1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /********************************************************************
4 * COPYRIGHT:
5 * Copyright (c) 2005-2016, International Business Machines Corporation and
6 * others. All Rights Reserved.
7 ********************************************************************/
8 /************************************************************************
9 * Tests for the UText and UTextIterator text abstraction classses
10 *
11 ************************************************************************/
12
13 #include <string.h>
14 #include <stdio.h>
15 #include <stdlib.h>
16 #include "unicode/utypes.h"
17 #include "unicode/utext.h"
18 #include "unicode/utf8.h"
19 #include "unicode/utf16.h"
20 #include "unicode/ustring.h"
21 #include "unicode/uchriter.h"
22 #include "cmemory.h"
23 #include "cstr.h"
24 #include "utxttest.h"
25
26 static UBool gFailed = FALSE;
27 static int gTestNum = 0;
28
29 // Forward decl
30 UText *openFragmentedUnicodeString(UText *ut, UnicodeString *s, UErrorCode *status);
31
32 #define TEST_ASSERT(x) \
33 { if ((x)==FALSE) {errln("Test #%d failure in file %s at line %d\n", gTestNum, __FILE__, __LINE__);\
34 gFailed = TRUE;\
35 }}
36
37
38 #define TEST_SUCCESS(status) \
39 { if (U_FAILURE(status)) {errln("Test #%d failure in file %s at line %d. Error = \"%s\"\n", \
40 gTestNum, __FILE__, __LINE__, u_errorName(status)); \
41 gFailed = TRUE;\
42 }}
43
UTextTest()44 UTextTest::UTextTest() {
45 }
46
~UTextTest()47 UTextTest::~UTextTest() {
48 }
49
50
51 void
runIndexedTest(int32_t index,UBool exec,const char * & name,char *)52 UTextTest::runIndexedTest(int32_t index, UBool exec,
53 const char* &name, char* /*par*/) {
54 TESTCASE_AUTO_BEGIN;
55 TESTCASE_AUTO(TextTest);
56 TESTCASE_AUTO(ErrorTest);
57 TESTCASE_AUTO(FreezeTest);
58 TESTCASE_AUTO(Ticket5560);
59 TESTCASE_AUTO(Ticket6847);
60 TESTCASE_AUTO(Ticket10562);
61 TESTCASE_AUTO(Ticket10983);
62 TESTCASE_AUTO(Ticket12130);
63 TESTCASE_AUTO(Ticket13344);
64 TESTCASE_AUTO_END;
65 }
66
67 //
68 // Quick and dirty random number generator.
69 // (don't use library so that results are portable.
70 static uint32_t m_seed = 1;
m_rand()71 static uint32_t m_rand()
72 {
73 m_seed = m_seed * 1103515245 + 12345;
74 return (uint32_t)(m_seed/65536) % 32768;
75 }
76
77
78 //
79 // TextTest()
80 //
81 // Top Level function for UText testing.
82 // Specifies the strings to be tested, with the acutal testing itself
83 // being carried out in another function, TestString().
84 //
TextTest()85 void UTextTest::TextTest() {
86 int32_t i, j;
87
88 TestString("abcd\\U00010001xyz");
89 TestString("");
90
91 // Supplementary chars at start or end
92 TestString("\\U00010001");
93 TestString("abc\\U00010001");
94 TestString("\\U00010001abc");
95
96 // Test simple strings of lengths 1 to 60, looking for glitches at buffer boundaries
97 UnicodeString s;
98 for (i=1; i<60; i++) {
99 s.truncate(0);
100 for (j=0; j<i; j++) {
101 if (j+0x30 == 0x5c) {
102 // backslash. Needs to be escaped
103 s.append((UChar)0x5c);
104 }
105 s.append(UChar(j+0x30));
106 }
107 TestString(s);
108 }
109
110 // Test strings with odd-aligned supplementary chars,
111 // looking for glitches at buffer boundaries
112 for (i=1; i<60; i++) {
113 s.truncate(0);
114 s.append((UChar)0x41);
115 for (j=0; j<i; j++) {
116 s.append(UChar32(j+0x11000));
117 }
118 TestString(s);
119 }
120
121 // String of chars of randomly varying size in utf-8 representation.
122 // Exercise the mapping, and the varying sized buffer.
123 //
124 s.truncate(0);
125 UChar32 c1 = 0;
126 UChar32 c2 = 0x100;
127 UChar32 c3 = 0xa000;
128 UChar32 c4 = 0x11000;
129 for (i=0; i<1000; i++) {
130 int len8 = m_rand()%4 + 1;
131 switch (len8) {
132 case 1:
133 c1 = (c1+1)%0x80;
134 // don't put 0 into string (0 terminated strings for some tests)
135 // don't put '\', will cause unescape() to fail.
136 if (c1==0x5c || c1==0) {
137 c1++;
138 }
139 s.append(c1);
140 break;
141 case 2:
142 s.append(c2++);
143 break;
144 case 3:
145 s.append(c3++);
146 break;
147 case 4:
148 s.append(c4++);
149 break;
150 }
151 }
152 TestString(s);
153 }
154
155
156 //
157 // TestString() Run a suite of UText tests on a string.
158 // The test string is unescaped before use.
159 //
TestString(const UnicodeString & s)160 void UTextTest::TestString(const UnicodeString &s) {
161 int32_t i;
162 int32_t j;
163 UChar32 c;
164 int32_t cpCount = 0;
165 UErrorCode status = U_ZERO_ERROR;
166 UText *ut = NULL;
167 int32_t saLen;
168
169 UnicodeString sa = s.unescape();
170 saLen = sa.length();
171
172 //
173 // Build up a mapping between code points and UTF-16 code unit indexes.
174 //
175 m *cpMap = new m[sa.length() + 1];
176 j = 0;
177 for (i=0; i<sa.length(); i=sa.moveIndex32(i, 1)) {
178 c = sa.char32At(i);
179 cpMap[j].nativeIdx = i;
180 cpMap[j].cp = c;
181 j++;
182 cpCount++;
183 }
184 cpMap[j].nativeIdx = i; // position following the last char in utf-16 string.
185
186
187 // UChar * test, null terminated
188 status = U_ZERO_ERROR;
189 UChar *buf = new UChar[saLen+1];
190 sa.extract(buf, saLen+1, status);
191 TEST_SUCCESS(status);
192 ut = utext_openUChars(NULL, buf, -1, &status);
193 TEST_SUCCESS(status);
194 TestAccess(sa, ut, cpCount, cpMap);
195 utext_close(ut);
196 delete [] buf;
197
198 // UChar * test, with length
199 status = U_ZERO_ERROR;
200 buf = new UChar[saLen+1];
201 sa.extract(buf, saLen+1, status);
202 TEST_SUCCESS(status);
203 ut = utext_openUChars(NULL, buf, saLen, &status);
204 TEST_SUCCESS(status);
205 TestAccess(sa, ut, cpCount, cpMap);
206 utext_close(ut);
207 delete [] buf;
208
209
210 // UnicodeString test
211 status = U_ZERO_ERROR;
212 ut = utext_openUnicodeString(NULL, &sa, &status);
213 TEST_SUCCESS(status);
214 TestAccess(sa, ut, cpCount, cpMap);
215 TestCMR(sa, ut, cpCount, cpMap, cpMap);
216 utext_close(ut);
217
218
219 // Const UnicodeString test
220 status = U_ZERO_ERROR;
221 ut = utext_openConstUnicodeString(NULL, &sa, &status);
222 TEST_SUCCESS(status);
223 TestAccess(sa, ut, cpCount, cpMap);
224 utext_close(ut);
225
226
227 // Replaceable test. (UnicodeString inherits Replaceable)
228 status = U_ZERO_ERROR;
229 ut = utext_openReplaceable(NULL, &sa, &status);
230 TEST_SUCCESS(status);
231 TestAccess(sa, ut, cpCount, cpMap);
232 TestCMR(sa, ut, cpCount, cpMap, cpMap);
233 utext_close(ut);
234
235 // Character Iterator Tests
236 status = U_ZERO_ERROR;
237 const UChar *cbuf = sa.getBuffer();
238 CharacterIterator *ci = new UCharCharacterIterator(cbuf, saLen, status);
239 TEST_SUCCESS(status);
240 ut = utext_openCharacterIterator(NULL, ci, &status);
241 TEST_SUCCESS(status);
242 TestAccess(sa, ut, cpCount, cpMap);
243 utext_close(ut);
244 delete ci;
245
246
247 // Fragmented UnicodeString (Chunk size of one)
248 //
249 status = U_ZERO_ERROR;
250 ut = openFragmentedUnicodeString(NULL, &sa, &status);
251 TEST_SUCCESS(status);
252 TestAccess(sa, ut, cpCount, cpMap);
253 utext_close(ut);
254
255 //
256 // UTF-8 test
257 //
258
259 // Convert the test string from UnicodeString to (char *) in utf-8 format
260 int32_t u8Len = sa.extract(0, sa.length(), NULL, 0, "utf-8");
261 char *u8String = new char[u8Len + 1];
262 sa.extract(0, sa.length(), u8String, u8Len+1, "utf-8");
263
264 // Build up the map of code point indices in the utf-8 string
265 m * u8Map = new m[sa.length() + 1];
266 i = 0; // native utf-8 index
267 for (j=0; j<cpCount ; j++) { // code point number
268 u8Map[j].nativeIdx = i;
269 U8_NEXT(u8String, i, u8Len, c)
270 u8Map[j].cp = c;
271 }
272 u8Map[cpCount].nativeIdx = u8Len; // position following the last char in utf-8 string.
273
274 // Do the test itself
275 status = U_ZERO_ERROR;
276 ut = utext_openUTF8(NULL, u8String, -1, &status);
277 TEST_SUCCESS(status);
278 TestAccess(sa, ut, cpCount, u8Map);
279 utext_close(ut);
280
281
282
283 delete []cpMap;
284 delete []u8Map;
285 delete []u8String;
286 }
287
288 // TestCMR test Copy, Move and Replace operations.
289 // us UnicodeString containing the test text.
290 // ut UText containing the same test text.
291 // cpCount number of code points in the test text.
292 // nativeMap Mapping from code points to native indexes for the UText.
293 // u16Map Mapping from code points to UTF-16 indexes, for use with the UnicodeString.
294 //
295 // This function runs a whole series of opertions on each incoming UText.
296 // The UText is deep-cloned prior to each operation, so that the original UText remains unchanged.
297 //
TestCMR(const UnicodeString & us,UText * ut,int cpCount,m * nativeMap,m * u16Map)298 void UTextTest::TestCMR(const UnicodeString &us, UText *ut, int cpCount, m *nativeMap, m *u16Map) {
299 TEST_ASSERT(utext_isWritable(ut) == TRUE);
300
301 int srcLengthType; // Loop variables for selecting the postion and length
302 int srcPosType; // of the block to operate on within the source text.
303 int destPosType;
304
305 int srcIndex = 0; // Code Point indexes of the block to operate on for
306 int srcLength = 0; // a specific test.
307
308 int destIndex = 0; // Code point index of the destination for a copy/move test.
309
310 int32_t nativeStart = 0; // Native unit indexes for a test.
311 int32_t nativeLimit = 0;
312 int32_t nativeDest = 0;
313
314 int32_t u16Start = 0; // UTF-16 indexes for a test.
315 int32_t u16Limit = 0; // used when performing the same operation in a Unicode String
316 int32_t u16Dest = 0;
317
318 // Iterate over a whole series of source index, length and a target indexes.
319 // This is done with code point indexes; these will be later translated to native
320 // indexes using the cpMap.
321 for (srcLengthType=1; srcLengthType<=3; srcLengthType++) {
322 switch (srcLengthType) {
323 case 1: srcLength = 1; break;
324 case 2: srcLength = 5; break;
325 case 3: srcLength = cpCount / 3;
326 }
327 for (srcPosType=1; srcPosType<=5; srcPosType++) {
328 switch (srcPosType) {
329 case 1: srcIndex = 0; break;
330 case 2: srcIndex = 1; break;
331 case 3: srcIndex = cpCount - srcLength; break;
332 case 4: srcIndex = cpCount - srcLength - 1; break;
333 case 5: srcIndex = cpCount / 2; break;
334 }
335 if (srcIndex < 0 || srcIndex + srcLength > cpCount) {
336 // filter out bogus test cases -
337 // those with a source range that falls of an edge of the string.
338 continue;
339 }
340
341 //
342 // Copy and move tests.
343 // iterate over a variety of destination positions.
344 //
345 for (destPosType=1; destPosType<=4; destPosType++) {
346 switch (destPosType) {
347 case 1: destIndex = 0; break;
348 case 2: destIndex = 1; break;
349 case 3: destIndex = srcIndex - 1; break;
350 case 4: destIndex = srcIndex + srcLength + 1; break;
351 case 5: destIndex = cpCount-1; break;
352 case 6: destIndex = cpCount; break;
353 }
354 if (destIndex<0 || destIndex>cpCount) {
355 // filter out bogus test cases.
356 continue;
357 }
358
359 nativeStart = nativeMap[srcIndex].nativeIdx;
360 nativeLimit = nativeMap[srcIndex+srcLength].nativeIdx;
361 nativeDest = nativeMap[destIndex].nativeIdx;
362
363 u16Start = u16Map[srcIndex].nativeIdx;
364 u16Limit = u16Map[srcIndex+srcLength].nativeIdx;
365 u16Dest = u16Map[destIndex].nativeIdx;
366
367 gFailed = FALSE;
368 TestCopyMove(us, ut, FALSE,
369 nativeStart, nativeLimit, nativeDest,
370 u16Start, u16Limit, u16Dest);
371
372 TestCopyMove(us, ut, TRUE,
373 nativeStart, nativeLimit, nativeDest,
374 u16Start, u16Limit, u16Dest);
375
376 if (gFailed) {
377 return;
378 }
379 }
380
381 //
382 // Replace tests.
383 //
384 UnicodeString fullRepString("This is an arbitrary string that will be used as replacement text");
385 for (int32_t replStrLen=0; replStrLen<20; replStrLen++) {
386 UnicodeString repStr(fullRepString, 0, replStrLen);
387 TestReplace(us, ut,
388 nativeStart, nativeLimit,
389 u16Start, u16Limit,
390 repStr);
391 if (gFailed) {
392 return;
393 }
394 }
395
396 }
397 }
398
399 }
400
401 //
402 // TestCopyMove run a single test case for utext_copy.
403 // Test cases are created in TestCMR and dispatched here for execution.
404 //
TestCopyMove(const UnicodeString & us,UText * ut,UBool move,int32_t nativeStart,int32_t nativeLimit,int32_t nativeDest,int32_t u16Start,int32_t u16Limit,int32_t u16Dest)405 void UTextTest::TestCopyMove(const UnicodeString &us, UText *ut, UBool move,
406 int32_t nativeStart, int32_t nativeLimit, int32_t nativeDest,
407 int32_t u16Start, int32_t u16Limit, int32_t u16Dest)
408 {
409 UErrorCode status = U_ZERO_ERROR;
410 UText *targetUT = NULL;
411 gTestNum++;
412 gFailed = FALSE;
413
414 //
415 // clone the UText. The test will be run in the cloned copy
416 // so that we don't alter the original.
417 //
418 targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status);
419 TEST_SUCCESS(status);
420 UnicodeString targetUS(us); // And copy the reference string.
421
422 // do the test operation first in the reference
423 targetUS.copy(u16Start, u16Limit, u16Dest);
424 if (move) {
425 // delete out the source range.
426 if (u16Limit < u16Dest) {
427 targetUS.removeBetween(u16Start, u16Limit);
428 } else {
429 int32_t amtCopied = u16Limit - u16Start;
430 targetUS.removeBetween(u16Start+amtCopied, u16Limit+amtCopied);
431 }
432 }
433
434 // Do the same operation in the UText under test
435 utext_copy(targetUT, nativeStart, nativeLimit, nativeDest, move, &status);
436 if (nativeDest > nativeStart && nativeDest < nativeLimit) {
437 TEST_ASSERT(status == U_INDEX_OUTOFBOUNDS_ERROR);
438 } else {
439 TEST_SUCCESS(status);
440
441 // Compare the results of the two parallel tests
442 int32_t usi = 0; // UnicodeString postion, utf-16 index.
443 int64_t uti = 0; // UText position, native index.
444 int32_t cpi; // char32 position (code point index)
445 UChar32 usc; // code point from Unicode String
446 UChar32 utc; // code point from UText
447 utext_setNativeIndex(targetUT, 0);
448 for (cpi=0; ; cpi++) {
449 usc = targetUS.char32At(usi);
450 utc = utext_next32(targetUT);
451 if (utc < 0) {
452 break;
453 }
454 TEST_ASSERT(uti == usi);
455 TEST_ASSERT(utc == usc);
456 usi = targetUS.moveIndex32(usi, 1);
457 uti = utext_getNativeIndex(targetUT);
458 if (gFailed) {
459 goto cleanupAndReturn;
460 }
461 }
462 int64_t expectedNativeLength = utext_nativeLength(ut);
463 if (move == FALSE) {
464 expectedNativeLength += nativeLimit - nativeStart;
465 }
466 uti = utext_getNativeIndex(targetUT);
467 TEST_ASSERT(uti == expectedNativeLength);
468 }
469
470 cleanupAndReturn:
471 utext_close(targetUT);
472 }
473
474
475 //
476 // TestReplace Test a single Replace operation.
477 //
TestReplace(const UnicodeString & us,UText * ut,int32_t nativeStart,int32_t nativeLimit,int32_t u16Start,int32_t u16Limit,const UnicodeString & repStr)478 void UTextTest::TestReplace(
479 const UnicodeString &us, // reference UnicodeString in which to do the replace
480 UText *ut, // UnicodeText object under test.
481 int32_t nativeStart, // Range to be replaced, in UText native units.
482 int32_t nativeLimit,
483 int32_t u16Start, // Range to be replaced, in UTF-16 units
484 int32_t u16Limit, // for use in the reference UnicodeString.
485 const UnicodeString &repStr) // The replacement string
486 {
487 UErrorCode status = U_ZERO_ERROR;
488 UText *targetUT = NULL;
489 gTestNum++;
490 gFailed = FALSE;
491
492 //
493 // clone the target UText. The test will be run in the cloned copy
494 // so that we don't alter the original.
495 //
496 targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status);
497 TEST_SUCCESS(status);
498 UnicodeString targetUS(us); // And copy the reference string.
499
500 //
501 // Do the replace operation in the Unicode String, to
502 // produce a reference result.
503 //
504 targetUS.replace(u16Start, u16Limit-u16Start, repStr);
505
506 //
507 // Do the replace on the UText under test
508 //
509 const UChar *rs = repStr.getBuffer();
510 int32_t rsLen = repStr.length();
511 int32_t actualDelta = utext_replace(targetUT, nativeStart, nativeLimit, rs, rsLen, &status);
512 int32_t expectedDelta = repStr.length() - (nativeLimit - nativeStart);
513 TEST_ASSERT(actualDelta == expectedDelta);
514
515 //
516 // Compare the results
517 //
518 int32_t usi = 0; // UnicodeString postion, utf-16 index.
519 int64_t uti = 0; // UText position, native index.
520 int32_t cpi; // char32 position (code point index)
521 UChar32 usc; // code point from Unicode String
522 UChar32 utc; // code point from UText
523 int64_t expectedNativeLength = 0;
524 utext_setNativeIndex(targetUT, 0);
525 for (cpi=0; ; cpi++) {
526 usc = targetUS.char32At(usi);
527 utc = utext_next32(targetUT);
528 if (utc < 0) {
529 break;
530 }
531 TEST_ASSERT(uti == usi);
532 TEST_ASSERT(utc == usc);
533 usi = targetUS.moveIndex32(usi, 1);
534 uti = utext_getNativeIndex(targetUT);
535 if (gFailed) {
536 goto cleanupAndReturn;
537 }
538 }
539 expectedNativeLength = utext_nativeLength(ut) + expectedDelta;
540 uti = utext_getNativeIndex(targetUT);
541 TEST_ASSERT(uti == expectedNativeLength);
542
543 cleanupAndReturn:
544 utext_close(targetUT);
545 }
546
547 //
548 // TestAccess Test the read only access functions on a UText, including cloning.
549 // The text is accessed in a variety of ways, and compared with
550 // the reference UnicodeString.
551 //
TestAccess(const UnicodeString & us,UText * ut,int cpCount,m * cpMap)552 void UTextTest::TestAccess(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) {
553 // Run the standard tests on the caller-supplied UText.
554 TestAccessNoClone(us, ut, cpCount, cpMap);
555
556 // Re-run tests on a shallow clone.
557 utext_setNativeIndex(ut, 0);
558 UErrorCode status = U_ZERO_ERROR;
559 UText *shallowClone = utext_clone(NULL, ut, FALSE /*deep*/, FALSE /*readOnly*/, &status);
560 TEST_SUCCESS(status);
561 TestAccessNoClone(us, shallowClone, cpCount, cpMap);
562
563 //
564 // Rerun again on a deep clone.
565 // Note that text providers are not required to provide deep cloning,
566 // so unsupported errors are ignored.
567 //
568 status = U_ZERO_ERROR;
569 utext_setNativeIndex(shallowClone, 0);
570 UText *deepClone = utext_clone(NULL, shallowClone, TRUE, FALSE, &status);
571 utext_close(shallowClone);
572 if (status != U_UNSUPPORTED_ERROR) {
573 TEST_SUCCESS(status);
574 TestAccessNoClone(us, deepClone, cpCount, cpMap);
575 }
576 utext_close(deepClone);
577 }
578
579
580 //
581 // TestAccessNoClone() Test the read only access functions on a UText.
582 // The text is accessed in a variety of ways, and compared with
583 // the reference UnicodeString.
584 //
TestAccessNoClone(const UnicodeString & us,UText * ut,int cpCount,m * cpMap)585 void UTextTest::TestAccessNoClone(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) {
586 UErrorCode status = U_ZERO_ERROR;
587 gTestNum++;
588
589 //
590 // Check the length from the UText
591 //
592 int64_t expectedLen = cpMap[cpCount].nativeIdx;
593 int64_t utlen = utext_nativeLength(ut);
594 TEST_ASSERT(expectedLen == utlen);
595
596 //
597 // Iterate forwards, verify that we get the correct code points
598 // at the correct native offsets.
599 //
600 int i = 0;
601 int64_t index;
602 int64_t expectedIndex = 0;
603 int64_t foundIndex = 0;
604 UChar32 expectedC;
605 UChar32 foundC;
606 int64_t len;
607
608 for (i=0; i<cpCount; i++) {
609 expectedIndex = cpMap[i].nativeIdx;
610 foundIndex = utext_getNativeIndex(ut);
611 TEST_ASSERT(expectedIndex == foundIndex);
612 expectedC = cpMap[i].cp;
613 foundC = utext_next32(ut);
614 TEST_ASSERT(expectedC == foundC);
615 foundIndex = utext_getPreviousNativeIndex(ut);
616 TEST_ASSERT(expectedIndex == foundIndex);
617 if (gFailed) {
618 return;
619 }
620 }
621 foundC = utext_next32(ut);
622 TEST_ASSERT(foundC == U_SENTINEL);
623
624 // Repeat above, using macros
625 utext_setNativeIndex(ut, 0);
626 for (i=0; i<cpCount; i++) {
627 expectedIndex = cpMap[i].nativeIdx;
628 foundIndex = UTEXT_GETNATIVEINDEX(ut);
629 TEST_ASSERT(expectedIndex == foundIndex);
630 expectedC = cpMap[i].cp;
631 foundC = UTEXT_NEXT32(ut);
632 TEST_ASSERT(expectedC == foundC);
633 if (gFailed) {
634 return;
635 }
636 }
637 foundC = UTEXT_NEXT32(ut);
638 TEST_ASSERT(foundC == U_SENTINEL);
639
640 //
641 // Forward iteration (above) should have left index at the
642 // end of the input, which should == length().
643 //
644 len = utext_nativeLength(ut);
645 foundIndex = utext_getNativeIndex(ut);
646 TEST_ASSERT(len == foundIndex);
647
648 //
649 // Iterate backwards over entire test string
650 //
651 len = utext_getNativeIndex(ut);
652 utext_setNativeIndex(ut, len);
653 for (i=cpCount-1; i>=0; i--) {
654 expectedC = cpMap[i].cp;
655 expectedIndex = cpMap[i].nativeIdx;
656 int64_t prevIndex = utext_getPreviousNativeIndex(ut);
657 foundC = utext_previous32(ut);
658 foundIndex = utext_getNativeIndex(ut);
659 TEST_ASSERT(expectedIndex == foundIndex);
660 TEST_ASSERT(expectedC == foundC);
661 TEST_ASSERT(prevIndex == foundIndex);
662 if (gFailed) {
663 return;
664 }
665 }
666
667 //
668 // Backwards iteration, above, should have left our iterator
669 // position at zero, and continued backwards iterationshould fail.
670 //
671 foundIndex = utext_getNativeIndex(ut);
672 TEST_ASSERT(foundIndex == 0);
673 foundIndex = utext_getPreviousNativeIndex(ut);
674 TEST_ASSERT(foundIndex == 0);
675
676
677 foundC = utext_previous32(ut);
678 TEST_ASSERT(foundC == U_SENTINEL);
679 foundIndex = utext_getNativeIndex(ut);
680 TEST_ASSERT(foundIndex == 0);
681 foundIndex = utext_getPreviousNativeIndex(ut);
682 TEST_ASSERT(foundIndex == 0);
683
684
685 // And again, with the macros
686 utext_setNativeIndex(ut, len);
687 for (i=cpCount-1; i>=0; i--) {
688 expectedC = cpMap[i].cp;
689 expectedIndex = cpMap[i].nativeIdx;
690 foundC = UTEXT_PREVIOUS32(ut);
691 foundIndex = UTEXT_GETNATIVEINDEX(ut);
692 TEST_ASSERT(expectedIndex == foundIndex);
693 TEST_ASSERT(expectedC == foundC);
694 if (gFailed) {
695 return;
696 }
697 }
698
699 //
700 // Backwards iteration, above, should have left our iterator
701 // position at zero, and continued backwards iterationshould fail.
702 //
703 foundIndex = UTEXT_GETNATIVEINDEX(ut);
704 TEST_ASSERT(foundIndex == 0);
705
706 foundC = UTEXT_PREVIOUS32(ut);
707 TEST_ASSERT(foundC == U_SENTINEL);
708 foundIndex = UTEXT_GETNATIVEINDEX(ut);
709 TEST_ASSERT(foundIndex == 0);
710 if (gFailed) {
711 return;
712 }
713
714 //
715 // next32From(), prevous32From(), Iterate in a somewhat random order.
716 //
717 int cpIndex = 0;
718 for (i=0; i<cpCount; i++) {
719 cpIndex = (cpIndex + 9973) % cpCount;
720 index = cpMap[cpIndex].nativeIdx;
721 expectedC = cpMap[cpIndex].cp;
722 foundC = utext_next32From(ut, index);
723 TEST_ASSERT(expectedC == foundC);
724 if (gFailed) {
725 return;
726 }
727 }
728
729 cpIndex = 0;
730 for (i=0; i<cpCount; i++) {
731 cpIndex = (cpIndex + 9973) % cpCount;
732 index = cpMap[cpIndex+1].nativeIdx;
733 expectedC = cpMap[cpIndex].cp;
734 foundC = utext_previous32From(ut, index);
735 TEST_ASSERT(expectedC == foundC);
736 if (gFailed) {
737 return;
738 }
739 }
740
741
742 //
743 // moveIndex(int32_t delta);
744 //
745
746 // Walk through frontwards, incrementing by one
747 utext_setNativeIndex(ut, 0);
748 for (i=1; i<=cpCount; i++) {
749 utext_moveIndex32(ut, 1);
750 index = utext_getNativeIndex(ut);
751 expectedIndex = cpMap[i].nativeIdx;
752 TEST_ASSERT(expectedIndex == index);
753 index = UTEXT_GETNATIVEINDEX(ut);
754 TEST_ASSERT(expectedIndex == index);
755 }
756
757 // Walk through frontwards, incrementing by two
758 utext_setNativeIndex(ut, 0);
759 for (i=2; i<cpCount; i+=2) {
760 utext_moveIndex32(ut, 2);
761 index = utext_getNativeIndex(ut);
762 expectedIndex = cpMap[i].nativeIdx;
763 TEST_ASSERT(expectedIndex == index);
764 index = UTEXT_GETNATIVEINDEX(ut);
765 TEST_ASSERT(expectedIndex == index);
766 }
767
768 // walk through the string backwards, decrementing by one.
769 i = cpMap[cpCount].nativeIdx;
770 utext_setNativeIndex(ut, i);
771 for (i=cpCount; i>=0; i--) {
772 expectedIndex = cpMap[i].nativeIdx;
773 index = utext_getNativeIndex(ut);
774 TEST_ASSERT(expectedIndex == index);
775 index = UTEXT_GETNATIVEINDEX(ut);
776 TEST_ASSERT(expectedIndex == index);
777 utext_moveIndex32(ut, -1);
778 }
779
780
781 // walk through backwards, decrementing by three
782 i = cpMap[cpCount].nativeIdx;
783 utext_setNativeIndex(ut, i);
784 for (i=cpCount; i>=0; i-=3) {
785 expectedIndex = cpMap[i].nativeIdx;
786 index = utext_getNativeIndex(ut);
787 TEST_ASSERT(expectedIndex == index);
788 index = UTEXT_GETNATIVEINDEX(ut);
789 TEST_ASSERT(expectedIndex == index);
790 utext_moveIndex32(ut, -3);
791 }
792
793
794 //
795 // Extract
796 //
797 int bufSize = us.length() + 10;
798 UChar *buf = new UChar[bufSize];
799 status = U_ZERO_ERROR;
800 expectedLen = us.length();
801 len = utext_extract(ut, 0, utlen, buf, bufSize, &status);
802 TEST_SUCCESS(status);
803 TEST_ASSERT(len == expectedLen);
804 int compareResult = us.compare(buf, -1);
805 TEST_ASSERT(compareResult == 0);
806
807 status = U_ZERO_ERROR;
808 len = utext_extract(ut, 0, utlen, NULL, 0, &status);
809 if (utlen == 0) {
810 TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING);
811 } else {
812 TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR);
813 }
814 TEST_ASSERT(len == expectedLen);
815
816 status = U_ZERO_ERROR;
817 u_memset(buf, 0x5555, bufSize);
818 len = utext_extract(ut, 0, utlen, buf, 1, &status);
819 if (us.length() == 0) {
820 TEST_SUCCESS(status);
821 TEST_ASSERT(buf[0] == 0);
822 } else {
823 // Buf len == 1, extracting a single 16 bit value.
824 // If the data char is supplementary, it doesn't matter whether the buffer remains unchanged,
825 // or whether the lead surrogate of the pair is extracted.
826 // It's a buffer overflow error in either case.
827 TEST_ASSERT(buf[0] == us.charAt(0) ||
828 (buf[0] == 0x5555 && U_IS_SUPPLEMENTARY(us.char32At(0))));
829 TEST_ASSERT(buf[1] == 0x5555);
830 if (us.length() == 1) {
831 TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING);
832 } else {
833 TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR);
834 }
835 }
836
837 delete []buf;
838 }
839
840 //
841 // ErrorTest() Check various error and edge cases.
842 //
ErrorTest()843 void UTextTest::ErrorTest()
844 {
845 // Close of an unitialized UText. Shouldn't blow up.
846 {
847 UText ut;
848 memset(&ut, 0, sizeof(UText));
849 utext_close(&ut);
850 utext_close(NULL);
851 }
852
853 // Double-close of a UText. Shouldn't blow up. UText should still be usable.
854 {
855 UErrorCode status = U_ZERO_ERROR;
856 UText ut = UTEXT_INITIALIZER;
857 UnicodeString s("Hello, World");
858 UText *ut2 = utext_openUnicodeString(&ut, &s, &status);
859 TEST_SUCCESS(status);
860 TEST_ASSERT(ut2 == &ut);
861
862 UText *ut3 = utext_close(&ut);
863 TEST_ASSERT(ut3 == &ut);
864
865 UText *ut4 = utext_close(&ut);
866 TEST_ASSERT(ut4 == &ut);
867
868 utext_openUnicodeString(&ut, &s, &status);
869 TEST_SUCCESS(status);
870 utext_close(&ut);
871 }
872
873 // Re-use of a UText, chaining through each of the types of UText
874 // (If it doesn't blow up, and doesn't leak, it's probably working fine)
875 {
876 UErrorCode status = U_ZERO_ERROR;
877 UText ut = UTEXT_INITIALIZER;
878 UText *utp;
879 UnicodeString s1("Hello, World");
880 UChar s2[] = {(UChar)0x41, (UChar)0x42, (UChar)0};
881 const char *s3 = "\x66\x67\x68";
882
883 utp = utext_openUnicodeString(&ut, &s1, &status);
884 TEST_SUCCESS(status);
885 TEST_ASSERT(utp == &ut);
886
887 utp = utext_openConstUnicodeString(&ut, &s1, &status);
888 TEST_SUCCESS(status);
889 TEST_ASSERT(utp == &ut);
890
891 utp = utext_openUTF8(&ut, s3, -1, &status);
892 TEST_SUCCESS(status);
893 TEST_ASSERT(utp == &ut);
894
895 utp = utext_openUChars(&ut, s2, -1, &status);
896 TEST_SUCCESS(status);
897 TEST_ASSERT(utp == &ut);
898
899 utp = utext_close(&ut);
900 TEST_ASSERT(utp == &ut);
901
902 utp = utext_openUnicodeString(&ut, &s1, &status);
903 TEST_SUCCESS(status);
904 TEST_ASSERT(utp == &ut);
905 }
906
907 // Invalid parameters on open
908 //
909 {
910 UErrorCode status = U_ZERO_ERROR;
911 UText ut = UTEXT_INITIALIZER;
912
913 utext_openUChars(&ut, NULL, 5, &status);
914 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
915
916 status = U_ZERO_ERROR;
917 utext_openUChars(&ut, NULL, -1, &status);
918 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
919
920 status = U_ZERO_ERROR;
921 utext_openUTF8(&ut, NULL, 4, &status);
922 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
923
924 status = U_ZERO_ERROR;
925 utext_openUTF8(&ut, NULL, -1, &status);
926 TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
927 }
928
929 //
930 // UTF-8 with malformed sequences.
931 // These should come through as the Unicode replacement char, \ufffd
932 //
933 {
934 UErrorCode status = U_ZERO_ERROR;
935 UText *ut = NULL;
936 const char *badUTF8 = "\x41\x81\x42\xf0\x81\x81\x43";
937 UChar32 c;
938
939 ut = utext_openUTF8(NULL, badUTF8, -1, &status);
940 TEST_SUCCESS(status);
941 c = utext_char32At(ut, 1);
942 TEST_ASSERT(c == 0xfffd);
943 c = utext_char32At(ut, 3);
944 TEST_ASSERT(c == 0xfffd);
945 c = utext_char32At(ut, 5);
946 TEST_ASSERT(c == 0xfffd);
947 c = utext_char32At(ut, 6);
948 TEST_ASSERT(c == 0x43);
949
950 UChar buf[10];
951 int n = utext_extract(ut, 0, 9, buf, 10, &status);
952 TEST_SUCCESS(status);
953 TEST_ASSERT(n==7);
954 TEST_ASSERT(buf[0] == 0x41);
955 TEST_ASSERT(buf[1] == 0xfffd);
956 TEST_ASSERT(buf[2] == 0x42);
957 TEST_ASSERT(buf[3] == 0xfffd);
958 TEST_ASSERT(buf[4] == 0xfffd);
959 TEST_ASSERT(buf[5] == 0xfffd);
960 TEST_ASSERT(buf[6] == 0x43);
961 utext_close(ut);
962 }
963
964
965 //
966 // isLengthExpensive - does it make the exptected transitions after
967 // getting the length of a nul terminated string?
968 //
969 {
970 UErrorCode status = U_ZERO_ERROR;
971 UnicodeString sa("Hello, this is a string");
972 UBool isExpensive;
973
974 UChar sb[100];
975 memset(sb, 0x20, sizeof(sb));
976 sb[99] = 0;
977
978 UText *uta = utext_openUnicodeString(NULL, &sa, &status);
979 TEST_SUCCESS(status);
980 isExpensive = utext_isLengthExpensive(uta);
981 TEST_ASSERT(isExpensive == FALSE);
982 utext_close(uta);
983
984 UText *utb = utext_openUChars(NULL, sb, -1, &status);
985 TEST_SUCCESS(status);
986 isExpensive = utext_isLengthExpensive(utb);
987 TEST_ASSERT(isExpensive == TRUE);
988 int64_t len = utext_nativeLength(utb);
989 TEST_ASSERT(len == 99);
990 isExpensive = utext_isLengthExpensive(utb);
991 TEST_ASSERT(isExpensive == FALSE);
992 utext_close(utb);
993 }
994
995 //
996 // Index to positions not on code point boundaries.
997 //
998 {
999 const char *u8str = "\xc8\x81\xe1\x82\x83\xf1\x84\x85\x86";
1000 int32_t startMap[] = { 0, 0, 2, 2, 2, 5, 5, 5, 5, 9, 9};
1001 int32_t nextMap[] = { 2, 2, 5, 5, 5, 9, 9, 9, 9, 9, 9};
1002 int32_t prevMap[] = { 0, 0, 0, 0, 0, 2, 2, 2, 2, 5, 5};
1003 UChar32 c32Map[] = {0x201, 0x201, 0x1083, 0x1083, 0x1083, 0x044146, 0x044146, 0x044146, 0x044146, -1, -1};
1004 UChar32 pr32Map[] = { -1, -1, 0x201, 0x201, 0x201, 0x1083, 0x1083, 0x1083, 0x1083, 0x044146, 0x044146};
1005
1006 // extractLen is the size, in UChars, of what will be extracted between index and index+1.
1007 // is zero when both index positions lie within the same code point.
1008 int32_t exLen[] = { 0, 1, 0, 0, 1, 0, 0, 0, 2, 0, 0};
1009
1010
1011 UErrorCode status = U_ZERO_ERROR;
1012 UText *ut = utext_openUTF8(NULL, u8str, -1, &status);
1013 TEST_SUCCESS(status);
1014
1015 // Check setIndex
1016 int32_t i;
1017 int32_t startMapLimit = UPRV_LENGTHOF(startMap);
1018 for (i=0; i<startMapLimit; i++) {
1019 utext_setNativeIndex(ut, i);
1020 int64_t cpIndex = utext_getNativeIndex(ut);
1021 TEST_ASSERT(cpIndex == startMap[i]);
1022 cpIndex = UTEXT_GETNATIVEINDEX(ut);
1023 TEST_ASSERT(cpIndex == startMap[i]);
1024 }
1025
1026 // Check char32At
1027 for (i=0; i<startMapLimit; i++) {
1028 UChar32 c32 = utext_char32At(ut, i);
1029 TEST_ASSERT(c32 == c32Map[i]);
1030 int64_t cpIndex = utext_getNativeIndex(ut);
1031 TEST_ASSERT(cpIndex == startMap[i]);
1032 }
1033
1034 // Check utext_next32From
1035 for (i=0; i<startMapLimit; i++) {
1036 UChar32 c32 = utext_next32From(ut, i);
1037 TEST_ASSERT(c32 == c32Map[i]);
1038 int64_t cpIndex = utext_getNativeIndex(ut);
1039 TEST_ASSERT(cpIndex == nextMap[i]);
1040 }
1041
1042 // check utext_previous32From
1043 for (i=0; i<startMapLimit; i++) {
1044 gTestNum++;
1045 UChar32 c32 = utext_previous32From(ut, i);
1046 TEST_ASSERT(c32 == pr32Map[i]);
1047 int64_t cpIndex = utext_getNativeIndex(ut);
1048 TEST_ASSERT(cpIndex == prevMap[i]);
1049 }
1050
1051 // check Extract
1052 // Extract from i to i+1, which may be zero or one code points,
1053 // depending on whether the indices straddle a cp boundary.
1054 for (i=0; i<startMapLimit; i++) {
1055 UChar buf[3];
1056 status = U_ZERO_ERROR;
1057 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1058 TEST_SUCCESS(status);
1059 TEST_ASSERT(extractedLen == exLen[i]);
1060 if (extractedLen > 0) {
1061 UChar32 c32;
1062 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1063 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1064 TEST_ASSERT(c32 == c32Map[i]);
1065 }
1066 }
1067
1068 utext_close(ut);
1069 }
1070
1071
1072 { // Similar test, with utf16 instead of utf8
1073 // TODO: merge the common parts of these tests.
1074
1075 UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV);
1076 int32_t startMap[] ={ 0, 1, 1, 3, 4, 4, 6, 6};
1077 int32_t nextMap[] = { 1, 3, 3, 4, 6, 6, 6, 6};
1078 int32_t prevMap[] = { 0, 0, 0, 1, 3, 3, 4, 4};
1079 UChar32 c32Map[] = {0x1000, 0x11000, 0x11000, 0x2000, 0x22000, 0x22000, -1, -1};
1080 UChar32 pr32Map[] = { -1, 0x1000, 0x1000, 0x11000, 0x2000, 0x2000, 0x22000, 0x22000};
1081 int32_t exLen[] = { 1, 0, 2, 1, 0, 2, 0, 0,};
1082
1083 u16str = u16str.unescape();
1084 UErrorCode status = U_ZERO_ERROR;
1085 UText *ut = utext_openUnicodeString(NULL, &u16str, &status);
1086 TEST_SUCCESS(status);
1087
1088 int32_t startMapLimit = UPRV_LENGTHOF(startMap);
1089 int i;
1090 for (i=0; i<startMapLimit; i++) {
1091 utext_setNativeIndex(ut, i);
1092 int64_t cpIndex = utext_getNativeIndex(ut);
1093 TEST_ASSERT(cpIndex == startMap[i]);
1094 }
1095
1096 // Check char32At
1097 for (i=0; i<startMapLimit; i++) {
1098 UChar32 c32 = utext_char32At(ut, i);
1099 TEST_ASSERT(c32 == c32Map[i]);
1100 int64_t cpIndex = utext_getNativeIndex(ut);
1101 TEST_ASSERT(cpIndex == startMap[i]);
1102 }
1103
1104 // Check utext_next32From
1105 for (i=0; i<startMapLimit; i++) {
1106 UChar32 c32 = utext_next32From(ut, i);
1107 TEST_ASSERT(c32 == c32Map[i]);
1108 int64_t cpIndex = utext_getNativeIndex(ut);
1109 TEST_ASSERT(cpIndex == nextMap[i]);
1110 }
1111
1112 // check utext_previous32From
1113 for (i=0; i<startMapLimit; i++) {
1114 UChar32 c32 = utext_previous32From(ut, i);
1115 TEST_ASSERT(c32 == pr32Map[i]);
1116 int64_t cpIndex = utext_getNativeIndex(ut);
1117 TEST_ASSERT(cpIndex == prevMap[i]);
1118 }
1119
1120 // check Extract
1121 // Extract from i to i+1, which may be zero or one code points,
1122 // depending on whether the indices straddle a cp boundary.
1123 for (i=0; i<startMapLimit; i++) {
1124 UChar buf[3];
1125 status = U_ZERO_ERROR;
1126 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1127 TEST_SUCCESS(status);
1128 TEST_ASSERT(extractedLen == exLen[i]);
1129 if (extractedLen > 0) {
1130 UChar32 c32;
1131 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1132 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1133 TEST_ASSERT(c32 == c32Map[i]);
1134 }
1135 }
1136
1137 utext_close(ut);
1138 }
1139
1140 { // Similar test, with UText over Replaceable
1141 // TODO: merge the common parts of these tests.
1142
1143 UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV);
1144 int32_t startMap[] ={ 0, 1, 1, 3, 4, 4, 6, 6};
1145 int32_t nextMap[] = { 1, 3, 3, 4, 6, 6, 6, 6};
1146 int32_t prevMap[] = { 0, 0, 0, 1, 3, 3, 4, 4};
1147 UChar32 c32Map[] = {0x1000, 0x11000, 0x11000, 0x2000, 0x22000, 0x22000, -1, -1};
1148 UChar32 pr32Map[] = { -1, 0x1000, 0x1000, 0x11000, 0x2000, 0x2000, 0x22000, 0x22000};
1149 int32_t exLen[] = { 1, 0, 2, 1, 0, 2, 0, 0,};
1150
1151 u16str = u16str.unescape();
1152 UErrorCode status = U_ZERO_ERROR;
1153 UText *ut = utext_openReplaceable(NULL, &u16str, &status);
1154 TEST_SUCCESS(status);
1155
1156 int32_t startMapLimit = UPRV_LENGTHOF(startMap);
1157 int i;
1158 for (i=0; i<startMapLimit; i++) {
1159 utext_setNativeIndex(ut, i);
1160 int64_t cpIndex = utext_getNativeIndex(ut);
1161 TEST_ASSERT(cpIndex == startMap[i]);
1162 }
1163
1164 // Check char32At
1165 for (i=0; i<startMapLimit; i++) {
1166 UChar32 c32 = utext_char32At(ut, i);
1167 TEST_ASSERT(c32 == c32Map[i]);
1168 int64_t cpIndex = utext_getNativeIndex(ut);
1169 TEST_ASSERT(cpIndex == startMap[i]);
1170 }
1171
1172 // Check utext_next32From
1173 for (i=0; i<startMapLimit; i++) {
1174 UChar32 c32 = utext_next32From(ut, i);
1175 TEST_ASSERT(c32 == c32Map[i]);
1176 int64_t cpIndex = utext_getNativeIndex(ut);
1177 TEST_ASSERT(cpIndex == nextMap[i]);
1178 }
1179
1180 // check utext_previous32From
1181 for (i=0; i<startMapLimit; i++) {
1182 UChar32 c32 = utext_previous32From(ut, i);
1183 TEST_ASSERT(c32 == pr32Map[i]);
1184 int64_t cpIndex = utext_getNativeIndex(ut);
1185 TEST_ASSERT(cpIndex == prevMap[i]);
1186 }
1187
1188 // check Extract
1189 // Extract from i to i+1, which may be zero or one code points,
1190 // depending on whether the indices straddle a cp boundary.
1191 for (i=0; i<startMapLimit; i++) {
1192 UChar buf[3];
1193 status = U_ZERO_ERROR;
1194 int32_t extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1195 TEST_SUCCESS(status);
1196 TEST_ASSERT(extractedLen == exLen[i]);
1197 if (extractedLen > 0) {
1198 UChar32 c32;
1199 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1200 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1201 TEST_ASSERT(c32 == c32Map[i]);
1202 }
1203 }
1204
1205 utext_close(ut);
1206 }
1207 }
1208
1209
FreezeTest()1210 void UTextTest::FreezeTest() {
1211 // Check isWritable() and freeze() behavior.
1212 //
1213
1214 UnicodeString ustr("Hello, World.");
1215 const char u8str[] = {char(0x31), (char)0x32, (char)0x33, 0};
1216 const UChar u16str[] = {(UChar)0x31, (UChar)0x32, (UChar)0x44, 0};
1217
1218 UErrorCode status = U_ZERO_ERROR;
1219 UText *ut = NULL;
1220 UText *ut2 = NULL;
1221
1222 ut = utext_openUTF8(ut, u8str, -1, &status);
1223 TEST_SUCCESS(status);
1224 UBool writable = utext_isWritable(ut);
1225 TEST_ASSERT(writable == FALSE);
1226 utext_copy(ut, 1, 2, 0, TRUE, &status);
1227 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1228
1229 status = U_ZERO_ERROR;
1230 ut = utext_openUChars(ut, u16str, -1, &status);
1231 TEST_SUCCESS(status);
1232 writable = utext_isWritable(ut);
1233 TEST_ASSERT(writable == FALSE);
1234 utext_copy(ut, 1, 2, 0, TRUE, &status);
1235 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1236
1237 status = U_ZERO_ERROR;
1238 ut = utext_openUnicodeString(ut, &ustr, &status);
1239 TEST_SUCCESS(status);
1240 writable = utext_isWritable(ut);
1241 TEST_ASSERT(writable == TRUE);
1242 utext_freeze(ut);
1243 writable = utext_isWritable(ut);
1244 TEST_ASSERT(writable == FALSE);
1245 utext_copy(ut, 1, 2, 0, TRUE, &status);
1246 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1247
1248 status = U_ZERO_ERROR;
1249 ut = utext_openUnicodeString(ut, &ustr, &status);
1250 TEST_SUCCESS(status);
1251 ut2 = utext_clone(ut2, ut, FALSE, FALSE, &status); // clone with readonly = false
1252 TEST_SUCCESS(status);
1253 writable = utext_isWritable(ut2);
1254 TEST_ASSERT(writable == TRUE);
1255 ut2 = utext_clone(ut2, ut, FALSE, TRUE, &status); // clone with readonly = true
1256 TEST_SUCCESS(status);
1257 writable = utext_isWritable(ut2);
1258 TEST_ASSERT(writable == FALSE);
1259 utext_copy(ut2, 1, 2, 0, TRUE, &status);
1260 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1261
1262 status = U_ZERO_ERROR;
1263 ut = utext_openConstUnicodeString(ut, (const UnicodeString *)&ustr, &status);
1264 TEST_SUCCESS(status);
1265 writable = utext_isWritable(ut);
1266 TEST_ASSERT(writable == FALSE);
1267 utext_copy(ut, 1, 2, 0, TRUE, &status);
1268 TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1269
1270 // Deep Clone of a frozen UText should re-enable writing in the copy.
1271 status = U_ZERO_ERROR;
1272 ut = utext_openUnicodeString(ut, &ustr, &status);
1273 TEST_SUCCESS(status);
1274 utext_freeze(ut);
1275 ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status); // deep clone
1276 TEST_SUCCESS(status);
1277 writable = utext_isWritable(ut2);
1278 TEST_ASSERT(writable == TRUE);
1279
1280
1281 // Deep clone of a frozen UText, where the base type is intrinsically non-writable,
1282 // should NOT enable writing in the copy.
1283 status = U_ZERO_ERROR;
1284 ut = utext_openUChars(ut, u16str, -1, &status);
1285 TEST_SUCCESS(status);
1286 utext_freeze(ut);
1287 ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status); // deep clone
1288 TEST_SUCCESS(status);
1289 writable = utext_isWritable(ut2);
1290 TEST_ASSERT(writable == FALSE);
1291
1292 // cleanup
1293 utext_close(ut);
1294 utext_close(ut2);
1295 }
1296
1297
1298 //
1299 // Fragmented UText
1300 // A UText type that works with a chunk size of 1.
1301 // Intended to test for edge cases.
1302 // Input comes from a UnicodeString.
1303 //
1304 // ut.b the character. Put into both halves.
1305 //
1306
1307 U_CDECL_BEGIN
1308 static UBool U_CALLCONV
fragTextAccess(UText * ut,int64_t index,UBool forward)1309 fragTextAccess(UText *ut, int64_t index, UBool forward) {
1310 const UnicodeString *us = (const UnicodeString *)ut->context;
1311 UChar c;
1312 int32_t length = us->length();
1313 if (forward && index>=0 && index<length) {
1314 c = us->charAt((int32_t)index);
1315 ut->b = c | c<<16;
1316 ut->chunkOffset = 0;
1317 ut->chunkLength = 1;
1318 ut->chunkNativeStart = index;
1319 ut->chunkNativeLimit = index+1;
1320 return true;
1321 }
1322 if (!forward && index>0 && index <=length) {
1323 c = us->charAt((int32_t)index-1);
1324 ut->b = c | c<<16;
1325 ut->chunkOffset = 1;
1326 ut->chunkLength = 1;
1327 ut->chunkNativeStart = index-1;
1328 ut->chunkNativeLimit = index;
1329 return true;
1330 }
1331 ut->b = 0;
1332 ut->chunkOffset = 0;
1333 ut->chunkLength = 0;
1334 if (index <= 0) {
1335 ut->chunkNativeStart = 0;
1336 ut->chunkNativeLimit = 0;
1337 } else {
1338 ut->chunkNativeStart = length;
1339 ut->chunkNativeLimit = length;
1340 }
1341 return false;
1342 }
1343
1344 // Function table to be used with this fragmented text provider.
1345 // Initialized in the open function.
1346 static UTextFuncs fragmentFuncs;
1347
1348 // Clone function for fragmented text provider.
1349 // Didn't really want to provide this, but it's easier to provide it than to keep it
1350 // out of the tests.
1351 //
1352 UText *
cloneFragmentedUnicodeString(UText * dest,const UText * src,UBool deep,UErrorCode * status)1353 cloneFragmentedUnicodeString(UText *dest, const UText *src, UBool deep, UErrorCode *status) {
1354 if (U_FAILURE(*status)) {
1355 return NULL;
1356 }
1357 if (deep) {
1358 *status = U_UNSUPPORTED_ERROR;
1359 return NULL;
1360 }
1361 dest = utext_openUnicodeString(dest, (UnicodeString *)src->context, status);
1362 utext_setNativeIndex(dest, utext_getNativeIndex(src));
1363 return dest;
1364 }
1365
1366 U_CDECL_END
1367
1368 // Open function for the fragmented text provider.
1369 UText *
openFragmentedUnicodeString(UText * ut,UnicodeString * s,UErrorCode * status)1370 openFragmentedUnicodeString(UText *ut, UnicodeString *s, UErrorCode *status) {
1371 ut = utext_openUnicodeString(ut, s, status);
1372 if (U_FAILURE(*status)) {
1373 return ut;
1374 }
1375
1376 // Copy of the function table from the stock UnicodeString UText,
1377 // and replace the entry for the access function.
1378 memcpy(&fragmentFuncs, ut->pFuncs, sizeof(fragmentFuncs));
1379 fragmentFuncs.access = fragTextAccess;
1380 fragmentFuncs.clone = cloneFragmentedUnicodeString;
1381 ut->pFuncs = &fragmentFuncs;
1382
1383 ut->chunkContents = (UChar *)&ut->b;
1384 ut->pFuncs->access(ut, 0, TRUE);
1385 return ut;
1386 }
1387
1388 // Regression test for Ticket 5560
1389 // Clone fails to update chunkContentPointer in the cloned copy.
1390 // This is only an issue for UText types that work in a local buffer,
1391 // (UTF-8 wrapper, for example)
1392 //
1393 // The test:
1394 // 1. Create an inital UText
1395 // 2. Deep clone it. Contents should match original.
1396 // 3. Reset original to something different.
1397 // 4. Check that clone contents did not change.
1398 //
Ticket5560()1399 void UTextTest::Ticket5560() {
1400 /* The following two strings are in UTF-8 even on EBCDIC platforms. */
1401 static const char s1[] = {0x41,0x42,0x43,0x44,0x45,0x46,0}; /* "ABCDEF" */
1402 static const char s2[] = {0x31,0x32,0x33,0x34,0x35,0x36,0}; /* "123456" */
1403 UErrorCode status = U_ZERO_ERROR;
1404
1405 UText ut1 = UTEXT_INITIALIZER;
1406 UText ut2 = UTEXT_INITIALIZER;
1407
1408 utext_openUTF8(&ut1, s1, -1, &status);
1409 UChar c = utext_next32(&ut1);
1410 TEST_ASSERT(c == 0x41); // c == 'A'
1411
1412 utext_clone(&ut2, &ut1, TRUE, FALSE, &status);
1413 TEST_SUCCESS(status);
1414 c = utext_next32(&ut2);
1415 TEST_ASSERT(c == 0x42); // c == 'B'
1416 c = utext_next32(&ut1);
1417 TEST_ASSERT(c == 0x42); // c == 'B'
1418
1419 utext_openUTF8(&ut1, s2, -1, &status);
1420 c = utext_next32(&ut1);
1421 TEST_ASSERT(c == 0x31); // c == '1'
1422 c = utext_next32(&ut2);
1423 TEST_ASSERT(c == 0x43); // c == 'C'
1424
1425 utext_close(&ut1);
1426 utext_close(&ut2);
1427 }
1428
1429
1430 // Test for Ticket 6847
1431 //
Ticket6847()1432 void UTextTest::Ticket6847() {
1433 const int STRLEN = 90;
1434 UChar s[STRLEN+1];
1435 u_memset(s, 0x41, STRLEN);
1436 s[STRLEN] = 0;
1437
1438 UErrorCode status = U_ZERO_ERROR;
1439 UText *ut = utext_openUChars(NULL, s, -1, &status);
1440
1441 utext_setNativeIndex(ut, 0);
1442 int32_t count = 0;
1443 UChar32 c = 0;
1444 int64_t nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1445 TEST_ASSERT(nativeIndex == 0);
1446 while ((c = utext_next32(ut)) != U_SENTINEL) {
1447 TEST_ASSERT(c == 0x41);
1448 TEST_ASSERT(count < STRLEN);
1449 if (count >= STRLEN) {
1450 break;
1451 }
1452 count++;
1453 nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1454 TEST_ASSERT(nativeIndex == count);
1455 }
1456 TEST_ASSERT(count == STRLEN);
1457 nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1458 TEST_ASSERT(nativeIndex == STRLEN);
1459 utext_close(ut);
1460 }
1461
1462
Ticket10562()1463 void UTextTest::Ticket10562() {
1464 // Note: failures show as a heap error when the test is run under valgrind.
1465 UErrorCode status = U_ZERO_ERROR;
1466
1467 const char *utf8_string = "\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41";
1468 UText *utf8Text = utext_openUTF8(NULL, utf8_string, -1, &status);
1469 TEST_SUCCESS(status);
1470 UText *deepClone = utext_clone(NULL, utf8Text, TRUE, FALSE, &status);
1471 TEST_SUCCESS(status);
1472 UText *shallowClone = utext_clone(NULL, deepClone, FALSE, FALSE, &status);
1473 TEST_SUCCESS(status);
1474 utext_close(shallowClone);
1475 utext_close(deepClone);
1476 utext_close(utf8Text);
1477
1478 status = U_ZERO_ERROR;
1479 UnicodeString usString("Hello, World.");
1480 UText *usText = utext_openUnicodeString(NULL, &usString, &status);
1481 TEST_SUCCESS(status);
1482 UText *usDeepClone = utext_clone(NULL, usText, TRUE, FALSE, &status);
1483 TEST_SUCCESS(status);
1484 UText *usShallowClone = utext_clone(NULL, usDeepClone, FALSE, FALSE, &status);
1485 TEST_SUCCESS(status);
1486 utext_close(usShallowClone);
1487 utext_close(usDeepClone);
1488 utext_close(usText);
1489 }
1490
1491
Ticket10983()1492 void UTextTest::Ticket10983() {
1493 // Note: failure shows as a seg fault when the defect is present.
1494
1495 UErrorCode status = U_ZERO_ERROR;
1496 UnicodeString s("Hello, World");
1497 UText *ut = utext_openConstUnicodeString(NULL, &s, &status);
1498 TEST_SUCCESS(status);
1499
1500 status = U_INVALID_STATE_ERROR;
1501 UText *cloned = utext_clone(NULL, ut, TRUE, TRUE, &status);
1502 TEST_ASSERT(cloned == NULL);
1503 TEST_ASSERT(status == U_INVALID_STATE_ERROR);
1504
1505 utext_close(ut);
1506 }
1507
1508 // Ticket 12130 - extract on a UText wrapping a null terminated UChar * string
1509 // leaves the iteration position set incorrectly when the
1510 // actual string length is not yet known.
1511 //
1512 // The test text needs to be long enough that UText defers getting the length.
1513
Ticket12130()1514 void UTextTest::Ticket12130() {
1515 UErrorCode status = U_ZERO_ERROR;
1516
1517 const char *text8 =
1518 "Fundamentally, computers just deal with numbers. They store letters and other characters "
1519 "by assigning a number for each one. Before Unicode was invented, there were hundreds "
1520 "of different encoding systems for assigning these numbers. No single encoding could "
1521 "contain enough characters: for example, the European Union alone requires several "
1522 "different encodings to cover all its languages. Even for a single language like "
1523 "English no single encoding was adequate for all the letters, punctuation, and technical "
1524 "symbols in common use.";
1525
1526 UnicodeString str(text8);
1527 const UChar *ustr = str.getTerminatedBuffer();
1528 UText ut = UTEXT_INITIALIZER;
1529 utext_openUChars(&ut, ustr, -1, &status);
1530 UChar extractBuffer[50];
1531
1532 for (int32_t startIdx = 0; startIdx<str.length(); ++startIdx) {
1533 int32_t endIdx = startIdx + 20;
1534
1535 u_memset(extractBuffer, 0, UPRV_LENGTHOF(extractBuffer));
1536 utext_extract(&ut, startIdx, endIdx, extractBuffer, UPRV_LENGTHOF(extractBuffer), &status);
1537 if (U_FAILURE(status)) {
1538 errln("%s:%d %s", __FILE__, __LINE__, u_errorName(status));
1539 return;
1540 }
1541 int64_t ni = utext_getNativeIndex(&ut);
1542 int64_t expectedni = startIdx + 20;
1543 if (expectedni > str.length()) {
1544 expectedni = str.length();
1545 }
1546 if (expectedni != ni) {
1547 errln("%s:%d utext_getNativeIndex() expected %d, got %d", __FILE__, __LINE__, expectedni, ni);
1548 }
1549 if (0 != str.tempSubString(startIdx, 20).compare(extractBuffer)) {
1550 errln("%s:%d utext_extract() failed. expected \"%s\", got \"%s\"",
1551 __FILE__, __LINE__, CStr(str.tempSubString(startIdx, 20))(), CStr(UnicodeString(extractBuffer))());
1552 }
1553 }
1554 utext_close(&ut);
1555
1556 // Similar utext extract, this time with the string length provided to the UText in advance,
1557 // and a buffer of larger than required capacity.
1558
1559 utext_openUChars(&ut, ustr, str.length(), &status);
1560 for (int32_t startIdx = 0; startIdx<str.length(); ++startIdx) {
1561 int32_t endIdx = startIdx + 20;
1562 u_memset(extractBuffer, 0, UPRV_LENGTHOF(extractBuffer));
1563 utext_extract(&ut, startIdx, endIdx, extractBuffer, UPRV_LENGTHOF(extractBuffer), &status);
1564 if (U_FAILURE(status)) {
1565 errln("%s:%d %s", __FILE__, __LINE__, u_errorName(status));
1566 return;
1567 }
1568 int64_t ni = utext_getNativeIndex(&ut);
1569 int64_t expectedni = startIdx + 20;
1570 if (expectedni > str.length()) {
1571 expectedni = str.length();
1572 }
1573 if (expectedni != ni) {
1574 errln("%s:%d utext_getNativeIndex() expected %d, got %d", __FILE__, __LINE__, expectedni, ni);
1575 }
1576 if (0 != str.tempSubString(startIdx, 20).compare(extractBuffer)) {
1577 errln("%s:%d utext_extract() failed. expected \"%s\", got \"%s\"",
1578 __FILE__, __LINE__, CStr(str.tempSubString(startIdx, 20))(), CStr(UnicodeString(extractBuffer))());
1579 }
1580 }
1581 utext_close(&ut);
1582 }
1583
1584 // Ticket 13344 The macro form of UTEXT_SETNATIVEINDEX failed when target was a trail surrogate
1585 // of a supplementary character.
1586
Ticket13344()1587 void UTextTest::Ticket13344() {
1588 UErrorCode status = U_ZERO_ERROR;
1589 const char16_t *str = u"abc\U0010abcd xyz";
1590 LocalUTextPointer ut(utext_openUChars(NULL, str, -1, &status));
1591
1592 assertSuccess("UTextTest::Ticket13344-status", status);
1593 UTEXT_SETNATIVEINDEX(ut.getAlias(), 3);
1594 assertEquals("UTextTest::Ticket13344-lead", (int64_t)3, utext_getNativeIndex(ut.getAlias()));
1595 UTEXT_SETNATIVEINDEX(ut.getAlias(), 4);
1596 assertEquals("UTextTest::Ticket13344-trail", (int64_t)3, utext_getNativeIndex(ut.getAlias()));
1597 UTEXT_SETNATIVEINDEX(ut.getAlias(), 5);
1598 assertEquals("UTextTest::Ticket13344-bmp", (int64_t)5, utext_getNativeIndex(ut.getAlias()));
1599
1600 utext_setNativeIndex(ut.getAlias(), 3);
1601 assertEquals("UTextTest::Ticket13344-lead-2", (int64_t)3, utext_getNativeIndex(ut.getAlias()));
1602 utext_setNativeIndex(ut.getAlias(), 4);
1603 assertEquals("UTextTest::Ticket13344-trail-2", (int64_t)3, utext_getNativeIndex(ut.getAlias()));
1604 utext_setNativeIndex(ut.getAlias(), 5);
1605 assertEquals("UTextTest::Ticket13344-bmp-2", (int64_t)5, utext_getNativeIndex(ut.getAlias()));
1606 }
1607
1608