]>
Commit | Line | Data |
---|---|---|
b75a7d8f A |
1 | /******************************************************************** |
2 | * COPYRIGHT: | |
46f4442e | 3 | * Copyright (c) 1997-2008, International Business Machines Corporation and |
b75a7d8f A |
4 | * others. All Rights Reserved. |
5 | ********************************************************************/ | |
6 | ||
7 | /** | |
8 | * IntlTestCollator is the medium level test class for everything in the directory "collate". | |
9 | */ | |
10 | ||
11 | /*********************************************************************** | |
12 | * Modification history | |
13 | * Date Name Description | |
14 | * 02/14/2001 synwee Compare with cintltst and commented away tests | |
15 | * that are not run. | |
16 | ***********************************************************************/ | |
17 | ||
18 | #include "unicode/utypes.h" | |
19 | ||
20 | #if !UCONFIG_NO_COLLATION | |
21 | ||
22 | #include "unicode/uchar.h" | |
374ca955 | 23 | #include "unicode/ustring.h" |
b75a7d8f A |
24 | |
25 | #include "dadrcoll.h" | |
26 | ||
27 | #include "encoll.h" | |
28 | #include "frcoll.h" | |
29 | #include "decoll.h" | |
b75a7d8f A |
30 | #include "escoll.h" |
31 | #include "ficoll.h" | |
32 | #include "jacoll.h" | |
33 | #include "trcoll.h" | |
34 | #include "allcoll.h" | |
35 | #include "g7coll.h" | |
36 | #include "mnkytst.h" | |
37 | #include "apicoll.h" | |
38 | #include "regcoll.h" | |
39 | #include "currcoll.h" | |
40 | #include "itercoll.h" | |
b75a7d8f A |
41 | #include "tstnorm.h" |
42 | #include "normconf.h" | |
43 | #include "thcoll.h" | |
44 | #include "srchtest.h" | |
46f4442e | 45 | #include "ssearch.h" |
b75a7d8f A |
46 | #include "cntabcol.h" |
47 | #include "lcukocol.h" | |
48 | #include "ucaconf.h" | |
49 | #include "svccoll.h" | |
374ca955 | 50 | #include "cmemory.h" |
46f4442e A |
51 | //#include "rndmcoll.h" |
52 | ||
53 | // Set to 1 to test offsets in backAndForth() | |
54 | #define TEST_OFFSETS 0 | |
374ca955 A |
55 | |
56 | #define TESTCLASS(n,classname) \ | |
57 | case n: \ | |
58 | name = #classname; \ | |
59 | if (exec) { \ | |
60 | logln(#classname "---"); \ | |
61 | logln(""); \ | |
62 | classname t; \ | |
63 | callTest(t, par); \ | |
64 | } \ | |
65 | break | |
b75a7d8f A |
66 | |
67 | void IntlTestCollator::runIndexedTest( int32_t index, UBool exec, const char* &name, char* par ) | |
68 | { | |
374ca955 | 69 | if (exec) { |
b75a7d8f A |
70 | logln("TestSuite Collator: "); |
71 | } | |
72 | ||
374ca955 A |
73 | switch (index) { |
74 | TESTCLASS(0, CollationEnglishTest); | |
75 | TESTCLASS(1, CollationFrenchTest); | |
76 | TESTCLASS(2, CollationGermanTest); | |
77 | TESTCLASS(3, CollationSpanishTest); | |
78 | TESTCLASS(4, CollationKanaTest); | |
79 | TESTCLASS(5, CollationTurkishTest); | |
80 | TESTCLASS(6, CollationDummyTest); | |
81 | TESTCLASS(7, G7CollationTest); | |
82 | TESTCLASS(8, CollationMonkeyTest); | |
83 | TESTCLASS(9, CollationAPITest); | |
84 | TESTCLASS(10, CollationRegressionTest); | |
85 | TESTCLASS(11, CollationCurrencyTest); | |
86 | TESTCLASS(12, CollationIteratorTest); | |
87 | TESTCLASS(13, CollationThaiTest); | |
88 | TESTCLASS(14, LotusCollationKoreanTest); | |
89 | TESTCLASS(15, StringSearchTest); | |
90 | TESTCLASS(16, ContractionTableTest); | |
91 | TESTCLASS(17, DataDrivenCollatorTest); | |
92 | TESTCLASS(18, UCAConformanceTest); | |
93 | TESTCLASS(19, CollationServiceTest); | |
374ca955 | 94 | TESTCLASS(20, CollationFinnishTest); // removed by weiv - we have changed Finnish collation |
46f4442e A |
95 | //TESTCLASS(21, RandomCollatorTest); // See ticket 5747 about reenabling this test. |
96 | TESTCLASS(21, SSearchTest); | |
73c04bcf | 97 | |
374ca955 | 98 | default: name = ""; break; |
b75a7d8f A |
99 | } |
100 | } | |
101 | ||
102 | UCollationResult | |
103 | IntlTestCollator::compareUsingPartials(UCollator *coll, const UChar source[], int32_t sLen, const UChar target[], int32_t tLen, int32_t pieceSize, UErrorCode &status) { | |
104 | int32_t partialSKResult = 0; | |
105 | uint8_t sBuf[512], tBuf[512]; | |
106 | UCharIterator sIter, tIter; | |
107 | uint32_t sState[2], tState[2]; | |
108 | int32_t sSize = pieceSize, tSize = pieceSize; | |
109 | int32_t i = 0; | |
110 | status = U_ZERO_ERROR; | |
111 | sState[0] = 0; sState[1] = 0; | |
112 | tState[0] = 0; tState[1] = 0; | |
113 | while(sSize == pieceSize && tSize == pieceSize && partialSKResult == 0) { | |
114 | uiter_setString(&sIter, source, sLen); | |
115 | uiter_setString(&tIter, target, tLen); | |
116 | sSize = ucol_nextSortKeyPart(coll, &sIter, sState, sBuf, pieceSize, &status); | |
117 | tSize = ucol_nextSortKeyPart(coll, &tIter, tState, tBuf, pieceSize, &status); | |
118 | ||
119 | if(sState[0] != 0 || tState[0] != 0) { | |
120 | log("State != 0 : %08X %08X\n", sState[0], tState[0]); | |
121 | } | |
122 | log("%i ", i++); | |
123 | ||
124 | partialSKResult = memcmp(sBuf, tBuf, pieceSize); | |
125 | } | |
126 | ||
127 | if(partialSKResult < 0) { | |
128 | return UCOL_LESS; | |
129 | } else if(partialSKResult > 0) { | |
130 | return UCOL_GREATER; | |
131 | } else { | |
132 | return UCOL_EQUAL; | |
133 | } | |
134 | } | |
135 | ||
136 | void | |
137 | IntlTestCollator::doTestVariant(Collator* col, const UnicodeString &source, const UnicodeString &target, Collator::EComparisonResult result) | |
138 | { | |
139 | UErrorCode status = U_ZERO_ERROR; | |
140 | ||
141 | UCollator *myCollation = (UCollator *)((RuleBasedCollator *)col)->getUCollator(); | |
142 | ||
143 | Collator::EComparisonResult compareResult = col->compare(source, target); | |
144 | ||
145 | CollationKey srckey, tgtkey; | |
146 | col->getCollationKey(source, srckey, status); | |
147 | col->getCollationKey(target, tgtkey, status); | |
148 | if (U_FAILURE(status)){ | |
149 | errln("Creation of collation keys failed\n"); | |
150 | } | |
151 | Collator::EComparisonResult keyResult = srckey.compareTo(tgtkey); | |
152 | ||
153 | reportCResult(source, target, srckey, tgtkey, compareResult, keyResult, result, result); | |
154 | ||
155 | UColAttributeValue norm = ucol_getAttribute(myCollation, UCOL_NORMALIZATION_MODE, &status); | |
156 | ||
157 | int32_t sLen = source.length(), tLen = target.length(); | |
158 | const UChar* src = source.getBuffer(); | |
159 | const UChar* trg = target.getBuffer(); | |
160 | UCollationResult compareResultIter = (UCollationResult)result; | |
161 | ||
46f4442e | 162 | { |
b75a7d8f A |
163 | UCharIterator sIter, tIter; |
164 | uiter_setString(&sIter, src, sLen); | |
165 | uiter_setString(&tIter, trg, tLen); | |
166 | compareResultIter = ucol_strcollIter(myCollation, &sIter, &tIter, &status); | |
167 | if(compareResultIter != (UCollationResult)result) { | |
168 | errln("Different result for iterative comparison "+source+" "+target); | |
169 | } | |
170 | } | |
171 | /* convert the strings to UTF-8 and do try comparing with char iterator */ | |
172 | if(!quick) { /*!QUICK*/ | |
173 | char utf8Source[256], utf8Target[256]; | |
174 | int32_t utf8SourceLen = 0, utf8TargetLen = 0; | |
175 | u_strToUTF8(utf8Source, 256, &utf8SourceLen, src, sLen, &status); | |
176 | if(U_FAILURE(status)) { /* probably buffer is not big enough */ | |
177 | log("Src UTF-8 buffer too small! Will not compare!\n"); | |
178 | } else { | |
179 | u_strToUTF8(utf8Target, 256, &utf8TargetLen, trg, tLen, &status); | |
180 | if(U_SUCCESS(status)) { /* probably buffer is not big enough */ | |
181 | UCollationResult compareResultUTF8 = (UCollationResult)result, compareResultUTF8Norm = (UCollationResult)result; | |
182 | UCharIterator sIter, tIter; | |
183 | /*log_verbose("Strings converted to UTF-8:%s, %s\n", aescstrdup(source,-1), aescstrdup(target,-1));*/ | |
184 | uiter_setUTF8(&sIter, utf8Source, utf8SourceLen); | |
185 | uiter_setUTF8(&tIter, utf8Target, utf8TargetLen); | |
186 | /*uiter_setString(&sIter, source, sLen); | |
187 | uiter_setString(&tIter, target, tLen);*/ | |
188 | compareResultUTF8 = ucol_strcollIter(myCollation, &sIter, &tIter, &status); | |
189 | ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, UCOL_ON, &status); | |
190 | sIter.move(&sIter, 0, UITER_START); | |
191 | tIter.move(&tIter, 0, UITER_START); | |
192 | compareResultUTF8Norm = ucol_strcollIter(myCollation, &sIter, &tIter, &status); | |
193 | ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, norm, &status); | |
194 | if(compareResultUTF8 != compareResultIter) { | |
195 | errln("different results in iterative comparison for UTF-16 and UTF-8 encoded strings. "+source+", "+target); | |
196 | } | |
197 | if(compareResultUTF8 != compareResultUTF8Norm) { | |
198 | errln("different results in iterative when normalization is turned on with UTF-8 strings. "+source+", "+target); | |
199 | } | |
200 | } else { | |
201 | log("Target UTF-8 buffer too small! Did not compare!\n"); | |
202 | } | |
203 | if(U_FAILURE(status)) { | |
204 | log("UTF-8 strcoll failed! Ignoring result\n"); | |
205 | } | |
206 | } | |
207 | } | |
208 | ||
209 | /* testing the partial sortkeys */ | |
46f4442e | 210 | { /*!QUICK*/ |
b75a7d8f A |
211 | int32_t partialSizes[] = { 3, 1, 2, 4, 8, 20, 80 }; /* just size 3 in the quick mode */ |
212 | int32_t partialSizesSize = 1; | |
213 | if(!quick) { | |
214 | partialSizesSize = 7; | |
215 | } | |
216 | int32_t i = 0; | |
217 | log("partial sortkey test piecesize="); | |
218 | for(i = 0; i < partialSizesSize; i++) { | |
219 | UCollationResult partialSKResult = (UCollationResult)result, partialNormalizedSKResult = (UCollationResult)result; | |
220 | log("%i ", partialSizes[i]); | |
221 | ||
222 | partialSKResult = compareUsingPartials(myCollation, src, sLen, trg, tLen, partialSizes[i], status); | |
223 | if(partialSKResult != (UCollationResult)result) { | |
224 | errln("Partial sortkey comparison returned wrong result: "+source+", "+target+" (size "+partialSizes[i]+")"); | |
225 | } | |
226 | ||
227 | if(norm != UCOL_ON && !quick) { | |
228 | log("N "); | |
229 | ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, UCOL_ON, &status); | |
230 | partialNormalizedSKResult = compareUsingPartials(myCollation, src, sLen, trg, tLen, partialSizes[i], status); | |
231 | ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, norm, &status); | |
232 | if(partialSKResult != partialNormalizedSKResult) { | |
233 | errln("Partial sortkey comparison gets different result when normalization is on: "+source+", "+target+" (size "+partialSizes[i]+")"); | |
234 | } | |
235 | } | |
236 | } | |
237 | log("\n"); | |
238 | } | |
239 | /* | |
240 | if (compareResult != result) { | |
241 | errln("String comparison failed in variant test\n"); | |
242 | } | |
243 | if (keyResult != result) { | |
244 | errln("Collation key comparison failed in variant test\n"); | |
245 | } | |
246 | */ | |
247 | } | |
248 | ||
249 | void | |
250 | IntlTestCollator::doTest(Collator* col, const UChar *source, const UChar *target, Collator::EComparisonResult result) { | |
251 | doTest(col, UnicodeString(source), UnicodeString(target), result); | |
252 | } | |
253 | ||
254 | void | |
255 | IntlTestCollator::doTest(Collator* col, const UnicodeString &source, const UnicodeString &target, Collator::EComparisonResult result) | |
256 | { | |
257 | if(col) { | |
258 | doTestVariant(col, source, target, result); | |
259 | if(result == Collator::LESS) { | |
260 | doTestVariant(col, target, source, Collator::GREATER); | |
261 | } else if (result == Collator::GREATER) { | |
262 | doTestVariant(col, target, source, Collator::LESS); | |
263 | } | |
264 | ||
265 | UErrorCode status = U_ZERO_ERROR; | |
266 | CollationElementIterator* c = ((RuleBasedCollator *)col)->createCollationElementIterator( source ); | |
267 | logln("Testing iterating source: "+source); | |
268 | backAndForth(*c); | |
269 | c->setText(target, status); | |
270 | logln("Testing iterating target: "+target); | |
271 | backAndForth(*c); | |
272 | delete c; | |
273 | } | |
274 | } | |
275 | ||
276 | ||
277 | // used for collation result reporting, defined here for convenience | |
278 | // (maybe moved later) | |
279 | void | |
280 | IntlTestCollator::reportCResult( const UnicodeString &source, const UnicodeString &target, | |
281 | CollationKey &sourceKey, CollationKey &targetKey, | |
282 | Collator::EComparisonResult compareResult, | |
283 | Collator::EComparisonResult keyResult, | |
284 | Collator::EComparisonResult incResult, | |
285 | Collator::EComparisonResult expectedResult ) | |
286 | { | |
287 | if (expectedResult < -1 || expectedResult > 1) | |
288 | { | |
289 | errln("***** invalid call to reportCResult ****"); | |
290 | return; | |
291 | } | |
292 | ||
293 | UBool ok1 = (compareResult == expectedResult); | |
294 | UBool ok2 = (keyResult == expectedResult); | |
295 | UBool ok3 = (incResult == expectedResult); | |
296 | ||
297 | ||
298 | if (ok1 && ok2 && ok3 && !verbose) { | |
299 | // Keep non-verbose, passing tests fast | |
300 | return; | |
301 | } else { | |
302 | UnicodeString msg1(ok1 ? "Ok: compare(" : "FAIL: compare("); | |
303 | UnicodeString msg2(", "), msg3(") returned "), msg4("; expected "); | |
304 | UnicodeString prettySource, prettyTarget, sExpect, sResult; | |
305 | ||
306 | IntlTest::prettify(source, prettySource); | |
307 | IntlTest::prettify(target, prettyTarget); | |
308 | appendCompareResult(compareResult, sResult); | |
309 | appendCompareResult(expectedResult, sExpect); | |
310 | ||
311 | if (ok1) { | |
312 | logln(msg1 + prettySource + msg2 + prettyTarget + msg3 + sResult); | |
313 | } else { | |
314 | errln(msg1 + prettySource + msg2 + prettyTarget + msg3 + sResult + msg4 + sExpect); | |
315 | } | |
316 | ||
317 | msg1 = UnicodeString(ok2 ? "Ok: key(" : "FAIL: key("); | |
318 | msg2 = ").compareTo(key("; | |
319 | msg3 = ")) returned "; | |
320 | ||
321 | appendCompareResult(keyResult, sResult); | |
322 | ||
323 | if (ok2) { | |
324 | logln(msg1 + prettySource + msg2 + prettyTarget + msg3 + sResult); | |
325 | } else { | |
326 | errln(msg1 + prettySource + msg2 + prettyTarget + msg3 + sResult + msg4 + sExpect); | |
327 | ||
328 | msg1 = " "; | |
329 | msg2 = " vs. "; | |
330 | ||
331 | prettify(sourceKey, prettySource); | |
332 | prettify(targetKey, prettyTarget); | |
333 | ||
334 | errln(msg1 + prettySource + msg2 + prettyTarget); | |
335 | } | |
336 | msg1 = UnicodeString (ok3 ? "Ok: incCompare(" : "FAIL: incCompare("); | |
337 | msg2 = ", "; | |
338 | msg3 = ") returned "; | |
339 | ||
340 | appendCompareResult(incResult, sResult); | |
341 | ||
342 | if (ok3) { | |
343 | logln(msg1 + prettySource + msg2 + prettyTarget + msg3 + sResult); | |
344 | } else { | |
345 | errln(msg1 + prettySource + msg2 + prettyTarget + msg3 + sResult + msg4 + sExpect); | |
346 | } | |
347 | } | |
348 | } | |
349 | ||
350 | UnicodeString& | |
351 | IntlTestCollator::appendCompareResult(Collator::EComparisonResult result, | |
352 | UnicodeString& target) | |
353 | { | |
354 | if (result == Collator::LESS) | |
355 | { | |
356 | target += "LESS"; | |
357 | } | |
358 | else if (result == Collator::EQUAL) | |
359 | { | |
360 | target += "EQUAL"; | |
361 | } | |
362 | else if (result == Collator::GREATER) | |
363 | { | |
364 | target += "GREATER"; | |
365 | } | |
366 | else | |
367 | { | |
368 | UnicodeString huh = "?"; | |
369 | ||
370 | target += (huh + (int32_t)result); | |
371 | } | |
372 | ||
373 | return target; | |
374 | } | |
375 | ||
376 | // Produce a printable representation of a CollationKey | |
377 | UnicodeString &IntlTestCollator::prettify(const CollationKey &source, UnicodeString &target) | |
378 | { | |
379 | int32_t i, byteCount; | |
380 | const uint8_t *bytes = source.getByteArray(byteCount); | |
381 | ||
382 | target.remove(); | |
383 | target += "["; | |
384 | ||
385 | for (i = 0; i < byteCount; i += 1) | |
386 | { | |
387 | appendHex(bytes[i], 2, target); | |
388 | target += " "; | |
389 | } | |
390 | ||
391 | target += "]"; | |
392 | ||
393 | return target; | |
394 | } | |
395 | ||
396 | void IntlTestCollator::backAndForth(CollationElementIterator &iter) | |
397 | { | |
398 | // Run through the iterator forwards and stick it into an array | |
399 | int32_t orderLength = 0; | |
46f4442e | 400 | Order *orders = getOrders(iter, orderLength); |
b75a7d8f A |
401 | UErrorCode status = U_ZERO_ERROR; |
402 | ||
403 | // Now go through it backwards and make sure we get the same values | |
404 | int32_t index = orderLength; | |
405 | int32_t o; | |
406 | ||
407 | // reset the iterator | |
408 | iter.reset(); | |
409 | ||
410 | while ((o = iter.previous(status)) != CollationElementIterator::NULLORDER) | |
411 | { | |
46f4442e A |
412 | int32_t offset = iter.getOffset(); |
413 | ||
374ca955 A |
414 | if (index == 0) { |
415 | if(o == 0) { | |
416 | continue; | |
417 | } else { // this is an error, orders exhausted but there are non-ignorable CEs from | |
418 | // going backwards | |
419 | errln("Backward iteration returned a non ignorable after orders are exhausted"); | |
420 | break; | |
46f4442e | 421 | } |
374ca955 | 422 | } |
46f4442e A |
423 | |
424 | index -= 1; | |
425 | if (o != orders[index].order) { | |
b75a7d8f | 426 | if (o == 0) |
46f4442e A |
427 | index += 1; |
428 | else { | |
429 | while (index > 0 && orders[--index].order == 0) { | |
430 | // nothing... | |
b75a7d8f | 431 | } |
46f4442e A |
432 | |
433 | if (o != orders[index].order) { | |
434 | errln("Mismatched order at index %d: 0x%0:8X vs. 0x%0:8X", index, | |
435 | orders[index].order, o); | |
436 | //break; | |
437 | goto bail; | |
b75a7d8f A |
438 | } |
439 | } | |
440 | } | |
46f4442e A |
441 | |
442 | #if TEST_OFFSETS | |
443 | if (offset != orders[index].offset) { | |
444 | errln("Mismatched offset at index %d: %d vs. %d", index, | |
445 | orders[index].offset, offset); | |
446 | //break; | |
447 | goto bail; | |
448 | } | |
449 | #endif | |
450 | ||
b75a7d8f A |
451 | } |
452 | ||
46f4442e | 453 | while (index != 0 && orders[index - 1].order == 0) |
b75a7d8f A |
454 | { |
455 | index --; | |
456 | } | |
457 | ||
458 | if (index != 0) | |
459 | { | |
460 | UnicodeString msg("Didn't get back to beginning - index is "); | |
461 | errln(msg + index); | |
462 | ||
463 | iter.reset(); | |
464 | err("next: "); | |
465 | while ((o = iter.next(status)) != CollationElementIterator::NULLORDER) | |
466 | { | |
467 | UnicodeString hexString("0x"); | |
468 | ||
469 | appendHex(o, 8, hexString); | |
470 | hexString += " "; | |
471 | err(hexString); | |
472 | } | |
473 | errln(""); | |
474 | ||
475 | err("prev: "); | |
476 | while ((o = iter.previous(status)) != CollationElementIterator::NULLORDER) | |
477 | { | |
478 | UnicodeString hexString("0x"); | |
479 | ||
480 | appendHex(o, 8, hexString); | |
481 | hexString += " "; | |
482 | err(hexString); | |
483 | } | |
484 | errln(""); | |
485 | } | |
486 | ||
46f4442e | 487 | bail: |
b75a7d8f A |
488 | delete[] orders; |
489 | } | |
490 | ||
491 | ||
492 | /** | |
493 | * Return an integer array containing all of the collation orders | |
494 | * returned by calls to next on the specified iterator | |
495 | */ | |
46f4442e | 496 | IntlTestCollator::Order *IntlTestCollator::getOrders(CollationElementIterator &iter, int32_t &orderLength) |
b75a7d8f A |
497 | { |
498 | int32_t maxSize = 100; | |
499 | int32_t size = 0; | |
46f4442e | 500 | Order *orders = new Order[maxSize]; |
b75a7d8f | 501 | UErrorCode status = U_ZERO_ERROR; |
46f4442e | 502 | int32_t offset = iter.getOffset(); |
b75a7d8f A |
503 | |
504 | int32_t order; | |
505 | while ((order = iter.next(status)) != CollationElementIterator::NULLORDER) | |
506 | { | |
507 | if (size == maxSize) | |
508 | { | |
509 | maxSize *= 2; | |
46f4442e | 510 | Order *temp = new Order[maxSize]; |
b75a7d8f | 511 | |
46f4442e | 512 | uprv_memcpy(temp, orders, size * sizeof(Order)); |
b75a7d8f A |
513 | delete[] orders; |
514 | orders = temp; | |
515 | } | |
516 | ||
46f4442e A |
517 | orders[size].order = order; |
518 | orders[size].offset = offset; | |
519 | ||
520 | offset = iter.getOffset(); | |
521 | size += 1; | |
b75a7d8f A |
522 | } |
523 | ||
524 | if (maxSize > size) | |
525 | { | |
46f4442e | 526 | Order *temp = new Order[size]; |
b75a7d8f | 527 | |
46f4442e | 528 | uprv_memcpy(temp, orders, size * sizeof(Order)); |
b75a7d8f A |
529 | delete[] orders; |
530 | orders = temp; | |
531 | } | |
532 | ||
533 | orderLength = size; | |
534 | return orders; | |
535 | } | |
536 | ||
537 | #endif /* #if !UCONFIG_NO_COLLATION */ |