1 /*
2 **********************************************************************
3 * Copyright (C) 2000-2011, 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 "unicode/utf16.h"
30 #include "putilimp.h"
31 #include "cmemory.h"
32 #include "transrt.h"
33 #include "testutil.h"
34 #include <string.h>
35 #include <stdio.h>
36
37 #define CASE(id,test) case id: \
38 name = #test; \
39 if (exec) { \
40 logln(#test "---"); \
41 logln((UnicodeString)""); \
42 UDate t = uprv_getUTCtime(); \
43 test(); \
44 t = uprv_getUTCtime() - t; \
45 logln((UnicodeString)#test " took " + t/U_MILLIS_PER_DAY + " seconds"); \
46 } \
47 break
48
49 #define EXHAUSTIVE(id,test) case id: \
50 if(quick==FALSE){ \
51 name = #test; \
52 if (exec){ \
53 logln(#test "---"); \
54 logln((UnicodeString)""); \
55 test(); \
56 } \
57 }else{ \
58 name=""; \
59 } \
60 break
61 void
runIndexedTest(int32_t index,UBool exec,const char * & name,char *)62 TransliteratorRoundTripTest::runIndexedTest(int32_t index, UBool exec,
63 const char* &name, char* /*par*/) {
64 switch (index) {
65 CASE(0, TestCyrillic);
66 // CASE(0,TestKana);
67 CASE(1,TestHiragana);
68 CASE(2,TestKatakana);
69 CASE(3,TestJamo);
70 CASE(4,TestHangul);
71 CASE(5,TestGreek);
72 CASE(6,TestGreekUNGEGN);
73 CASE(7,Testel);
74 CASE(8,TestDevanagariLatin);
75 CASE(9,TestInterIndic);
76 CASE(10, TestHebrew);
77 CASE(11, TestArabic);
78 CASE(12, TestHan);
79 default: name = ""; break;
80 }
81 }
82
83
84 //--------------------------------------------------------------------
85 // TransliteratorPointer
86 //--------------------------------------------------------------------
87
88 /**
89 * A transliterator pointer wrapper that deletes the contained
90 * pointer automatically when the wrapper goes out of scope.
91 * Sometimes called a "janitor" or "smart pointer".
92 */
93 class TransliteratorPointer {
94 Transliterator* t;
95 // disallowed:
96 TransliteratorPointer(const TransliteratorPointer& rhs);
97 TransliteratorPointer& operator=(const TransliteratorPointer& rhs);
98 public:
TransliteratorPointer(Transliterator * adopted)99 TransliteratorPointer(Transliterator* adopted) {
100 t = adopted;
101 }
~TransliteratorPointer()102 ~TransliteratorPointer() {
103 delete t;
104 }
operator ->()105 inline Transliterator* operator->() { return t; }
operator const Transliterator*() const106 inline operator const Transliterator*() const { return t; }
operator Transliterator*()107 inline operator Transliterator*() { return t; }
108 };
109
110 //--------------------------------------------------------------------
111 // Legal
112 //--------------------------------------------------------------------
113
114 class Legal {
115 public:
Legal()116 Legal() {}
~Legal()117 virtual ~Legal() {}
is(const UnicodeString &) const118 virtual UBool is(const UnicodeString& /*sourceString*/) const {return TRUE;}
119 };
120
121 class LegalJamo : public Legal {
122 // any initial must be followed by a medial (or initial)
123 // any medial must follow an initial (or medial)
124 // any final must follow a medial (or final)
125 public:
LegalJamo()126 LegalJamo() {}
~LegalJamo()127 virtual ~LegalJamo() {}
128 virtual UBool is(const UnicodeString& sourceString) const;
129 int getType(UChar c) const;
130 };
131
is(const UnicodeString & sourceString) const132 UBool LegalJamo::is(const UnicodeString& sourceString) const {
133 int t;
134 UnicodeString decomp;
135 UErrorCode ec = U_ZERO_ERROR;
136 Normalizer::decompose(sourceString, FALSE, 0, decomp, ec);
137 if (U_FAILURE(ec)) {
138 return FALSE;
139 }
140 for (int i = 0; i < decomp.length(); ++i) { // don't worry about surrogates
141 switch (getType(decomp.charAt(i))) {
142 case 0: t = getType(decomp.charAt(i+1));
143 if (t != 0 && t != 1) { return FALSE; }
144 break;
145 case 1: t = getType(decomp.charAt(i-1));
146 if (t != 0 && t != 1) { return FALSE; }
147 break;
148 case 2: t = getType(decomp.charAt(i-1));
149 if (t != 1 && t != 2) { return FALSE; }
150 break;
151 }
152 }
153 return TRUE;
154 }
155
getType(UChar c) const156 int LegalJamo::getType(UChar c) const {
157 if (0x1100 <= c && c <= 0x1112)
158 return 0;
159 else if (0x1161 <= c && c <= 0x1175)
160 return 1;
161 else if (0x11A8 <= c && c <= 0x11C2)
162 return 2;
163 return -1; // other
164 }
165
166 class LegalGreek : public Legal {
167 UBool full;
168 public:
LegalGreek(UBool _full)169 LegalGreek(UBool _full) { full = _full; }
~LegalGreek()170 virtual ~LegalGreek() {}
171
172 virtual UBool is(const UnicodeString& sourceString) const;
173
174 static UBool isVowel(UChar c);
175
176 static UBool isRho(UChar c);
177 };
178
is(const UnicodeString & sourceString) const179 UBool LegalGreek::is(const UnicodeString& sourceString) const {
180 UnicodeString decomp;
181 UErrorCode ec = U_ZERO_ERROR;
182 Normalizer::decompose(sourceString, FALSE, 0, decomp, ec);
183
184 // modern is simpler: don't care about anything but a grave
185 if (full == FALSE) {
186 // A special case which is legal but should be
187 // excluded from round trip
188 // if (sourceString == UnicodeString("\\u039C\\u03C0", "")) {
189 // return FALSE;
190 // }
191 for (int32_t i = 0; i < decomp.length(); ++i) {
192 UChar c = decomp.charAt(i);
193 // exclude all the accents
194 if (c == 0x0313 || c == 0x0314 || c == 0x0300 || c == 0x0302
195 || c == 0x0342 || c == 0x0345
196 ) return FALSE;
197 }
198 return TRUE;
199 }
200
201 // Legal greek has breathing marks IFF there is a vowel or RHO at the start
202 // IF it has them, it has exactly one.
203 // IF it starts with a RHO, then the breathing mark must come before the second letter.
204 // Since there are no surrogates in greek, don't worry about them
205 UBool firstIsVowel = FALSE;
206 UBool firstIsRho = FALSE;
207 UBool noLetterYet = TRUE;
208 int32_t breathingCount = 0;
209 int32_t letterCount = 0;
210 for (int32_t i = 0; i < decomp.length(); ++i) {
211 UChar c = decomp.charAt(i);
212 if (u_isalpha(c)) {
213 ++letterCount;
214 if (noLetterYet) {
215 noLetterYet = FALSE;
216 firstIsVowel = isVowel(c);
217 firstIsRho = isRho(c);
218 }
219 if (firstIsRho && letterCount == 2 && breathingCount == 0) {
220 return FALSE;
221 }
222 }
223 if (c == 0x0313 || c == 0x0314) {
224 ++breathingCount;
225 }
226 }
227
228 if (firstIsVowel || firstIsRho) return breathingCount == 1;
229 return breathingCount == 0;
230 }
231
isVowel(UChar c)232 UBool LegalGreek::isVowel(UChar c) {
233 switch (c) {
234 case 0x03B1:
235 case 0x03B5:
236 case 0x03B7:
237 case 0x03B9:
238 case 0x03BF:
239 case 0x03C5:
240 case 0x03C9:
241 case 0x0391:
242 case 0x0395:
243 case 0x0397:
244 case 0x0399:
245 case 0x039F:
246 case 0x03A5:
247 case 0x03A9:
248 return TRUE;
249 }
250 return FALSE;
251 }
252
isRho(UChar c)253 UBool LegalGreek::isRho(UChar c) {
254 switch (c) {
255 case 0x03C1:
256 case 0x03A1:
257 return TRUE;
258 }
259 return FALSE;
260 }
261
262 // AbbreviatedUnicodeSetIterator Interface ---------------------------------------------
263 //
264 // Iterate over a UnicodeSet, only returning a sampling of the contained code points.
265 // density is the approximate total number of code points to returned for the entire set.
266 //
267
268 class AbbreviatedUnicodeSetIterator : public UnicodeSetIterator {
269 public :
270
271 AbbreviatedUnicodeSetIterator();
272 virtual ~AbbreviatedUnicodeSetIterator();
273 void reset(UnicodeSet& set, UBool abb = FALSE, int32_t density = 100);
274
275 /**
276 * ICU "poor man's RTTI", returns a UClassID for this class.
277 */
getStaticClassID()278 static inline UClassID getStaticClassID() { return (UClassID)&fgClassID; }
279
280 /**
281 * ICU "poor man's RTTI", returns a UClassID for the actual class.
282 */
getDynamicClassID() const283 virtual inline UClassID getDynamicClassID() const { return getStaticClassID(); }
284
285 private :
286 UBool abbreviated;
287 int32_t perRange; // The maximum number of code points to be returned from each range
288 virtual void loadRange(int32_t range);
289
290 /**
291 * The address of this static class variable serves as this class's ID
292 * for ICU "poor man's RTTI".
293 */
294 static const char fgClassID;
295 };
296
297 // AbbreviatedUnicodeSetIterator Implementation ---------------------------------------
298
299 const char AbbreviatedUnicodeSetIterator::fgClassID=0;
300
AbbreviatedUnicodeSetIterator()301 AbbreviatedUnicodeSetIterator::AbbreviatedUnicodeSetIterator() :
302 UnicodeSetIterator(), abbreviated(FALSE) {
303 }
304
~AbbreviatedUnicodeSetIterator()305 AbbreviatedUnicodeSetIterator::~AbbreviatedUnicodeSetIterator() {
306 }
307
reset(UnicodeSet & newSet,UBool abb,int32_t density)308 void AbbreviatedUnicodeSetIterator::reset(UnicodeSet& newSet, UBool abb, int32_t density) {
309 UnicodeSetIterator::reset(newSet);
310 abbreviated = abb;
311 perRange = newSet.getRangeCount();
312 if (perRange != 0) {
313 perRange = density / perRange;
314 }
315 }
316
loadRange(int32_t myRange)317 void AbbreviatedUnicodeSetIterator::loadRange(int32_t myRange) {
318 UnicodeSetIterator::loadRange(myRange);
319 if (abbreviated && (endElement > nextElement + perRange)) {
320 endElement = nextElement + perRange;
321 }
322 }
323
324 //--------------------------------------------------------------------
325 // RTTest Interface
326 //--------------------------------------------------------------------
327
328 class RTTest : public IntlTest {
329
330 // PrintWriter out;
331
332 UnicodeString transliteratorID;
333 int32_t errorLimit;
334 int32_t errorCount;
335 int32_t pairLimit;
336 UnicodeSet sourceRange;
337 UnicodeSet targetRange;
338 UnicodeSet toSource;
339 UnicodeSet toTarget;
340 UnicodeSet roundtripExclusionsSet;
341 IntlTest* parent;
342 Legal* legalSource; // NOT owned
343 UnicodeSet badCharacters;
344
345 public:
346
347 /*
348 * create a test for the given script transliterator.
349 */
350 RTTest(const UnicodeString& transliteratorIDStr);
351
352 virtual ~RTTest();
353
354 void setErrorLimit(int32_t limit);
355
356 void setPairLimit(int32_t limit);
357
358 void test(const UnicodeString& sourceRange,
359 const UnicodeString& targetRange,
360 const char* roundtripExclusions,
361 IntlTest* parent,
362 UBool quick,
363 Legal* adoptedLegal,
364 int32_t density = 100);
365
366 private:
367
368 // Added to do better equality check.
369
370 static UBool isSame(const UnicodeString& a, const UnicodeString& b);
371
372 static UBool isCamel(const UnicodeString& a);
373
374 UBool checkIrrelevants(Transliterator *t, const UnicodeString& irrelevants);
375
376 void test2(UBool quick, int32_t density);
377
378 void logWrongScript(const UnicodeString& label,
379 const UnicodeString& from,
380 const UnicodeString& to);
381
382 void logNotCanonical(const UnicodeString& label,
383 const UnicodeString& from,
384 const UnicodeString& to,
385 const UnicodeString& fromCan,
386 const UnicodeString& toCan);
387
388 void logFails(const UnicodeString& label);
389
390 void logToRulesFails(const UnicodeString& label,
391 const UnicodeString& from,
392 const UnicodeString& to,
393 const UnicodeString& toCan);
394
395 void logRoundTripFailure(const UnicodeString& from,
396 const UnicodeString& toID,
397 const UnicodeString& to,
398 const UnicodeString& backID,
399 const UnicodeString& back);
400 };
401
402 //--------------------------------------------------------------------
403 // RTTest Implementation
404 //--------------------------------------------------------------------
405
406 /*
407 * create a test for the given script transliterator.
408 */
RTTest(const UnicodeString & transliteratorIDStr)409 RTTest::RTTest(const UnicodeString& transliteratorIDStr) {
410 transliteratorID = transliteratorIDStr;
411 errorLimit = 500;
412 errorCount = 0;
413 pairLimit = 0x10000;
414 }
415
~RTTest()416 RTTest::~RTTest() {
417 }
418
setErrorLimit(int32_t limit)419 void RTTest::setErrorLimit(int32_t limit) {
420 errorLimit = limit;
421 }
422
setPairLimit(int32_t limit)423 void RTTest::setPairLimit(int32_t limit) {
424 pairLimit = limit;
425 }
426
isSame(const UnicodeString & a,const UnicodeString & b)427 UBool RTTest::isSame(const UnicodeString& a, const UnicodeString& b) {
428 if (a == b) return TRUE;
429 if (a.caseCompare(b, U_FOLD_CASE_DEFAULT)==0 && isCamel(a)) return TRUE;
430 UnicodeString aa, bb;
431 UErrorCode ec = U_ZERO_ERROR;
432 Normalizer::decompose(a, FALSE, 0, aa, ec);
433 Normalizer::decompose(b, FALSE, 0, bb, ec);
434 if (aa == bb) return TRUE;
435 if (aa.caseCompare(bb, U_FOLD_CASE_DEFAULT)==0 && isCamel(aa)) return TRUE;
436 return FALSE;
437 }
438
isCamel(const UnicodeString & a)439 UBool RTTest::isCamel(const UnicodeString& a) {
440 // see if string is of the form aB; e.g. lower, then upper or title
441 UChar32 cp;
442 UBool haveLower = FALSE;
443 for (int32_t i = 0; i < a.length(); i += U16_LENGTH(cp)) {
444 cp = a.char32At(i);
445 int8_t t = u_charType(cp);
446 switch (t) {
447 case U_UPPERCASE_LETTER:
448 if (haveLower) return TRUE;
449 break;
450 case U_TITLECASE_LETTER:
451 if (haveLower) return TRUE;
452 // drop through, since second letter is lower.
453 case U_LOWERCASE_LETTER:
454 haveLower = TRUE;
455 break;
456 }
457 }
458 return FALSE;
459 }
460
test(const UnicodeString & sourceRangeVal,const UnicodeString & targetRangeVal,const char * roundtripExclusions,IntlTest * logVal,UBool quickRt,Legal * adoptedLegal,int32_t density)461 void RTTest::test(const UnicodeString& sourceRangeVal,
462 const UnicodeString& targetRangeVal,
463 const char* roundtripExclusions,
464 IntlTest* logVal, UBool quickRt,
465 Legal* adoptedLegal,
466 int32_t density)
467 {
468
469 UErrorCode status = U_ZERO_ERROR;
470
471 this->parent = logVal;
472 this->legalSource = adoptedLegal;
473
474 UnicodeSet neverOk("[:Other:]", status);
475 UnicodeSet okAnyway("[^[:Letter:]]", status);
476
477 if (U_FAILURE(status)) {
478 parent->dataerrln("FAIL: Initializing UnicodeSet with [:Other:] or [^[:Letter:]] - Error: %s", u_errorName(status));
479 return;
480 }
481
482 this->sourceRange.clear();
483 this->sourceRange.applyPattern(sourceRangeVal, status);
484 if (U_FAILURE(status)) {
485 parent->errln("FAIL: UnicodeSet::applyPattern(" +
486 sourceRangeVal + ")");
487 return;
488 }
489 this->sourceRange.removeAll(neverOk);
490
491 this->targetRange.clear();
492 this->targetRange.applyPattern(targetRangeVal, status);
493 if (U_FAILURE(status)) {
494 parent->errln("FAIL: UnicodeSet::applyPattern(" +
495 targetRangeVal + ")");
496 return;
497 }
498 this->targetRange.removeAll(neverOk);
499
500 this->toSource.clear();
501 this->toSource.applyPattern(sourceRangeVal, status);
502 if (U_FAILURE(status)) {
503 parent->errln("FAIL: UnicodeSet::applyPattern(" +
504 sourceRangeVal + ")");
505 return;
506 }
507 this->toSource.addAll(okAnyway);
508
509 this->toTarget.clear();
510 this->toTarget.applyPattern(targetRangeVal, status);
511 if (U_FAILURE(status)) {
512 parent->errln("FAIL: UnicodeSet::applyPattern(" +
513 targetRangeVal + ")");
514 return;
515 }
516 this->toTarget.addAll(okAnyway);
517
518 this->roundtripExclusionsSet.clear();
519 if (roundtripExclusions != NULL && strlen(roundtripExclusions) > 0) {
520 this->roundtripExclusionsSet.applyPattern(UnicodeString(roundtripExclusions, -1, US_INV), status);
521 if (U_FAILURE(status)) {
522 parent->errln("FAIL: UnicodeSet::applyPattern(%s)", roundtripExclusions);
523 return;
524 }
525 }
526
527 badCharacters.clear();
528 badCharacters.applyPattern("[:Other:]", status);
529 if (U_FAILURE(status)) {
530 parent->errln("FAIL: UnicodeSet::applyPattern([:Other:])");
531 return;
532 }
533
534 test2(quickRt, density);
535
536 if (errorCount > 0) {
537 char str[100];
538 int32_t length = transliteratorID.extract(str, 100, NULL, status);
539 str[length] = 0;
540 parent->errln("FAIL: %s errors: %d %s", str, errorCount, (errorCount > errorLimit ? " (at least!)" : " ")); // + ", see " + logFileName);
541 } else {
542 char str[100];
543 int32_t length = transliteratorID.extract(str, 100, NULL, status);
544 str[length] = 0;
545 parent->logln("%s ok", str);
546 }
547 }
548
checkIrrelevants(Transliterator * t,const UnicodeString & irrelevants)549 UBool RTTest::checkIrrelevants(Transliterator *t,
550 const UnicodeString& irrelevants) {
551 for (int i = 0; i < irrelevants.length(); ++i) {
552 UChar c = irrelevants.charAt(i);
553 UnicodeString srcStr(c);
554 UnicodeString targ = srcStr;
555 t->transliterate(targ);
556 if (srcStr == targ) return TRUE;
557 }
558 return FALSE;
559 }
560
test2(UBool quickRt,int32_t density)561 void RTTest::test2(UBool quickRt, int32_t density) {
562
563 UnicodeString srcStr, targ, reverse;
564 UErrorCode status = U_ZERO_ERROR;
565 UParseError parseError ;
566 TransliteratorPointer sourceToTarget(
567 Transliterator::createInstance(transliteratorID, UTRANS_FORWARD, parseError,
568 status));
569 if ((Transliterator *)sourceToTarget == NULL) {
570 parent->dataerrln("FAIL: createInstance(" + transliteratorID +
571 ") returned NULL. Error: " + u_errorName(status)
572 + "\n\tpreContext : " + prettify(parseError.preContext)
573 + "\n\tpostContext : " + prettify(parseError.postContext));
574
575 return;
576 }
577 TransliteratorPointer targetToSource(sourceToTarget->createInverse(status));
578 if ((Transliterator *)targetToSource == NULL) {
579 parent->errln("FAIL: " + transliteratorID +
580 ".createInverse() returned NULL. Error:" + u_errorName(status)
581 + "\n\tpreContext : " + prettify(parseError.preContext)
582 + "\n\tpostContext : " + prettify(parseError.postContext));
583 return;
584 }
585
586 AbbreviatedUnicodeSetIterator usi;
587 AbbreviatedUnicodeSetIterator usi2;
588
589 parent->logln("Checking that at least one irrelevant character is not NFC'ed");
590 // string is from NFC_NO in the UCD
591 UnicodeString irrelevants = CharsToUnicodeString("\\u2000\\u2001\\u2126\\u212A\\u212B\\u2329");
592
593 if (checkIrrelevants(sourceToTarget, irrelevants) == FALSE) {
594 logFails("Source-Target, irrelevants");
595 }
596 if (checkIrrelevants(targetToSource, irrelevants) == FALSE) {
597 logFails("Target-Source, irrelevants");
598 }
599
600 if (!quickRt){
601 parent->logln("Checking that toRules works");
602 UnicodeString rules = "";
603
604 UParseError parseError;
605 rules = sourceToTarget->toRules(rules, TRUE);
606 // parent->logln((UnicodeString)"toRules => " + rules);
607 TransliteratorPointer sourceToTarget2(Transliterator::createFromRules(
608 "s2t2", rules,
609 UTRANS_FORWARD,
610 parseError, status));
611 if (U_FAILURE(status)) {
612 parent->errln("FAIL: createFromRules %s\n", u_errorName(status));
613 return;
614 }
615
616 rules = targetToSource->toRules(rules, FALSE);
617 TransliteratorPointer targetToSource2(Transliterator::createFromRules(
618 "t2s2", rules,
619 UTRANS_FORWARD,
620 parseError, status));
621 if (U_FAILURE(status)) {
622 parent->errln("FAIL: createFromRules %s\n", u_errorName(status));
623 return;
624 }
625
626 usi.reset(sourceRange);
627 for (;;) {
628 if (!usi.next() || usi.isString()) break;
629 UChar32 c = usi.getCodepoint();
630
631 UnicodeString srcStr((UChar32)c);
632 UnicodeString targ = srcStr;
633 sourceToTarget->transliterate(targ);
634 UnicodeString targ2 = srcStr;
635 sourceToTarget2->transliterate(targ2);
636 if (targ != targ2) {
637 logToRulesFails("Source-Target, toRules", srcStr, targ, targ2);
638 }
639 }
640
641 usi.reset(targetRange);
642 for (;;) {
643 if (!usi.next() || usi.isString()) break;
644 UChar32 c = usi.getCodepoint();
645
646 UnicodeString srcStr((UChar32)c);
647 UnicodeString targ = srcStr;
648 targetToSource->transliterate(targ);
649 UnicodeString targ2 = srcStr;
650 targetToSource2->transliterate(targ2);
651 if (targ != targ2) {
652 logToRulesFails("Target-Source, toRules", srcStr, targ, targ2);
653 }
654 }
655 }
656
657 parent->logln("Checking that all source characters convert to target - Singles");
658
659 UnicodeSet failSourceTarg;
660 usi.reset(sourceRange);
661 for (;;) {
662 if (!usi.next() || usi.isString()) break;
663 UChar32 c = usi.getCodepoint();
664
665 UnicodeString srcStr((UChar32)c);
666 UnicodeString targ = srcStr;
667 sourceToTarget->transliterate(targ);
668 if (toTarget.containsAll(targ) == FALSE
669 || badCharacters.containsSome(targ) == TRUE) {
670 UnicodeString targD;
671 Normalizer::decompose(targ, FALSE, 0, targD, status);
672 if (U_FAILURE(status)) {
673 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
674 return;
675 }
676 if (toTarget.containsAll(targD) == FALSE ||
677 badCharacters.containsSome(targD) == TRUE) {
678 logWrongScript("Source-Target", srcStr, targ);
679 failSourceTarg.add(c);
680 continue;
681 }
682 }
683
684 UnicodeString cs2;
685 Normalizer::decompose(srcStr, FALSE, 0, cs2, status);
686 if (U_FAILURE(status)) {
687 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
688 return;
689 }
690 UnicodeString targ2 = cs2;
691 sourceToTarget->transliterate(targ2);
692 if (targ != targ2) {
693 logNotCanonical("Source-Target", srcStr, targ,cs2, targ2);
694 }
695 }
696
697 parent->logln("Checking that all source characters convert to target - Doubles");
698
699 UnicodeSet sourceRangeMinusFailures(sourceRange);
700 sourceRangeMinusFailures.removeAll(failSourceTarg);
701
702 usi.reset(sourceRangeMinusFailures, quickRt, density);
703 for (;;) {
704 if (!usi.next() || usi.isString()) break;
705 UChar32 c = usi.getCodepoint();
706
707 usi2.reset(sourceRangeMinusFailures, quickRt, density);
708 for (;;) {
709 if (!usi2.next() || usi2.isString()) break;
710 UChar32 d = usi2.getCodepoint();
711
712 UnicodeString srcStr;
713 srcStr += (UChar32)c;
714 srcStr += (UChar32)d;
715 UnicodeString targ = srcStr;
716 sourceToTarget->transliterate(targ);
717 if (toTarget.containsAll(targ) == FALSE ||
718 badCharacters.containsSome(targ) == TRUE)
719 {
720 UnicodeString targD;
721 Normalizer::decompose(targ, FALSE, 0, targD, status);
722 if (U_FAILURE(status)) {
723 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
724 return;
725 }
726 if (toTarget.containsAll(targD) == FALSE ||
727 badCharacters.containsSome(targD) == TRUE) {
728 logWrongScript("Source-Target", srcStr, targ);
729 continue;
730 }
731 }
732 UnicodeString cs2;
733 Normalizer::decompose(srcStr, FALSE, 0, cs2, status);
734 if (U_FAILURE(status)) {
735 parent->errln("FAIL: Internal error during decomposition %s\n", u_errorName(status));
736 return;
737 }
738 UnicodeString targ2 = cs2;
739 sourceToTarget->transliterate(targ2);
740 if (targ != targ2) {
741 logNotCanonical("Source-Target", srcStr, targ, cs2,targ2);
742 }
743 }
744 }
745
746 parent->logln("Checking that target characters convert to source and back - Singles");
747
748 UnicodeSet failTargSource;
749 UnicodeSet failRound;
750
751 usi.reset(targetRange);
752 for (;;) {
753 if (!usi.next()) break;
754
755 if(usi.isString()){
756 srcStr = usi.getString();
757 }else{
758 srcStr = (UnicodeString)usi.getCodepoint();
759 }
760
761 UChar32 c = srcStr.char32At(0);
762
763 targ = srcStr;
764 targetToSource->transliterate(targ);
765 reverse = targ;
766 sourceToTarget->transliterate(reverse);
767
768 if (toSource.containsAll(targ) == FALSE ||
769 badCharacters.containsSome(targ) == TRUE) {
770 UnicodeString targD;
771 Normalizer::decompose(targ, FALSE, 0, targD, status);
772 if (U_FAILURE(status)) {
773 parent->errln("FAIL: Internal error during decomposition%s\n", u_errorName(status));
774 return;
775 }
776 if (toSource.containsAll(targD) == FALSE) {
777 logWrongScript("Target-Source", srcStr, targ);
778 failTargSource.add(c);
779 continue;
780 }
781 if (badCharacters.containsSome(targD) == TRUE) {
782 logWrongScript("Target-Source*", srcStr, targ);
783 failTargSource.add(c);
784 continue;
785 }
786 }
787 if (isSame(srcStr, reverse) == FALSE &&
788 roundtripExclusionsSet.contains(c) == FALSE
789 && roundtripExclusionsSet.contains(srcStr)==FALSE) {
790 logRoundTripFailure(srcStr,targetToSource->getID(), targ,sourceToTarget->getID(), reverse);
791 failRound.add(c);
792 continue;
793 }
794
795 UnicodeString targ2;
796 Normalizer::decompose(targ, FALSE, 0, targ2, status);
797 if (U_FAILURE(status)) {
798 parent->errln("FAIL: Internal error during decomposition%s\n", u_errorName(status));
799 return;
800 }
801 UnicodeString reverse2 = targ2;
802 sourceToTarget->transliterate(reverse2);
803 if (reverse != reverse2) {
804 logNotCanonical("Target-Source", targ, reverse, targ2, reverse2);
805 }
806 }
807
808 parent->logln("Checking that target characters convert to source and back - Doubles");
809 int32_t count = 0;
810
811 UnicodeSet targetRangeMinusFailures(targetRange);
812 targetRangeMinusFailures.removeAll(failTargSource);
813 targetRangeMinusFailures.removeAll(failRound);
814
815 usi.reset(targetRangeMinusFailures, quickRt, density);
816 UnicodeString targ2;
817 UnicodeString reverse2;
818 UnicodeString targD;
819 for (;;) {
820 if (!usi.next() || usi.isString()) break;
821 UChar32 c = usi.getCodepoint();
822 if (++count > pairLimit) {
823 //throw new TestTruncated("Test truncated at " + pairLimit + " x 64k pairs");
824 parent->logln("");
825 parent->logln((UnicodeString)"Test truncated at " + pairLimit + " x 64k pairs");
826 return;
827 }
828
829 usi2.reset(targetRangeMinusFailures, quickRt, density);
830 for (;;) {
831 if (!usi2.next() || usi2.isString())
832 break;
833 UChar32 d = usi2.getCodepoint();
834 srcStr.truncate(0); // empty the variable without construction/destruction
835 srcStr += c;
836 srcStr += d;
837
838 targ = srcStr;
839 targetToSource->transliterate(targ);
840 reverse = targ;
841 sourceToTarget->transliterate(reverse);
842
843 if (toSource.containsAll(targ) == FALSE ||
844 badCharacters.containsSome(targ) == TRUE)
845 {
846 targD.truncate(0); // empty the variable without construction/destruction
847 Normalizer::decompose(targ, FALSE, 0, targD, status);
848 if (U_FAILURE(status)) {
849 parent->errln("FAIL: Internal error during decomposition%s\n",
850 u_errorName(status));
851 return;
852 }
853 if (toSource.containsAll(targD) == FALSE
854 || badCharacters.containsSome(targD) == TRUE)
855 {
856 logWrongScript("Target-Source", srcStr, targ);
857 continue;
858 }
859 }
860 if (isSame(srcStr, reverse) == FALSE &&
861 roundtripExclusionsSet.contains(c) == FALSE&&
862 roundtripExclusionsSet.contains(d) == FALSE &&
863 roundtripExclusionsSet.contains(srcStr)== FALSE)
864 {
865 logRoundTripFailure(srcStr,targetToSource->getID(), targ, sourceToTarget->getID(),reverse);
866 continue;
867 }
868
869 targ2.truncate(0); // empty the variable without construction/destruction
870 Normalizer::decompose(targ, FALSE, 0, targ2, status);
871 if (U_FAILURE(status)) {
872 parent->errln("FAIL: Internal error during decomposition%s\n", u_errorName(status));
873 return;
874 }
875 reverse2 = targ2;
876 sourceToTarget->transliterate(reverse2);
877 if (reverse != reverse2) {
878 logNotCanonical("Target-Source", targ,reverse, targ2, reverse2);
879 }
880 }
881 }
882 parent->logln("");
883 }
884
logWrongScript(const UnicodeString & label,const UnicodeString & from,const UnicodeString & to)885 void RTTest::logWrongScript(const UnicodeString& label,
886 const UnicodeString& from,
887 const UnicodeString& to) {
888 parent->errln((UnicodeString)"FAIL " +
889 label + ": " +
890 from + "(" + TestUtility::hex(from) + ") => " +
891 to + "(" + TestUtility::hex(to) + ")");
892 ++errorCount;
893 }
894
logNotCanonical(const UnicodeString & label,const UnicodeString & from,const UnicodeString & to,const UnicodeString & fromCan,const UnicodeString & toCan)895 void RTTest::logNotCanonical(const UnicodeString& label,
896 const UnicodeString& from,
897 const UnicodeString& to,
898 const UnicodeString& fromCan,
899 const UnicodeString& toCan) {
900 parent->errln((UnicodeString)"FAIL (can.equiv)" +
901 label + ": " +
902 from + "(" + TestUtility::hex(from) + ") => " +
903 to + "(" + TestUtility::hex(to) + ")" +
904 fromCan + "(" + TestUtility::hex(fromCan) + ") => " +
905 toCan + " (" +
906 TestUtility::hex(toCan) + ")"
907 );
908 ++errorCount;
909 }
910
logFails(const UnicodeString & label)911 void RTTest::logFails(const UnicodeString& label) {
912 parent->errln((UnicodeString)"<br>FAIL " + label);
913 ++errorCount;
914 }
915
logToRulesFails(const UnicodeString & label,const UnicodeString & from,const UnicodeString & to,const UnicodeString & otherTo)916 void RTTest::logToRulesFails(const UnicodeString& label,
917 const UnicodeString& from,
918 const UnicodeString& to,
919 const UnicodeString& otherTo)
920 {
921 parent->errln((UnicodeString)"FAIL: " +
922 label + ": " +
923 from + "(" + TestUtility::hex(from) + ") => " +
924 to + "(" + TestUtility::hex(to) + ")" +
925 "!=" +
926 otherTo + " (" +
927 TestUtility::hex(otherTo) + ")"
928 );
929 ++errorCount;
930 }
931
932
logRoundTripFailure(const UnicodeString & from,const UnicodeString & toID,const UnicodeString & to,const UnicodeString & backID,const UnicodeString & back)933 void RTTest::logRoundTripFailure(const UnicodeString& from,
934 const UnicodeString& toID,
935 const UnicodeString& to,
936 const UnicodeString& backID,
937 const UnicodeString& back) {
938 if (legalSource->is(from) == FALSE) return; // skip illegals
939
940 parent->errln((UnicodeString)"FAIL Roundtrip: " +
941 from + "(" + TestUtility::hex(from) + ") => " +
942 to + "(" + TestUtility::hex(to) + ") "+toID+" => " +
943 back + "(" + TestUtility::hex(back) + ") "+backID+" => ");
944 ++errorCount;
945 }
946
947 //--------------------------------------------------------------------
948 // Specific Tests
949 //--------------------------------------------------------------------
950
951 /*
952 Note: Unicode 3.2 added new Hiragana/Katakana characters:
953
954 3095..3096 ; 3.2 # [2] HIRAGANA LETTER SMALL KA..HIRAGANA LETTER SMALL KE
955 309F..30A0 ; 3.2 # [2] HIRAGANA DIGRAPH YORI..KATAKANA-HIRAGANA DOUBLE HYPHEN
956 30FF ; 3.2 # KATAKANA DIGRAPH KOTO
957 31F0..31FF ; 3.2 # [16] KATAKANA LETTER SMALL KU..KATAKANA LETTER SMALL RO
958
959 Unicode 5.2 added another Hiragana character:
960 1F200 ; 5.2 # SQUARE HIRAGANA HOKA
961
962 We will not add them to the rules until they are more supported (e.g. in fonts on Windows)
963 A bug has been filed to remind us to do this: #1979.
964 */
965
966 static const char KATAKANA[] = "[[[:katakana:][\\u30A1-\\u30FA\\u30FC]]-[\\u30FF\\u31F0-\\u31FF]-[:^age=5.2:]]";
967 static const char HIRAGANA[] = "[[[:hiragana:][\\u3040-\\u3094]]-[\\u3095-\\u3096\\u309F-\\u30A0\\U0001F200-\\U0001F2FF]-[:^age=5.2:]]";
968 static const char LENGTH[] = "[\\u30FC]";
969 static const char HALFWIDTH_KATAKANA[] = "[\\uFF65-\\uFF9D]";
970 static const char KATAKANA_ITERATION[] = "[\\u30FD\\u30FE]";
971 static const char HIRAGANA_ITERATION[] = "[\\u309D\\u309E]";
972 static const int32_t TEMP_MAX=256;
973
TestKana()974 void TransliteratorRoundTripTest::TestKana() {
975 RTTest test("Katakana-Hiragana");
976 Legal *legal = new Legal();
977 char temp[TEMP_MAX];
978 strcpy(temp, "[");
979 strcat(temp, HALFWIDTH_KATAKANA);
980 strcat(temp, LENGTH);
981 strcat(temp, "]");
982 test.test(KATAKANA, UnicodeString("[") + HIRAGANA + LENGTH + UnicodeString("]"),
983 temp,
984 this, quick, legal);
985 delete legal;
986 }
987
TestHiragana()988 void TransliteratorRoundTripTest::TestHiragana() {
989 RTTest test("Latin-Hiragana");
990 Legal *legal = new Legal();
991 test.test(UnicodeString("[a-zA-Z]", ""),
992 UnicodeString(HIRAGANA, -1, US_INV),
993 HIRAGANA_ITERATION, this, quick, legal);
994 delete legal;
995 }
996
TestKatakana()997 void TransliteratorRoundTripTest::TestKatakana() {
998 RTTest test("Latin-Katakana");
999 Legal *legal = new Legal();
1000 char temp[TEMP_MAX];
1001 strcpy(temp, "[");
1002 strcat(temp, KATAKANA_ITERATION);
1003 strcat(temp, HALFWIDTH_KATAKANA);
1004 strcat(temp, "]");
1005 test.test(UnicodeString("[a-zA-Z]", ""),
1006 UnicodeString(KATAKANA, -1, US_INV),
1007 temp,
1008 this, quick, legal);
1009 delete legal;
1010 }
1011
TestJamo()1012 void TransliteratorRoundTripTest::TestJamo() {
1013 RTTest t("Latin-Jamo");
1014 Legal *legal = new LegalJamo();
1015 t.test(UnicodeString("[a-zA-Z]", ""),
1016 UnicodeString("[\\u1100-\\u1112 \\u1161-\\u1175 \\u11A8-\\u11C2]",
1017 ""),
1018 NULL, this, quick, legal);
1019 delete legal;
1020 }
1021
TestHangul()1022 void TransliteratorRoundTripTest::TestHangul() {
1023 RTTest t("Latin-Hangul");
1024 Legal *legal = new Legal();
1025 if (quick) t.setPairLimit(1000);
1026 t.test(UnicodeString("[a-zA-Z]", ""),
1027 UnicodeString("[\\uAC00-\\uD7A4]", ""),
1028 NULL, this, quick, legal, 1);
1029 delete legal;
1030 }
1031
1032
1033 #define ASSERT_SUCCESS(status) {if (U_FAILURE(status)) { \
1034 errcheckln(status, "error at file %s, line %d, status = %s", __FILE__, __LINE__, \
1035 u_errorName(status)); \
1036 return;}}
1037
1038
writeStringInU8(FILE * out,const UnicodeString & s)1039 static void writeStringInU8(FILE *out, const UnicodeString &s) {
1040 int i;
1041 for (i=0; i<s.length(); i=s.moveIndex32(i, 1)) {
1042 UChar32 c = s.char32At(i);
1043 uint8_t bufForOneChar[10];
1044 UBool isError = FALSE;
1045 int32_t destIdx = 0;
1046 U8_APPEND(bufForOneChar, destIdx, (int32_t)sizeof(bufForOneChar), c, isError);
1047 fwrite(bufForOneChar, 1, destIdx, out);
1048 }
1049 }
1050
1051
1052
1053
TestHan()1054 void TransliteratorRoundTripTest::TestHan() {
1055 UErrorCode status = U_ZERO_ERROR;
1056 LocalULocaleDataPointer uld(ulocdata_open("zh",&status));
1057 LocalUSetPointer USetExemplars(ulocdata_getExemplarSet(uld.getAlias(), uset_openEmpty(), 0, ULOCDATA_ES_STANDARD, &status));
1058 ASSERT_SUCCESS(status);
1059
1060 UnicodeString source;
1061 UChar32 c;
1062 int i;
1063 for (i=0; ;i++) {
1064 // Add all of the Chinese exemplar chars to the string "source".
1065 c = uset_charAt(USetExemplars.getAlias(), i);
1066 if (c == (UChar32)-1) {
1067 break;
1068 }
1069 source.append(c);
1070 }
1071
1072 // transform with Han translit
1073 Transliterator *hanTL = Transliterator::createInstance("Han-Latin", UTRANS_FORWARD, status);
1074 ASSERT_SUCCESS(status);
1075 UnicodeString target=source;
1076 hanTL->transliterate(target);
1077 // now verify that there are no Han characters left
1078 UnicodeSet allHan("[:han:]", status);
1079 ASSERT_SUCCESS(status);
1080 if (allHan.containsSome(target)) {
1081 errln("file %s, line %d, No Han must be left after Han-Latin transliteration",
1082 __FILE__, __LINE__);
1083 }
1084
1085 // check the pinyin translit
1086 Transliterator *pn = Transliterator::createInstance("Latin-NumericPinyin", UTRANS_FORWARD, status);
1087 ASSERT_SUCCESS(status);
1088 UnicodeString target2 = target;
1089 pn->transliterate(target2);
1090
1091 // verify that there are no marks
1092 Transliterator *nfd = Transliterator::createInstance("nfd", UTRANS_FORWARD, status);
1093 ASSERT_SUCCESS(status);
1094
1095 UnicodeString nfded = target2;
1096 nfd->transliterate(nfded);
1097 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)
1098 ASSERT_SUCCESS(status);
1099 assertFalse("NumericPinyin must contain no marks", allMarks.containsSome(nfded));
1100
1101 // verify roundtrip
1102 Transliterator *np = pn->createInverse(status);
1103 ASSERT_SUCCESS(status);
1104 UnicodeString target3 = target2;
1105 np->transliterate(target3);
1106 UBool roundtripOK = (target3.compare(target) == 0);
1107 assertTrue("NumericPinyin must roundtrip", roundtripOK);
1108 if (!roundtripOK) {
1109 const char *filename = "numeric-pinyin.log.txt";
1110 FILE *out = fopen(filename, "w");
1111 errln("Creating log file %s\n", filename);
1112 fprintf(out, "Pinyin: ");
1113 writeStringInU8(out, target);
1114 fprintf(out, "\nPinyin-Numeric-Pinyin: ");
1115 writeStringInU8(out, target2);
1116 fprintf(out, "\nNumeric-Pinyin-Pinyin: ");
1117 writeStringInU8(out, target3);
1118 fprintf(out, "\n");
1119 fclose(out);
1120 }
1121
1122 delete hanTL;
1123 delete pn;
1124 delete nfd;
1125 delete np;
1126 }
1127
1128
TestGreek()1129 void TransliteratorRoundTripTest::TestGreek() {
1130
1131 // CLDR bug #1911: This test should be moved into CLDR.
1132 // It is left in its current state as a regression test.
1133
1134 // if (isICUVersionAtLeast(ICU_39)) {
1135 // // We temporarily filter against Unicode 4.1, but we only do this
1136 // // before version 3.4.
1137 // errln("FAIL: TestGreek needs to be updated to remove delete the [:Age=4.0:] filter ");
1138 // return;
1139 // } else {
1140 // logln("Warning: TestGreek needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1141 // }
1142
1143 RTTest test("Latin-Greek");
1144 LegalGreek *legal = new LegalGreek(TRUE);
1145
1146 test.test(UnicodeString("[a-zA-Z]", ""),
1147 UnicodeString("[\\u003B\\u00B7[[:Greek:]&[:Letter:]]-["
1148 "\\u1D26-\\u1D2A" // L& [5] GREEK LETTER SMALL CAPITAL GAMMA..GREEK LETTER SMALL CAPITAL PSI
1149 "\\u1D5D-\\u1D61" // Lm [5] MODIFIER LETTER SMALL BETA..MODIFIER LETTER SMALL CHI
1150 "\\u1D66-\\u1D6A" // L& [5] GREEK SUBSCRIPT SMALL LETTER BETA..GREEK SUBSCRIPT SMALL LETTER CHI
1151 "\\u03D7-\\u03EF" // \N{GREEK KAI SYMBOL}..\N{COPTIC SMALL LETTER DEI}
1152 "] & [:Age=4.0:]]",
1153
1154 //UnicodeString("[[\\u003B\\u00B7[:Greek:]-[\\u0374\\u0385\\u1fcd\\u1fce\\u1fdd\\u1fde\\u1fed-\\u1fef\\u1ffd\\u03D7-\\u03EF]]&[:Age=3.2:]]",
1155 ""),
1156 "[\\u00B5\\u037A\\u03D0-\\u03F5\\u03f9]", /* exclusions */
1157 this, quick, legal, 50);
1158
1159
1160 delete legal;
1161 }
1162
1163
TestGreekUNGEGN()1164 void TransliteratorRoundTripTest::TestGreekUNGEGN() {
1165
1166 // CLDR bug #1911: This test should be moved into CLDR.
1167 // It is left in its current state as a regression test.
1168
1169 // if (isICUVersionAtLeast(ICU_39)) {
1170 // // We temporarily filter against Unicode 4.1, but we only do this
1171 // // before version 3.4.
1172 // errln("FAIL: TestGreek needs to be updated to remove delete the [:Age=4.0:] filter ");
1173 // return;
1174 // } else {
1175 // logln("Warning: TestGreek needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1176 // }
1177
1178 RTTest test("Latin-Greek/UNGEGN");
1179 LegalGreek *legal = new LegalGreek(FALSE);
1180
1181 test.test(UnicodeString("[a-zA-Z]", ""),
1182 UnicodeString("[\\u003B\\u00B7[[:Greek:]&[:Letter:]]-["
1183 "\\u1D26-\\u1D2A" // L& [5] GREEK LETTER SMALL CAPITAL GAMMA..GREEK LETTER SMALL CAPITAL PSI
1184 "\\u1D5D-\\u1D61" // Lm [5] MODIFIER LETTER SMALL BETA..MODIFIER LETTER SMALL CHI
1185 "\\u1D66-\\u1D6A" // L& [5] GREEK SUBSCRIPT SMALL LETTER BETA..GREEK SUBSCRIPT SMALL LETTER CHI
1186 "\\u03D7-\\u03EF" // \N{GREEK KAI SYMBOL}..\N{COPTIC SMALL LETTER DEI}
1187 "] & [:Age=4.0:]]",
1188 //UnicodeString("[[\\u003B\\u00B7[:Greek:]-[\\u0374\\u0385\\u1fce\\u1fde\\u03D7-\\u03EF]]&[:Age=3.2:]]",
1189 ""),
1190 "[\\u0385\\u00B5\\u037A\\u03D0-\\uFFFF {\\u039C\\u03C0}]", /* roundtrip exclusions */
1191 this, quick, legal);
1192
1193 delete legal;
1194 }
1195
Testel()1196 void TransliteratorRoundTripTest::Testel() {
1197
1198 // CLDR bug #1911: This test should be moved into CLDR.
1199 // It is left in its current state as a regression test.
1200
1201 // if (isICUVersionAtLeast(ICU_39)) {
1202 // // We temporarily filter against Unicode 4.1, but we only do this
1203 // // before version 3.4.
1204 // errln("FAIL: TestGreek needs to be updated to remove delete the [:Age=4.0:] filter ");
1205 // return;
1206 // } else {
1207 // logln("Warning: TestGreek needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1208 // }
1209
1210 RTTest test("Latin-el");
1211 LegalGreek *legal = new LegalGreek(FALSE);
1212
1213 test.test(UnicodeString("[a-zA-Z]", ""),
1214 UnicodeString("[\\u003B\\u00B7[[:Greek:]&[:Letter:]]-["
1215 "\\u1D26-\\u1D2A" // L& [5] GREEK LETTER SMALL CAPITAL GAMMA..GREEK LETTER SMALL CAPITAL PSI
1216 "\\u1D5D-\\u1D61" // Lm [5] MODIFIER LETTER SMALL BETA..MODIFIER LETTER SMALL CHI
1217 "\\u1D66-\\u1D6A" // L& [5] GREEK SUBSCRIPT SMALL LETTER BETA..GREEK SUBSCRIPT SMALL LETTER CHI
1218 "\\u03D7-\\u03EF" // \N{GREEK KAI SYMBOL}..\N{COPTIC SMALL LETTER DEI}
1219 "] & [:Age=4.0:]]",
1220 //UnicodeString("[[\\u003B\\u00B7[:Greek:]-[\\u0374\\u0385\\u1fce\\u1fde\\u03D7-\\u03EF]]&[:Age=3.2:]]",
1221 ""),
1222 "[\\u00B5\\u037A\\u03D0-\\uFFFF {\\u039C\\u03C0}]", /* exclusions */
1223 this, quick, legal);
1224
1225
1226 delete legal;
1227 }
1228
1229
TestArabic()1230 void TransliteratorRoundTripTest::TestArabic() {
1231 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);
1232 Legal *legal = new Legal();
1233 RTTest test("Latin-Arabic");
1234 test.test(UNICODE_STRING_SIMPLE("[a-zA-Z\\u02BE\\u02BF\\u207F]"), ARABIC, "[a-zA-Z\\u02BE\\u02BF\\u207F]",this, quick, legal); //
1235 delete legal;
1236 }
1237 class LegalHebrew : public Legal {
1238 private:
1239 UnicodeSet FINAL;
1240 UnicodeSet NON_FINAL;
1241 UnicodeSet LETTER;
1242 public:
1243 LegalHebrew(UErrorCode& error);
~LegalHebrew()1244 virtual ~LegalHebrew() {}
1245 virtual UBool is(const UnicodeString& sourceString) const;
1246 };
1247
LegalHebrew(UErrorCode & error)1248 LegalHebrew::LegalHebrew(UErrorCode& error){
1249 FINAL.applyPattern(UNICODE_STRING_SIMPLE("[\\u05DA\\u05DD\\u05DF\\u05E3\\u05E5]"), error);
1250 NON_FINAL.applyPattern(UNICODE_STRING_SIMPLE("[\\u05DB\\u05DE\\u05E0\\u05E4\\u05E6]"), error);
1251 LETTER.applyPattern("[:letter:]", error);
1252 }
is(const UnicodeString & sourceString) const1253 UBool LegalHebrew::is(const UnicodeString& sourceString)const{
1254
1255 if (sourceString.length() == 0) return TRUE;
1256 // don't worry about surrogates.
1257 for (int i = 0; i < sourceString.length(); ++i) {
1258 UChar ch = sourceString.charAt(i);
1259 UChar next = i+1 == sourceString.length() ? 0x0000 : sourceString.charAt(i);
1260 if (FINAL.contains(ch)) {
1261 if (LETTER.contains(next)) return FALSE;
1262 } else if (NON_FINAL.contains(ch)) {
1263 if (!LETTER.contains(next)) return FALSE;
1264 }
1265 }
1266 return TRUE;
1267 }
TestHebrew()1268 void TransliteratorRoundTripTest::TestHebrew() {
1269 // CLDR bug #1911: This test should be moved into CLDR.
1270 // It is left in its current state as a regression test.
1271 // if (isICUVersionAtLeast(ICU_39)) {
1272 // // We temporarily filter against Unicode 4.1, but we only do this
1273 // // before version 3.4.
1274 // errln("FAIL: TestHebrew needs to be updated to remove delete the [:Age=4.0:] filter ");
1275 // return;
1276 // } else {
1277 // logln("Warning: TestHebrew needs to be updated to remove delete the section marked [:Age=4.0:] filter");
1278 // }
1279 //long start = System.currentTimeMillis();
1280 UErrorCode error = U_ZERO_ERROR;
1281 LegalHebrew* legal = new LegalHebrew(error);
1282 if(U_FAILURE(error)){
1283 dataerrln("Could not construct LegalHebrew object. Error: %s", u_errorName(error));
1284 return;
1285 }
1286 RTTest test("Latin-Hebrew");
1287 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);
1288
1289 //showElapsed(start, "TestHebrew");
1290 delete legal;
1291 }
TestCyrillic()1292 void TransliteratorRoundTripTest::TestCyrillic() {
1293 RTTest test("Latin-Cyrillic");
1294 Legal *legal = new Legal();
1295
1296 test.test(UnicodeString("[a-zA-Z\\u0110\\u0111\\u02BA\\u02B9]", ""),
1297 UnicodeString("[[\\u0400-\\u045F] & [:Age=3.2:]]", ""), NULL, this, quick,
1298 legal);
1299
1300 delete legal;
1301 }
1302
1303
1304 // Inter-Indic Tests ----------------------------------
1305 class LegalIndic :public Legal{
1306 UnicodeSet vowelSignSet;
1307 UnicodeSet avagraha;
1308 UnicodeSet nukta;
1309 UnicodeSet virama;
1310 UnicodeSet sanskritStressSigns;
1311 UnicodeSet chandrabindu;
1312
1313 public:
1314 LegalIndic();
1315 virtual UBool is(const UnicodeString& sourceString) const;
~LegalIndic()1316 virtual ~LegalIndic() {};
1317 };
is(const UnicodeString & sourceString) const1318 UBool LegalIndic::is(const UnicodeString& sourceString) const{
1319 int cp=sourceString.charAt(0);
1320
1321 // A vowel sign cannot be the first char
1322 if(vowelSignSet.contains(cp)){
1323 return FALSE;
1324 }else if(avagraha.contains(cp)){
1325 return FALSE;
1326 }else if(virama.contains(cp)){
1327 return FALSE;
1328 }else if(nukta.contains(cp)){
1329 return FALSE;
1330 }else if(sanskritStressSigns.contains(cp)){
1331 return FALSE;
1332 }else if(chandrabindu.contains(cp) &&
1333 ((sourceString.length()>1) &&
1334 vowelSignSet.contains(sourceString.charAt(1)))){
1335 return FALSE;
1336 }
1337 return TRUE;
1338 }
LegalIndic()1339 LegalIndic::LegalIndic(){
1340 UErrorCode status = U_ZERO_ERROR;
1341 vowelSignSet.addAll( UnicodeSet("[\\u0902\\u0903\\u0904\\u093e-\\u094c\\u0962\\u0963]",status));/* Devanagari */
1342 vowelSignSet.addAll( UnicodeSet("[\\u0982\\u0983\\u09be-\\u09cc\\u09e2\\u09e3\\u09D7]",status));/* Bengali */
1343 vowelSignSet.addAll( UnicodeSet("[\\u0a02\\u0a03\\u0a3e-\\u0a4c\\u0a62\\u0a63\\u0a70\\u0a71]",status));/* Gurmukhi */
1344 vowelSignSet.addAll( UnicodeSet("[\\u0a82\\u0a83\\u0abe-\\u0acc\\u0ae2\\u0ae3]",status));/* Gujarati */
1345 vowelSignSet.addAll( UnicodeSet("[\\u0b02\\u0b03\\u0b3e-\\u0b4c\\u0b62\\u0b63\\u0b56\\u0b57]",status));/* Oriya */
1346 vowelSignSet.addAll( UnicodeSet("[\\u0b82\\u0b83\\u0bbe-\\u0bcc\\u0be2\\u0be3\\u0bd7]",status));/* Tamil */
1347 vowelSignSet.addAll( UnicodeSet("[\\u0c02\\u0c03\\u0c3e-\\u0c4c\\u0c62\\u0c63\\u0c55\\u0c56]",status));/* Telugu */
1348 vowelSignSet.addAll( UnicodeSet("[\\u0c82\\u0c83\\u0cbe-\\u0ccc\\u0ce2\\u0ce3\\u0cd5\\u0cd6]",status));/* Kannada */
1349 vowelSignSet.addAll( UnicodeSet("[\\u0d02\\u0d03\\u0d3e-\\u0d4c\\u0d62\\u0d63\\u0d57]",status));/* Malayalam */
1350
1351 avagraha.addAll(UnicodeSet("[\\u093d\\u09bd\\u0abd\\u0b3d\\u0cbd]",status));
1352 nukta.addAll(UnicodeSet("[\\u093c\\u09bc\\u0a3c\\u0abc\\u0b3c\\u0cbc]",status));
1353 virama.addAll(UnicodeSet("[\\u094d\\u09cd\\u0a4d\\u0acd\\u0b4d\\u0bcd\\u0c4d\\u0ccd\\u0d4d]",status));
1354 sanskritStressSigns.addAll(UnicodeSet("[\\u0951\\u0952\\u0953\\u0954\\u097d]",status));
1355 chandrabindu.addAll(UnicodeSet("[\\u0901\\u0981\\u0A81\\u0b01\\u0c01]",status));
1356
1357 }
1358
1359 static const char latinForIndic[] = "[['.0-9A-Za-z~\\u00C0-\\u00C5\\u00C7-\\u00CF\\u00D1-\\u00D6\\u00D9-\\u00DD"
1360 "\\u00E0-\\u00E5\\u00E7-\\u00EF\\u00F1-\\u00F6\\u00F9-\\u00FD\\u00FF-\\u010F"
1361 "\\u0112-\\u0125\\u0128-\\u0130\\u0134-\\u0137\\u0139-\\u013E\\u0143-\\u0148"
1362 "\\u014C-\\u0151\\u0154-\\u0165\\u0168-\\u017E\\u01A0-\\u01A1\\u01AF-\\u01B0"
1363 "\\u01CD-\\u01DC\\u01DE-\\u01E3\\u01E6-\\u01ED\\u01F0\\u01F4-\\u01F5\\u01F8-\\u01FB"
1364 "\\u0200-\\u021B\\u021E-\\u021F\\u0226-\\u0233\\u0294\\u0303-\\u0304\\u0306\\u0314-\\u0315"
1365 "\\u0325\\u040E\\u0419\\u0439\\u045E\\u04C1-\\u04C2\\u04D0-\\u04D1\\u04D6-\\u04D7"
1366 "\\u04E2-\\u04E3\\u04EE-\\u04EF\\u1E00-\\u1E99\\u1EA0-\\u1EF9\\u1F01\\u1F03\\u1F05"
1367 "\\u1F07\\u1F09\\u1F0B\\u1F0D\\u1F0F\\u1F11\\u1F13\\u1F15\\u1F19\\u1F1B\\u1F1D\\u1F21"
1368 "\\u1F23\\u1F25\\u1F27\\u1F29\\u1F2B\\u1F2D\\u1F2F\\u1F31\\u1F33\\u1F35\\u1F37\\u1F39"
1369 "\\u1F3B\\u1F3D\\u1F3F\\u1F41\\u1F43\\u1F45\\u1F49\\u1F4B\\u1F4D\\u1F51\\u1F53\\u1F55"
1370 "\\u1F57\\u1F59\\u1F5B\\u1F5D\\u1F5F\\u1F61\\u1F63\\u1F65\\u1F67\\u1F69\\u1F6B\\u1F6D"
1371 "\\u1F6F\\u1F81\\u1F83\\u1F85\\u1F87\\u1F89\\u1F8B\\u1F8D\\u1F8F\\u1F91\\u1F93\\u1F95"
1372 "\\u1F97\\u1F99\\u1F9B\\u1F9D\\u1F9F\\u1FA1\\u1FA3\\u1FA5\\u1FA7\\u1FA9\\u1FAB\\u1FAD"
1373 "\\u1FAF-\\u1FB1\\u1FB8-\\u1FB9\\u1FD0-\\u1FD1\\u1FD8-\\u1FD9\\u1FE0-\\u1FE1\\u1FE5"
1374 "\\u1FE8-\\u1FE9\\u1FEC\\u212A-\\u212B\\uE04D\\uE064]"
1375 "-[\\uE000-\\uE080 \\u01E2\\u01E3]& [[:latin:][:mark:]]]";
1376
TestDevanagariLatin()1377 void TransliteratorRoundTripTest::TestDevanagariLatin() {
1378 {
1379 UErrorCode status = U_ZERO_ERROR;
1380 UParseError parseError;
1381 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));
1382 if((Transliterator *)t1 != NULL){
1383 TransliteratorPointer t2(t1->createInverse(status));
1384 if(U_FAILURE(status)){
1385 errln("FAIL: could not create the Inverse:-( \n");
1386 }
1387 }else {
1388 dataerrln("FAIL: could not create the transliterator. Error: %s\n", u_errorName(status));
1389 }
1390
1391 }
1392 RTTest test("Latin-Devanagari");
1393 Legal *legal = new LegalIndic();
1394 // CLDR bug #1911: This test should be moved into CLDR.
1395 // It is left in its current state as a regression test.
1396 // if (isICUVersionAtLeast(ICU_39)) {
1397 // // We temporarily filter against Unicode 4.1, but we only do this
1398 // // before version 3.4.
1399 // errln("FAIL: TestDevanagariLatin needs to be updated to remove delete the [:Age=4.1:] filter ");
1400 // return;
1401 // } else {
1402 // logln("Warning: TestDevanagariLatin needs to be updated to remove delete the section marked [:Age=4.1:] filter");
1403 // }
1404 test.test(UnicodeString(latinForIndic, ""),
1405 UnicodeString("[[[:Devanagari:][\\u094d][\\u0964\\u0965]]&[:Age=4.1:]-[\\u0970]]", ""), "[\\u0965\\u0904]", this, quick,
1406 legal, 50);
1407
1408 delete legal;
1409 }
1410
1411 /* Defined this way for HP/UX11CC :-( */
1412 static const int32_t INTER_INDIC_ARRAY_WIDTH = 4;
1413 static const char * const interIndicArray[] = {
1414
1415 "BENGALI-DEVANAGARI", "[:BENGALI:]", "[[:Devanagari:]-[\\u0970]]",
1416 "[\\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*/
1417
1418 "DEVANAGARI-BENGALI", "[[:Devanagari:]-[\\u0970]]", "[:BENGALI:]",
1419 "[\\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*/
1420
1421 "GURMUKHI-DEVANAGARI", "[:GURMUKHI:]", "[[:Devanagari:]-[\\u0970]]",
1422 "[\\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*/
1423
1424 "DEVANAGARI-GURMUKHI", "[[:Devanagari:]-[\\u0970]]", "[:GURMUKHI:]",
1425 "[\\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*/
1426
1427 "GUJARATI-DEVANAGARI", "[:GUJARATI:]", "[[:Devanagari:]-[\\u0970]]",
1428 "[\\u0946\\u094A\\u0962\\u0963\\u0951-\\u0954\\u0961\\u090c\\u090e\\u0912\\u097d]", /*roundtrip exclusions*/
1429
1430 "DEVANAGARI-GUJARATI", "[[:Devanagari:]-[\\u0970]]", "[:GUJARATI:]",
1431 "[\\u0951-\\u0954\\u0961\\u090c\\u090e\\u0912]", /*roundtrip exclusions*/
1432
1433 "ORIYA-DEVANAGARI", "[:ORIYA:]", "[[:Devanagari:]-[\\u0970]]",
1434 "[\\u0904\\u0943-\\u094a\\u0962\\u0963\\u0951-\\u0954\\u0950\\u090D\\u090e\\u0912\\u0911\\u0931\\u0935\\u097d]", /*roundtrip exclusions*/
1435
1436 "DEVANAGARI-ORIYA", "[[:Devanagari:]-[\\u0970]]", "[:ORIYA:]",
1437 "[\\u0b5f\\u0b56\\u0b57\\u0b70\\u0b71\\u0950\\u090D\\u090e\\u0912\\u0911\\u0931]", /*roundtrip exclusions*/
1438
1439 "Tamil-DEVANAGARI", "[:tamil:]", "[[:Devanagari:]-[\\u0970]]",
1440 "[\\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*/
1441
1442 "DEVANAGARI-Tamil", "[[:Devanagari:]-[\\u0970]]", "[:tamil:]",
1443 "[\\u0bd7\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1444
1445 "Telugu-DEVANAGARI", "[:telugu:]", "[[:Devanagari:]-[\\u0970]]",
1446 "[\\u0904\\u093c\\u0950\\u0945\\u0949\\u0951-\\u0954\\u0962\\u0963\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]\\u097d]", /*roundtrip exclusions*/
1447
1448 "DEVANAGARI-TELUGU", "[[:Devanagari:]-[\\u0970]]", "[:TELUGU:]",
1449 "[\\u0c55\\u0c56\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]]", /*roundtrip exclusions*/
1450
1451 "KANNADA-DEVANAGARI", "[:KANNADA:]", "[[:Devanagari:]-[\\u0970]]",
1452 "[\\u0901\\u0904\\u0946\\u093c\\u0950\\u0945\\u0949\\u0951-\\u0954\\u0962\\u0963\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]\\u097d]", /*roundtrip exclusions*/
1453
1454 "DEVANAGARI-KANNADA", "[[:Devanagari:]-[\\u0970]]", "[:KANNADA:]",
1455 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cde\\u0cd5\\u0cd6\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]]", /*roundtrip exclusions*/
1456
1457 "MALAYALAM-DEVANAGARI", "[:MALAYALAM:]", "[[:Devanagari:]-[\\u0970]]",
1458 "[\\u0901\\u0904\\u094a\\u094b\\u094c\\u093c\\u0950\\u0944\\u0945\\u0949\\u0951-\\u0954\\u0962\\u0963\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]\\u097d]", /*roundtrip exclusions*/
1459
1460 "DEVANAGARI-MALAYALAM", "[[:Devanagari:]-[\\u0970]]", "[:MALAYALAM:]",
1461 "[\\u0d4c\\u0d57\\u0950\\u090D\\u0911\\u093d\\u0929\\u0934[\\u0958-\\u095f]]", /*roundtrip exclusions*/
1462
1463 "GURMUKHI-BENGALI", "[:GURMUKHI:]", "[:BENGALI:]",
1464 "[\\u0981\\u0982\\u09b6\\u09e2\\u09e3\\u09c3\\u09c4\\u09d7\\u098B\\u098C\\u09B7\\u09E0\\u09E1\\u09F0\\u09F1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1465
1466 "BENGALI-GURMUKHI", "[:BENGALI:]", "[:GURMUKHI:]",
1467 "[\\u0A02\\u0a5c\\u0a47\\u0a70\\u0a71\\u0A33\\u0A35\\u0A59\\u0A5A\\u0A5B\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1468
1469 "GUJARATI-BENGALI", "[:GUJARATI:]", "[:BENGALI:]",
1470 "[\\u09d7\\u09e2\\u09e3\\u098c\\u09e1\\u09f0\\u09f1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1471
1472 "BENGALI-GUJARATI", "[:BENGALI:]", "[:GUJARATI:]",
1473 "[\\u0A82\\u0a83\\u0Ac9\\u0Ac5\\u0ac7\\u0A8D\\u0A91\\u0AB3\\u0AB5\\u0ABD\\u0AD0]", /*roundtrip exclusions*/
1474
1475 "ORIYA-BENGALI", "[:ORIYA:]", "[:BENGALI:]",
1476 "[\\u09c4\\u09e2\\u09e3\\u09f0\\u09f1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1477
1478 "BENGALI-ORIYA", "[:BENGALI:]", "[:ORIYA:]",
1479 "[\\u0b35\\u0b71\\u0b5f\\u0b56\\u0b33\\u0b3d]", /*roundtrip exclusions*/
1480
1481 "Tamil-BENGALI", "[:tamil:]", "[:BENGALI:]",
1482 "[\\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*/
1483
1484 "BENGALI-Tamil", "[:BENGALI:]", "[:tamil:]",
1485 "[\\u0bc6\\u0bc7\\u0bca\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB3\\u0BB4\\u0BB5\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1486
1487 "Telugu-BENGALI", "[:telugu:]", "[:BENGALI:]",
1488 "[\\u09e2\\u09e3\\u09bc\\u09d7\\u09f0\\u09f1\\u09dc\\u09dd\\u09df\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1489
1490 "BENGALI-TELUGU", "[:BENGALI:]", "[:TELUGU:]",
1491 "[\\u0c55\\u0c56\\u0c47\\u0c46\\u0c4a\\u0C0E\\u0C12\\u0C31\\u0C33\\u0C35]", /*roundtrip exclusions*/
1492
1493 "KANNADA-BENGALI", "[:KANNADA:]", "[:BENGALI:]",
1494 "[\\u0981\\u09e2\\u09e3\\u09bc\\u09d7\\u09dc\\u09dd\\u09df\\u09f0\\u09f1\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1495
1496 "BENGALI-KANNADA", "[:BENGALI:]", "[:KANNADA:]",
1497 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cc6\\u0cca\\u0cd5\\u0cd6\\u0cc7\\u0C8E\\u0C92\\u0CB1\\u0cb3\\u0cb5\\u0cde]", /*roundtrip exclusions*/
1498
1499 "MALAYALAM-BENGALI", "[:MALAYALAM:]", "[:BENGALI:]",
1500 "[\\u0981\\u09e2\\u09e3\\u09bc\\u09c4\\u09f0\\u09f1\\u09dc\\u09dd\\u09df\\u09dc\\u09dd\\u09df\\u09f2-\\u09fa\\u09ce]", /*roundtrip exclusions*/
1501
1502 "BENGALI-MALAYALAM", "[:BENGALI:]", "[:MALAYALAM:]",
1503 "[\\u0d46\\u0d4a\\u0d47\\u0d31-\\u0d35\\u0d0e\\u0d12]", /*roundtrip exclusions*/
1504
1505 "GUJARATI-GURMUKHI", "[:GUJARATI:]", "[:GURMUKHI:]",
1506 "[\\u0A02\\u0ab3\\u0ab6\\u0A70\\u0a71\\u0a82\\u0a83\\u0ac3\\u0ac4\\u0ac5\\u0ac9\\u0a5c\\u0a72\\u0a73\\u0a74\\u0a8b\\u0a8d\\u0a91\\u0abd]", /*roundtrip exclusions*/
1507
1508 "GURMUKHI-GUJARATI", "[:GURMUKHI:]", "[:GUJARATI:]",
1509 "[\\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*/
1510
1511 "ORIYA-GURMUKHI", "[:ORIYA:]", "[:GURMUKHI:]",
1512 "[\\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*/
1513
1514 "GURMUKHI-ORIYA", "[:GURMUKHI:]", "[:ORIYA:]",
1515 "[\\u0b01\\u0b02\\u0b03\\u0b33\\u0b36\\u0b43\\u0b56\\u0b57\\u0B0B\\u0B0C\\u0B37\\u0B3D\\u0B5F\\u0B60\\u0B61\\u0b70\\u0b71]", /*roundtrip exclusions*/
1516
1517 "TAMIL-GURMUKHI", "[:TAMIL:]", "[:GURMUKHI:]",
1518 "[\\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*/
1519
1520 "GURMUKHI-TAMIL", "[:GURMUKHI:]", "[:TAMIL:]",
1521 "[\\u0b82\\u0bc6\\u0bca\\u0bd7\\u0bb7\\u0bb3\\u0b83\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB4\\u0bb6\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1522
1523 "TELUGU-GURMUKHI", "[:TELUGU:]", "[:GURMUKHI:]",
1524 "[\\u0A02\\u0a33\\u0a36\\u0a3c\\u0a70\\u0a71\\u0A59\\u0A5A\\u0A5B\\u0A5C\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1525
1526 "GURMUKHI-TELUGU", "[:GURMUKHI:]", "[:TELUGU:]",
1527 "[\\u0c01\\u0c02\\u0c03\\u0c33\\u0c36\\u0c44\\u0c43\\u0c46\\u0c4a\\u0c56\\u0c55\\u0C0B\\u0C0C\\u0C0E\\u0C12\\u0C31\\u0C37\\u0C60\\u0C61]", /*roundtrip exclusions*/
1528
1529 "KANNADA-GURMUKHI", "[:KANNADA:]", "[:GURMUKHI:]",
1530 "[\\u0A01\\u0A02\\u0a33\\u0a36\\u0a3c\\u0a70\\u0a71\\u0A59\\u0A5A\\u0A5B\\u0A5C\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1531
1532 "GURMUKHI-KANNADA", "[:GURMUKHI:]", "[:KANNADA:]",
1533 "[{\\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*/
1534
1535 "MALAYALAM-GURMUKHI", "[:MALAYALAM:]", "[:GURMUKHI:]",
1536 "[\\u0A01\\u0A02\\u0a4b\\u0a4c\\u0a33\\u0a36\\u0a3c\\u0a70\\u0a71\\u0A59\\u0A5A\\u0A5B\\u0A5C\\u0A5E\\u0A72\\u0A73\\u0A74]", /*roundtrip exclusions*/
1537
1538 "GURMUKHI-MALAYALAM", "[:GURMUKHI:]", "[:MALAYALAM:]",
1539 "[\\u0d02\\u0d03\\u0d33\\u0d36\\u0d43\\u0d46\\u0d4a\\u0d4c\\u0d57\\u0D0B\\u0D0C\\u0D0E\\u0D12\\u0D31\\u0D34\\u0D37\\u0D60\\u0D61]", /*roundtrip exclusions*/
1540
1541 "GUJARATI-ORIYA", "[:GUJARATI:]", "[:ORIYA:]",
1542 "[\\u0b56\\u0b57\\u0B0C\\u0B5F\\u0B61\\u0b70\\u0b71]", /*roundtrip exclusions*/
1543
1544 "ORIYA-GUJARATI", "[:ORIYA:]", "[:GUJARATI:]",
1545 "[\\u0Ac4\\u0Ac5\\u0Ac9\\u0Ac7\\u0A8D\\u0A91\\u0AB5\\u0Ad0]", /*roundtrip exclusions*/
1546
1547 "TAMIL-GUJARATI", "[:TAMIL:]", "[:GUJARATI:]",
1548 "[\\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*/
1549
1550 "GUJARATI-TAMIL", "[:GUJARATI:]", "[:TAMIL:]",
1551 "[\\u0Bc6\\u0Bca\\u0Bd7\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB4\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1552
1553 "TELUGU-GUJARATI", "[:TELUGU:]", "[:GUJARATI:]",
1554 "[\\u0abc\\u0Ac5\\u0Ac9\\u0A8D\\u0A91\\u0ABD\\u0Ad0]", /*roundtrip exclusions*/
1555
1556 "GUJARATI-TELUGU", "[:GUJARATI:]", "[:TELUGU:]",
1557 "[\\u0c46\\u0c4a\\u0c55\\u0c56\\u0C0C\\u0C0E\\u0C12\\u0C31\\u0C61]", /*roundtrip exclusions*/
1558
1559 "KANNADA-GUJARATI", "[:KANNADA:]", "[:GUJARATI:]",
1560 "[\\u0A81\\u0abc\\u0Ac5\\u0Ac9\\u0A8D\\u0A91\\u0ABD\\u0Ad0]", /*roundtrip exclusions*/
1561
1562 "GUJARATI-KANNADA", "[:GUJARATI:]", "[:KANNADA:]",
1563 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cc6\\u0cca\\u0cd5\\u0cd6\\u0C8C\\u0C8E\\u0C92\\u0CB1\\u0CDE\\u0CE1]", /*roundtrip exclusions*/
1564
1565 "MALAYALAM-GUJARATI", "[:MALAYALAM:]", "[:GUJARATI:]",
1566 "[\\u0A81\\u0ac4\\u0acb\\u0acc\\u0abc\\u0Ac5\\u0Ac9\\u0A8D\\u0A91\\u0ABD\\u0Ad0]", /*roundtrip exclusions*/
1567
1568 "GUJARATI-MALAYALAM", "[:GUJARATI:]", "[:MALAYALAM:]",
1569 "[\\u0d46\\u0d4a\\u0d4c\\u0d55\\u0d57\\u0D0C\\u0D0E\\u0D12\\u0D31\\u0D34\\u0D61]", /*roundtrip exclusions*/
1570
1571 "TAMIL-ORIYA", "[:TAMIL:]", "[:ORIYA:]",
1572 "[\\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*/
1573
1574 "ORIYA-TAMIL", "[:ORIYA:]", "[:TAMIL:]",
1575 "[\\u0bc6\\u0bca\\u0bc7\\u0B8E\\u0B92\\u0BA9\\u0BB1\\u0BB4\\u0BB5\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1576
1577 "TELUGU-ORIYA", "[:TELUGU:]", "[:ORIYA:]",
1578 "[\\u0b3c\\u0b57\\u0b56\\u0B3D\\u0B5C\\u0B5D\\u0B5F\\u0b70\\u0b71]", /*roundtrip exclusions*/
1579
1580 "ORIYA-TELUGU", "[:ORIYA:]", "[:TELUGU:]",
1581 "[\\u0c44\\u0c46\\u0c4a\\u0c55\\u0c47\\u0C0E\\u0C12\\u0C31\\u0C35]", /*roundtrip exclusions*/
1582
1583 "KANNADA-ORIYA", "[:KANNADA:]", "[:ORIYA:]",
1584 "[\\u0B01\\u0b3c\\u0b57\\u0B3D\\u0B5C\\u0B5D\\u0B5F\\u0b70\\u0b71]", /*roundtrip exclusions*/
1585
1586 "ORIYA-KANNADA", "[:ORIYA:]", "[:KANNADA:]",
1587 "[{\\u0cb0\\u0cbc}{\\u0cb3\\u0cbc}\\u0cc4\\u0cc6\\u0cca\\u0cd5\\u0cc7\\u0C8E\\u0C92\\u0CB1\\u0CB5\\u0CDE]", /*roundtrip exclusions*/
1588
1589 "MALAYALAM-ORIYA", "[:MALAYALAM:]", "[:ORIYA:]",
1590 "[\\u0B01\\u0b3c\\u0b56\\u0B3D\\u0B5C\\u0B5D\\u0B5F\\u0b70\\u0b71]", /*roundtrip exclusions*/
1591
1592 "ORIYA-MALAYALAM", "[:ORIYA:]", "[:MALAYALAM:]",
1593 "[\\u0D47\\u0D46\\u0D4a\\u0D0E\\u0D12\\u0D31\\u0D34\\u0D35]", /*roundtrip exclusions*/
1594
1595 "TELUGU-TAMIL", "[:TELUGU:]", "[:TAMIL:]",
1596 "[\\u0bd7\\u0ba9\\u0bb4\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1597
1598 "TAMIL-TELUGU", "[:TAMIL:]", "[:TELUGU:]",
1599 "[\\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*/
1600
1601 "KANNADA-TAMIL", "[:KANNADA:]", "[:TAMIL:]",
1602 "[\\u0bd7\\u0bc6\\u0ba9\\u0bb4\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1603
1604 "TAMIL-KANNADA", "[:TAMIL:]", "[:KANNADA:]",
1605 "[\\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*/
1606
1607 "MALAYALAM-TAMIL", "[:MALAYALAM:]", "[:TAMIL:]",
1608 "[\\u0ba9\\u0BF0\\u0BF1\\u0BF2]", /*roundtrip exclusions*/
1609
1610 "TAMIL-MALAYALAM", "[:TAMIL:]", "[:MALAYALAM:]",
1611 "[\\u0d43\\u0d12\\u0D0B\\u0D0C\\u0D16\\u0D17\\u0D18\\u0D1B\\u0D1D\\u0D20\\u0D21\\u0D22\\u0D25\\u0D26\\u0D27\\u0D2B\\u0D2C\\u0D2D\\u0D36\\u0D60\\u0D61]", /*roundtrip exclusions*/
1612
1613 "KANNADA-TELUGU", "[:KANNADA:]", "[:TELUGU:]",
1614 "[\\u0C01\\u0c3f\\u0c46\\u0c48\\u0c4a]", /*roundtrip exclusions*/
1615
1616 "TELUGU-KANNADA", "[:TELUGU:]", "[:KANNADA:]",
1617 "[\\u0cc8\\u0cd5\\u0cd6\\u0cbc\\u0cbd\\u0CDE]", /*roundtrip exclusions*/
1618
1619 "MALAYALAM-TELUGU", "[:MALAYALAM:]", "[:TELUGU:]",
1620 "[\\u0C01\\u0c44\\u0c4a\\u0c4c\\u0c4b\\u0c55\\u0c56]", /*roundtrip exclusions*/
1621
1622 "TELUGU-MALAYALAM", "[:TELUGU:]", "[:MALAYALAM:]",
1623 "[\\u0d4c\\u0d57\\u0D34]", /*roundtrip exclusions*/
1624
1625 "MALAYALAM-KANNADA", "[:MALAYALAM:]", "[:KANNADA:]",
1626 "[\\u0cbc\\u0cbd\\u0cc4\\u0cc6\\u0cca\\u0ccc\\u0ccb\\u0cd5\\u0cd6\\u0cDe]", /*roundtrip exclusions*/
1627
1628 "KANNADA-MALAYALAM", "[:KANNADA:]", "[:MALAYALAM:]",
1629 "[\\u0d4c\\u0d57\\u0d46\\u0D34]", /*roundtrip exclusions*/
1630
1631 "Latin-Bengali",latinForIndic, "[[:Bengali:][\\u0964\\u0965]]",
1632 "[\\u0965\\u09f0-\\u09fa\\u09ce]" /*roundtrip exclusions*/ ,
1633
1634 "Latin-Gurmukhi", latinForIndic, "[[:Gurmukhi:][\\u0964\\u0965]]",
1635 "[\\u0a01\\u0965\\u0a02\\u0a72\\u0a73\\u0a74]" /*roundtrip exclusions*/,
1636
1637 "Latin-Gujarati",latinForIndic, "[[:Gujarati:][\\u0964\\u0965]]",
1638 "[\\u0965]" /*roundtrip exclusions*/,
1639
1640 "Latin-Oriya",latinForIndic, "[[:Oriya:][\\u0964\\u0965]]",
1641 "[\\u0965\\u0b70]" /*roundtrip exclusions*/,
1642
1643 "Latin-Tamil",latinForIndic, "[:Tamil:]",
1644 "[\\u0BF0\\u0BF1\\u0BF2]" /*roundtrip exclusions*/,
1645
1646 "Latin-Telugu",latinForIndic, "[:Telugu:]",
1647 NULL /*roundtrip exclusions*/,
1648
1649 "Latin-Kannada",latinForIndic, "[:Kannada:]",
1650 NULL /*roundtrip exclusions*/,
1651
1652 "Latin-Malayalam",latinForIndic, "[:Malayalam:]",
1653 NULL /*roundtrip exclusions*/
1654 };
1655
TestDebug(const char * name,const char fromSet[],const char * toSet,const char * exclusions)1656 void TransliteratorRoundTripTest::TestDebug(const char* name,const char fromSet[],
1657 const char* toSet,const char* exclusions){
1658
1659 RTTest test(name);
1660 Legal *legal = new LegalIndic();
1661 test.test(UnicodeString(fromSet,""),UnicodeString(toSet,""),exclusions,this,quick,legal);
1662 }
1663
TestInterIndic()1664 void TransliteratorRoundTripTest::TestInterIndic() {
1665 //TestDebug("Latin-Gurmukhi", latinForIndic, "[:Gurmukhi:]","[\\u0965\\u0a02\\u0a72\\u0a73\\u0a74]",TRUE);
1666 int32_t num = (int32_t)(sizeof(interIndicArray)/(INTER_INDIC_ARRAY_WIDTH*sizeof(char*)));
1667 if(quick){
1668 logln("Testing only 5 of %i. Skipping rest (use -e for exhaustive)",num);
1669 num = 5;
1670 }
1671 // CLDR bug #1911: This test should be moved into CLDR.
1672 // It is left in its current state as a regression test.
1673 // if (isICUVersionAtLeast(ICU_39)) {
1674 // // We temporarily filter against Unicode 4.1, but we only do this
1675 // // before version 3.4.
1676 // errln("FAIL: TestInterIndic needs to be updated to remove delete the [:Age=4.1:] filter ");
1677 // return;
1678 // } else {
1679 // logln("Warning: TestInterIndic needs to be updated to remove delete the section marked [:Age=4.1:] filter");
1680 // }
1681 for(int i = 0; i < num;i++){
1682 RTTest test(interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 0]);
1683 Legal *legal = new LegalIndic();
1684 logln(UnicodeString("Stress testing ") + interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 0]);
1685 /* Uncomment lines below when transliterator is fixed */
1686 /*
1687 test.test( interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 1],
1688 interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 2],
1689 interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 3], // roundtrip exclusions
1690 this, quick, legal, 50);
1691 */
1692 /* comment lines below when transliterator is fixed */
1693 // start
1694 UnicodeString source("[");
1695 source.append(interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 1]);
1696 source.append(" & [:Age=4.1:]]");
1697 UnicodeString target("[");
1698 target.append(interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 2]);
1699 target.append(" & [:Age=4.1:]]");
1700 test.test( source,
1701 target,
1702 interIndicArray[i*INTER_INDIC_ARRAY_WIDTH + 3], // roundtrip exclusions
1703 this, quick, legal, 50);
1704 // end
1705 delete legal;
1706 }
1707 }
1708
1709 // end indic tests ----------------------------------------------------------
1710
1711 #endif /* #if !UCONFIG_NO_TRANSLITERATION */
1712