1 /********************************************************************
2  * COPYRIGHT:
3  * Copyright (c) 2005-2014, 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 
UTextTest()39 UTextTest::UTextTest() {
40 }
41 
~UTextTest()42 UTextTest::~UTextTest() {
43 }
44 
45 
46 void
runIndexedTest(int32_t index,UBool exec,const char * & name,char *)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         case 6: name = "Ticket10983";
63             if (exec) Ticket10983();  break;
64         default: name = "";          break;
65     }
66 }
67 
68 //
69 // Quick and dirty random number generator.
70 //   (don't use library so that results are portable.
71 static uint32_t m_seed = 1;
m_rand()72 static uint32_t m_rand()
73 {
74     m_seed = m_seed * 1103515245 + 12345;
75     return (uint32_t)(m_seed/65536) % 32768;
76 }
77 
78 
79 //
80 //   TextTest()
81 //
82 //       Top Level function for UText testing.
83 //       Specifies the strings to be tested, with the acutal testing itself
84 //       being carried out in another function, TestString().
85 //
TextTest()86 void  UTextTest::TextTest() {
87     int32_t i, j;
88 
89     TestString("abcd\\U00010001xyz");
90     TestString("");
91 
92     // Supplementary chars at start or end
93     TestString("\\U00010001");
94     TestString("abc\\U00010001");
95     TestString("\\U00010001abc");
96 
97     // Test simple strings of lengths 1 to 60, looking for glitches at buffer boundaries
98     UnicodeString s;
99     for (i=1; i<60; i++) {
100         s.truncate(0);
101         for (j=0; j<i; j++) {
102             if (j+0x30 == 0x5c) {
103                 // backslash.  Needs to be escaped
104                 s.append((UChar)0x5c);
105             }
106             s.append(UChar(j+0x30));
107         }
108         TestString(s);
109     }
110 
111    // Test strings with odd-aligned supplementary chars,
112    //    looking for glitches at buffer boundaries
113     for (i=1; i<60; i++) {
114         s.truncate(0);
115         s.append((UChar)0x41);
116         for (j=0; j<i; j++) {
117             s.append(UChar32(j+0x11000));
118         }
119         TestString(s);
120     }
121 
122     // String of chars of randomly varying size in utf-8 representation.
123     //   Exercise the mapping, and the varying sized buffer.
124     //
125     s.truncate(0);
126     UChar32  c1 = 0;
127     UChar32  c2 = 0x100;
128     UChar32  c3 = 0xa000;
129     UChar32  c4 = 0x11000;
130     for (i=0; i<1000; i++) {
131         int len8 = m_rand()%4 + 1;
132         switch (len8) {
133             case 1:
134                 c1 = (c1+1)%0x80;
135                 // don't put 0 into string (0 terminated strings for some tests)
136                 // don't put '\', will cause unescape() to fail.
137                 if (c1==0x5c || c1==0) {
138                     c1++;
139                 }
140                 s.append(c1);
141                 break;
142             case 2:
143                 s.append(c2++);
144                 break;
145             case 3:
146                 s.append(c3++);
147                 break;
148             case 4:
149                 s.append(c4++);
150                 break;
151         }
152     }
153     TestString(s);
154 }
155 
156 
157 //
158 //  TestString()     Run a suite of UText tests on a string.
159 //                   The test string is unescaped before use.
160 //
TestString(const UnicodeString & s)161 void UTextTest::TestString(const UnicodeString &s) {
162     int32_t       i;
163     int32_t       j;
164     UChar32       c;
165     int32_t       cpCount = 0;
166     UErrorCode    status  = U_ZERO_ERROR;
167     UText        *ut      = NULL;
168     int32_t       saLen;
169 
170     UnicodeString sa = s.unescape();
171     saLen = sa.length();
172 
173     //
174     // Build up a mapping between code points and UTF-16 code unit indexes.
175     //
176     m *cpMap = new m[sa.length() + 1];
177     j = 0;
178     for (i=0; i<sa.length(); i=sa.moveIndex32(i, 1)) {
179         c = sa.char32At(i);
180         cpMap[j].nativeIdx = i;
181         cpMap[j].cp = c;
182         j++;
183         cpCount++;
184     }
185     cpMap[j].nativeIdx = i;   // position following the last char in utf-16 string.
186 
187 
188     // UChar * test, null terminated
189     status = U_ZERO_ERROR;
190     UChar *buf = new UChar[saLen+1];
191     sa.extract(buf, saLen+1, status);
192     TEST_SUCCESS(status);
193     ut = utext_openUChars(NULL, buf, -1, &status);
194     TEST_SUCCESS(status);
195     TestAccess(sa, ut, cpCount, cpMap);
196     utext_close(ut);
197     delete [] buf;
198 
199     // UChar * test, with length
200     status = U_ZERO_ERROR;
201     buf = new UChar[saLen+1];
202     sa.extract(buf, saLen+1, status);
203     TEST_SUCCESS(status);
204     ut = utext_openUChars(NULL, buf, saLen, &status);
205     TEST_SUCCESS(status);
206     TestAccess(sa, ut, cpCount, cpMap);
207     utext_close(ut);
208     delete [] buf;
209 
210 
211     // UnicodeString test
212     status = U_ZERO_ERROR;
213     ut = utext_openUnicodeString(NULL, &sa, &status);
214     TEST_SUCCESS(status);
215     TestAccess(sa, ut, cpCount, cpMap);
216     TestCMR(sa, ut, cpCount, cpMap, cpMap);
217     utext_close(ut);
218 
219 
220     // Const UnicodeString test
221     status = U_ZERO_ERROR;
222     ut = utext_openConstUnicodeString(NULL, &sa, &status);
223     TEST_SUCCESS(status);
224     TestAccess(sa, ut, cpCount, cpMap);
225     utext_close(ut);
226 
227 
228     // Replaceable test.  (UnicodeString inherits Replaceable)
229     status = U_ZERO_ERROR;
230     ut = utext_openReplaceable(NULL, &sa, &status);
231     TEST_SUCCESS(status);
232     TestAccess(sa, ut, cpCount, cpMap);
233     TestCMR(sa, ut, cpCount, cpMap, cpMap);
234     utext_close(ut);
235 
236     // Character Iterator Tests
237     status = U_ZERO_ERROR;
238     const UChar *cbuf = sa.getBuffer();
239     CharacterIterator *ci = new UCharCharacterIterator(cbuf, saLen, status);
240     TEST_SUCCESS(status);
241     ut = utext_openCharacterIterator(NULL, ci, &status);
242     TEST_SUCCESS(status);
243     TestAccess(sa, ut, cpCount, cpMap);
244     utext_close(ut);
245     delete ci;
246 
247 
248     // Fragmented UnicodeString  (Chunk size of one)
249     //
250     status = U_ZERO_ERROR;
251     ut = openFragmentedUnicodeString(NULL, &sa, &status);
252     TEST_SUCCESS(status);
253     TestAccess(sa, ut, cpCount, cpMap);
254     utext_close(ut);
255 
256     //
257     // UTF-8 test
258     //
259 
260     // Convert the test string from UnicodeString to (char *) in utf-8 format
261     int32_t u8Len = sa.extract(0, sa.length(), NULL, 0, "utf-8");
262     char *u8String = new char[u8Len + 1];
263     sa.extract(0, sa.length(), u8String, u8Len+1, "utf-8");
264 
265     // Build up the map of code point indices in the utf-8 string
266     m * u8Map = new m[sa.length() + 1];
267     i = 0;   // native utf-8 index
268     for (j=0; j<cpCount ; j++) {  // code point number
269         u8Map[j].nativeIdx = i;
270         U8_NEXT(u8String, i, u8Len, c)
271         u8Map[j].cp = c;
272     }
273     u8Map[cpCount].nativeIdx = u8Len;   // position following the last char in utf-8 string.
274 
275     // Do the test itself
276     status = U_ZERO_ERROR;
277     ut = utext_openUTF8(NULL, u8String, -1, &status);
278     TEST_SUCCESS(status);
279     TestAccess(sa, ut, cpCount, u8Map);
280     utext_close(ut);
281 
282 
283 
284     delete []cpMap;
285     delete []u8Map;
286     delete []u8String;
287 }
288 
289 //  TestCMR   test Copy, Move and Replace operations.
290 //              us         UnicodeString containing the test text.
291 //              ut         UText containing the same test text.
292 //              cpCount    number of code points in the test text.
293 //              nativeMap  Mapping from code points to native indexes for the UText.
294 //              u16Map     Mapping from code points to UTF-16 indexes, for use with the UnicodeString.
295 //
296 //     This function runs a whole series of opertions on each incoming UText.
297 //     The UText is deep-cloned prior to each operation, so that the original UText remains unchanged.
298 //
TestCMR(const UnicodeString & us,UText * ut,int cpCount,m * nativeMap,m * u16Map)299 void UTextTest::TestCMR(const UnicodeString &us, UText *ut, int cpCount, m *nativeMap, m *u16Map) {
300     TEST_ASSERT(utext_isWritable(ut) == TRUE);
301 
302     int  srcLengthType;       // Loop variables for selecting the postion and length
303     int  srcPosType;          //   of the block to operate on within the source text.
304     int  destPosType;
305 
306     int  srcIndex  = 0;       // Code Point indexes of the block to operate on for
307     int  srcLength = 0;       //   a specific test.
308 
309     int  destIndex = 0;       // Code point index of the destination for a copy/move test.
310 
311     int32_t  nativeStart = 0; // Native unit indexes for a test.
312     int32_t  nativeLimit = 0;
313     int32_t  nativeDest  = 0;
314 
315     int32_t  u16Start    = 0; // UTF-16 indexes for a test.
316     int32_t  u16Limit    = 0; //   used when performing the same operation in a Unicode String
317     int32_t  u16Dest     = 0;
318 
319     // Iterate over a whole series of source index, length and a target indexes.
320     // This is done with code point indexes; these will be later translated to native
321     //   indexes using the cpMap.
322     for (srcLengthType=1; srcLengthType<=3; srcLengthType++) {
323         switch (srcLengthType) {
324             case 1: srcLength = 1; break;
325             case 2: srcLength = 5; break;
326             case 3: srcLength = cpCount / 3;
327         }
328         for (srcPosType=1; srcPosType<=5; srcPosType++) {
329             switch (srcPosType) {
330                 case 1: srcIndex = 0; break;
331                 case 2: srcIndex = 1; break;
332                 case 3: srcIndex = cpCount - srcLength; break;
333                 case 4: srcIndex = cpCount - srcLength - 1; break;
334                 case 5: srcIndex = cpCount / 2; break;
335             }
336             if (srcIndex < 0 || srcIndex + srcLength > cpCount) {
337                 // filter out bogus test cases -
338                 //   those with a source range that falls of an edge of the string.
339                 continue;
340             }
341 
342             //
343             // Copy and move tests.
344             //   iterate over a variety of destination positions.
345             //
346             for (destPosType=1; destPosType<=4; destPosType++) {
347                 switch (destPosType) {
348                     case 1: destIndex = 0; break;
349                     case 2: destIndex = 1; break;
350                     case 3: destIndex = srcIndex - 1; break;
351                     case 4: destIndex = srcIndex + srcLength + 1; break;
352                     case 5: destIndex = cpCount-1; break;
353                     case 6: destIndex = cpCount; break;
354                 }
355                 if (destIndex<0 || destIndex>cpCount) {
356                     // filter out bogus test cases.
357                     continue;
358                 }
359 
360                 nativeStart = nativeMap[srcIndex].nativeIdx;
361                 nativeLimit = nativeMap[srcIndex+srcLength].nativeIdx;
362                 nativeDest  = nativeMap[destIndex].nativeIdx;
363 
364                 u16Start    = u16Map[srcIndex].nativeIdx;
365                 u16Limit    = u16Map[srcIndex+srcLength].nativeIdx;
366                 u16Dest     = u16Map[destIndex].nativeIdx;
367 
368                 gFailed = FALSE;
369                 TestCopyMove(us, ut, FALSE,
370                     nativeStart, nativeLimit, nativeDest,
371                     u16Start, u16Limit, u16Dest);
372 
373                 TestCopyMove(us, ut, TRUE,
374                     nativeStart, nativeLimit, nativeDest,
375                     u16Start, u16Limit, u16Dest);
376 
377                 if (gFailed) {
378                     return;
379                 }
380             }
381 
382             //
383             //  Replace tests.
384             //
385             UnicodeString fullRepString("This is an arbitrary string that will be used as replacement text");
386             for (int32_t replStrLen=0; replStrLen<20; replStrLen++) {
387                 UnicodeString repStr(fullRepString, 0, replStrLen);
388                 TestReplace(us, ut,
389                     nativeStart, nativeLimit,
390                     u16Start, u16Limit,
391                     repStr);
392                 if (gFailed) {
393                     return;
394                 }
395             }
396 
397         }
398     }
399 
400 }
401 
402 //
403 //   TestCopyMove    run a single test case for utext_copy.
404 //                   Test cases are created in TestCMR and dispatched here for execution.
405 //
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)406 void UTextTest::TestCopyMove(const UnicodeString &us, UText *ut, UBool move,
407                     int32_t nativeStart, int32_t nativeLimit, int32_t nativeDest,
408                     int32_t u16Start, int32_t u16Limit, int32_t u16Dest)
409 {
410     UErrorCode      status   = U_ZERO_ERROR;
411     UText          *targetUT = NULL;
412     gTestNum++;
413     gFailed = FALSE;
414 
415     //
416     //  clone the UText.  The test will be run in the cloned copy
417     //  so that we don't alter the original.
418     //
419     targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status);
420     TEST_SUCCESS(status);
421     UnicodeString targetUS(us);    // And copy the reference string.
422 
423     // do the test operation first in the reference
424     targetUS.copy(u16Start, u16Limit, u16Dest);
425     if (move) {
426         // delete out the source range.
427         if (u16Limit < u16Dest) {
428             targetUS.removeBetween(u16Start, u16Limit);
429         } else {
430             int32_t amtCopied = u16Limit - u16Start;
431             targetUS.removeBetween(u16Start+amtCopied, u16Limit+amtCopied);
432         }
433     }
434 
435     // Do the same operation in the UText under test
436     utext_copy(targetUT, nativeStart, nativeLimit, nativeDest, move, &status);
437     if (nativeDest > nativeStart && nativeDest < nativeLimit) {
438         TEST_ASSERT(status == U_INDEX_OUTOFBOUNDS_ERROR);
439     } else {
440         TEST_SUCCESS(status);
441 
442         // Compare the results of the two parallel tests
443         int32_t  usi = 0;    // UnicodeString postion, utf-16 index.
444         int64_t  uti = 0;    // UText position, native index.
445         int32_t  cpi;        // char32 position (code point index)
446         UChar32  usc;        // code point from Unicode String
447         UChar32  utc;        // code point from UText
448         utext_setNativeIndex(targetUT, 0);
449         for (cpi=0; ; cpi++) {
450             usc = targetUS.char32At(usi);
451             utc = utext_next32(targetUT);
452             if (utc < 0) {
453                 break;
454             }
455             TEST_ASSERT(uti == usi);
456             TEST_ASSERT(utc == usc);
457             usi = targetUS.moveIndex32(usi, 1);
458             uti = utext_getNativeIndex(targetUT);
459             if (gFailed) {
460                 goto cleanupAndReturn;
461             }
462         }
463         int64_t expectedNativeLength = utext_nativeLength(ut);
464         if (move == FALSE) {
465             expectedNativeLength += nativeLimit - nativeStart;
466         }
467         uti = utext_getNativeIndex(targetUT);
468         TEST_ASSERT(uti == expectedNativeLength);
469     }
470 
471 cleanupAndReturn:
472     utext_close(targetUT);
473 }
474 
475 
476 //
477 //  TestReplace   Test a single Replace operation.
478 //
TestReplace(const UnicodeString & us,UText * ut,int32_t nativeStart,int32_t nativeLimit,int32_t u16Start,int32_t u16Limit,const UnicodeString & repStr)479 void UTextTest::TestReplace(
480             const UnicodeString &us,     // reference UnicodeString in which to do the replace
481             UText         *ut,                // UnicodeText object under test.
482             int32_t       nativeStart,        // Range to be replaced, in UText native units.
483             int32_t       nativeLimit,
484             int32_t       u16Start,           // Range to be replaced, in UTF-16 units
485             int32_t       u16Limit,           //    for use in the reference UnicodeString.
486             const UnicodeString &repStr)      // The replacement string
487 {
488     UErrorCode      status   = U_ZERO_ERROR;
489     UText          *targetUT = NULL;
490     gTestNum++;
491     gFailed = FALSE;
492 
493     //
494     //  clone the target UText.  The test will be run in the cloned copy
495     //  so that we don't alter the original.
496     //
497     targetUT = utext_clone(NULL, ut, TRUE, FALSE, &status);
498     TEST_SUCCESS(status);
499     UnicodeString targetUS(us);    // And copy the reference string.
500 
501     //
502     // Do the replace operation in the Unicode String, to
503     //   produce a reference result.
504     //
505     targetUS.replace(u16Start, u16Limit-u16Start, repStr);
506 
507     //
508     // Do the replace on the UText under test
509     //
510     const UChar *rs = repStr.getBuffer();
511     int32_t  rsLen = repStr.length();
512     int32_t actualDelta = utext_replace(targetUT, nativeStart, nativeLimit, rs, rsLen, &status);
513     int32_t expectedDelta = repStr.length() - (nativeLimit - nativeStart);
514     TEST_ASSERT(actualDelta == expectedDelta);
515 
516     //
517     // Compare the results
518     //
519     int32_t  usi = 0;    // UnicodeString postion, utf-16 index.
520     int64_t  uti = 0;    // UText position, native index.
521     int32_t  cpi;        // char32 position (code point index)
522     UChar32  usc;        // code point from Unicode String
523     UChar32  utc;        // code point from UText
524     int64_t  expectedNativeLength = 0;
525     utext_setNativeIndex(targetUT, 0);
526     for (cpi=0; ; cpi++) {
527         usc = targetUS.char32At(usi);
528         utc = utext_next32(targetUT);
529         if (utc < 0) {
530             break;
531         }
532         TEST_ASSERT(uti == usi);
533         TEST_ASSERT(utc == usc);
534         usi = targetUS.moveIndex32(usi, 1);
535         uti = utext_getNativeIndex(targetUT);
536         if (gFailed) {
537             goto cleanupAndReturn;
538         }
539     }
540     expectedNativeLength = utext_nativeLength(ut) + expectedDelta;
541     uti = utext_getNativeIndex(targetUT);
542     TEST_ASSERT(uti == expectedNativeLength);
543 
544 cleanupAndReturn:
545     utext_close(targetUT);
546 }
547 
548 //
549 //  TestAccess      Test the read only access functions on a UText, including cloning.
550 //                  The text is accessed in a variety of ways, and compared with
551 //                  the reference UnicodeString.
552 //
TestAccess(const UnicodeString & us,UText * ut,int cpCount,m * cpMap)553 void UTextTest::TestAccess(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) {
554     // Run the standard tests on the caller-supplied UText.
555     TestAccessNoClone(us, ut, cpCount, cpMap);
556 
557     // Re-run tests on a shallow clone.
558     utext_setNativeIndex(ut, 0);
559     UErrorCode status = U_ZERO_ERROR;
560     UText *shallowClone = utext_clone(NULL, ut, FALSE /*deep*/, FALSE /*readOnly*/, &status);
561     TEST_SUCCESS(status);
562     TestAccessNoClone(us, shallowClone, cpCount, cpMap);
563 
564     //
565     // Rerun again on a deep clone.
566     // Note that text providers are not required to provide deep cloning,
567     //   so unsupported errors are ignored.
568     //
569     status = U_ZERO_ERROR;
570     utext_setNativeIndex(shallowClone, 0);
571     UText *deepClone = utext_clone(NULL, shallowClone, TRUE, FALSE, &status);
572     utext_close(shallowClone);
573     if (status != U_UNSUPPORTED_ERROR) {
574         TEST_SUCCESS(status);
575         TestAccessNoClone(us, deepClone, cpCount, cpMap);
576     }
577     utext_close(deepClone);
578 }
579 
580 
581 //
582 //  TestAccessNoClone()    Test the read only access functions on a UText.
583 //                         The text is accessed in a variety of ways, and compared with
584 //                         the reference UnicodeString.
585 //
TestAccessNoClone(const UnicodeString & us,UText * ut,int cpCount,m * cpMap)586 void UTextTest::TestAccessNoClone(const UnicodeString &us, UText *ut, int cpCount, m *cpMap) {
587     UErrorCode  status = U_ZERO_ERROR;
588     gTestNum++;
589 
590     //
591     //  Check the length from the UText
592     //
593     int64_t expectedLen = cpMap[cpCount].nativeIdx;
594     int64_t utlen = utext_nativeLength(ut);
595     TEST_ASSERT(expectedLen == utlen);
596 
597     //
598     //  Iterate forwards, verify that we get the correct code points
599     //   at the correct native offsets.
600     //
601     int         i = 0;
602     int64_t     index;
603     int64_t     expectedIndex = 0;
604     int64_t     foundIndex = 0;
605     UChar32     expectedC;
606     UChar32     foundC;
607     int64_t     len;
608 
609     for (i=0; i<cpCount; i++) {
610         expectedIndex = cpMap[i].nativeIdx;
611         foundIndex    = utext_getNativeIndex(ut);
612         TEST_ASSERT(expectedIndex == foundIndex);
613         expectedC     = cpMap[i].cp;
614         foundC        = utext_next32(ut);
615         TEST_ASSERT(expectedC == foundC);
616         foundIndex    = utext_getPreviousNativeIndex(ut);
617         TEST_ASSERT(expectedIndex == foundIndex);
618         if (gFailed) {
619             return;
620         }
621     }
622     foundC = utext_next32(ut);
623     TEST_ASSERT(foundC == U_SENTINEL);
624 
625     // Repeat above, using macros
626     utext_setNativeIndex(ut, 0);
627     for (i=0; i<cpCount; i++) {
628         expectedIndex = cpMap[i].nativeIdx;
629         foundIndex    = UTEXT_GETNATIVEINDEX(ut);
630         TEST_ASSERT(expectedIndex == foundIndex);
631         expectedC     = cpMap[i].cp;
632         foundC        = UTEXT_NEXT32(ut);
633         TEST_ASSERT(expectedC == foundC);
634         if (gFailed) {
635             return;
636         }
637     }
638     foundC = UTEXT_NEXT32(ut);
639     TEST_ASSERT(foundC == U_SENTINEL);
640 
641     //
642     //  Forward iteration (above) should have left index at the
643     //   end of the input, which should == length().
644     //
645     len = utext_nativeLength(ut);
646     foundIndex  = utext_getNativeIndex(ut);
647     TEST_ASSERT(len == foundIndex);
648 
649     //
650     // Iterate backwards over entire test string
651     //
652     len = utext_getNativeIndex(ut);
653     utext_setNativeIndex(ut, len);
654     for (i=cpCount-1; i>=0; i--) {
655         expectedC     = cpMap[i].cp;
656         expectedIndex = cpMap[i].nativeIdx;
657         int64_t prevIndex = utext_getPreviousNativeIndex(ut);
658         foundC        = utext_previous32(ut);
659         foundIndex    = utext_getNativeIndex(ut);
660         TEST_ASSERT(expectedIndex == foundIndex);
661         TEST_ASSERT(expectedC == foundC);
662         TEST_ASSERT(prevIndex == foundIndex);
663         if (gFailed) {
664             return;
665         }
666     }
667 
668     //
669     //  Backwards iteration, above, should have left our iterator
670     //   position at zero, and continued backwards iterationshould fail.
671     //
672     foundIndex = utext_getNativeIndex(ut);
673     TEST_ASSERT(foundIndex == 0);
674     foundIndex = utext_getPreviousNativeIndex(ut);
675     TEST_ASSERT(foundIndex == 0);
676 
677 
678     foundC = utext_previous32(ut);
679     TEST_ASSERT(foundC == U_SENTINEL);
680     foundIndex = utext_getNativeIndex(ut);
681     TEST_ASSERT(foundIndex == 0);
682     foundIndex = utext_getPreviousNativeIndex(ut);
683     TEST_ASSERT(foundIndex == 0);
684 
685 
686     // And again, with the macros
687     utext_setNativeIndex(ut, len);
688     for (i=cpCount-1; i>=0; i--) {
689         expectedC     = cpMap[i].cp;
690         expectedIndex = cpMap[i].nativeIdx;
691         foundC        = UTEXT_PREVIOUS32(ut);
692         foundIndex    = UTEXT_GETNATIVEINDEX(ut);
693         TEST_ASSERT(expectedIndex == foundIndex);
694         TEST_ASSERT(expectedC == foundC);
695         if (gFailed) {
696             return;
697         }
698     }
699 
700     //
701     //  Backwards iteration, above, should have left our iterator
702     //   position at zero, and continued backwards iterationshould fail.
703     //
704     foundIndex = UTEXT_GETNATIVEINDEX(ut);
705     TEST_ASSERT(foundIndex == 0);
706 
707     foundC = UTEXT_PREVIOUS32(ut);
708     TEST_ASSERT(foundC == U_SENTINEL);
709     foundIndex = UTEXT_GETNATIVEINDEX(ut);
710     TEST_ASSERT(foundIndex == 0);
711     if (gFailed) {
712         return;
713     }
714 
715     //
716     //  next32From(), prevous32From(), Iterate in a somewhat random order.
717     //
718     int  cpIndex = 0;
719     for (i=0; i<cpCount; i++) {
720         cpIndex = (cpIndex + 9973) % cpCount;
721         index         = cpMap[cpIndex].nativeIdx;
722         expectedC     = cpMap[cpIndex].cp;
723         foundC        = utext_next32From(ut, index);
724         TEST_ASSERT(expectedC == foundC);
725         if (gFailed) {
726             return;
727         }
728     }
729 
730     cpIndex = 0;
731     for (i=0; i<cpCount; i++) {
732         cpIndex = (cpIndex + 9973) % cpCount;
733         index         = cpMap[cpIndex+1].nativeIdx;
734         expectedC     = cpMap[cpIndex].cp;
735         foundC        = utext_previous32From(ut, index);
736         TEST_ASSERT(expectedC == foundC);
737         if (gFailed) {
738             return;
739         }
740     }
741 
742 
743     //
744     // moveIndex(int32_t delta);
745     //
746 
747     // Walk through frontwards, incrementing by one
748     utext_setNativeIndex(ut, 0);
749     for (i=1; i<=cpCount; i++) {
750         utext_moveIndex32(ut, 1);
751         index = utext_getNativeIndex(ut);
752         expectedIndex = cpMap[i].nativeIdx;
753         TEST_ASSERT(expectedIndex == index);
754         index = UTEXT_GETNATIVEINDEX(ut);
755         TEST_ASSERT(expectedIndex == index);
756     }
757 
758     // Walk through frontwards, incrementing by two
759     utext_setNativeIndex(ut, 0);
760     for (i=2; i<cpCount; i+=2) {
761         utext_moveIndex32(ut, 2);
762         index = utext_getNativeIndex(ut);
763         expectedIndex = cpMap[i].nativeIdx;
764         TEST_ASSERT(expectedIndex == index);
765         index = UTEXT_GETNATIVEINDEX(ut);
766         TEST_ASSERT(expectedIndex == index);
767     }
768 
769     // walk through the string backwards, decrementing by one.
770     i = cpMap[cpCount].nativeIdx;
771     utext_setNativeIndex(ut, i);
772     for (i=cpCount; i>=0; i--) {
773         expectedIndex = cpMap[i].nativeIdx;
774         index = utext_getNativeIndex(ut);
775         TEST_ASSERT(expectedIndex == index);
776         index = UTEXT_GETNATIVEINDEX(ut);
777         TEST_ASSERT(expectedIndex == index);
778         utext_moveIndex32(ut, -1);
779     }
780 
781 
782     // walk through backwards, decrementing by three
783     i = cpMap[cpCount].nativeIdx;
784     utext_setNativeIndex(ut, i);
785     for (i=cpCount; i>=0; i-=3) {
786         expectedIndex = cpMap[i].nativeIdx;
787         index = utext_getNativeIndex(ut);
788         TEST_ASSERT(expectedIndex == index);
789         index = UTEXT_GETNATIVEINDEX(ut);
790         TEST_ASSERT(expectedIndex == index);
791         utext_moveIndex32(ut, -3);
792     }
793 
794 
795     //
796     // Extract
797     //
798     int bufSize = us.length() + 10;
799     UChar *buf = new UChar[bufSize];
800     status = U_ZERO_ERROR;
801     expectedLen = us.length();
802     len = utext_extract(ut, 0, utlen, buf, bufSize, &status);
803     TEST_SUCCESS(status);
804     TEST_ASSERT(len == expectedLen);
805     int compareResult = us.compare(buf, -1);
806     TEST_ASSERT(compareResult == 0);
807 
808     status = U_ZERO_ERROR;
809     len = utext_extract(ut, 0, utlen, NULL, 0, &status);
810     if (utlen == 0) {
811         TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING);
812     } else {
813         TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR);
814     }
815     TEST_ASSERT(len == expectedLen);
816 
817     status = U_ZERO_ERROR;
818     u_memset(buf, 0x5555, bufSize);
819     len = utext_extract(ut, 0, utlen, buf, 1, &status);
820     if (us.length() == 0) {
821         TEST_SUCCESS(status);
822         TEST_ASSERT(buf[0] == 0);
823     } else {
824         // Buf len == 1, extracting a single 16 bit value.
825         // If the data char is supplementary, it doesn't matter whether the buffer remains unchanged,
826         //   or whether the lead surrogate of the pair is extracted.
827         //   It's a buffer overflow error in either case.
828         TEST_ASSERT(buf[0] == us.charAt(0) ||
829                     (buf[0] == 0x5555 && U_IS_SUPPLEMENTARY(us.char32At(0))));
830         TEST_ASSERT(buf[1] == 0x5555);
831         if (us.length() == 1) {
832             TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING);
833         } else {
834             TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR);
835         }
836     }
837 
838     delete []buf;
839 }
840 
841 //
842 //  ErrorTest()    Check various error and edge cases.
843 //
ErrorTest()844 void UTextTest::ErrorTest()
845 {
846     // Close of an unitialized UText.  Shouldn't blow up.
847     {
848         UText  ut;
849         memset(&ut, 0, sizeof(UText));
850         utext_close(&ut);
851         utext_close(NULL);
852     }
853 
854     // Double-close of a UText.  Shouldn't blow up.  UText should still be usable.
855     {
856         UErrorCode status = U_ZERO_ERROR;
857         UText ut = UTEXT_INITIALIZER;
858         UnicodeString s("Hello, World");
859         UText *ut2 = utext_openUnicodeString(&ut, &s, &status);
860         TEST_SUCCESS(status);
861         TEST_ASSERT(ut2 == &ut);
862 
863         UText *ut3 = utext_close(&ut);
864         TEST_ASSERT(ut3 == &ut);
865 
866         UText *ut4 = utext_close(&ut);
867         TEST_ASSERT(ut4 == &ut);
868 
869         utext_openUnicodeString(&ut, &s, &status);
870         TEST_SUCCESS(status);
871         utext_close(&ut);
872     }
873 
874     // Re-use of a UText, chaining through each of the types of UText
875     //   (If it doesn't blow up, and doesn't leak, it's probably working fine)
876     {
877         UErrorCode status = U_ZERO_ERROR;
878         UText ut = UTEXT_INITIALIZER;
879         UText  *utp;
880         UnicodeString s1("Hello, World");
881         UChar s2[] = {(UChar)0x41, (UChar)0x42, (UChar)0};
882         const char  *s3 = "\x66\x67\x68";
883 
884         utp = utext_openUnicodeString(&ut, &s1, &status);
885         TEST_SUCCESS(status);
886         TEST_ASSERT(utp == &ut);
887 
888         utp = utext_openConstUnicodeString(&ut, &s1, &status);
889         TEST_SUCCESS(status);
890         TEST_ASSERT(utp == &ut);
891 
892         utp = utext_openUTF8(&ut, s3, -1, &status);
893         TEST_SUCCESS(status);
894         TEST_ASSERT(utp == &ut);
895 
896         utp = utext_openUChars(&ut, s2, -1, &status);
897         TEST_SUCCESS(status);
898         TEST_ASSERT(utp == &ut);
899 
900         utp = utext_close(&ut);
901         TEST_ASSERT(utp == &ut);
902 
903         utp = utext_openUnicodeString(&ut, &s1, &status);
904         TEST_SUCCESS(status);
905         TEST_ASSERT(utp == &ut);
906     }
907 
908     // Invalid parameters on open
909     //
910     {
911         UErrorCode status = U_ZERO_ERROR;
912         UText ut = UTEXT_INITIALIZER;
913 
914         utext_openUChars(&ut, NULL, 5, &status);
915         TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
916 
917         status = U_ZERO_ERROR;
918         utext_openUChars(&ut, NULL, -1, &status);
919         TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
920 
921         status = U_ZERO_ERROR;
922         utext_openUTF8(&ut, NULL, 4, &status);
923         TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
924 
925         status = U_ZERO_ERROR;
926         utext_openUTF8(&ut, NULL, -1, &status);
927         TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR);
928     }
929 
930     //
931     //  UTF-8 with malformed sequences.
932     //    These should come through as the Unicode replacement char, \ufffd
933     //
934     {
935         UErrorCode status = U_ZERO_ERROR;
936         UText *ut = NULL;
937         const char *badUTF8 = "\x41\x81\x42\xf0\x81\x81\x43";
938         UChar32  c;
939 
940         ut = utext_openUTF8(NULL, badUTF8, -1, &status);
941         TEST_SUCCESS(status);
942         c = utext_char32At(ut, 1);
943         TEST_ASSERT(c == 0xfffd);
944         c = utext_char32At(ut, 3);
945         TEST_ASSERT(c == 0xfffd);
946         c = utext_char32At(ut, 5);
947         TEST_ASSERT(c == 0xfffd);
948         c = utext_char32At(ut, 6);
949         TEST_ASSERT(c == 0x43);
950 
951         UChar buf[10];
952         int n = utext_extract(ut, 0, 9, buf, 10, &status);
953         TEST_SUCCESS(status);
954         TEST_ASSERT(n==5);
955         TEST_ASSERT(buf[1] == 0xfffd);
956         TEST_ASSERT(buf[3] == 0xfffd);
957         TEST_ASSERT(buf[2] == 0x42);
958         utext_close(ut);
959     }
960 
961 
962     //
963     //  isLengthExpensive - does it make the exptected transitions after
964     //                      getting the length of a nul terminated string?
965     //
966     {
967         UErrorCode status = U_ZERO_ERROR;
968         UnicodeString sa("Hello, this is a string");
969         UBool  isExpensive;
970 
971         UChar sb[100];
972         memset(sb, 0x20, sizeof(sb));
973         sb[99] = 0;
974 
975         UText *uta = utext_openUnicodeString(NULL, &sa, &status);
976         TEST_SUCCESS(status);
977         isExpensive = utext_isLengthExpensive(uta);
978         TEST_ASSERT(isExpensive == FALSE);
979         utext_close(uta);
980 
981         UText *utb = utext_openUChars(NULL, sb, -1, &status);
982         TEST_SUCCESS(status);
983         isExpensive = utext_isLengthExpensive(utb);
984         TEST_ASSERT(isExpensive == TRUE);
985         int64_t  len = utext_nativeLength(utb);
986         TEST_ASSERT(len == 99);
987         isExpensive = utext_isLengthExpensive(utb);
988         TEST_ASSERT(isExpensive == FALSE);
989         utext_close(utb);
990     }
991 
992     //
993     // Index to positions not on code point boundaries.
994     //
995     {
996         const char *u8str =         "\xc8\x81\xe1\x82\x83\xf1\x84\x85\x86";
997         int32_t startMap[] =        {   0,  0,  2,  2,  2,  5,  5,  5,  5,  9,  9};
998         int32_t nextMap[]  =        {   2,  2,  5,  5,  5,  9,  9,  9,  9,  9,  9};
999         int32_t prevMap[]  =        {   0,  0,  0,  0,  0,  2,  2,  2,  2,  5,  5};
1000         UChar32  c32Map[] =    {0x201, 0x201, 0x1083, 0x1083, 0x1083, 0x044146, 0x044146, 0x044146, 0x044146, -1, -1};
1001         UChar32  pr32Map[] =   {    -1,   -1,  0x201,  0x201,  0x201,   0x1083,   0x1083,   0x1083,   0x1083, 0x044146, 0x044146};
1002 
1003         // extractLen is the size, in UChars, of what will be extracted between index and index+1.
1004         //  is zero when both index positions lie within the same code point.
1005         int32_t  exLen[] =          {   0,  1,   0,  0,  1,  0,  0,  0,  2,  0,  0};
1006 
1007 
1008         UErrorCode status = U_ZERO_ERROR;
1009         UText *ut = utext_openUTF8(NULL, u8str, -1, &status);
1010         TEST_SUCCESS(status);
1011 
1012         // Check setIndex
1013         int32_t i;
1014         int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t);
1015         for (i=0; i<startMapLimit; i++) {
1016             utext_setNativeIndex(ut, i);
1017             int64_t cpIndex = utext_getNativeIndex(ut);
1018             TEST_ASSERT(cpIndex == startMap[i]);
1019             cpIndex = UTEXT_GETNATIVEINDEX(ut);
1020             TEST_ASSERT(cpIndex == startMap[i]);
1021         }
1022 
1023         // Check char32At
1024         for (i=0; i<startMapLimit; i++) {
1025             UChar32 c32 = utext_char32At(ut, i);
1026             TEST_ASSERT(c32 == c32Map[i]);
1027             int64_t cpIndex = utext_getNativeIndex(ut);
1028             TEST_ASSERT(cpIndex == startMap[i]);
1029         }
1030 
1031         // Check utext_next32From
1032         for (i=0; i<startMapLimit; i++) {
1033             UChar32 c32 = utext_next32From(ut, i);
1034             TEST_ASSERT(c32 == c32Map[i]);
1035             int64_t cpIndex = utext_getNativeIndex(ut);
1036             TEST_ASSERT(cpIndex == nextMap[i]);
1037         }
1038 
1039         // check utext_previous32From
1040         for (i=0; i<startMapLimit; i++) {
1041             gTestNum++;
1042             UChar32 c32 = utext_previous32From(ut, i);
1043             TEST_ASSERT(c32 == pr32Map[i]);
1044             int64_t cpIndex = utext_getNativeIndex(ut);
1045             TEST_ASSERT(cpIndex == prevMap[i]);
1046         }
1047 
1048         // check Extract
1049         //   Extract from i to i+1, which may be zero or one code points,
1050         //     depending on whether the indices straddle a cp boundary.
1051         for (i=0; i<startMapLimit; i++) {
1052             UChar buf[3];
1053             status = U_ZERO_ERROR;
1054             int32_t  extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1055             TEST_SUCCESS(status);
1056             TEST_ASSERT(extractedLen == exLen[i]);
1057             if (extractedLen > 0) {
1058                 UChar32  c32;
1059                 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1060                 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1061                 TEST_ASSERT(c32 == c32Map[i]);
1062             }
1063         }
1064 
1065         utext_close(ut);
1066     }
1067 
1068 
1069     {    //  Similar test, with utf16 instead of utf8
1070          //  TODO:  merge the common parts of these tests.
1071 
1072         UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV);
1073         int32_t startMap[]  ={ 0,     1,   1,    3,     4,  4,     6,  6};
1074         int32_t nextMap[]  = { 1,     3,   3,    4,     6,  6,     6,  6};
1075         int32_t prevMap[]  = { 0,     0,   0,    1,     3,  3,     4,  4};
1076         UChar32  c32Map[] =  {0x1000, 0x11000, 0x11000, 0x2000,  0x22000, 0x22000, -1, -1};
1077         UChar32  pr32Map[] = {    -1, 0x1000,  0x1000,  0x11000, 0x2000,  0x2000,   0x22000,   0x22000};
1078         int32_t  exLen[] =   {   1,  0,   2,  1,  0,  2,  0,  0,};
1079 
1080         u16str = u16str.unescape();
1081         UErrorCode status = U_ZERO_ERROR;
1082         UText *ut = utext_openUnicodeString(NULL, &u16str, &status);
1083         TEST_SUCCESS(status);
1084 
1085         int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t);
1086         int i;
1087         for (i=0; i<startMapLimit; i++) {
1088             utext_setNativeIndex(ut, i);
1089             int64_t cpIndex = utext_getNativeIndex(ut);
1090             TEST_ASSERT(cpIndex == startMap[i]);
1091         }
1092 
1093         // Check char32At
1094         for (i=0; i<startMapLimit; i++) {
1095             UChar32 c32 = utext_char32At(ut, i);
1096             TEST_ASSERT(c32 == c32Map[i]);
1097             int64_t cpIndex = utext_getNativeIndex(ut);
1098             TEST_ASSERT(cpIndex == startMap[i]);
1099         }
1100 
1101         // Check utext_next32From
1102         for (i=0; i<startMapLimit; i++) {
1103             UChar32 c32 = utext_next32From(ut, i);
1104             TEST_ASSERT(c32 == c32Map[i]);
1105             int64_t cpIndex = utext_getNativeIndex(ut);
1106             TEST_ASSERT(cpIndex == nextMap[i]);
1107         }
1108 
1109         // check utext_previous32From
1110         for (i=0; i<startMapLimit; i++) {
1111             UChar32 c32 = utext_previous32From(ut, i);
1112             TEST_ASSERT(c32 == pr32Map[i]);
1113             int64_t cpIndex = utext_getNativeIndex(ut);
1114             TEST_ASSERT(cpIndex == prevMap[i]);
1115         }
1116 
1117         // check Extract
1118         //   Extract from i to i+1, which may be zero or one code points,
1119         //     depending on whether the indices straddle a cp boundary.
1120         for (i=0; i<startMapLimit; i++) {
1121             UChar buf[3];
1122             status = U_ZERO_ERROR;
1123             int32_t  extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1124             TEST_SUCCESS(status);
1125             TEST_ASSERT(extractedLen == exLen[i]);
1126             if (extractedLen > 0) {
1127                 UChar32  c32;
1128                 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1129                 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1130                 TEST_ASSERT(c32 == c32Map[i]);
1131             }
1132         }
1133 
1134         utext_close(ut);
1135     }
1136 
1137     {    //  Similar test, with UText over Replaceable
1138          //  TODO:  merge the common parts of these tests.
1139 
1140         UnicodeString u16str("\\u1000\\U00011000\\u2000\\U00022000", -1, US_INV);
1141         int32_t startMap[]  ={ 0,     1,   1,    3,     4,  4,     6,  6};
1142         int32_t nextMap[]  = { 1,     3,   3,    4,     6,  6,     6,  6};
1143         int32_t prevMap[]  = { 0,     0,   0,    1,     3,  3,     4,  4};
1144         UChar32  c32Map[] =  {0x1000, 0x11000, 0x11000, 0x2000,  0x22000, 0x22000, -1, -1};
1145         UChar32  pr32Map[] = {    -1, 0x1000,  0x1000,  0x11000, 0x2000,  0x2000,   0x22000,   0x22000};
1146         int32_t  exLen[] =   {   1,  0,   2,  1,  0,  2,  0,  0,};
1147 
1148         u16str = u16str.unescape();
1149         UErrorCode status = U_ZERO_ERROR;
1150         UText *ut = utext_openReplaceable(NULL, &u16str, &status);
1151         TEST_SUCCESS(status);
1152 
1153         int32_t startMapLimit = sizeof(startMap) / sizeof(int32_t);
1154         int i;
1155         for (i=0; i<startMapLimit; i++) {
1156             utext_setNativeIndex(ut, i);
1157             int64_t cpIndex = utext_getNativeIndex(ut);
1158             TEST_ASSERT(cpIndex == startMap[i]);
1159         }
1160 
1161         // Check char32At
1162         for (i=0; i<startMapLimit; i++) {
1163             UChar32 c32 = utext_char32At(ut, i);
1164             TEST_ASSERT(c32 == c32Map[i]);
1165             int64_t cpIndex = utext_getNativeIndex(ut);
1166             TEST_ASSERT(cpIndex == startMap[i]);
1167         }
1168 
1169         // Check utext_next32From
1170         for (i=0; i<startMapLimit; i++) {
1171             UChar32 c32 = utext_next32From(ut, i);
1172             TEST_ASSERT(c32 == c32Map[i]);
1173             int64_t cpIndex = utext_getNativeIndex(ut);
1174             TEST_ASSERT(cpIndex == nextMap[i]);
1175         }
1176 
1177         // check utext_previous32From
1178         for (i=0; i<startMapLimit; i++) {
1179             UChar32 c32 = utext_previous32From(ut, i);
1180             TEST_ASSERT(c32 == pr32Map[i]);
1181             int64_t cpIndex = utext_getNativeIndex(ut);
1182             TEST_ASSERT(cpIndex == prevMap[i]);
1183         }
1184 
1185         // check Extract
1186         //   Extract from i to i+1, which may be zero or one code points,
1187         //     depending on whether the indices straddle a cp boundary.
1188         for (i=0; i<startMapLimit; i++) {
1189             UChar buf[3];
1190             status = U_ZERO_ERROR;
1191             int32_t  extractedLen = utext_extract(ut, i, i+1, buf, 3, &status);
1192             TEST_SUCCESS(status);
1193             TEST_ASSERT(extractedLen == exLen[i]);
1194             if (extractedLen > 0) {
1195                 UChar32  c32;
1196                 /* extractedLen-extractedLen == 0 is used to get around a compiler warning. */
1197                 U16_GET(buf, 0, extractedLen-extractedLen, extractedLen, c32);
1198                 TEST_ASSERT(c32 == c32Map[i]);
1199             }
1200         }
1201 
1202         utext_close(ut);
1203     }
1204 }
1205 
1206 
FreezeTest()1207 void UTextTest::FreezeTest() {
1208     // Check isWritable() and freeze() behavior.
1209     //
1210 
1211     UnicodeString  ustr("Hello, World.");
1212     const char u8str[] = {char(0x31), (char)0x32, (char)0x33, 0};
1213     const UChar u16str[] = {(UChar)0x31, (UChar)0x32, (UChar)0x44, 0};
1214 
1215     UErrorCode status = U_ZERO_ERROR;
1216     UText  *ut        = NULL;
1217     UText  *ut2       = NULL;
1218 
1219     ut = utext_openUTF8(ut, u8str, -1, &status);
1220     TEST_SUCCESS(status);
1221     UBool writable = utext_isWritable(ut);
1222     TEST_ASSERT(writable == FALSE);
1223     utext_copy(ut, 1, 2, 0, TRUE, &status);
1224     TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1225 
1226     status = U_ZERO_ERROR;
1227     ut = utext_openUChars(ut, u16str, -1, &status);
1228     TEST_SUCCESS(status);
1229     writable = utext_isWritable(ut);
1230     TEST_ASSERT(writable == FALSE);
1231     utext_copy(ut, 1, 2, 0, TRUE, &status);
1232     TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1233 
1234     status = U_ZERO_ERROR;
1235     ut = utext_openUnicodeString(ut, &ustr, &status);
1236     TEST_SUCCESS(status);
1237     writable = utext_isWritable(ut);
1238     TEST_ASSERT(writable == TRUE);
1239     utext_freeze(ut);
1240     writable = utext_isWritable(ut);
1241     TEST_ASSERT(writable == FALSE);
1242     utext_copy(ut, 1, 2, 0, TRUE, &status);
1243     TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1244 
1245     status = U_ZERO_ERROR;
1246     ut = utext_openUnicodeString(ut, &ustr, &status);
1247     TEST_SUCCESS(status);
1248     ut2 = utext_clone(ut2, ut, FALSE, FALSE, &status);  // clone with readonly = false
1249     TEST_SUCCESS(status);
1250     writable = utext_isWritable(ut2);
1251     TEST_ASSERT(writable == TRUE);
1252     ut2 = utext_clone(ut2, ut, FALSE, TRUE, &status);  // clone with readonly = true
1253     TEST_SUCCESS(status);
1254     writable = utext_isWritable(ut2);
1255     TEST_ASSERT(writable == FALSE);
1256     utext_copy(ut2, 1, 2, 0, TRUE, &status);
1257     TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1258 
1259     status = U_ZERO_ERROR;
1260     ut = utext_openConstUnicodeString(ut, (const UnicodeString *)&ustr, &status);
1261     TEST_SUCCESS(status);
1262     writable = utext_isWritable(ut);
1263     TEST_ASSERT(writable == FALSE);
1264     utext_copy(ut, 1, 2, 0, TRUE, &status);
1265     TEST_ASSERT(status == U_NO_WRITE_PERMISSION);
1266 
1267     // Deep Clone of a frozen UText should re-enable writing in the copy.
1268     status = U_ZERO_ERROR;
1269     ut = utext_openUnicodeString(ut, &ustr, &status);
1270     TEST_SUCCESS(status);
1271     utext_freeze(ut);
1272     ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status);   // deep clone
1273     TEST_SUCCESS(status);
1274     writable = utext_isWritable(ut2);
1275     TEST_ASSERT(writable == TRUE);
1276 
1277 
1278     // Deep clone of a frozen UText, where the base type is intrinsically non-writable,
1279     //  should NOT enable writing in the copy.
1280     status = U_ZERO_ERROR;
1281     ut = utext_openUChars(ut, u16str, -1, &status);
1282     TEST_SUCCESS(status);
1283     utext_freeze(ut);
1284     ut2 = utext_clone(ut2, ut, TRUE, FALSE, &status);   // deep clone
1285     TEST_SUCCESS(status);
1286     writable = utext_isWritable(ut2);
1287     TEST_ASSERT(writable == FALSE);
1288 
1289     // cleanup
1290     utext_close(ut);
1291     utext_close(ut2);
1292 }
1293 
1294 
1295 //
1296 //  Fragmented UText
1297 //      A UText type that works with a chunk size of 1.
1298 //      Intended to test for edge cases.
1299 //      Input comes from a UnicodeString.
1300 //
1301 //       ut.b    the character.  Put into both halves.
1302 //
1303 
1304 U_CDECL_BEGIN
1305 static UBool U_CALLCONV
fragTextAccess(UText * ut,int64_t index,UBool forward)1306 fragTextAccess(UText *ut, int64_t index, UBool forward) {
1307     const UnicodeString *us = (const UnicodeString *)ut->context;
1308     UChar  c;
1309     int32_t length = us->length();
1310     if (forward && index>=0 && index<length) {
1311         c = us->charAt((int32_t)index);
1312         ut->b = c | c<<16;
1313         ut->chunkOffset = 0;
1314         ut->chunkLength = 1;
1315         ut->chunkNativeStart = index;
1316         ut->chunkNativeLimit = index+1;
1317         return true;
1318     }
1319     if (!forward && index>0 && index <=length) {
1320         c = us->charAt((int32_t)index-1);
1321         ut->b = c | c<<16;
1322         ut->chunkOffset = 1;
1323         ut->chunkLength = 1;
1324         ut->chunkNativeStart = index-1;
1325         ut->chunkNativeLimit = index;
1326         return true;
1327     }
1328     ut->b = 0;
1329     ut->chunkOffset = 0;
1330     ut->chunkLength = 0;
1331     if (index <= 0) {
1332         ut->chunkNativeStart = 0;
1333         ut->chunkNativeLimit = 0;
1334     } else {
1335         ut->chunkNativeStart = length;
1336         ut->chunkNativeLimit = length;
1337     }
1338     return false;
1339 }
1340 
1341 // Function table to be used with this fragmented text provider.
1342 //   Initialized in the open function.
1343 static UTextFuncs  fragmentFuncs;
1344 
1345 // Clone function for fragmented text provider.
1346 //   Didn't really want to provide this, but it's easier to provide it than to keep it
1347 //   out of the tests.
1348 //
1349 UText *
cloneFragmentedUnicodeString(UText * dest,const UText * src,UBool deep,UErrorCode * status)1350 cloneFragmentedUnicodeString(UText *dest, const UText *src, UBool deep, UErrorCode *status) {
1351     if (U_FAILURE(*status)) {
1352         return NULL;
1353     }
1354     if (deep) {
1355         *status = U_UNSUPPORTED_ERROR;
1356         return NULL;
1357     }
1358     dest = utext_openUnicodeString(dest, (UnicodeString *)src->context, status);
1359     utext_setNativeIndex(dest, utext_getNativeIndex(src));
1360     return dest;
1361 }
1362 
1363 U_CDECL_END
1364 
1365 // Open function for the fragmented text provider.
1366 UText *
openFragmentedUnicodeString(UText * ut,UnicodeString * s,UErrorCode * status)1367 openFragmentedUnicodeString(UText *ut, UnicodeString *s, UErrorCode *status) {
1368     ut = utext_openUnicodeString(ut, s, status);
1369     if (U_FAILURE(*status)) {
1370         return ut;
1371     }
1372 
1373     // Copy of the function table from the stock UnicodeString UText,
1374     //   and replace the entry for the access function.
1375     memcpy(&fragmentFuncs, ut->pFuncs, sizeof(fragmentFuncs));
1376     fragmentFuncs.access = fragTextAccess;
1377     fragmentFuncs.clone  = cloneFragmentedUnicodeString;
1378     ut->pFuncs = &fragmentFuncs;
1379 
1380     ut->chunkContents = (UChar *)&ut->b;
1381     ut->pFuncs->access(ut, 0, TRUE);
1382     return ut;
1383 }
1384 
1385 // Regression test for Ticket 5560
1386 //   Clone fails to update chunkContentPointer in the cloned copy.
1387 //   This is only an issue for UText types that work in a local buffer,
1388 //      (UTF-8 wrapper, for example)
1389 //
1390 //   The test:
1391 //     1.  Create an inital UText
1392 //     2.  Deep clone it.  Contents should match original.
1393 //     3.  Reset original to something different.
1394 //     4.  Check that clone contents did not change.
1395 //
Ticket5560()1396 void UTextTest::Ticket5560() {
1397     /* The following two strings are in UTF-8 even on EBCDIC platforms. */
1398     static const char s1[] = {0x41,0x42,0x43,0x44,0x45,0x46,0}; /* "ABCDEF" */
1399     static const char s2[] = {0x31,0x32,0x33,0x34,0x35,0x36,0}; /* "123456" */
1400 	UErrorCode status = U_ZERO_ERROR;
1401 
1402 	UText ut1 = UTEXT_INITIALIZER;
1403 	UText ut2 = UTEXT_INITIALIZER;
1404 
1405 	utext_openUTF8(&ut1, s1, -1, &status);
1406 	UChar c = utext_next32(&ut1);
1407 	TEST_ASSERT(c == 0x41);  // c == 'A'
1408 
1409 	utext_clone(&ut2, &ut1, TRUE, FALSE, &status);
1410 	TEST_SUCCESS(status);
1411     c = utext_next32(&ut2);
1412 	TEST_ASSERT(c == 0x42);  // c == 'B'
1413     c = utext_next32(&ut1);
1414 	TEST_ASSERT(c == 0x42);  // c == 'B'
1415 
1416 	utext_openUTF8(&ut1, s2, -1, &status);
1417 	c = utext_next32(&ut1);
1418 	TEST_ASSERT(c == 0x31);  // c == '1'
1419     c = utext_next32(&ut2);
1420 	TEST_ASSERT(c == 0x43);  // c == 'C'
1421 
1422     utext_close(&ut1);
1423     utext_close(&ut2);
1424 }
1425 
1426 
1427 // Test for Ticket 6847
1428 //
Ticket6847()1429 void UTextTest::Ticket6847() {
1430     const int STRLEN = 90;
1431     UChar s[STRLEN+1];
1432     u_memset(s, 0x41, STRLEN);
1433     s[STRLEN] = 0;
1434 
1435     UErrorCode status = U_ZERO_ERROR;
1436     UText *ut = utext_openUChars(NULL, s, -1, &status);
1437 
1438     utext_setNativeIndex(ut, 0);
1439     int32_t count = 0;
1440     UChar32 c = 0;
1441     int64_t nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1442     TEST_ASSERT(nativeIndex == 0);
1443     while ((c = utext_next32(ut)) != U_SENTINEL) {
1444         TEST_ASSERT(c == 0x41);
1445         TEST_ASSERT(count < STRLEN);
1446         if (count >= STRLEN) {
1447             break;
1448         }
1449         count++;
1450         nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1451         TEST_ASSERT(nativeIndex == count);
1452     }
1453     TEST_ASSERT(count == STRLEN);
1454     nativeIndex = UTEXT_GETNATIVEINDEX(ut);
1455     TEST_ASSERT(nativeIndex == STRLEN);
1456     utext_close(ut);
1457 }
1458 
1459 
Ticket10562()1460 void UTextTest::Ticket10562() {
1461     // Note: failures show as a heap error when the test is run under valgrind.
1462     UErrorCode status = U_ZERO_ERROR;
1463 
1464     const char *utf8_string = "\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41\x41";
1465     UText *utf8Text = utext_openUTF8(NULL, utf8_string, -1, &status);
1466     TEST_SUCCESS(status);
1467     UText *deepClone = utext_clone(NULL, utf8Text, TRUE, FALSE, &status);
1468     TEST_SUCCESS(status);
1469     UText *shallowClone = utext_clone(NULL, deepClone, FALSE, FALSE, &status);
1470     TEST_SUCCESS(status);
1471     utext_close(shallowClone);
1472     utext_close(deepClone);
1473     utext_close(utf8Text);
1474 
1475     status = U_ZERO_ERROR;
1476     UnicodeString usString("Hello, World.");
1477     UText *usText = utext_openUnicodeString(NULL, &usString, &status);
1478     TEST_SUCCESS(status);
1479     UText *usDeepClone = utext_clone(NULL, usText, TRUE, FALSE, &status);
1480     TEST_SUCCESS(status);
1481     UText *usShallowClone = utext_clone(NULL, usDeepClone, FALSE, FALSE, &status);
1482     TEST_SUCCESS(status);
1483     utext_close(usShallowClone);
1484     utext_close(usDeepClone);
1485     utext_close(usText);
1486 }
1487 
1488 
Ticket10983()1489 void UTextTest::Ticket10983() {
1490     // Note: failure shows as a seg fault when the defect is present.
1491 
1492     UErrorCode status = U_ZERO_ERROR;
1493     UnicodeString s("Hello, World");
1494     UText *ut = utext_openConstUnicodeString(NULL, &s, &status);
1495     TEST_SUCCESS(status);
1496 
1497     status = U_INVALID_STATE_ERROR;
1498     UText *cloned = utext_clone(NULL, ut, TRUE, TRUE, &status);
1499     TEST_ASSERT(cloned == NULL);
1500     TEST_ASSERT(status == U_INVALID_STATE_ERROR);
1501 
1502     utext_close(ut);
1503 }
1504