 0, 1, 1, 3, 4, 4, 6, 6};
1140 int32_t nextMap[] = { 1, 3, 3, 4, 6, 6, 6, 6};
1141 int32_t prevMap[] = { 0, 0, 0, 1, 3, 3, 4, 4};
1142 UChar32 c32Map[] = {0x1000, 0x11000, 0x11000, 0x2000, 0x22000, 0x22000, -1, -1};
1143 UChar32 pr32Map[] = { -1, 0x1000, 0x1000, 0x11000, 0x2000, 0x2000, 0x22000, 0x22000};
1144 int32_t exLen[] = { 1, 0, 2, 1, 0, 2, 0, 0,};
1145
1146 u16str = u16str.unescape();
1147 UErrorCode status = U_ZERO_ERROR;
1148 UText *ut = utext_openReplaceable(NULL, &u16str, &status);
1149 TEST_SUCCESS(status);
1150
1151 int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t);
1152 int i;
1153 for (i=0; i<startMapLimit; i++) {
1154 utext_setNativeIndex(ut, i);
1155 int64_t cpIndex = utext_getNativeIndex(ut);
1156 TEST_ASSERT(cpIndex == startMap[i]);
1157 }
1158
1159 // Check char32At
1160 for (i=0; i<startMapLimit; i++) {
1161 UChar32 c32 = utext_char32At(ut, i);
1162 TEST_ASSERT(c32 == c32Map[i]);
1163 int64_t cpIndex = utext_getNativeIndex(ut);
1164 TEST_ASSERT(cpIndex == startMap[i]);
1165 }
1166
1167 // Check utext_next32From
1168 for (i=0; i<startMapLimit; i++) {
1169 UChar32 c32 = utext_next32From(ut, i);
1170 TEST_ASSERT(c32 == c32Map[i]);
1171 int64_t cpIndex = utext_getNativeIndex(ut);
1172 TEST_ASSERT(cpIndex == nextMap[i]);
1173 }
1174
1175 // check utext_previous32From
1176 for (i=0; i<startMapLimit; i++) {
1177 UChar32 c32 = utext_previous32From(ut, i);
1178 TEST_ASSERT(c32 == pr32Map[i]);
1179 int64_t cpIndex = utext_getNativeIndex(ut);
1180 TEST_ASSERT(cpIndex == prevMap[i]);
1181 }
1182
1183 // check Extract
1184 // Extract from i to i+1, which may be zero or one code points,
1185 // depending on whether the indices straddle a cp boundary.
1186 for (i=0; i<startMapLimit; i++) {
1187 UChar buf[3];
1188 status = U_ZERO_ERROR;
1189 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1190 TEST_SUCCESS(status);
1191 TEST_ASSERT(extractedLen == exLen[i]);
1192 if (extractedLen > 0) {
1193 UChar32 c32;
1194 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1195 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1196 TEST_ASSERT(c32 == c32Map[i]);
1197 }
1198 }
1199
1200 utext_close(ut);
1201 }
1202 }
1203
1204
1205 void UTextTest::FreezeTest() {
1206 // Check isWritable() and freeze() behavior.
1207 //
1208
1209 UnicodeString ustr("Hello, World.");
1210 const char u8str[] = {char(0x31), (char)0x32, (char)0x33, 0};
1211 const UChar u16str[] = {(UChar)0x31, (UChar)0x32, (UChar)0x44, 0};
1212
1213 UErrorCode status = U_ZERO_ERROR;
1214 UText *ut = NULL;
1215 UText *ut2 = NULL;
1216
1217 ut = utext_openUTF8(ut, u8str, -1, &status);
1218 TEST_SUCCESS(status);
1219 UBool writable = utext_isWritable(ut);
1220 TEST_ASSERT(writable == FALSE);
1221 utext_copy(ut, 1, 2, 0, TRUE, &status);
1222 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1223
1224 status = U_ZERO_ERROR;
1225 ut = utext_openUChars(ut, u16str, -1, &status);
1226 TEST_SUCCESS(status);
1227 writable = utext_isWritable(ut);
1228 TEST_ASSERT(writable == FALSE);
1229 utext_copy(ut, 1, 2, 0, TRUE, &status);
1230 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1231
1232 status = U_ZERO_ERROR;
1233 ut = utext_openUnicodeString(ut, &ustr, &status);
1234 TEST_SUCCESS(status);
1235 writable = utext_isWritable(ut);
1236 TEST_ASSERT(writable == TRUE);
1237 utext_freeze(ut);
1238 writable = utext_isWritable(ut);
1239 TEST_ASSERT(writable == FALSE);
1240 utext_copy(ut, 1, 2, 0, TRUE, &status);
1241 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1242
1243 status = U_ZERO_ERROR;
1244 ut = utext_openUnicodeString(ut, &ustr, &status);
1245 TEST_SUCCESS(status);
1246 ut2 = utext_clone(ut2, ut, FALSE, FALSE, &status); // clone with readonly = false
1247 TEST_SUCCESS(status);
1248 writable = utext_isWritable(ut2);
1249 TEST_ASSERT(writable == TRUE);
1250 ut2 = utext_clone(ut2, ut, FALSE, TRUE, &status); // clone with readonly = true
1251 TEST_SUCCESS(status);
1252 writable = utext_isWritable(ut2);
1253 TEST_ASSERT(writable == FALSE);
1254 utext_copy(ut2, 1, 2, 0, TRUE, &status);
1255 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1256
1257 status = U_ZERO_ERROR;
1258 ut = utext_openConstUnicodeString(ut, (const UnicodeString *)&ustr, &status);
1259 TEST_SUCCESS(status);
1260 writable = utext_isWritable(ut);
1261 TEST_ASSERT(writable == FALSE);
1262 utext_copy(ut, 1, 2, 0, TRUE, &status);
1263 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1264
1265 // Deep Clone of a frozen UText should re-enable writing in the copy.
1266 status = U_ZERO_ERROR;
1267 ut = utext_openUnicodeString(ut, &ustr, &status);
1268 TEST_SUCCESS(status);
1269 utext_freeze(ut);
1270 ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status); // deep clone
1271 TEST_SUCCESS(status);
1272 writable = utext_isWritable(ut2);
1273 TEST_ASSERT(writable == TRUE);
1274
1275
1276 // Deep clone of a frozen UText, where the base type is intrinsically non-writable,
1277 // should NOT enable writing in the copy.
1278 status = U_ZERO_ERROR;
1279 ut = utext_openUChars(ut, u16str, -1, &status);
1280 TEST_SUCCESS(status);
1281 utext_freeze(ut);
1282 ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status); // deep clone
1283 TEST_SUCCESS(status);
1284 writable = utext_isWritable(ut2);
1285 TEST_ASSERT(writable == FALSE);
1286
1287 // cleanup
1288 utext_close(ut);
1289 utext_close(ut2);
1290 }
1291
1292
1293 //
1294 // Fragmented UText
1295 // A UText type that works with a chunk size of 1.
1296 // Intended to test for edge cases.
1297 // Input comes from a UnicodeString.
1298 //
1299 // ut.b the character. Put into both halves.
1300 //
1301
1302 U_CDECL_BEGIN
1303 static UBool U_CALLCONV
1304 fragTextAccess(UText *ut, int64_t index, UBool forward) {
1305 const UnicodeString *us = (const UnicodeString *)ut->context;
1306 UChar c;
1307 int32_t length = us->length();
1308 if (forward && index>=0 && index<length) {
1309 c = us->charAt((int32_t)index);
1310 ut->b = c | c<<16;
1311 ut->chunkOffset = 0;
1312 ut->chunkLength = 1;
1313 ut->chunkNativeStart = index;
1314 ut->chunkNativeLimit = index+1;
1315 return true;
1316 }
1317 if (!forward && index>0 && index <=length) {
1318 c = us->charAt((int32_t)index-1);
1319 ut->b = c | c<<16;
1320 ut->chunkOffset = 1;
1321 ut->chunkLength = 1;
1322 ut->chunkNativeStart = index-1;
1323 ut->chunkNativeLimit = index;
1324 return true;
1325 }
1326 ut->b = 0;
1327 ut->chunkOffset = 0;
1328 ut->chunkLength = 0;
1329 if (index <= 0) {
1330 ut->chunkNativeStart = 0;
1331 ut->chunkNativeLimit = 0;
1332 } else {
1333 ut->chunkNativeStart = length;
1334 ut->chunkNativeLimit = length;
1335 }
1336 return false;
1337 }
1338
1339 // Function table to be used with this fragmented text provider.
1340 // Initialized in the open function.
1341 static UTextFuncs fragmentFuncs;
1342
1343 // Clone function for fragmented text provider.
1344 // Didn't really want to provide this, but it's easier to provide it than to keep it
1345 // out of the tests.
1346 //
1347 UText *
1348 cloneFragmentedUnicodeString(UText *dest, const UText *src, UBool deep, UErrorCode *status) {
1349 if (U_FAILURE(*status)) {
1350 return NULL;
1351 }
1352 if (deep) {
1353 *status = U_UNSUPPORTED_ERROR;
1354 return NULL;
1355 }
1356 dest = utext_openUnicodeString(dest, (UnicodeString *)src->context, status);
1357 utext_setNativeIndex(dest, utext_getNativeIndex(src));
1358 return dest;
1359 }
1360
1361 U_CDECL_END
1362
1363 // Open function for the fragmented text provider.
1364 UText *
1365 openFragmentedUnicodeString(UText *ut, UnicodeString *s, UErrorCode *status) {
1366 ut = utext_openUnicodeString(ut, s, status);
1367 if (U_FAILURE(*status)) {
1368 return ut;
1369 }
1370
1371 // Copy of the function table from the stock UnicodeString UText,
1372 // and replace the entry for the access function.
1373 memcpy(&fragmentFuncs, ut->pFuncs, sizeof(fragmentFuncs));
1374 fragmentFuncs.access = fragTextAccess;
1375 fragmentFuncs.clone = cloneFragmentedUnicodeString;
1376 ut->pFuncs = &fragmentFuncs;
1377
1378 ut->chunkContents = (UChar *)&ut->b;
1379 ut->pFuncs->access(ut, 0, TRUE);
1380 return ut;
1381 }
1382
1383 // Regression test for Ticket 5560
1384 // Clone fails to update chunkContentPointer in the cloned copy.
1385 // This is only an issue for UText types that work in a local buffer,
1386 // (UTF-8 wrapper, for example)
1387 //
1388 // The test:
1389 // 1. Create an inital UText
1390 // 2. Deep clone it. Contents should match original.
1391 // 3. Reset original to something different.
1392 // 4. Check that clone contents did not change.
1393 //
1394 void UTextTest::Ticket5560() {
1395 /* The following two strings are in UTF-8 even on EBCDIC platforms. */
1396 static const char s1[] = {0x41,0x42,0x43,0x44,0x45,0x46,0}; /* "ABCDEF" */
1397 static const char s2[] = {0x31,0x32,0x33,0x34,0x35,0x36,0}; /* "123456" */
1398 UErrorCode status = U_ZERO_ERROR;
1399
1400 UText ut1 = UTEXT_INITIALIZER;
1401 UText ut2 = UTEXT_INITIALIZER;
1402
1403 utext_openUTF8(&ut1, s1, -1, &status);
1404 UChar c = utext_next32(&ut1);
1405 TEST_ASSERT(c == 0x41); // c == 'A'
1406
1407 utext_clone(&ut2, &ut1, TRUE, FALSE, &status);
1408 TEST_SUCCESS(status);
1409 c = utext_next32(&ut2);
1410 TEST_ASSERT(c == 0x42); // c == 'B'
1411 c = utext_next32(&ut1);
1412 TEST_ASSERT(c == 0x42); // c == 'B'
1413
1414 utext_openUTF8(&ut1, s2, -1, &status);
1415 c = utext_next32(&ut1);
1416 TEST_ASSERT(c == 0x31); // c == '1'
1417 c = utext_next32(&ut2);
1418 TEST_ASSERT(c == 0x43); // c == 'C'
1419
1420 utext_close(&ut1);
1421 utext_close(&ut2);
1422 }
1423
1424
1425 // Test for Ticket 6847
1426 //
1427 void UTextTest::Ticket6847() {
1428 const int STRLEN = 90;
1429 UChar s[STRLEN+1];
1430 u_memset(s, 0x41, STRLEN);
1431 s[STRLEN] = 0;
1432
1433 UErrorCode status = U_ZERO_ERROR;
1434 UText *ut = utext_openUChars(NULL, s, -1, &status);
1435
1436 utext_setNativeIndex(ut, 0);
1437 int32_t count = 0;
1438 UChar32 c = 0;
1439 int64_t nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1440 TEST_ASSERT(nativeIndex == 0);
1441 while ((c = utext_next32(ut)) != U_SENTINEL) {
1442 TEST_ASSERT(c == 0x41);
1443 TEST_ASSERT(count < STRLEN);
1444 if (count >= STRLEN) {
1445 break;
1446 }
1447 count++;
1448 nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1449 TEST_ASSERT(nativeIndex == count);
1450 }
1451 TEST_ASSERT(count == STRLEN);
1452 nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1453 TEST_ASSERT(nativeIndex == STRLEN);
1454 utext_close(ut);
1455 }
1456
1457
1458 void UTextTest::Ticket10562() {
1459 // Note: failures show as a heap error when the test is run under valgrind.
1460 UErrorCode status = U_ZERO_ERROR;
1461
1462 const char *utf8_string = "\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41";
1463 UText *utf8Text = utext_openUTF8(NULL, utf8_string, -1, &status);
1464 TEST_SUCCESS(status);
1465 UText *deepClone = utext_clone(NULL, utf8Text, TRUE, FALSE, &status);
1466 TEST_SUCCESS(status);
1467 UText *shallowClone = utext_clone(NULL, deepClone, FALSE, FALSE, &status);
1468 TEST_SUCCESS(status);
1469 utext_close(shallowClone);
1470 utext_close(deepClone);
1471 utext_close(utf8Text);
1472
1473 status = U_ZERO_ERROR;
1474 UnicodeString usString("Hello, World.");
1475 UText *usText = utext_openUnicodeString(NULL, &usString, &status);
1476 TEST_SUCCESS(status);
1477 UText *usDeepClone = utext_clone(NULL, usText, TRUE, FALSE, &status);
1478 TEST_SUCCESS(status);
1479 UText *usShallowClone = utext_clone(NULL, usDeepClone, FALSE, FALSE, &status);
1480 TEST_SUCCESS(status);
1481 utext_close(usShallowClone);
1482 utext_close(usDeepClone);
1483 utext_close(usText);
1484 }
1485
mirror of ICU