1 /*
2 **********************************************************************
3 * Copyright (C) 2000-2009, International Business Machines
4 * Corporation and others. All Rights Reserved.
5 **********************************************************************
6 * Date Name Description
7 * 05/23/00 aliu Creation.
8 **********************************************************************
9 */
10
11 #include "unicode/utypes.h"
12
13 #if !UCONFIG_NO_TRANSLITERATION
14
15 #include "unicode/translit.h"
16 #include "rbt.h"
17 #include "unicode/calendar.h"
18 #include "unicode/uniset.h"
19 #include "unicode/uchar.h"
20 #include "unicode/normlzr.h"
21 #include "unicode/uchar.h"
22 #include "unicode/parseerr.h"
23 #include "unicode/usetiter.h"
24 #include "unicode/putil.h"
25 #include "unicode/uversion.h"
26 #include "unicode/locid.h"
27 #include "unicode/ulocdata.h"
28 #include "unicode/utf8.h"
29 #include "putilimp.h"
30 #include "cmemory.h"
31 #include "transrt.h"
32 #include "testutil.h"
33 #include <string.h>
34 #include <stdio.h>
35
36 #define CASE(id,test) case id: \
37 name = #test; \
38 if (exec) { \
39 logln(#test "---"); \
40 logln((UnicodeString)""); \
41 UDate t = uprv_getUTCtime(); \
42 test(); \
43 t = uprv_getUTCtime() - t; \
44 logln((UnicodeString)#test " took " + t/U_MILLIS_PER_DAY + " seconds"); \
45 } \
46 break
47
48 #define EXHAUSTIVE(id,test) case id: \
49 if(quick==FALSE){ \
50 name = #test; \
51 if (exec){ \
52 logln(#test "---"); \
53 logln((UnicodeString)""); \
54 test(); \
55 } \
56 }else{ \
57 name=""; \
58 } \
59 break
60 void
runIndexedTest(int32_t index,UBool exec,const char * & name,char *)61 TransliteratorRoundTripTest::runIndexedTest(int32_t index, UBool exec,
62 const char* &name, char* /*par*/) {
63 switch (index) {
64 CASE(0, TestCyrillic);
65 // CASE(0,TestKana);
66 CASE(1,TestHiragana);
67 CASE(2,TestKatakana);
68 CASE(3,TestJamo);
69 CASE(4,TestHangul);
70 CASE(5,TestGreek);
71 CASE(6,TestGreekUNGEGN);
72 CASE(7,Testel);
73 CASE(8,TestDevanagariLatin);
74 CASE(9,TestInterIndic);
75 CASE(10, TestHebrew);
76 CASE(11, TestArabic);
77 CASE(12, TestHan);
78 default: name = ""; break;
79 }
80 }
81
82
83 //--------------------------------------------------------------------
84 // TransliteratorPointer
85 //--------------------------------------------------------------------
86
87 /**
88 * A transliterator pointer wrapper that deletes the contained
89 * pointer automatically when the wrapper goes out of scope.
90 * Sometimes called a "janitor" or "smart pointer".
91 */
92 class TransliteratorPointer {
93 Transliterator* t;
94 // disallowed:
95 TransliteratorPointer(const TransliteratorPointer& rhs);
96 TransliteratorPointer& operator=(const TransliteratorPointer& rhs);
97 public:
TransliteratorPointer(Transliterator * adopted)98 TransliteratorPointer(Transliterator* adopted) {
99 t = adopted;
100 }
~TransliteratorPointer()101 ~TransliteratorPointer() {
102 delete t;
103 }
operator ->()104 inline Transliterator* operator->() { return t; }
operator const Transliterator*() const105 inline operator const Transliterator*() const { return t; }
operator Transliterator*()106 inline operator Transliterator*() { return t; }
107 };
108
109 //--------------------------------------------------------------------
110 // Legal
111 //--------------------------------------------------------------------
112
113 class Legal {
114 public:
Legal()115 Legal() {}
~Legal()116 virtual ~Legal() {}
is(const UnicodeString &) const117 virtual UBool is(const UnicodeString& /*sourceString*/) const {return TRUE;}
118 };
119
120 class LegalJamo : public Legal {
121 // any initial must be followed by a medial (or initial)
122 // any medial must follow an initial (or medial)
123 // any final must follow a medial (or final)
124 public:
LegalJamo()125 LegalJamo() {}
~LegalJamo()126 virtual ~LegalJamo() {}
127 virtual UBool is(const UnicodeString& sourceString) const;
128 int getType(UChar c) const;
129 };
130
is(const UnicodeString & sourceString) const131 UBool LegalJamo::is(const UnicodeString& sourceString) const {
132 int t;
133 UnicodeString decomp;
134 UErrorCode ec = U_ZERO_ERROR;
135 Normalizer::decompose(sourceString, FALSE, 0, decomp, ec);
136 if (U_FAILURE(ec)) {
137 return FALSE;
138 }
139 for (int i = 0; i < decomp.length(); ++i) { // don't worry about surrogates
140 switch (getType(decomp.charAt(i))) {
141 case 0: t = getType(decomp.charAt(i+1));
142 if (t != 0 && t != 1) { return FALSE; }
143 break;
144 case 1: t = getType(decomp.charAt(i-1));
145 if (t != 0 && t != 1) { return FALSE; }
146 break;
147 case 2: t = getType(decomp.charAt(i-1));
148 if (t != 1 && t != 2) { return FALSE; }
149 break;
150 }
151 }
152 return TRUE;
153 }
154
getType(UChar c) const155 int LegalJamo::getType(UChar c) const {
156 if (0x1100 <= c && c <= 0x1112)
157 return 0;
158 else if (0x1161 <= c && c <= 0x1175)
159 return 1;
160 else if (0x11A8 <= c && c <= 0x11C2)
161 return 2;
162 return -1; // other
163 }
164
165 class LegalGreek : public Legal {
166 UBool full;
167 public:
LegalGreek(UBool _full)168 LegalGreek(UBool _full) { full = _full; }
~LegalGreek()169 virtual ~LegalGreek() {}
170
171 virtual UBool is(const UnicodeString& sourceString) const;
172
173 static UBool isVowel(UChar c);
174
175 static UBool isRho(UChar c);
176 };
177
is(const UnicodeString & sourceString) const178 UBool LegalGreek::is(const UnicodeString& sourceString) const {
179 UnicodeString decomp;
180 UErrorCode ec = U_ZERO_ERROR;
181 Normalizer::decompose(sourceString, FALSE, 0, decomp, ec);
182
183 // modern is simpler: don't care about anything but a grave
184 if (full == FALSE) {
185 // A special case which is legal but should be
186 // excluded from round trip
187 // if (sourceString == UnicodeString("\\u039C\\u03C0", "")) {
188 // return FALSE;
189 // }
190 for (int32_t i = 0; i < decomp.length(); ++i) {
191 UChar c = decomp.charAt(i);
192 // exclude all the accents
193 if (c == 0x0313 || c == 0x0314 || c == 0x0300 || c == 0x0302
194 || c == 0x0342 || c == 0x0345
195 ) return FALSE;
196 }
197 return TRUE;
198 }
199
200 // Legal greek has breathing marks IFF there is a vowel or RHO at the start
201 // IF it has them, it has exactly one.
202 // IF it starts with a RHO, then the breathing mark must come before the second letter.
203 // Since there are no surrogates in greek, don't worry about them
204 UBool firstIsVowel = FALSE;
205 UBool firstIsRho = FALSE;
206 UBool noLetterYet = TRUE;
207 int32_t breathingCount = 0;
208 int32_t letterCount = 0;
209 for (int32_t i = 0; i < decomp.length(); ++i) {
210 UChar c = decomp.charAt(i);
211 if (u_isalpha(c)) {
212 ++letterCount;
213 if (noLetterYet) {
214 noLetterYet = FALSE;
215 firstIsVowel = isVowel(c);
216 firstIsRho = isRho(c);
217 }
218 if (firstIsRho && letterCount == 2 && breathingCount == 0) {
219 return FALSE;
220 }
221 }
222 if (c == 0x0313 || c == 0x0314) {
223 ++breathingCount;
224 }
225 }
226
227 if (firstIsVowel || firstIsRho) return breathingCount == 1;
228 return breathingCount == 0;
229 }
230
isVowel(UChar c)231 UBool LegalGreek::isVowel(UChar c) {
232 switch (c) {
233 case 0x03B1:
234 case 0x03B5:
235 case 0x03B7:
236 case 0x03B9:
237 case 0x03BF:
238 case 0x03C5:
239 case 0x03C9:
240 case 0x0391:
241 case 0x0395:
242 case 0x0397:
243 case 0x0399:
244 case 0x039F:
245 case 0x03A5:
246 case 0x03A9:
247 return TRUE;
248 }
249 return FALSE;
250 }
251
isRho(UChar c)252 UBool LegalGreek::isRho(UChar c) {
253 switch (c) {
254 case 0x03C1:
255 case 0x03A1:
256 return TRUE;
257 }
258 return FALSE;
259 }
260
261 // AbbreviatedUnicodeSetIterator Interface ---------------------------------------------
262 //
263 // Iterate over a UnicodeSet, only returning a sampling of the contained code points.
264 // density is the approximate total number of code points to returned for the entire set.
265 //
266
267 class AbbreviatedUnicodeSetIterator : public UnicodeSetIterator {
268 public :
269
270 AbbreviatedUnicodeSetIterator();
271 virtual ~AbbreviatedUnicodeSetIterator();
272 void reset(UnicodeSet& set, UBool abb = FALSE, int32_t density = 100);
273
274 /**
275 * ICU "poor man's RTTI", returns a UClassID for this class.
276 */
getStaticClassID()277 static inline UClassID getStaticClassID() { return (UClassID)&fgClassID; }
278
279 /**
280 * ICU "poor man's RTTI", returns a UClassID for the actual class.
281 */
getDynamicClassID() const282 virtual inline UClassID getDynamicClassID() const { return getStaticClassID(); }
283
284 private :
285 UBool abbreviated;
286 int32_t perRange; // The maximum number of code points to be returned from each range
287 virtual void loadRange(int32_t range);
288
289 /**
290 * The address of this static class variable serves as this class's ID
291 * for ICU "poor man's RTTI".
292 */
293 static const char fgClassID;
294 };
295
296 // AbbreviatedUnicodeSetIterator Implementation ---------------------------------------
297
298 const char AbbreviatedUnicodeSetIterator::fgClassID=0;
299
AbbreviatedUnicodeSetIterator()300 AbbreviatedUnicodeSetIterator::AbbreviatedUnicodeSetIterator() :
301 UnicodeSetIterator(), abbreviated(FALSE) {
302 }
303
~AbbreviatedUnicodeSetIterator()304 AbbreviatedUnicodeSetIterator::~AbbreviatedUnicodeSetIterator() {
305 }
306
reset(UnicodeSet & newSet,UBool abb,int32_t density)307 void AbbreviatedUnicodeSetIterator::reset(UnicodeSet& newSet, UBool abb, int32_t density) {
308 UnicodeSetIterator::reset(newSet);
309 abbreviated = abb;
310 perRange = newSet.getRangeCount();
311 if (perRange != 0) {
312 perRange = density / perRange;
313 }
314 }
315
loadRange(int32_t myRange)316 void AbbreviatedUnicodeSetIterator::loadRange(int32_t myRange) {
317 UnicodeSetIterator::loadRange(myRange);
318 if (abbreviated && (endElement > nextElement + perRange)) {
319 endElement = nextElement + perRange;
320 }
321 }
322
323 //--------------------------------------------------------------------
324 // RTTest Interface
325 //--------------------------------------------------------------------
326
327 class RTTest : public IntlTest {
328
329 // PrintWriter out;
330
331 UnicodeString transliteratorID;
332 int32_t errorLimit;
333 int32_t errorCount;
334 int32_t pairLimit;
335 UnicodeSet sourceRange;
336 UnicodeSet targetRange;
337 UnicodeSet toSource;
338 UnicodeSet toTarget;
339 UnicodeSet roundtripExclusionsSet;
340 IntlTest* parent;
341 Legal* legalSource; // NOT owned
342 UnicodeSet badCharacters;
343
344 public:
345
346 /*
347 * create a test for the given script transliterator.
348 */
349 RTTest(const UnicodeString& transliteratorIDStr);
350
351 virtual ~RTTest();
352
353 void setErrorLimit(int32_t limit);
354
355 void setPairLimit(int32_t limit);
356
357 void test(const UnicodeString& sourceRange,
358 const UnicodeString& targetRange,
359 const char* roundtripExclusions,
360 IntlTest* parent,
361 UBool quick,
362 Legal* adoptedLegal,
363 int32_t density = 100);
364
365 private:
366
367 // Added to do better equality check.
368
369 static UBool isSame(const UnicodeString& a, const UnicodeString& b);
370
371 static UBool isCamel(const UnicodeString& a);
372
373 UBool checkIrrelevants(Transliterator *t, const UnicodeString& irrelevants);
374
375 void test2(UBool quick, int32_t density);
376
377 void logWrongScript(const UnicodeString& label,
378 const UnicodeString& from,
379 const UnicodeString& to);
380
381 void logNotCanonical(const UnicodeString& label,
382 const UnicodeString& from,
383 const UnicodeString& to,
384 const UnicodeString& fromCan,
385 const UnicodeString& toCan);
386
387 void logFails(const UnicodeString& label);
388
389 void logToRulesFails(const UnicodeString& label,
390 const UnicodeString& from,
391 const UnicodeString& to,
392 const UnicodeString& toCan);
393
394 void logRoundTripFailure(const UnicodeString& from,
395 const UnicodeString& toID,
396 const UnicodeString& to,
397 const UnicodeString& backID,
398 const UnicodeString& back);
399 };
400
401 //--------------------------------------------------------------------
402 // RTTest Implementation
403 //--------------------------------------------------------------------
404
405 /*
406 * create a test for the given script transliterator.
407 */
RTTest(const UnicodeString & transliteratorIDStr)408 RTTest::RTTest(const UnicodeString& transliteratorIDStr) {
409 transliteratorID = transliteratorIDStr;
410 errorLimit = 500;
411 errorCount = 0;
412 pairLimit = 0x10000;
413 }
414
~RTTest()415 RTTest::~RTTest() {
416 }
417
setErrorLimit(int32_t limit)418 void RTTest::setErrorLimit(int32_t limit) {
419 errorLimit = limit;
420 }
421
setPairLimit(int32_t limit)422 void RTTest::setPairLimit(int32_t limit) {
423 pairLimit = limit;
424 }
425
isSame(const UnicodeString & a,const UnicodeString & b)426 UBool RTTest::isSame(const UnicodeString& a, const UnicodeString& b) {
427 if (a == b) return TRUE;
428 if (a.caseCompare(b, U_FOLD_CASE_DEFAULT)==0 && isCamel(a)) return TRUE;
429 UnicodeString aa, bb;
430 UErrorCode ec = U_ZERO_ERROR;
431 Normalizer::decompose(a, FALSE, 0, aa, ec);
432 Normalizer::decompose(b, FALSE, 0, bb, ec);
433 if (aa == bb) return TRUE;
434 if (aa.caseCompare(bb, U_FOLD_CASE_DEFAULT)==0 && isCamel(aa)) return TRUE;
435 return FALSE;
436 }
437
isCamel(const UnicodeString & a)438 UBool RTTest::isCamel(const UnicodeString& a) {
439 // see if string is of the form aB; e.g. lower, then upper or title
440 UChar32 cp;
441 UBool haveLower = FALSE;
442 for (int32_t i = 0; i < a.length(); i += UTF_CHAR_LENGTH(cp)) {
443 cp = a.char32At(i);
444 int8_t t = u_charType(cp);
445 switch (t) {
446 case U_UPPERCASE_LETTER:
447 if (haveLower) return TRUE;
448 break;
449 case U_TITLECASE_LETTER:
450 if (haveLower) return TRUE;
451 // drop through, since second letter is lower.
452 case U_LOWERCASE_LETTER:
453 haveLower = TRUE;
454 break;
455 }
456 }
457 return FALSE;
458 }
459
test(const UnicodeString & sourceRangeVal,const UnicodeString & targetRangeVal,const char * roundtripExclusions,IntlTest * logVal,UBool quickRt,Legal * adoptedLegal,int32_t density)460 void RTTest::test(const UnicodeString& sourceRangeVal,
461 const UnicodeString& targetRangeVal,
462 const char* roundtripExclusions,
463 IntlTest* logVal, UBool quickRt,
464 Legal* adoptedLegal,
465 int32_t density)
466 {
467
468 UErrorCode status = U_ZERO_ERROR;
469
470 this->parent = logVal;
471 this->legalSource = adoptedLegal;
472
473 UnicodeSet neverOk("[:Other:]", status);
474 UnicodeSet okAnyway("[^[:Letter:]]", status);
475
476 if (U_FAILURE(status)) {
477 parent->dataerrln("FAIL: Initializing UnicodeSet with [:Other:] or [^[:Letter:]] - Error: %s", u_errorName(status));
478 return;
479 }
480
481 this->sourceRange.clear();
482 this->sourceRange.applyPattern(sourceRangeVal, status);
483 if (U_FAILURE(status)) {
484 parent->errln("FAIL: UnicodeSet::applyPattern(" +
485 sourceRangeVal + ")");
486 return;
487 }
488 this->sourceRange.removeAll(neverOk);
489
490 this->targetRange.clear();
491 this->targetRange.applyPattern(targetRangeVal, status);
492 if (U_FAILURE(status)) {
493 parent->errln("FAIL: UnicodeSet::applyPattern(" +
494 targetRangeVal + ")");
495 return;
496 }
497 this->targetRange.removeAll(neverOk);
498
499 this->toSource.clear();
500 this->toSource.applyPattern(sourceRangeVal, status);
501 if (U_FAILURE(status)) {
502 parent->errln("FAIL: UnicodeSet::applyPattern(" +
503 sourceRangeVal + ")");
504 return;
505 }
506 this->toSource.addAll(okAnyway);
507
508 this->toTarget.clear();
509 this->toTarget.applyPattern(targetRangeVal, status);
510 if (U_FAILURE(status)) {
511 parent->errln("FAIL: UnicodeSet::applyPattern(" +
512 targetRangeVal + ")");
513 return;
514 }
515 this->toTarget.addAll(okAnyway);
516
517 this->roundtripExclusionsSet.clear();
518 if (roundtripExclusions != NULL && strlen(roundtripExclusions) > 0) {
519 this->roundtripExclusionsSet.applyPattern(UnicodeString(roundtripExclusions, -1, US_INV), status);
520 if (U_FAILURE(status)) {
521 parent->errln("FAIL: UnicodeSet::applyPattern(%s)", roundtripExclusions);
522 return;
523 }
524 }
525
526 badCharacters.clear();
527 badCharacters.applyPattern("[:Other:]", status);
528 if (U_FAILURE(status)) {
529 parent->errln("FAIL: UnicodeSet::applyPattern([:Other:])");
530 return;
531 }
532
533 test2(quickRt, density);
534
535 if (errorCount > 0) {
536 char str[100];
537 int32_t length = transliteratorID.extract(str, 100, NULL, status);
538 str[length] = 0;
539 parent->errln("FAIL: %s errors: %d %s", str, errorCount, (errorCount > errorLimit ? " (at least!)" : " ")); // + ", see " + logFileName);
540 } else {
541 char str[100];
542 int32_t length = transliteratorID.extract(str, 100, NULL, status);
543 str[length] = 0;
544 parent->logln("%s ok", str);
545 }
546 }
547
checkIrrelevants(Transliterator * t,const UnicodeString & irrelevants)548 UBool RTTest::checkIrrelevants(Transliterator *t,
549 const UnicodeString& irrelevants) {
550 for (int i = 0; i < irrelevants.length(); ++i) {
551 UChar c = irrelevants.charAt(i);
552 UnicodeString srcStr(c);
553 UnicodeString targ = srcStr;
554 t->transliterate(targ);
555 if (srcStr == targ) return TRUE;
556 }
557 return FALSE;
558 }
559
test2(UBool quickRt,int32_t density)560 void RTTest::test2(UBool quickRt, int32_t density) {
561
562 UnicodeString srcStr, targ, reverse;
563 UErrorCode status = U_ZERO_ERROR;
564 UParseError parseError ;
565 TransliteratorPointer sourceToTarget(
566 Transliterator::createInstance(transliteratorID, UTRANS_FORWARD, parseError,
567 status));
568 if ((Transliterator *)sourceToTarget == NULL) {
569 parent->errln("FAIL: createInstance(" + transliteratorID +
570 ") returned NULL. Error: " + u_errorName(status)
571 + "\n\tpreContext : " + prettify(parseError.preContext)
572 + "\n\tpostContext : " + prettify(parseError.postContext));
573
574 return;
575 }
576 TransliteratorPointer targetToSource(sourceToTarget->createInverse(status));
577 if ((Transliterator *)targetToSource == NULL) {
578 parent->errln("FAIL: " + transliteratorID +
579 ".createInverse() returned NULL. Error:" + u_errorName(status)
580 + "\n\tpreContext : " + prettify(parseError.preContext)
581 + "\n\tpostContext : " + prettify(parseError.postContext));
582 return;
583 }
584
585 AbbreviatedUnicodeSetIterator usi;
586 AbbreviatedUnicodeSetIterator usi2;
587
588 parent->logln("Checking that at least one irrelevant character is not NFC'ed");
589 // string is from NFC_NO in the UCD
590 UnicodeString irrelevants = CharsToUnicodeString("\\u2000\\u2001\\u2126\\u212A\\u212B\\u2329");
591
592 if (checkIrrelevants(sourceToTarget, irrelevants) == FALSE) {
593 logFails("Source-Target, irrelevants");
594 }
595 if (checkIrrelevants(targetToSource, irrelevants) == FALSE) {
596 logFails("Target-Source, irrelevants");
597 }
598
599 if (!quickRt){
600 parent->logln("Checking that toRules works");
601 UnicodeString rules = "";
602
603 UParseError parseError;
604 rules = sourceToTarget->toRules(rules, TRUE);
605 // parent->logln((UnicodeString)"toRules => " + rules);
606 TransliteratorPointer sourceToTarget2(Transliterator::createFromRules(
607 "s2t2", rules,
608 UTRANS_FORWARD,
609 parseError, status));
610 if (U_FAILURE(status)) {
611 parent->errln("FAIL: createFromRules %s\n", u_errorName(status));
612 return;
613 }
614
615 rules = targetToSource->toRules(rules, FALSE);
616 TransliteratorPointer targetToSource2(Transliterator::createFromRules(
617 "t2s2", rules,
618 UTRANS_FORWARD,
619 parseError, status));
620 if (U_FAILURE(status)) {
621 parent->errln("FAIL: createFromRules %s\n", u_errorName(status));
622 return;
623 }
624
625 usi.reset(sourceRange);
626 for (;;) {
627 if (!usi.next() || usi.isString()) break;
628 UChar32 c = usi.getCodepoint();
629
630 UnicodeString srcStr((UChar32)c);
631 UnicodeString targ = srcStr;
632 sourceToTarget->transliterate(targ);
633 UnicodeString targ2 = srcStr;
634 sourceToTarget2->transliterate(targ2);
635 if (targ != targ2) {
636 logToRulesFails("Source-Target, toRules", srcStr, targ, targ2);
637 }
638 }
639
640 usi.reset(targetRange);
641 for (;;) {
642 if (!usi.next() || usi.isString()) break;
643 UChar32 c = usi.getCodepoint();
644
645 UnicodeString srcStr((UChar32)c);
646 UnicodeString targ = srcStr;
647 targetToSource->transliterate(targ);
648 UnicodeString targ2 = srcStr;
649 targetToSource2->transliterate(targ2);
650 if (targ != targ2) {
651 logToRulesFails("Target-Source, toRules", srcStr, targ, targ2);
652 }
653 }
654 }
655
656 parent->logln("Checking that all source characters convert to target - Singles");
657
658 UnicodeSet failSourceTarg;
659 usi.reset(sourceRange);
660 for (;;) {
661 if (!usi.next() || usi.isString()) break;
662 UChar32 c = usi.getCodepoint();
663
664 UnicodeString srcStr((UChar32)c);
665 UnicodeString targ = srcStr;
666 sourceToTarget->transliterate(targ);
667 if (toTarget.containsAll(targ) == FALSE
668 || badCharacters.containsSome(targ) == TRUE) {
669 UnicodeString targD;
670 Normalizer::decompose(targ, FALSE, 0, targD, status);
671 if (U_FAILURE(status)) {
672 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
673 return;
674 }
675 if (toTarget.containsAll(targD) == FALSE ||
676 badCharacters.containsSome(targD) == TRUE) {
677 logWrongScript("Source-Target", srcStr, targ);
678 failSourceTarg.add(c);
679 continue;
680 }
681 }
682
683 UnicodeString cs2;
684 Normalizer::decompose(srcStr, FALSE, 0, cs2, status);
685 if (U_FAILURE(status)) {
686 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
687 return;
688 }
689 UnicodeString targ2 = cs2;
690 sourceToTarget->transliterate(targ2);
691 if (targ != targ2) {
692 logNotCanonical("Source-Target", srcStr, targ,cs2, targ2);
693 }
694 }
695
696 parent->logln("Checking that all source characters convert to target - Doubles");
697
698 UnicodeSet sourceRangeMinusFailures(sourceRange);
699 sourceRangeMinusFailures.removeAll(failSourceTarg);
700
701 usi.reset(sourceRangeMinusFailures, quickRt, density);
702 for (;;) {
703 if (!usi.next() || usi.isString()) break;
704 UChar32 c = usi.getCodepoint();
705
706 usi2.reset(sourceRangeMinusFailures, quickRt, density);
707 for (;;) {
708 if (!usi2.next() || usi2.isString()) break;
709 UChar32 d = usi2.getCodepoint();
710
711 UnicodeString srcStr;
712 srcStr += (UChar32)c;
713 srcStr += (UChar32)d;
714 UnicodeString targ = srcStr;
715 sourceToTarget->transliterate(targ);
716 if (toTarget.containsAll(targ) == FALSE ||
717 badCharacters.containsSome(targ) == TRUE)
718 {
719 UnicodeString targD;
720 Normalizer::decompose(targ, FALSE, 0, targD, status);
721 if (U_FAILURE(status)) {
722 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
723 return;
724 }
725 if (toTarget.containsAll(targD) == FALSE ||
726 badCharacters.containsSome(targD) == TRUE) {
727 logWrongScript("Source-Target", srcStr, targ);
728 continue;
729 }
730 }
731 UnicodeString cs2;
732 Normalizer::decompose(srcStr, FALSE, 0, cs2, status);
733 if (U_FAILURE(status)) {
734 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
735 return;
736 }
737 UnicodeString targ2 = cs2;
738 sourceToTarget->transliterate(targ2);
739 if (targ != targ2) {
740 logNotCanonical("Source-Target", srcStr, targ, cs2,targ2);
741 }
742 }
743 }
744
745 parent->logln("Checking that target characters convert to source and back - Singles");
746
747 UnicodeSet failTargSource;
748 UnicodeSet failRound;
749
750 usi.reset(targetRange);
751 for (;;) {
752 if (!usi.next()) break;
753
754 if(usi.isString()){
755 srcStr = usi.getString();
756 }else{
757 srcStr = (UnicodeString)usi.getCodepoint();
758 }
759
760 UChar32 c = srcStr.char32At(0);
761
762 targ = srcStr;
763 targetToSource->transliterate(targ);
764 reverse = targ;
765 sourceToTarget->transliterate(reverse);
766
767 if (toSource.containsAll(targ) == FALSE ||
768 badCharacters.containsSome(targ) == TRUE) {
769 UnicodeString targD;
770 Normalizer::decompose(targ, FALSE, 0, targD, status);
771 if (U_FAILURE(status)) {
772 parent->errln("FAIL: Internal error during decomposition%s\n", u_errorName(status));
773 return;
774 }
775 if (toSource.containsAll(targD) == FALSE) {
776 logWrongScript("Target-Source", srcStr, targ);
777 failTargSource.add(c);
778 continue;
779 }
780 if (badCharacters.containsSome(targD) == TRUE) {
781 logWrongScript("Target-Source*", srcStr, targ);
782 failTargSource.add(c);
783 continue;
784 }
785 }
786 if (isSame(srcStr, reverse) == FALSE &&
787 roundtripExclusionsSet.contains(c) == FALSE
788 && roundtripExclusionsSet.contains(srcStr)==FALSE) {
789 logRoundTripFailure(srcStr,targetToSource->getID(), targ,sourceToTarget->getID(), reverse);
790 failRound.add(c);
791 continue;
792 }
793
794 UnicodeString targ2;
795 Normalizer::decompose(targ, FALSE, 0, targ2, status);
796 if (U_FAILURE(status)) {
797 parent->errln("FAIL: Internal error during decomposition%s\n", u_errorName(status));
798 return;
799 }
800 UnicodeString reverse2 = targ2;
801 sourceToTarget->transliterate(reverse2);
802 if (reverse != reverse2) {
803 logNotCanonical("Target-Source", targ, reverse, targ2, reverse2);
804 }
805 }
806
807 parent->logln("Checking that target characters convert to source and back - Doubles");
808 int32_t count = 0;
809
810 UnicodeSet targetRangeMinusFailures(targetRange);
811 targetRangeMinusFailures.removeAll(failTargSource);
812 targetRangeMinusFailures.removeAll(failRound);
813
814 usi.reset(targetRangeMinusFailures, quickRt, density);
815 UnicodeString targ2;
816 UnicodeString reverse2;
817 UnicodeString targD;
818 for (;;) {
819 if (!usi.next() || usi.isString()) break;
820 UChar32 c = usi.getCodepoint();
821 if (++count > pairLimit) {
822 //throw new TestTruncated("Test truncated at " + pairLimit + " x 64k pairs");
823 parent->logln("");
824 parent->logln((UnicodeString)"Test truncated at " + pairLimit + " x 64k pairs");
825 return;
826 }
827
828 usi2.reset(targetRangeMinusFailures, quickRt, density);
829 for (;;) {
830 if (!usi2.next() || usi2.isString())
831 break;
832 UChar32 d = usi2.getCodepoint();
833 srcStr.truncate(0); // empty the variable without construction/destruction
834 srcStr += c;
835 srcStr += d;
836
837 targ = srcStr;
838 targetToSource->transliterate(targ);
839 reverse = targ;
840 sourceToTarget->transliterate(reverse);
841
842 if (toSource.containsAll(targ) == FALSE ||
843 badCharacters.containsSome(targ) == TRUE)
844 {
845 targD.truncate(0); // empty the variable without construction/destruction
846 Normalizer::decompose(targ, FALSE, 0, targD, status);
847 if (U_FAILURE(status)) {
848 parent->errln("FAIL: Internal error during decomposition%s\n",
849 u_errorName(status));
850 return;
851 }
852 if (toSource.containsAll(targD) == FALSE
853 || badCharacters.containsSome(targD) == TRUE)
854 {
855 logWrongScript("Target-Source", srcStr, targ);
856 continue;
857 }
858 }
859 if (isSame(srcStr, reverse) == FALSE &&
860 roundtripExclusionsSet.contains(c) == FALSE&&
861 roundtripExclusionsSet.contains(d) == FALSE &&
862 roundtripExclusionsSet.contains(srcStr)== FALSE)
863 {
864 logRoundTripFailure(srcStr,targetToSource->getID(), targ, sourceToTarget->getID(),reverse);
865 continue;
866 }
867
868 targ2.truncate(0); // empty the variable without construction/destruction
869 Normalizer::decompose(targ, FALSE, 0, targ2, status);
870 if (U_FAILURE(status)) {
871 parent->errln("FAIL: Internal error during decomposition%s\n", u_errorName(status));
872 return;
873 }
874 reverse2 = targ2;
875 sourceToTarget->transliterate(reverse2);
876 if (reverse != reverse2) {
877 logNotCanonical("Target-Source", targ,reverse, targ2, reverse2);
878 }
879 }
880 }
881 parent->logln("");
882 }
883
logWrongScript(const UnicodeString & label,const UnicodeString & from,const UnicodeString & to)884 void RTTest::logWrongScript(const UnicodeString& label,
885 const UnicodeString& from,
886 const UnicodeString& to) {
887 parent->errln((UnicodeString)"FAIL " +
888 label + ": " +
889 from + "(" + TestUtility::hex(from) + ") => " +
890 to + "(" + TestUtility::hex(to) + ")");
891 ++errorCount;
892 }
893
logNotCanonical(const UnicodeString & label,const UnicodeString & from,const UnicodeString & to,const UnicodeString & fromCan,const UnicodeString & toCan)894 void RTTest::logNotCanonical(const UnicodeString& label,
895 const UnicodeString& from,
896 const UnicodeString& to,
897 const UnicodeString& fromCan,
898 const UnicodeString& toCan) {
899 parent->errln((UnicodeString)"FAIL (can.equiv)" +
900 label + ": " +
901 from + "(" + TestUtility::hex(from) + ") => " +
902 to + "(" + TestUtility::hex(to) + ")" +
903 fromCan + "(" + TestUtility::hex(fromCan) + ") => " +
904 toCan + " (" +
905 TestUtility::hex(toCan) + ")"
906 );
907 ++errorCount;
908 }
909
logFails(const UnicodeString & label)910 void RTTest::logFails(const UnicodeString& label) {
911 parent->errln((UnicodeString)"<br>FAIL " + label);
912 ++errorCount;
913 }
914
logToRulesFails(const UnicodeString & label,const UnicodeString & from,const UnicodeString & to,const UnicodeString & otherTo)915 void RTTest::logToRulesFails(const UnicodeString& label,
916 const UnicodeString& from,
917 const UnicodeString& to,
918 const UnicodeString& otherTo)
919 {
920 parent->errln((UnicodeString)"FAIL: " +
921 label + ": " +
922 from + "(" + TestUtility::hex(from) + ") => " +
923 to + "(" + TestUtility::hex(to) + ")" +
924 "!=" +
925 otherTo + " (" +
926 TestUtility::hex(otherTo) + ")"
927 );
928 ++errorCount;
929 }
930
931
logRoundTripFailure(const UnicodeString & from,const UnicodeString & toID,const UnicodeString & to,const UnicodeString & backID,const UnicodeString & back)932 void RTTest::logRoundTripFailure(const UnicodeString& from,
933 const UnicodeString& toID,
934 const UnicodeString& to,
935 const UnicodeString& backID,
936 const UnicodeString& back) {
937 if (legalSource->is(from) == FALSE) return; // skip illegals
938
939 parent->errln((UnicodeString)"FAIL Roundtrip: " +
940 from + "(" + TestUtility::hex(from) + ") => " +
941 to + "(" + TestUtility::hex(to) + ") "+toID+" => " +
942 back + "(" + TestUtility::hex(back) + ") "+backID+" => ");
943 ++errorCount;
944 }
945
946 //--------------------------------------------------------------------
947 // Specific Tests
948 //--------------------------------------------------------------------
949
950 /*
951 Note: Unicode 3.2 added new Hiragana/Katakana characters:
952
953 3095..3096 ; 3.2 # [2] HIRAGANA LETTER SMALL KA..HIRAGANA LETTER SMALL KE
954 309F..30A0 ; 3.2 # [2] HIRAGANA DIGRAPH YORI..KATAKANA-HIRAGANA DOUBLE HYPHEN
955 30FF ; 3.2 # KATAKANA DIGRAPH KOTO
956 31F0..31FF ; 3.2 # [16] KATAKANA LETTER SMALL KU..KATAKANA LETTER SMALL RO
957
958 We will not add them to the rules until they are more supported (e.g. in fonts on Windows)
959 A bug has been filed to remind us to do this: #1979.
960 */
961
962 static const char KATAKANA[] = "[[[:katakana:][\\u30A1-\\u30FA\\u30FC]]-[\\u30FF\\u31F0-\\u31FF]]";
963 static const char HIRAGANA[] = "[[[:hiragana:][\\u3040-\\u3094]]-[\\u3095-\\u3096\\u309F-\\u30A0]]";
964 static const char LENGTH[] = "[\\u30FC]";
965 static const char HALFWIDTH_KATAKANA[] = "[\\uFF65-\\uFF9D]";
966 static const char KATAKANA_ITERATION[] = "[\\u30FD\\u30FE]";
967 static const char HIRAGANA_ITERATION[] = "[\\u309D\\u309E]";
968 static const int32_t TEMP_MAX=256;
969
TestKana()970 void TransliteratorRoundTripTest::TestKana() {
971 RTTest test("Katakana-Hiragana");
972 Legal *legal = new Legal();
973 char temp[TEMP_MAX];
974 strcpy(temp, "[");
975 strcat(temp, HALFWIDTH_KATAKANA);
976 strcat(temp, LENGTH);
977 strcat(temp, "]");
978 test.test(KATAKANA, UnicodeString("[") + HIRAGANA + LENGTH + UnicodeString("]"),
979 temp,
980 this, quick, legal);
981 delete legal;
982 }
983
TestHiragana()984 void TransliteratorRoundTripTest::TestHiragana() {
985 RTTest test("Latin-Hiragana");
986 Legal *legal = new Legal();
987 test.test(UnicodeString("[a-zA-Z]", ""),
988 UnicodeString(HIRAGANA, -1, US_INV),
989 HIRAGANA_ITERATION, this, quick, legal);
990 delete legal;
991 }
992
TestKatakana()993 void TransliteratorRoundTripTest::TestKatakana() {
994 RTTest test("Latin-Katakana");
995 Legal *legal = new Legal();
996 char temp[TEMP_MAX];
997 strcpy(temp, "[");
998 strcat(temp, KATAKANA_ITERATION);
999 strcat(temp, HALFWIDTH_KATAKANA);
1000 strcat(temp, "]");
1001 test.test(UnicodeString("[a-zA-Z]", ""),
1002 UnicodeString(KATAKANA, -1, US_INV),
1003 temp,
1004 this, quick, legal);
1005 delete legal;
1006 }
1007
TestJamo()1008 void TransliteratorRoundTripTest::TestJamo() {
1009 RTTest t("Latin-Jamo");
1010 Legal *legal = new LegalJamo();
1011 t.test(UnicodeString("[a-zA-Z]", ""),
1012 UnicodeString("[\\u1100-\\u1112 \\u1161-\\u1175 \\u11A8-\\u11C2]",
1013 ""),
1014 NULL, this, quick, legal);
1015 delete legal;
1016 }
1017
TestHangul()1018 void TransliteratorRoundTripTest::TestHangul() {
1019 RTTest t("Latin-Hangul");
1020 Legal *legal = new Legal();
1021 if (quick) t.setPairLimit(1000);
1022 t.test(UnicodeString("[a-zA-Z]", ""),
1023 UnicodeString("[\\uAC00-\\uD7A4]", ""),
1024 NULL, this, quick, legal, 1);
1025 delete legal;
1026 }
1027
1028
1029 #define ASSERT_SUCCESS(status) {if (U_FAILURE(status)) { \
1030 errcheckln(status, "error at file %s, line %d, status = %s", __FILE__, __LINE__, \
1031 u_errorName(status)); \
1032 return;}}
1033
1034
writeStringInU8(FILE * out,const UnicodeString & s)1035 static void writeStringInU8(FILE *out, const UnicodeString &s) {
1036 int i;
1037 for (i=0; i<s.length(); i=s.moveIndex32(i, 1)) {
1038 UChar32 c = s.char32At(i);
1039 uint8_t bufForOneChar[10];
1040 UBool isError = FALSE;
1041 int32_t destIdx = 0;
1042 U8_APPEND(bufForOneChar, destIdx, (int32_t)sizeof(bufForOneChar), c, isError);
1043 fwrite(bufForOneChar, 1, destIdx, out);
1044 }
1045 }
1046
1047
1048
1049
TestHan()1050 void TransliteratorRoundTripTest::TestHan() {
1051 UErrorCode status = U_ZERO_ERROR;
1052
1053 // TODO: getExemplars() exists only as a C API, taking a USet.
1054 // The set API we need to use exists only on UnicodeSet, not on USet.
1055 // Do a hacky cast, knowing that a USet is really a UnicodeSet in
1056 // the implementation. Once USet gets the missing API, switch back
1057 // to using that.
1058 USet *USetExemplars = NULL;
1059 ULocaleData *uld = ulocdata_open("zh",&status);
1060 USetExemplars = uset_open(0, 0);
1061 USetExemplars = ulocdata_getExemplarSet(uld, USetExemplars, 0, ULOCDATA_ES_STANDARD, &status);
1062 ASSERT_SUCCESS(status);
1063 ulocdata_close(uld);
1064
1065 UnicodeString source;
1066 UChar32 c;
1067 int i;
1068 for (i=0; ;i++) {
1069 // Add all of the Chinese exemplar chars to the string "source".
1070 c = uset_charAt(USetExemplars, i);
1071 if (c == (UChar32)-1) {
1072 break;
1073 }
1074 source.append(c);
1075 }
1076
1077 // transform with Han translit
1078 Transliterator *hanTL = Transliterator::createInstance("Han-Latin", UTRANS_FORWARD, status);
1079 ASSERT_SUCCESS(status);
1080 UnicodeString target=source;
1081 hanTL->transliterate(target);
1082 // now verify that there are no Han characters left
1083 UnicodeSet allHan("[:han:]", status);
1084 ASSERT_SUCCESS(status);
1085 if (allHan.containsSome(target)) {
1086 errln("file %s, line %d, No Han must be left after Han-Latin transliteration",
1087 __FILE__, __LINE__);
1088 }
1089
1090 // check the pinyin translit
1091 Transliterator *pn = Transliterator::createInstance("Latin-NumericPinyin", UTRANS_FORWARD, status);
1092 ASSERT_SUCCESS(status);
1093 UnicodeString target2 = target;
1094 pn->transliterate(target2);
1095
1096 // verify that there are no marks
1097 Transliterator *nfd = Transliterator::createInstance("nfd", UTRANS_FORWARD, status);
1098 ASSERT_SUCCESS(status);
1099
1100 UnicodeString nfded = target2;
1101 nfd->transliterate(nfded);
1102 UnicodeSet allMarks(UNICODE_STRING_SIMPLE("[\\u0304\\u0301\\u030C\\u0300\\u0306]"), status); // look only for Pinyin tone marks, not all marks (there are some others in there)
1103 ASSERT_SUCCESS(status);
1104 assertFalse("NumericPinyin must contain no marks", allMarks.containsSome(nfded));
1105
1106 // verify roundtrip
1107 Transliterator *np = pn->createInverse(status);
1108 ASSERT_SUCCESS(status);
1109 UnicodeString target3 = target2;
1110 np->transliterate(target3);
1111 UBool roundtripOK = (target3.compare(target) == 0);
1112 assertTrue("NumericPinyin must roundtrip", roundtripOK);
1113 if (!roundtripOK) {
1114 const char *filename = "numeric-pinyin.log.txt";
1115 FILE *out = fopen(filename, "w");
1116 errln("Creating log file %s\n", filename);
1117 fprintf(out, "Pinyin: ");
1118 writeStringInU8(out, target);
1119 fprintf(out, "\nPinyin-Numeric-Pinyin: ");
1120 writeStringInU8(out, target2);
1121 fprintf(out, "\nNumeric-Pinyin-Pinyin: ");
1122 writeStringInU8(out, target3);
1123 fprintf(out, "\n");
1124 fclose(out);
1125 }
1126
1127 delete hanTL;
1128 delete pn;
1129 delete nfd;
1130 delete np;
1131 uset_close(USetExemplars);
1132 }
1133
1134
TestGreek()1135 void TransliteratorRoundTripTest::TestGreek() {
1136
1137 // CLDR bug #1911: This test should be moved into CLDR.
1138 // It is left in its current state as a regression test.
1139
1140 // if (isICUVersionAtLeast(ICU_39)) {
1141 // // We temporarily filter against Unicode 4.1, but we only do this
1142 // // before version 3.4.
1143 // errln("FAIL: TestGreek needs to be updated to remove delete the [:Age=4.0:] filter ");
1144 // return;
1145 // } else {
1146 // logln("Warning: TestGreek needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1147 // }
1148
1149 RTTest test("Latin-Greek");
1150 LegalGreek *legal = new LegalGreek(TRUE);
1151
1152 test.test(UnicodeString("[a-zA-Z]", ""),
1153 UnicodeString("[\\u003B\\u00B7[[:Greek:]&[:Letter:]]-["
1154 "\\u1D26-\\u1D2A" // L& [5] GREEK LETTER SMALL CAPITAL GAMMA..GREEK LETTER SMALL CAPITAL PSI
1155 "\\u1D5D-\\u1D61" // Lm [5] MODIFIER LETTER SMALL BETA..MODIFIER LETTER SMALL CHI
1156 "\\u1D66-\\u1D6A" // L& [5] GREEK SUBSCRIPT SMALL LETTER BETA..GREEK SUBSCRIPT SMALL LETTER CHI
1157 "\\u03D7-\\u03EF" // \N{GREEK KAI SYMBOL}..\N{COPTIC SMALL LETTER DEI}
1158 "] & [:Age=4.0:]]",
1159
1160 //UnicodeString("[[\\u003B\\u00B7[:Greek:]-[\\u0374\\u0385\\u1fcd\\u1fce\\u1fdd\\u1fde\\u1fed-\\u1fef\\u1ffd\\u03D7-\\u03EF]]&[:Age=3.2:]]",
1161 ""),
1162 "[\\u00B5\\u037A\\u03D0-\\u03F5\\u03f9]", /* exclusions */
1163 this, quick, legal, 50);
1164
1165
1166 delete legal;
1167 }
1168
1169
TestGreekUNGEGN()1170 void TransliteratorRoundTripTest::TestGreekUNGEGN() {
1171
1172 // CLDR bug #1911: This test should be moved into CLDR.
1173 // It is left in its current state as a regression test.
1174
1175 // if (isICUVersionAtLeast(ICU_39)) {
1176 // // We temporarily filter against Unicode 4.1, but we only do this
1177 // // before version 3.4.
1178 // errln("FAIL: TestGreek needs to be updated to remove delete the [:Age=4.0:] filter ");
1179 // return;
1180 // } else {
1181 // logln("Warning: TestGreek needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1182 // }
1183
1184 RTTest test("Latin-Greek/UNGEGN");
1185 LegalGreek *legal = new LegalGreek(FALSE);
1186
1187 test.test(UnicodeString("[a-zA-Z]", ""),
1188 UnicodeString("[\\u003B\\u00B7[[:Greek:]&[:Letter:]]-["
1189 "\\u1D26-\\u1D2A" // L& [5] GREEK LETTER SMALL CAPITAL GAMMA..GREEK LETTER SMALL CAPITAL PSI
1190 "\\u1D5D-\\u1D61" // Lm [5] MODIFIER LETTER SMALL BETA..MODIFIER LETTER SMALL CHI
1191 "\\u1D66-\\u1D6A" // L& [5] GREEK SUBSCRIPT SMALL LETTER BETA..GREEK SUBSCRIPT SMALL LETTER CHI
1192 "\\u03D7-\\u03EF" // \N{GREEK KAI SYMBOL}..\N{COPTIC SMALL LETTER DEI}
1193 "] & [:Age=4.0:]]",
1194 //UnicodeString("[[\\u003B\\u00B7[:Greek:]-[\\u0374\\u0385\\u1fce\\u1fde\\u03D7-\\u03EF]]&[:Age=3.2:]]",
1195 ""),
1196 "[\\u0385\\u00B5\\u037A\\u03D0-\\uFFFF {\\u039C\\u03C0}]", /* roundtrip exclusions */
1197 this, quick, legal);
1198
1199 delete legal;
1200 }
1201
Testel()1202 void TransliteratorRoundTripTest::Testel() {
1203
1204 // CLDR bug #1911: This test should be moved into CLDR.
1205 // It is left in its current state as a regression test.
1206
1207 // if (isICUVersionAtLeast(ICU_39)) {
1208 // // We temporarily filter against Unicode 4.1, but we only do this
1209 // // before version 3.4.
1210 // errln("FAIL: TestGreek needs to be updated to remove delete the [:Age=4.0:] filter ");
1211 // return;
1212 // } else {
1213 // logln("Warning: TestGreek needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1214 // }
1215
1216 RTTest test("Latin-el");
1217 LegalGreek *legal = new LegalGreek(FALSE);
1218
1219 test.test(UnicodeString("[a-zA-Z]", ""),
1220 UnicodeString("[\\u003B\\u00B7[[:Greek:]&[:Letter:]]-["
1221 "\\u1D26-\\u1D2A" // L& [5] GREEK LETTER SMALL CAPITAL GAMMA..GREEK LETTER SMALL CAPITAL PSI
1222 "\\u1D5D-\\u1D61" // Lm [5] MODIFIER LETTER SMALL BETA..MODIFIER LETTER SMALL CHI
1223 "\\u1D66-\\u1D6A" // L& [5] GREEK SUBSCRIPT SMALL LETTER BETA..GREEK SUBSCRIPT SMALL LETTER CHI
1224 "\\u03D7-\\u03EF" // \N{GREEK KAI SYMBOL}..\N{COPTIC SMALL LETTER DEI}
1225 "] & [:Age=4.0:]]",
1226 //UnicodeString("[[\\u003B\\u00B7[:Greek:]-[\\u0374\\u0385\\u1fce\\u1fde\\u03D7-\\u03EF]]&[:Age=3.2:]]",
1227 ""),
1228 "[\\u00B5\\u037A\\u03D0-\\uFFFF {\\u039C\\u03C0}]", /* exclusions */
1229 this, quick, legal);
1230
1231
1232 delete legal;
1233 }
1234
1235
TestArabic()1236 void TransliteratorRoundTripTest::TestArabic() {
1237 UnicodeString ARABIC("[\\u060C\\u061B\\u061F\\u0621\\u0627-\\u063A\\u0641-\\u0655\\u0660-\\u066C\\u067E\\u0686\\u0698\\u06A4\\u06AD\\u06AF\\u06CB-\\u06CC\\u06F0-\\u06F9]", -1, US_INV);
1238 Legal *legal = new Legal();
1239 RTTest test("Latin-Arabic");
1240 test.test(UNICODE_STRING_SIMPLE("[a-zA-Z\\u02BE\\u02BF\\u207F]"), ARABIC, "[a-zA-Z\\u02BE\\u02BF\\u207F]",this, quick, legal); //
1241 delete legal;
1242 }
1243 class LegalHebrew : public Legal {
1244 private:
1245 UnicodeSet FINAL;
1246 UnicodeSet NON_FINAL;
1247 UnicodeSet LETTER;
1248 public:
1249 LegalHebrew(UErrorCode& error);
~LegalHebrew()1250 virtual ~LegalHebrew() {}
1251 virtual UBool is(const UnicodeString& sourceString) const;
1252 };
1253
LegalHebrew(UErrorCode & error)1254 LegalHebrew::LegalHebrew(UErrorCode& error){
1255 FINAL.applyPattern(UNICODE_STRING_SIMPLE("[\\u05DA\\u05DD\\u05DF\\u05E3\\u05E5]"), error);
1256 NON_FINAL.applyPattern(UNICODE_STRING_SIMPLE("[\\u05DB\\u05DE\\u05E0\\u05E4\\u05E6]"), error);
1257 LETTER.applyPattern("[:letter:]", error);
1258 }
is(const UnicodeString & sourceString) const1259 UBool LegalHebrew::is(const UnicodeString& sourceString)const{
1260
1261 if (sourceString.length() == 0) return TRUE;
1262 // don't worry about surrogates.
1263 for (int i = 0; i < sourceString.length(); ++i) {
1264 UChar ch = sourceString.charAt(i);
1265 UChar next = i+1 == sourceString.length() ? 0x0000 : sourceString.charAt(i);
1266 if (FINAL.contains(ch)) {
1267 if (LETTER.contains(next)) return FALSE;
1268 } else if (NON_FINAL.contains(ch)) {
1269 if (!LETTER.contains(next)) return FALSE;
1270 }
1271 }
1272 return TRUE;
1273 }
TestHebrew()1274 void TransliteratorRoundTripTest::TestHebrew() {
1275 // CLDR bug #1911: This test should be moved into CLDR.
1276 // It is left in its current state as a regression test.
1277 // if (isICUVersionAtLeast(ICU_39)) {
1278 // // We temporarily filter against Unicode 4.1, but we only do this
1279 // // before version 3.4.
1280 // errln("FAIL: TestHebrew needs to be updated to remove delete the [:Age=4.0:] filter ");
1281 // return;
1282 // } else {
1283 // logln("Warning: TestHebrew needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1284 // }
1285 //long start = System.currentTimeMillis();
1286 UErrorCode error = U_ZERO_ERROR;
1287 LegalHebrew* legal = new LegalHebrew(error);
1288 if(U_FAILURE(error)){
1289 dataerrln("Could not construct LegalHebrew object. Error: %s", u_errorName(error));
1290 return;
1291 }
1292 RTTest test("Latin-Hebrew");
1293 test.test(UNICODE_STRING_SIMPLE("[a-zA-Z\\u02BC\\u02BB]"), UNICODE_STRING_SIMPLE("[[[:hebrew:]-[\\u05BD\\uFB00-\\uFBFF]]&[:Age=4.0:]]"), "[\\u05F0\\u05F1\\u05F2]", this, quick, legal);
1294
1295 //showElapsed(start, "TestHebrew");
1296 delete legal;
1297 }
TestCyrillic()1298 void TransliteratorRoundTripTest::TestCyrillic() {
1299 RTTest test("Latin-Cyrillic");
1300 Legal *legal = new Legal();
1301
1302 test.test(UnicodeString("[a-zA-Z\\u0110\\u0111\\u02BA\\u02B9]", ""),
1303 UnicodeString("[[\\u0400-\\u045F] & [:Age=3.2:]]", ""), NULL, this, quick,
1304 legal);
1305
1306 delete legal;
1307 }
1308
1309
1310 // Inter-Indic Tests ----------------------------------
1311 class LegalIndic :public Legal{
1312 UnicodeSet vowelSignSet;
1313 UnicodeSet avagraha;
1314 UnicodeSet nukta;
1315 UnicodeSet virama;
1316 UnicodeSet sanskritStressSigns;
1317 UnicodeSet chandrabindu;
1318
1319 public:
1320 LegalIndic();
1321 virtual UBool is(const UnicodeString& sourceString) const;
~LegalIndic()1322 virtual ~LegalIndic() {};
1323 };
is(const UnicodeString & sourceString) const1324 UBool LegalIndic::is(const UnicodeString& sourceString) const{
1325 int cp=sourceString.charAt(0);
1326
1327 // A vowel sign cannot be the first char
1328 if(vowelSignSet.contains(cp)){
1329 return FALSE;
1330 }else if(avagraha.contains(cp)){
1331 return FALSE;
1332 }else if(virama.contains(cp)){
1333 return FALSE;
1334 }else if(nukta.contains(cp)){
1335 return FALSE;
1336 }else if(sanskritStressSigns.contains(cp)){
1337 return FALSE;
1338 }else if(chandrabindu.contains(cp) &&
1339 ((sourceString.length()>1) &&
1340 vowelSignSet.contains(sourceString.charAt(1)))){
1341 return FALSE;
1342 }
1343 return TRUE;
1344 }
LegalIndic()1345 LegalIndic::LegalIndic(){
1346 UErrorCode status = U_ZERO_ERROR;
1347 vowelSignSet.addAll( UnicodeSet("[\\u0902\\u0903\\u0904\\u093e-\\u094c\\u0962\\u0963]",status));/* Devanagari */
1348 vowelSignSet.addAll( UnicodeSet("[\\u0982\\u0983\\u09be-\\u09cc\\u09e2\\u09e3\\u09D7]",status));/* Bengali */
1349 vowelSignSet.addAll( UnicodeSet("[\\u0a02\\u0a03\\u0a3e-\\u0a4c\\u0a62\\u0a63\\u0a70\\u0a71]",status));/* Gurmukhi */
1350 vowelSignSet.addAll( UnicodeSet("[\\u0a82\\u0a83\\u0abe-\\u0acc\\u0ae2\\u0ae3]",status));/* Gujarati */
1351 vowelSignSet.addAll( UnicodeSet("[\\u0b02\\u0b03\\u0b3e-\\u0b4c\\u0b62\\u0b63\\u0b56\\u0b57]",status));/* Oriya */
1352 vowelSignSet.addAll( UnicodeSet("[\\u0b82\\u0b83\\u0bbe-\\u0bcc\\u0be2\\u0be3\\u0bd7]",status));/* Tamil */
1353 vowelSignSet.addAll( UnicodeSet("[\\u0c02\\u0c03\\u0c3e-\\u0c4c\\u0c62\\u0c63\\u0c55\\u0c56]",status));/* Telugu */
1354 vowelSignSet.addAll( UnicodeSet("[\\u0c82\\u0c83\\u0cbe-\\u0ccc\\u0ce2\\u0ce3\\u0cd5\\u0cd6]",status));/* Kannada */
1355 vowelSignSet.addAll( UnicodeSet("[\\u0d02\\u0d03\\u0d3e-\\u0d4c\\u0d62\\u0d63\\u0d57]",status));/* Malayalam */
1356
1357 avagraha.addAll(UnicodeSet("[\\u093d\\u09bd\\u0abd\\u0b3d\\u0cbd]",status));
1358 nukta.addAll(UnicodeSet("[\\u093c\\u09bc\\u0a3c\\u0abc\\u0b3c\\u0cbc]",status));
1359 virama.addAll(UnicodeSet("[\\u094d\\u09cd\\u0a4d\\u0acd\\u0b4d\\u0bcd\\u0c4d\\u0ccd\\u0d4d]",status));
1360 sanskritStressSigns.addAll(UnicodeSet("[\\u0951\\u0952\\u0953\\u0954\\u097d]",status));
1361 chandrabindu.addAll(UnicodeSet("[\\u0901\\u0981\\u0A81\\u0b01\\u0c01]",status));
1362
1363 }
1364
1365 static const char latinForIndic[] = "[['.0-9A-Za-z~\\u00C0-\\u00C5\\u00C7-\\u00CF\\u00D1-\\u00D6\\u00D9-\\u00DD"
1366 "\\u00E0-\\u00E5\\u00E7-\\u00EF\\u00F1-\\u00F6\\u00F9-\\u00FD\\u00FF-\\u010F"
1367 "\\u0112-\\u0125\\u0128-\\u0130\\u0134-\\u0137\\u0139-\\u013E\\u0143-\\u0148"
1368 "\\u014C-\\u0151\\u0154-\\u0165\\u0168-\\u017E\\u01A0-\\u01A1\\u01AF-\\u01B0"
1369 "\\u01CD-\\u01DC\\u01DE-\\u01E3\\u01E6-\\u01ED\\u01F0\\u01F4-\\u01F5\\u01F8-\\u01FB"
1370 "\\u0200-\\u021B\\u021E-\\u021F\\u0226-\\u0233\\u0294\\u0303-\\u0304\\u0306\\u0314-\\u0315"
1371 "\\u0325\\u040E\\u0419\\u0439\\u045E\\u04C1-\\u04C2\\u04D0-\\u04D1\\u04D6-\\u04D7"
1372 "\\u04E2-\\u04E3\\u04EE-\\u04EF\\u1E00-\\u1E99\\u1EA0-\\u1EF9\\u1F01\\u1F03\\u1F05"
1373 "\\u1F07\\u1F09\\u1F0B\\u1F0D\\u1F0F\\u1F11\\u1F13\\u1F15\\u1F19\\u1F1B\\u1F1D\\u1F21"
1374 "\\u1F23\\u1F25\\u1F27\\u1F29\\u1F2B\\u1F2D\\u1F2F\\u1F31\\u1F33\\u1F35\\u1F37\\u1F39"
1375 "\\u1F3B\\u1F3D\\u1F3F\\u1F41\\u1F43\\u1F45\\u1F49\\u1F4B\\u1F4D\\u1F51\\u1F53\\u1F55"
1376 "\\u1F57\\u1F59\\u1F5B\\u1F5D\\u1F5F\\u1F61\\u1F63\\u1F65\\u1F67\\u1F69\\u1F6B\\u1F6D"
1377 "\\u1F6F\\u1F81\\u1F83\\u1F85\\u1F87\\u1F89\\u1F8B\\u1F8D\\u1F8F\\u1F91\\u1F93\\u1F95"
1378 "\\u1F97\\u1F99\\u1F9B\\u1F9D\\u1F9F\\u1FA1\\u1FA3\\u1FA5\\u1FA7\\u1FA9\\u1FAB\\u1FAD"
1379 "\\u1FAF-\\u1FB1\\u1FB8-\\u1FB9\\u1FD0-\\u1FD1\\u1FD8-\\u1FD9\\u1FE0-\\u1FE1\\u1FE5"
1380 "\\u1FE8-\\u1FE9\\u1FEC\\u212A-\\u212B\\uE04D\\uE064]"
1381 "-[\\uE000-\\uE080 \\u01E2\\u01E3]& [[:latin:][:mark:]]]";
1382
TestDevanagariLatin()1383 void TransliteratorRoundTripTest::TestDevanagariLatin() {
1384 {
1385 UErrorCode status = U_ZERO_ERROR;
1386 UParseError parseError;
1387 TransliteratorPointer t1(Transliterator::createInstance("[\\u0964-\\u0965\\u0981-\\u0983\\u0985-\\u098C\\u098F-\\u0990\\u0993-\\u09A8\\u09AA-\\u09B0\\u09B2\\u09B6-\\u09B9\\u09BC\\u09BE-\\u09C4\\u09C7-\\u09C8\\u09CB-\\u09CD\\u09D7\\u09DC-\\u09DD\\u09DF-\\u09E3\\u09E6-\\u09FA];NFD;Bengali-InterIndic;InterIndic-Gujarati;NFC;",UTRANS_FORWARD, parseError, status));
1388 if((Transliterator *)t1 != NULL){
1389 TransliteratorPointer t2(t1->createInverse(status));
1390 if(U_FAILURE(status)){
1391 errln("FAIL: could not create the Inverse:-( \n");
1392 }
1393 }else {
1394 dataerrln("FAIL: could not create the transliterator. Error: %s\n", u_errorName(status));
1395 }
1396
1397 }
1398 RTTest test("Latin-Devanagari");
1399 Legal *legal = new LegalIndic();
1400 // CLDR bug #1911: This test should be moved into CLDR.
1401 // It is left in its current state as a regression test.
1402 // if (isICUVersionAtLeast(ICU_39)) {
1403 // // We temporarily filter against Unicode 4.1, but we only do this
1404 // // before version 3.4.
1405 // errln("FAIL: TestDevanagariLatin needs to be updated to remove delete the [:Age=4.1:] filter ");
1406 // return;
1407 // } else {
1408 // logln("Warning: TestDevanagariLatin needs to be updated to remove delete the section marked [:Age=4.1:] filter");
1409 // }
1410 test.test(UnicodeString(latinForIndic, ""),
1411 UnicodeString("[[[:Devanagari:][\\u094d][\\u0964\\u0965]]&[:Age=4.1:]]", ""), "[\\u0965\\u0904]", this, quick,
1412 legal, 50);
1413
1414 delete legal;
1415 }
1416
1417 /* Defined this way for HP/UX11CC :-( */
1418 static const int32_t INTER_INDIC_ARRAY_WIDTH = 4;
1419 static const char * const interIndicArray[] = {
1420
1421 "BENGALI-DEVANAGARI", "[:BENGALI:]", "[:Devanagari:]",
1422 "[\\u0904\\u0951-\\u0954\\u0943-\\u0949\\u094a\\u0962\\u0963\\u090D\\u090e\\u0911\\u0912\\u0929\\u0933\\u0934\\u0935\\u093d\\u0950\\u0958\\u0959\\u095a\\u095b\\u095e\\u097d]", /*roundtrip exclusions*/
1423
1424 "DEVANAGARI-BENGALI", "[:Devanagari:]", "[:BENGALI:]",
1425 "[\\u0951-\\u0954\\u0951-\\u0954\\u09D7\\u090D\\u090e\\u0911\\u0912\\u0929\\u0933\\u0934\\u0935\\u093d\\u0950\\u0958\\u0959\\u095a\\u095b\\u095e\\u09f0\\u09f1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1426
1427 "GURMUKHI-DEVANAGARI", "[:GURMUKHI:]", "[:Devanagari:]",
1428 "[\\u0904\\u0901\\u0902\\u0936\\u0933\\u0951-\\u0954\\u0902\\u0903\\u0943-\\u0949\\u094a\\u0962\\u0963\\u090B\\u090C\\u090D\\u090e\\u0911\\u0912\\u0934\\u0937\\u093D\\u0950\\u0960\\u0961\\u097d]", /*roundtrip exclusions*/
1429
1430 "DEVANAGARI-GURMUKHI", "[:Devanagari:]", "[:GURMUKHI:]",
1431 "[\\u0904\\u0A02\\u0946\\u0A5C\\u0951-\\u0954\\u0A70\\u0A71\\u090B\\u090C\\u090D\\u090e\\u0911\\u0912\\u0934\\u0937\\u093D\\u0950\\u0960\\u0961\\u0a72\\u0a73\\u0a74]", /*roundtrip exclusions*/
1432
1433 "GUJARATI-DEVANAGARI", "[:GUJARATI:]", "[:Devanagari:]",
1434 "[\\u0946\\u094A\\u0962\\u0963\\u0951-\\u0954\\u0961\\u090c\\u090e\\u0912\\u097d]", /*roundtrip exclusions*/
1435
1436 "DEVANAGARI-GUJARATI", "[:Devanagari:]", "[:GUJARATI:]",
1437 "[\\u0951-\\u0954\\u0961\\u090c\\u090e\\u0912]", /*roundtrip exclusions*/
1438
1439 "ORIYA-DEVANAGARI", "[:ORIYA:]", "[:Devanagari:]",
1440 "[\\u0904\\u0943-\\u094a\\u0962\\u0963\\u0951-\\u0954\\u0950\\u090D\\u090e\\u0912\\u0911\\u0931\\u0935\\u097d]", /*roundtrip exclusions*/
1441
1442 "DEVANAGARI-ORIYA", "[:Devanagari:]", "[:ORIYA:]",
1443 "[\\u0b5f\\u0b56\\u0b57\\u0b70\\u0b71\\u0950\\u090D\\u090e\\u0912\\u0911\\u0931]", /*roundtrip exclusions*/
1444
1445 "Tamil-DEVANAGARI", "[:tamil:]", "[:Devanagari:]",
1446 "[\\u0901\\u0904\\u093c\\u0943-\\u094a\\u0951-\\u0954\\u0962\\u0963\\u090B\\u090C\\u090D\\u0911\\u0916\\u0917\\u0918\\u091B\\u091D\\u0920\\u0921\\u0922\\u0925\\u0926\\u0927\\u092B\\u092C\\u092D\\u0936\\u093d\\u0950[\\u0958-\\u0961]\\u097d]", /*roundtrip exclusions*/
1447
1448 "DEVANAGARI-Tamil", "[:Devanagari:]", "[:tamil:]",
1449 "[\\u0bd7\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1450
1451 "Telugu-DEVANAGARI", "[:telugu:]", "[:Devanagari:]",
1452 "[\\u0904\\u093c\\u0950\\u0945\\u0949\\u0951-\\u0954\\u0962\\u0963\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]\\u097d]", /*roundtrip exclusions*/
1453
1454 "DEVANAGARI-TELUGU", "[:Devanagari:]", "[:TELUGU:]",
1455 "[\\u0c55\\u0c56\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]]", /*roundtrip exclusions*/
1456
1457 "KANNADA-DEVANAGARI", "[:KANNADA:]", "[:Devanagari:]",
1458 "[\\u0901\\u0904\\u0946\\u093c\\u0950\\u0945\\u0949\\u0951-\\u0954\\u0962\\u0963\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]\\u097d]", /*roundtrip exclusions*/
1459
1460 "DEVANAGARI-KANNADA", "[:Devanagari:]", "[:KANNADA:]",
1461 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cde\\u0cd5\\u0cd6\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]]", /*roundtrip exclusions*/
1462
1463 "MALAYALAM-DEVANAGARI", "[:MALAYALAM:]", "[:Devanagari:]",
1464 "[\\u0901\\u0904\\u094a\\u094b\\u094c\\u093c\\u0950\\u0944\\u0945\\u0949\\u0951-\\u0954\\u0962\\u0963\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]\\u097d]", /*roundtrip exclusions*/
1465
1466 "DEVANAGARI-MALAYALAM", "[:Devanagari:]", "[:MALAYALAM:]",
1467 "[\\u0d4c\\u0d57\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]]", /*roundtrip exclusions*/
1468
1469 "GURMUKHI-BENGALI", "[:GURMUKHI:]", "[:BENGALI:]",
1470 "[\\u0981\\u0982\\u09b6\\u09e2\\u09e3\\u09c3\\u09c4\\u09d7\\u098B\\u098C\\u09B7\\u09E0\\u09E1\\u09F0\\u09F1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1471
1472 "BENGALI-GURMUKHI", "[:BENGALI:]", "[:GURMUKHI:]",
1473 "[\\u0A02\\u0a5c\\u0a47\\u0a70\\u0a71\\u0A33\\u0A35\\u0A59\\u0A5A\\u0A5B\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1474
1475 "GUJARATI-BENGALI", "[:GUJARATI:]", "[:BENGALI:]",
1476 "[\\u09d7\\u09e2\\u09e3\\u098c\\u09e1\\u09f0\\u09f1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1477
1478 "BENGALI-GUJARATI", "[:BENGALI:]", "[:GUJARATI:]",
1479 "[\\u0A82\\u0a83\\u0Ac9\\u0Ac5\\u0ac7\\u0A8D\\u0A91\\u0AB3\\u0AB5\\u0ABD\\u0AD0]", /*roundtrip exclusions*/
1480
1481 "ORIYA-BENGALI", "[:ORIYA:]", "[:BENGALI:]",
1482 "[\\u09c4\\u09e2\\u09e3\\u09f0\\u09f1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1483
1484 "BENGALI-ORIYA", "[:BENGALI:]", "[:ORIYA:]",
1485 "[\\u0b35\\u0b71\\u0b5f\\u0b56\\u0b33\\u0b3d]", /*roundtrip exclusions*/
1486
1487 "Tamil-BENGALI", "[:tamil:]", "[:BENGALI:]",
1488 "[\\u0981\\u09bc\\u09c3\\u09c4\\u09e2\\u09e3\\u09f0\\u09f1\\u098B\\u098C\\u0996\\u0997\\u0998\\u099B\\u099D\\u09A0\\u09A1\\u09A2\\u09A5\\u09A6\\u09A7\\u09AB\\u09AC\\u09AD\\u09B6\\u09DC\\u09DD\\u09DF\\u09E0\\u09E1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1489
1490 "BENGALI-Tamil", "[:BENGALI:]", "[:tamil:]",
1491 "[\\u0bc6\\u0bc7\\u0bca\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB3\\u0BB4\\u0BB5\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1492
1493 "Telugu-BENGALI", "[:telugu:]", "[:BENGALI:]",
1494 "[\\u09e2\\u09e3\\u09bc\\u09d7\\u09f0\\u09f1\\u09dc\\u09dd\\u09df\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1495
1496 "BENGALI-TELUGU", "[:BENGALI:]", "[:TELUGU:]",
1497 "[\\u0c55\\u0c56\\u0c47\\u0c46\\u0c4a\\u0C0E\\u0C12\\u0C31\\u0C33\\u0C35]", /*roundtrip exclusions*/
1498
1499 "KANNADA-BENGALI", "[:KANNADA:]", "[:BENGALI:]",
1500 "[\\u0981\\u09e2\\u09e3\\u09bc\\u09d7\\u09dc\\u09dd\\u09df\\u09f0\\u09f1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1501
1502 "BENGALI-KANNADA", "[:BENGALI:]", "[:KANNADA:]",
1503 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cc6\\u0cca\\u0cd5\\u0cd6\\u0cc7\\u0C8E\\u0C92\\u0CB1\\u0cb3\\u0cb5\\u0cde]", /*roundtrip exclusions*/
1504
1505 "MALAYALAM-BENGALI", "[:MALAYALAM:]", "[:BENGALI:]",
1506 "[\\u0981\\u09e2\\u09e3\\u09bc\\u09c4\\u09f0\\u09f1\\u09dc\\u09dd\\u09df\\u09dc\\u09dd\\u09df\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1507
1508 "BENGALI-MALAYALAM", "[:BENGALI:]", "[:MALAYALAM:]",
1509 "[\\u0d46\\u0d4a\\u0d47\\u0d31-\\u0d35\\u0d0e\\u0d12]", /*roundtrip exclusions*/
1510
1511 "GUJARATI-GURMUKHI", "[:GUJARATI:]", "[:GURMUKHI:]",
1512 "[\\u0A02\\u0ab3\\u0ab6\\u0A70\\u0a71\\u0a82\\u0a83\\u0ac3\\u0ac4\\u0ac5\\u0ac9\\u0a5c\\u0a72\\u0a73\\u0a74\\u0a8b\\u0a8d\\u0a91\\u0abd]", /*roundtrip exclusions*/
1513
1514 "GURMUKHI-GUJARATI", "[:GURMUKHI:]", "[:GUJARATI:]",
1515 "[\\u0a5c\\u0A70\\u0a71\\u0a72\\u0a73\\u0a74\\u0a82\\u0a83\\u0a8b\\u0a8c\\u0a8d\\u0a91\\u0ab3\\u0ab6\\u0ab7\\u0abd\\u0ac3\\u0ac4\\u0ac5\\u0ac9\\u0ad0\\u0ae0\\u0ae1]", /*roundtrip exclusions*/
1516
1517 "ORIYA-GURMUKHI", "[:ORIYA:]", "[:GURMUKHI:]",
1518 "[\\u0A01\\u0A02\\u0a5c\\u0a21\\u0a47\\u0a71\\u0b02\\u0b03\\u0b33\\u0b36\\u0b43\\u0b56\\u0b57\\u0B0B\\u0B0C\\u0B37\\u0B3D\\u0B5F\\u0B60\\u0B61\\u0a35\\u0a72\\u0a73\\u0a74]", /*roundtrip exclusions*/
1519
1520 "GURMUKHI-ORIYA", "[:GURMUKHI:]", "[:ORIYA:]",
1521 "[\\u0b01\\u0b02\\u0b03\\u0b33\\u0b36\\u0b43\\u0b56\\u0b57\\u0B0B\\u0B0C\\u0B37\\u0B3D\\u0B5F\\u0B60\\u0B61\\u0b70\\u0b71]", /*roundtrip exclusions*/
1522
1523 "TAMIL-GURMUKHI", "[:TAMIL:]", "[:GURMUKHI:]",
1524 "[\\u0A01\\u0A02\\u0a33\\u0a36\\u0a3c\\u0a70\\u0a71\\u0a47\\u0A16\\u0A17\\u0A18\\u0A1B\\u0A1D\\u0A20\\u0A21\\u0A22\\u0A25\\u0A26\\u0A27\\u0A2B\\u0A2C\\u0A2D\\u0A59\\u0A5A\\u0A5B\\u0A5C\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1525
1526 "GURMUKHI-TAMIL", "[:GURMUKHI:]", "[:TAMIL:]",
1527 "[\\u0b82\\u0bc6\\u0bca\\u0bd7\\u0bb7\\u0bb3\\u0b83\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB4\\u0bb6\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1528
1529 "TELUGU-GURMUKHI", "[:TELUGU:]", "[:GURMUKHI:]",
1530 "[\\u0A02\\u0a33\\u0a36\\u0a3c\\u0a70\\u0a71\\u0A59\\u0A5A\\u0A5B\\u0A5C\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1531
1532 "GURMUKHI-TELUGU", "[:GURMUKHI:]", "[:TELUGU:]",
1533 "[\\u0c01\\u0c02\\u0c03\\u0c33\\u0c36\\u0c44\\u0c43\\u0c46\\u0c4a\\u0c56\\u0c55\\u0C0B\\u0C0C\\u0C0E\\u0C12\\u0C31\\u0C37\\u0C60\\u0C61]", /*roundtrip exclusions*/
1534
1535 "KANNADA-GURMUKHI", "[:KANNADA:]", "[:GURMUKHI:]",
1536 "[\\u0A01\\u0A02\\u0a33\\u0a36\\u0a3c\\u0a70\\u0a71\\u0A59\\u0A5A\\u0A5B\\u0A5C\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1537
1538 "GURMUKHI-KANNADA", "[:GURMUKHI:]", "[:KANNADA:]",
1539 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0c82\\u0c83\\u0cb3\\u0cb6\\u0cc4\\u0cc3\\u0cc6\\u0cca\\u0cd5\\u0cd6\\u0C8B\\u0C8C\\u0C8E\\u0C92\\u0CB1\\u0CB7\\u0cbd\\u0CE0\\u0CE1\\u0cde]", /*roundtrip exclusions*/
1540
1541 "MALAYALAM-GURMUKHI", "[:MALAYALAM:]", "[:GURMUKHI:]",
1542 "[\\u0A01\\u0A02\\u0a4b\\u0a4c\\u0a33\\u0a36\\u0a3c\\u0a70\\u0a71\\u0A59\\u0A5A\\u0A5B\\u0A5C\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1543
1544 "GURMUKHI-MALAYALAM", "[:GURMUKHI:]", "[:MALAYALAM:]",
1545 "[\\u0d02\\u0d03\\u0d33\\u0d36\\u0d43\\u0d46\\u0d4a\\u0d4c\\u0d57\\u0D0B\\u0D0C\\u0D0E\\u0D12\\u0D31\\u0D34\\u0D37\\u0D60\\u0D61]", /*roundtrip exclusions*/
1546
1547 "GUJARATI-ORIYA", "[:GUJARATI:]", "[:ORIYA:]",
1548 "[\\u0b56\\u0b57\\u0B0C\\u0B5F\\u0B61\\u0b70\\u0b71]", /*roundtrip exclusions*/
1549
1550 "ORIYA-GUJARATI", "[:ORIYA:]", "[:GUJARATI:]",
1551 "[\\u0Ac4\\u0Ac5\\u0Ac9\\u0Ac7\\u0A8D\\u0A91\\u0AB5\\u0Ad0]", /*roundtrip exclusions*/
1552
1553 "TAMIL-GUJARATI", "[:TAMIL:]", "[:GUJARATI:]",
1554 "[\\u0A81\\u0a8c\\u0abc\\u0ac3\\u0Ac4\\u0Ac5\\u0Ac9\\u0Ac7\\u0A8B\\u0A8D\\u0A91\\u0A96\\u0A97\\u0A98\\u0A9B\\u0A9D\\u0AA0\\u0AA1\\u0AA2\\u0AA5\\u0AA6\\u0AA7\\u0AAB\\u0AAC\\u0AAD\\u0AB6\\u0ABD\\u0AD0\\u0AE0\\u0AE1]", /*roundtrip exclusions*/
1555
1556 "GUJARATI-TAMIL", "[:GUJARATI:]", "[:TAMIL:]",
1557 "[\\u0Bc6\\u0Bca\\u0Bd7\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB4\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1558
1559 "TELUGU-GUJARATI", "[:TELUGU:]", "[:GUJARATI:]",
1560 "[\\u0abc\\u0Ac5\\u0Ac9\\u0A8D\\u0A91\\u0ABD\\u0Ad0]", /*roundtrip exclusions*/
1561
1562 "GUJARATI-TELUGU", "[:GUJARATI:]", "[:TELUGU:]",
1563 "[\\u0c46\\u0c4a\\u0c55\\u0c56\\u0C0C\\u0C0E\\u0C12\\u0C31\\u0C61]", /*roundtrip exclusions*/
1564
1565 "KANNADA-GUJARATI", "[:KANNADA:]", "[:GUJARATI:]",
1566 "[\\u0A81\\u0abc\\u0Ac5\\u0Ac9\\u0A8D\\u0A91\\u0ABD\\u0Ad0]", /*roundtrip exclusions*/
1567
1568 "GUJARATI-KANNADA", "[:GUJARATI:]", "[:KANNADA:]",
1569 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cc6\\u0cca\\u0cd5\\u0cd6\\u0C8C\\u0C8E\\u0C92\\u0CB1\\u0CDE\\u0CE1]", /*roundtrip exclusions*/
1570
1571 "MALAYALAM-GUJARATI", "[:MALAYALAM:]", "[:GUJARATI:]",
1572 "[\\u0A81\\u0ac4\\u0acb\\u0acc\\u0abc\\u0Ac5\\u0Ac9\\u0A8D\\u0A91\\u0ABD\\u0Ad0]", /*roundtrip exclusions*/
1573
1574 "GUJARATI-MALAYALAM", "[:GUJARATI:]", "[:MALAYALAM:]",
1575 "[\\u0d46\\u0d4a\\u0d4c\\u0d55\\u0d57\\u0D0C\\u0D0E\\u0D12\\u0D31\\u0D34\\u0D61]", /*roundtrip exclusions*/
1576
1577 "TAMIL-ORIYA", "[:TAMIL:]", "[:ORIYA:]",
1578 "[\\u0B01\\u0b3c\\u0b43\\u0b56\\u0B0B\\u0B0C\\u0B16\\u0B17\\u0B18\\u0B1B\\u0B1D\\u0B20\\u0B21\\u0B22\\u0B25\\u0B26\\u0B27\\u0B2B\\u0B2C\\u0B2D\\u0B36\\u0B3D\\u0B5C\\u0B5D\\u0B5F\\u0B60\\u0B61\\u0b70\\u0b71]", /*roundtrip exclusions*/
1579
1580 "ORIYA-TAMIL", "[:ORIYA:]", "[:TAMIL:]",
1581 "[\\u0bc6\\u0bca\\u0bc7\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB4\\u0BB5\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1582
1583 "TELUGU-ORIYA", "[:TELUGU:]", "[:ORIYA:]",
1584 "[\\u0b3c\\u0b57\\u0b56\\u0B3D\\u0B5C\\u0B5D\\u0B5F\\u0b70\\u0b71]", /*roundtrip exclusions*/
1585
1586 "ORIYA-TELUGU", "[:ORIYA:]", "[:TELUGU:]",
1587 "[\\u0c44\\u0c46\\u0c4a\\u0c55\\u0c47\\u0C0E\\u0C12\\u0C31\\u0C35]", /*roundtrip exclusions*/
1588
1589 "KANNADA-ORIYA", "[:KANNADA:]", "[:ORIYA:]",
1590 "[\\u0B01\\u0b3c\\u0b57\\u0B3D\\u0B5C\\u0B5D\\u0B5F\\u0b70\\u0b71]", /*roundtrip exclusions*/
1591
1592 "ORIYA-KANNADA", "[:ORIYA:]", "[:KANNADA:]",
1593 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cc4\\u0cc6\\u0cca\\u0cd5\\u0cc7\\u0C8E\\u0C92\\u0CB1\\u0CB5\\u0CDE]", /*roundtrip exclusions*/
1594
1595 "MALAYALAM-ORIYA", "[:MALAYALAM:]", "[:ORIYA:]",
1596 "[\\u0B01\\u0b3c\\u0b56\\u0B3D\\u0B5C\\u0B5D\\u0B5F\\u0b70\\u0b71]", /*roundtrip exclusions*/
1597
1598 "ORIYA-MALAYALAM", "[:ORIYA:]", "[:MALAYALAM:]",
1599 "[\\u0D47\\u0D46\\u0D4a\\u0D0E\\u0D12\\u0D31\\u0D34\\u0D35]", /*roundtrip exclusions*/
1600
1601 "TELUGU-TAMIL", "[:TELUGU:]", "[:TAMIL:]",
1602 "[\\u0bd7\\u0ba9\\u0bb4\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1603
1604 "TAMIL-TELUGU", "[:TAMIL:]", "[:TELUGU:]",
1605 "[\\u0C01\\u0c43\\u0c44\\u0c46\\u0c47\\u0c55\\u0c56\\u0c66\\u0C0B\\u0C0C\\u0C16\\u0C17\\u0C18\\u0C1B\\u0C1D\\u0C20\\u0C21\\u0C22\\u0C25\\u0C26\\u0C27\\u0C2B\\u0C2C\\u0C2D\\u0C36\\u0C60\\u0C61]", /*roundtrip exclusions*/
1606
1607 "KANNADA-TAMIL", "[:KANNADA:]", "[:TAMIL:]",
1608 "[\\u0bd7\\u0bc6\\u0ba9\\u0bb4\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1609
1610 "TAMIL-KANNADA", "[:TAMIL:]", "[:KANNADA:]",
1611 "[\\u0cc3\\u0cc4\\u0cc6\\u0cc7\\u0cd5\\u0cd6\\u0C8B\\u0C8C\\u0C96\\u0C97\\u0C98\\u0C9B\\u0C9D\\u0CA0\\u0CA1\\u0CA2\\u0CA5\\u0CA6\\u0CA7\\u0CAB\\u0CAC\\u0CAD\\u0CB6\\u0cbc\\u0cbd\\u0CDE\\u0CE0\\u0CE1]", /*roundtrip exclusions*/
1612
1613 "MALAYALAM-TAMIL", "[:MALAYALAM:]", "[:TAMIL:]",
1614 "[\\u0ba9\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1615
1616 "TAMIL-MALAYALAM", "[:TAMIL:]", "[:MALAYALAM:]",
1617 "[\\u0d43\\u0d12\\u0D0B\\u0D0C\\u0D16\\u0D17\\u0D18\\u0D1B\\u0D1D\\u0D20\\u0D21\\u0D22\\u0D25\\u0D26\\u0D27\\u0D2B\\u0D2C\\u0D2D\\u0D36\\u0D60\\u0D61]", /*roundtrip exclusions*/
1618
1619 "KANNADA-TELUGU", "[:KANNADA:]", "[:TELUGU:]",
1620 "[\\u0C01\\u0c3f\\u0c46\\u0c48\\u0c4a]", /*roundtrip exclusions*/
1621
1622 "TELUGU-KANNADA", "[:TELUGU:]", "[:KANNADA:]",
1623 "[\\u0cc8\\u0cd5\\u0cd6\\u0cbc\\u0cbd\\u0CDE]", /*roundtrip exclusions*/
1624
1625 "MALAYALAM-TELUGU", "[:MALAYALAM:]", "[:TELUGU:]",
1626 "[\\u0C01\\u0c44\\u0c4a\\u0c4c\\u0c4b\\u0c55\\u0c56]", /*roundtrip exclusions*/
1627
1628 "TELUGU-MALAYALAM", "[:TELUGU:]", "[:MALAYALAM:]",
1629 "[\\u0d4c\\u0d57\\u0D34]", /*roundtrip exclusions*/
1630
1631 "MALAYALAM-KANNADA", "[:MALAYALAM:]", "[:KANNADA:]",
1632 "[\\u0cbc\\u0cbd\\u0cc4\\u0cc6\\u0cca\\u0ccc\\u0ccb\\u0cd5\\u0cd6\\u0cDe]", /*roundtrip exclusions*/
1633
1634 "KANNADA-MALAYALAM", "[:KANNADA:]", "[:MALAYALAM:]",
1635 "[\\u0d4c\\u0d57\\u0d46\\u0D34]", /*roundtrip exclusions*/
1636
1637 "Latin-Bengali",latinForIndic, "[[:Bengali:][\\u0964\\u0965]]",
1638 "[\\u0965\\u09f0-\\u09fa\\u09ce]" /*roundtrip exclusions*/ ,
1639
1640 "Latin-Gurmukhi", latinForIndic, "[[:Gurmukhi:][\\u0964\\u0965]]",
1641 "[\\u0a01\\u0965\\u0a02\\u0a72\\u0a73\\u0a74]" /*roundtrip exclusions*/,
1642
1643 "Latin-Gujarati",latinForIndic, "[[:Gujarati:][\\u0964\\u0965]]",
1644 "[\\u0965]" /*roundtrip exclusions*/,
1645
1646 "Latin-Oriya",latinForIndic, "[[:Oriya:][\\u0964\\u0965]]",
1647 "[\\u0965\\u0b70]" /*roundtrip exclusions*/,
1648
1649 "Latin-Tamil",latinForIndic, "[:Tamil:]",
1650 "[\\u0BF0\\u0BF1\\u0BF2]" /*roundtrip exclusions*/,
1651
1652 "Latin-Telugu",latinForIndic, "[:Telugu:]",
1653 NULL /*roundtrip exclusions*/,
1654
1655 "Latin-Kannada",latinForIndic, "[:Kannada:]",
1656 NULL /*roundtrip exclusions*/,
1657
1658 "Latin-Malayalam",latinForIndic, "[:Malayalam:]",
1659 NULL /*roundtrip exclusions*/
1660 };
1661
TestDebug(const char * name,const char fromSet[],const char * toSet,const char * exclusions)1662 void TransliteratorRoundTripTest::TestDebug(const char* name,const char fromSet[],
1663 const char* toSet,const char* exclusions){
1664
1665 RTTest test(name);
1666 Legal *legal = new LegalIndic();
1667 test.test(UnicodeString(fromSet,""),UnicodeString(toSet,""),exclusions,this,quick,legal);
1668 }
1669
TestInterIndic()1670 void TransliteratorRoundTripTest::TestInterIndic() {
1671 //TestDebug("Latin-Gurmukhi", latinForIndic, "[:Gurmukhi:]","[\\u0965\\u0a02\\u0a72\\u0a73\\u0a74]",TRUE);
1672 int32_t num = (int32_t)(sizeof(interIndicArray)/(INTER_INDIC_ARRAY_WIDTH*sizeof(char*)));
1673 if(quick){
1674 logln("Testing only 5 of %i. Skipping rest (use -e for exhaustive)",num);
1675 num = 5;
1676 }
1677 // CLDR bug #1911: This test should be moved into CLDR.
1678 // It is left in its current state as a regression test.
1679 // if (isICUVersionAtLeast(ICU_39)) {
1680 // // We temporarily filter against Unicode 4.1, but we only do this
1681 // // before version 3.4.
1682 // errln("FAIL: TestInterIndic needs to be updated to remove delete the [:Age=4.1:] filter ");
1683 // return;
1684 // } else {
1685 // logln("Warning: TestInterIndic needs to be updated to remove delete the section marked [:Age=4.1:] filter");
1686 // }
1687 for(int i = 0; i < num;i++){
1688 RTTest test(interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 0]);
1689 Legal *legal = new LegalIndic();
1690 logln(UnicodeString("Stress testing ") + interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 0]);
1691 /* Uncomment lines below when transliterator is fixed */
1692 /*
1693 test.test( interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 1],
1694 interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 2],
1695 interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 3], // roundtrip exclusions
1696 this, quick, legal, 50);
1697 */
1698 /* comment lines below when transliterator is fixed */
1699 // start
1700 UnicodeString source("[");
1701 source.append(interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 1]);
1702 source.append(" & [:Age=4.1:]]");
1703 UnicodeString target("[");
1704 target.append(interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 2]);
1705 target.append(" & [:Age=4.1:]]");
1706 test.test( source,
1707 target,
1708 interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 3], // roundtrip exclusions
1709 this, quick, legal, 50);
1710 // end
1711 delete legal;
1712 }
1713 }
1714
1715 // end indic tests ----------------------------------------------------------
1716
1717 #endif /* #if !UCONFIG_NO_TRANSLITERATION */
1718