1 /********************************************************************
2 * Copyright (c) 1997-2011 International Business Machines
3 * Corporation and others. All Rights Reserved.
4 ********************************************************************/
5 /*****************************************************************************
6 *
7 * File CAPITEST.C
8 *
9 * Modification History:
10 * Name Description
11 * Madhu Katragadda Ported for C API
12 * Brian Rower Added TestOpenVsOpenRules
13 ******************************************************************************
14 *//* C API TEST For COLLATOR */
15
16 #include "unicode/utypes.h"
17
18 #if !UCONFIG_NO_COLLATION
19
20 #include <stdio.h>
21 #include <stdlib.h>
22 #include <string.h>
23 #include "unicode/uloc.h"
24 #include "unicode/ulocdata.h"
25 #include "unicode/ustring.h"
26 #include "unicode/ures.h"
27 #include "unicode/ucoleitr.h"
28 #include "cintltst.h"
29 #include "capitst.h"
30 #include "ccolltst.h"
31 #include "putilimp.h"
32 #include "cmemory.h"
33 #include "cstring.h"
34
35 static void TestAttribute(void);
36 static void TestDefault(void);
37 static void TestDefaultKeyword(void);
38 static void TestBengaliSortKey(void);
39 int TestBufferSize(); /* defined in "colutil.c" */
40
41
42
43
44 /* next two function is modified from "i18n/ucol.cpp" to avoid include "ucol_imp.h" */
uprv_appendByteToHexString(char * dst,uint8_t val)45 static void uprv_appendByteToHexString(char *dst, uint8_t val) {
46 uint32_t len = (uint32_t)strlen(dst);
47 sprintf(dst+len, "%02X", val);
48 }
49
ucol_sortKeyToString(const UCollator * coll,const uint8_t * sortkey,char * buffer,uint32_t * len)50 static char* U_EXPORT2 ucol_sortKeyToString(const UCollator *coll, const uint8_t *sortkey, char *buffer, uint32_t *len) {
51 int32_t strength = UCOL_PRIMARY;
52 uint32_t res_size = 0;
53 UBool doneCase = FALSE;
54
55 char *current = buffer;
56 const uint8_t *currentSk = sortkey;
57
58 UErrorCode error_code = U_ZERO_ERROR;
59
60 strcpy(current, "[");
61
62 while(strength <= UCOL_QUATERNARY && strength <= ucol_getAttribute(coll,UCOL_STRENGTH, &error_code)) {
63 if(U_FAILURE(error_code)) {
64 log_err("ucol_getAttribute returned error: %s\n", u_errorName(error_code));
65 }
66 if(strength > UCOL_PRIMARY) {
67 strcat(current, " . ");
68 }
69 while(*currentSk != 0x01 && *currentSk != 0x00) { /* print a level */
70 uprv_appendByteToHexString(current, *currentSk++);
71 strcat(current, " ");
72 }
73 if(ucol_getAttribute(coll,UCOL_CASE_LEVEL, &error_code) == UCOL_ON && strength == UCOL_SECONDARY && doneCase == FALSE) {
74 doneCase = TRUE;
75 } else if(ucol_getAttribute(coll,UCOL_CASE_LEVEL, &error_code) == UCOL_OFF || doneCase == TRUE || strength != UCOL_SECONDARY) {
76 strength ++;
77 }
78 if(U_FAILURE(error_code)) {
79 log_err("ucol_getAttribute returned error: %s\n", u_errorName(error_code));
80 }
81 uprv_appendByteToHexString(current, *currentSk++); /* This should print '01' */
82 if(strength == UCOL_QUATERNARY && ucol_getAttribute(coll,UCOL_ALTERNATE_HANDLING, &error_code) == UCOL_NON_IGNORABLE) {
83 break;
84 }
85 }
86
87 if(ucol_getAttribute(coll,UCOL_STRENGTH, &error_code) == UCOL_IDENTICAL) {
88 strcat(current, " . ");
89 while(*currentSk != 0) {
90 uprv_appendByteToHexString(current, *currentSk++);
91 strcat(current, " ");
92 }
93
94 uprv_appendByteToHexString(current, *currentSk++);
95 }
96 if(U_FAILURE(error_code)) {
97 log_err("ucol_getAttribute returned error: %s\n", u_errorName(error_code));
98 }
99 strcat(current, "]");
100
101 if(res_size > *len) {
102 return NULL;
103 }
104
105 return buffer;
106 }
107 /* end of avoid include "ucol_imp.h" */
108
109
addCollAPITest(TestNode ** root)110 void addCollAPITest(TestNode** root)
111 {
112 /* WEIVTODO: return tests here */
113 addTest(root, &TestProperty, "tscoll/capitst/TestProperty");
114 addTest(root, &TestRuleBasedColl, "tscoll/capitst/TestRuleBasedColl");
115 addTest(root, &TestCompare, "tscoll/capitst/TestCompare");
116 addTest(root, &TestSortKey, "tscoll/capitst/TestSortKey");
117 addTest(root, &TestHashCode, "tscoll/capitst/TestHashCode");
118 addTest(root, &TestElemIter, "tscoll/capitst/TestElemIter");
119 addTest(root, &TestGetAll, "tscoll/capitst/TestGetAll");
120 /*addTest(root, &TestGetDefaultRules, "tscoll/capitst/TestGetDefaultRules");*/
121 addTest(root, &TestDecomposition, "tscoll/capitst/TestDecomposition");
122 addTest(root, &TestSafeClone, "tscoll/capitst/TestSafeClone");
123 addTest(root, &TestCloneBinary, "tscoll/capitst/TestCloneBinary");
124 addTest(root, &TestGetSetAttr, "tscoll/capitst/TestGetSetAttr");
125 addTest(root, &TestBounds, "tscoll/capitst/TestBounds");
126 addTest(root, &TestGetLocale, "tscoll/capitst/TestGetLocale");
127 addTest(root, &TestSortKeyBufferOverrun, "tscoll/capitst/TestSortKeyBufferOverrun");
128 addTest(root, &TestAttribute, "tscoll/capitst/TestAttribute");
129 addTest(root, &TestGetTailoredSet, "tscoll/capitst/TestGetTailoredSet");
130 addTest(root, &TestMergeSortKeys, "tscoll/capitst/TestMergeSortKeys");
131 addTest(root, &TestShortString, "tscoll/capitst/TestShortString");
132 /* BEGIN android-removed
133 To save space, Android does not include the collation tailoring rules.
134 We skip the tailing tests for collations. */
135 /* addTest(root, &TestGetContractionsAndUnsafes, "tscoll/capitst/TestGetContractionsAndUnsafes"); */
136 /* END android-removed */
137 addTest(root, &TestOpenBinary, "tscoll/capitst/TestOpenBinary");
138 addTest(root, &TestDefault, "tscoll/capitst/TestDefault");
139 addTest(root, &TestDefaultKeyword, "tscoll/capitst/TestDefaultKeyword");
140 /* BEGIN android-removed
141 To save space, Android does not build full collation tables and tailing rules.
142 We skip the related tests. */
143 /* addTest(root, &TestOpenVsOpenRules, "tscoll/capitst/TestOpenVsOpenRules"); */
144 /* addTest(root, &TestGetKeywordValuesForLocale, "tscoll/capitst/TestGetKeywordValuesForLocale"); */
145 /* END android-removed */
146 addTest(root, &TestBengaliSortKey, "tscoll/capitst/TestBengaliSortKey");
147 }
148
TestGetSetAttr(void)149 void TestGetSetAttr(void) {
150 UErrorCode status = U_ZERO_ERROR;
151 UCollator *coll = ucol_open(NULL, &status);
152 struct attrTest {
153 UColAttribute att;
154 UColAttributeValue val[5];
155 uint32_t valueSize;
156 UColAttributeValue nonValue;
157 } attrs[] = {
158 {UCOL_FRENCH_COLLATION, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},
159 {UCOL_ALTERNATE_HANDLING, {UCOL_NON_IGNORABLE, UCOL_SHIFTED}, 2, UCOL_OFF},/* attribute for handling variable elements*/
160 {UCOL_CASE_FIRST, {UCOL_OFF, UCOL_LOWER_FIRST, UCOL_UPPER_FIRST}, 3, UCOL_SHIFTED},/* who goes first, lower case or uppercase */
161 {UCOL_CASE_LEVEL, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* do we have an extra case level */
162 {UCOL_NORMALIZATION_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* attribute for normalization */
163 {UCOL_DECOMPOSITION_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},
164 {UCOL_STRENGTH, {UCOL_PRIMARY, UCOL_SECONDARY, UCOL_TERTIARY, UCOL_QUATERNARY, UCOL_IDENTICAL}, 5, UCOL_SHIFTED},/* attribute for strength */
165 {UCOL_HIRAGANA_QUATERNARY_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* when turned on, this attribute */
166 };
167 UColAttribute currAttr;
168 UColAttributeValue value;
169 uint32_t i = 0, j = 0;
170
171 if (coll == NULL) {
172 log_err_status(status, "Unable to open collator. %s\n", u_errorName(status));
173 return;
174 }
175 for(i = 0; i<sizeof(attrs)/sizeof(attrs[0]); i++) {
176 currAttr = attrs[i].att;
177 ucol_setAttribute(coll, currAttr, UCOL_DEFAULT, &status);
178 if(U_FAILURE(status)) {
179 log_err_status(status, "ucol_setAttribute with the default value returned error: %s\n", u_errorName(status));
180 break;
181 }
182 value = ucol_getAttribute(coll, currAttr, &status);
183 if(U_FAILURE(status)) {
184 log_err("ucol_getAttribute returned error: %s\n", u_errorName(status));
185 break;
186 }
187 for(j = 0; j<attrs[i].valueSize; j++) {
188 ucol_setAttribute(coll, currAttr, attrs[i].val[j], &status);
189 if(U_FAILURE(status)) {
190 log_err("ucol_setAttribute with the value %i returned error: %s\n", attrs[i].val[j], u_errorName(status));
191 break;
192 }
193 }
194 status = U_ZERO_ERROR;
195 ucol_setAttribute(coll, currAttr, attrs[i].nonValue, &status);
196 if(U_SUCCESS(status)) {
197 log_err("ucol_setAttribute with the bad value didn't return an error\n");
198 break;
199 }
200 status = U_ZERO_ERROR;
201
202 ucol_setAttribute(coll, currAttr, value, &status);
203 if(U_FAILURE(status)) {
204 log_err("ucol_setAttribute with the default valuereturned error: %s\n", u_errorName(status));
205 break;
206 }
207 }
208 status = U_ZERO_ERROR;
209 value = ucol_getAttribute(coll, UCOL_ATTRIBUTE_COUNT, &status);
210 if(U_SUCCESS(status)) {
211 log_err("ucol_getAttribute for UCOL_ATTRIBUTE_COUNT didn't return an error\n");
212 }
213 status = U_ZERO_ERROR;
214 ucol_setAttribute(coll, UCOL_ATTRIBUTE_COUNT, UCOL_DEFAULT, &status);
215 if(U_SUCCESS(status)) {
216 log_err("ucol_setAttribute for UCOL_ATTRIBUTE_COUNT didn't return an error\n");
217 }
218 status = U_ZERO_ERROR;
219 ucol_close(coll);
220 }
221
222
doAssert(int condition,const char * message)223 static void doAssert(int condition, const char *message)
224 {
225 if (condition==0) {
226 log_err("ERROR : %s\n", message);
227 }
228 }
229
230 #if 0
231 /* We don't have default rules, at least not in the previous sense */
232 void TestGetDefaultRules(){
233 uint32_t size=0;
234 UErrorCode status=U_ZERO_ERROR;
235 UCollator *coll=NULL;
236 int32_t len1 = 0, len2=0;
237 uint8_t *binColData = NULL;
238
239 UResourceBundle *res = NULL;
240 UResourceBundle *binColl = NULL;
241 uint8_t *binResult = NULL;
242
243
244 const UChar * defaultRulesArray=ucol_getDefaultRulesArray(&size);
245 log_verbose("Test the function ucol_getDefaultRulesArray()\n");
246
247 coll = ucol_openRules(defaultRulesArray, size, UCOL_ON, UCOL_PRIMARY, &status);
248 if(U_SUCCESS(status) && coll !=NULL) {
249 binColData = (uint8_t*)ucol_cloneRuleData(coll, &len1, &status);
250
251 }
252
253
254 status=U_ZERO_ERROR;
255 res=ures_open(NULL, "root", &status);
256 if(U_FAILURE(status)){
257 log_err("ERROR: Failed to get resource for \"root Locale\" with %s", myErrorName(status));
258 return;
259 }
260 binColl=ures_getByKey(res, "%%Collation", binColl, &status);
261 if(U_SUCCESS(status)){
262 binResult=(uint8_t*)ures_getBinary(binColl, &len2, &status);
263 if(U_FAILURE(status)){
264 log_err("ERROR: ures_getBinary() failed\n");
265 }
266 }else{
267 log_err("ERROR: ures_getByKey(locale(default), %%Collation) failed");
268 }
269
270
271 if(len1 != len2){
272 log_err("Error: ucol_getDefaultRulesArray() failed to return the correct length.\n");
273 }
274 if(memcmp(binColData, binResult, len1) != 0){
275 log_err("Error: ucol_getDefaultRulesArray() failed\n");
276 }
277
278 free(binColData);
279 ures_close(binColl);
280 ures_close(res);
281 ucol_close(coll);
282
283 }
284 #endif
285
286 /* Collator Properties
287 ucol_open, ucol_strcoll, getStrength/setStrength
288 getDecomposition/setDecomposition, getDisplayName*/
TestProperty()289 void TestProperty()
290 {
291 UCollator *col, *ruled;
292 UChar *disName;
293 int32_t len = 0;
294 UChar *source, *target;
295 int32_t tempLength;
296 UErrorCode status = U_ZERO_ERROR;
297 /*
298 * Expected version of the English collator.
299 * Currently, the major/minor version numbers change when the builder code
300 * changes,
301 * number 2 is from the tailoring data version and
302 * number 3 is the UCA version.
303 * This changes with every UCA version change, and the expected value
304 * needs to be adjusted.
305 * Same in intltest/apicoll.cpp.
306 */
307 UVersionInfo currVersionArray = {0x31, 0xC0, 0x05, 0x2A}; /* from ICU 4.4/UCA 5.2 */
308 UVersionInfo versionArray = {0, 0, 0, 0};
309 UVersionInfo versionUCAArray = {0, 0, 0, 0};
310 UVersionInfo versionUCDArray = {0, 0, 0, 0};
311
312 log_verbose("The property tests begin : \n");
313 log_verbose("Test ucol_strcoll : \n");
314 col = ucol_open("en_US", &status);
315 if (U_FAILURE(status)) {
316 log_err_status(status, "Default Collator creation failed.: %s\n", myErrorName(status));
317 return;
318 }
319
320 ucol_getVersion(col, versionArray);
321 /* Check for a version greater than some value rather than equality
322 * so that we need not update the expected version each time. */
323 if (uprv_memcmp(versionArray, currVersionArray, 4)<0) {
324 log_err("Testing ucol_getVersion() - unexpected result: %02x.%02x.%02x.%02x\n",
325 versionArray[0], versionArray[1], versionArray[2], versionArray[3]);
326 } else {
327 log_verbose("ucol_getVersion() result: %02x.%02x.%02x.%02x\n",
328 versionArray[0], versionArray[1], versionArray[2], versionArray[3]);
329 }
330
331 /* Assume that the UCD and UCA versions are the same,
332 * rather than hardcoding (and updating each time) a particular UCA version. */
333 u_getUnicodeVersion(versionUCDArray);
334 ucol_getUCAVersion(col, versionUCAArray);
335 if (0!=uprv_memcmp(versionUCAArray, versionUCDArray, 4)) {
336 log_err("Testing ucol_getUCAVersion() - unexpected result: %hu.%hu.%hu.%hu\n",
337 versionUCAArray[0], versionUCAArray[1], versionUCAArray[2], versionUCAArray[3]);
338 }
339
340 source=(UChar*)malloc(sizeof(UChar) * 12);
341 target=(UChar*)malloc(sizeof(UChar) * 12);
342
343
344 u_uastrcpy(source, "ab");
345 u_uastrcpy(target, "abc");
346
347 doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_LESS), "ab < abc comparison failed");
348
349 u_uastrcpy(source, "ab");
350 u_uastrcpy(target, "AB");
351
352 doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_LESS), "ab < AB comparison failed");
353 /* u_uastrcpy(source, "black-bird");
354 u_uastrcpy(target, "blackbird"); */
355 u_uastrcpy(target, "black-bird");
356 u_uastrcpy(source, "blackbird");
357
358 doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_GREATER),
359 "black-bird > blackbird comparison failed");
360 u_uastrcpy(source, "black bird");
361 u_uastrcpy(target, "black-bird");
362 doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_LESS),
363 "black bird < black-bird comparison failed");
364 u_uastrcpy(source, "Hello");
365 u_uastrcpy(target, "hello");
366
367 doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_GREATER),
368 "Hello > hello comparison failed");
369 free(source);
370 free(target);
371 log_verbose("Test ucol_strcoll ends.\n");
372
373 log_verbose("testing ucol_getStrength() method ...\n");
374 doAssert( (ucol_getStrength(col) == UCOL_TERTIARY), "collation object has the wrong strength");
375 doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
376
377 log_verbose("testing ucol_setStrength() method ...\n");
378 ucol_setStrength(col, UCOL_SECONDARY);
379 doAssert( (ucol_getStrength(col) != UCOL_TERTIARY), "collation object's strength is secondary difference");
380 doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
381 doAssert( (ucol_getStrength(col) == UCOL_SECONDARY), "collation object has the wrong strength");
382
383
384 log_verbose("Get display name for the default collation in German : \n");
385
386 len=ucol_getDisplayName("en_US", "de_DE", NULL, 0, &status);
387 if(status==U_BUFFER_OVERFLOW_ERROR){
388 status=U_ZERO_ERROR;
389 disName=(UChar*)malloc(sizeof(UChar) * (len+1));
390 ucol_getDisplayName("en_US", "de_DE", disName, len+1, &status);
391 log_verbose("the display name for default collation in german: %s\n", austrdup(disName) );
392 free(disName);
393 }
394 if(U_FAILURE(status)){
395 log_err("ERROR: in getDisplayName: %s\n", myErrorName(status));
396 return;
397 }
398 log_verbose("Default collation getDisplayName ended.\n");
399
400 /* BEGIN android-removed
401 To save space, Android does not include the collation tailoring rules.
402 Skip the related tests.
403
404 ruled = ucol_open("da_DK", &status);
405 log_verbose("ucol_getRules() testing ...\n");
406 ucol_getRules(ruled, &tempLength);
407 doAssert( tempLength != 0, "getRules() result incorrect" );
408 log_verbose("getRules tests end.\n");
409 {
410 UChar *buffer = (UChar *)malloc(200000*sizeof(UChar));
411 int32_t bufLen = 200000;
412 buffer[0] = '\0';
413 log_verbose("ucol_getRulesEx() testing ...\n");
414 tempLength = ucol_getRulesEx(col,UCOL_TAILORING_ONLY,buffer,bufLen );
415 doAssert( tempLength == 0x00, "getRulesEx() result incorrect" );
416 log_verbose("getRules tests end.\n");
417
418 log_verbose("ucol_getRulesEx() testing ...\n");
419 tempLength=ucol_getRulesEx(col,UCOL_FULL_RULES,buffer,bufLen );
420 doAssert( tempLength != 0, "getRulesEx() result incorrect" );
421 log_verbose("getRules tests end.\n");
422 free(buffer);
423 }
424 ucol_close(ruled);
425 ucol_close(col);
426
427 END android-removed */
428
429 log_verbose("open an collator for french locale");
430 col = ucol_open("fr_FR", &status);
431 if (U_FAILURE(status)) {
432 log_err("ERROR: Creating French collation failed.: %s\n", myErrorName(status));
433 return;
434 }
435 ucol_setStrength(col, UCOL_PRIMARY);
436 log_verbose("testing ucol_getStrength() method again ...\n");
437 doAssert( (ucol_getStrength(col) != UCOL_TERTIARY), "collation object has the wrong strength");
438 doAssert( (ucol_getStrength(col) == UCOL_PRIMARY), "collation object's strength is not primary difference");
439
440 log_verbose("testing French ucol_setStrength() method ...\n");
441 ucol_setStrength(col, UCOL_TERTIARY);
442 doAssert( (ucol_getStrength(col) == UCOL_TERTIARY), "collation object's strength is not tertiary difference");
443 doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
444 doAssert( (ucol_getStrength(col) != UCOL_SECONDARY), "collation object's strength is secondary difference");
445 ucol_close(col);
446
447 log_verbose("Get display name for the french collation in english : \n");
448 len=ucol_getDisplayName("fr_FR", "en_US", NULL, 0, &status);
449 if(status==U_BUFFER_OVERFLOW_ERROR){
450 status=U_ZERO_ERROR;
451 disName=(UChar*)malloc(sizeof(UChar) * (len+1));
452 ucol_getDisplayName("fr_FR", "en_US", disName, len+1, &status);
453 log_verbose("the display name for french collation in english: %s\n", austrdup(disName) );
454 free(disName);
455 }
456 if(U_FAILURE(status)){
457 log_err("ERROR: in getDisplayName: %s\n", myErrorName(status));
458 return;
459 }
460 log_verbose("Default collation getDisplayName ended.\n");
461
462 }
463
464 /* Test RuleBasedCollator and getRules*/
TestRuleBasedColl()465 void TestRuleBasedColl()
466 {
467 UCollator *col1, *col2, *col3, *col4;
468 UCollationElements *iter1, *iter2;
469 UChar ruleset1[60];
470 UChar ruleset2[50];
471 UChar teststr[10];
472 UChar teststr2[10];
473 const UChar *rule1, *rule2, *rule3, *rule4;
474 int32_t tempLength;
475 UErrorCode status = U_ZERO_ERROR;
476 u_uastrcpy(ruleset1, "&9 < a, A < b, B < c, C; ch, cH, Ch, CH < d, D, e, E");
477 u_uastrcpy(ruleset2, "&9 < a, A < b, B < c, C < d, D, e, E");
478
479
480 col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL,&status);
481 if (U_FAILURE(status)) {
482 log_err_status(status, "RuleBased Collator creation failed.: %s\n", myErrorName(status));
483 return;
484 }
485 else
486 log_verbose("PASS: RuleBased Collator creation passed\n");
487
488 status = U_ZERO_ERROR;
489 col2 = ucol_openRules(ruleset2, u_strlen(ruleset2), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
490 if (U_FAILURE(status)) {
491 log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
492 return;
493 }
494 else
495 log_verbose("PASS: RuleBased Collator creation passed\n");
496
497
498 status = U_ZERO_ERROR;
499 col3= ucol_open(NULL, &status);
500 if (U_FAILURE(status)) {
501 log_err("Default Collator creation failed.: %s\n", myErrorName(status));
502 return;
503 }
504 else
505 log_verbose("PASS: Default Collator creation passed\n");
506
507 rule1 = ucol_getRules(col1, &tempLength);
508 rule2 = ucol_getRules(col2, &tempLength);
509 rule3 = ucol_getRules(col3, &tempLength);
510
511 doAssert((u_strcmp(rule1, rule2) != 0), "Default collator getRules failed");
512 doAssert((u_strcmp(rule2, rule3) != 0), "Default collator getRules failed");
513 doAssert((u_strcmp(rule1, rule3) != 0), "Default collator getRules failed");
514
515 col4=ucol_openRules(rule2, u_strlen(rule2), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
516 if (U_FAILURE(status)) {
517 log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
518 return;
519 }
520 rule4= ucol_getRules(col4, &tempLength);
521 doAssert((u_strcmp(rule2, rule4) == 0), "Default collator getRules failed");
522
523 ucol_close(col1);
524 ucol_close(col2);
525 ucol_close(col3);
526 ucol_close(col4);
527
528 /* tests that modifier ! is always ignored */
529 u_uastrcpy(ruleset1, "!&a<b");
530 teststr[0] = 0x0e40;
531 teststr[1] = 0x0e01;
532 teststr[2] = 0x0e2d;
533 col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
534 if (U_FAILURE(status)) {
535 log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
536 return;
537 }
538 col2 = ucol_open("en_US", &status);
539 if (U_FAILURE(status)) {
540 log_err("en_US Collator creation failed.: %s\n", myErrorName(status));
541 return;
542 }
543 iter1 = ucol_openElements(col1, teststr, 3, &status);
544 iter2 = ucol_openElements(col2, teststr, 3, &status);
545 if(U_FAILURE(status)) {
546 log_err("ERROR: CollationElement iterator creation failed.: %s\n", myErrorName(status));
547 return;
548 }
549 while (TRUE) {
550 /* testing with en since thai has its own tailoring */
551 uint32_t ce = ucol_next(iter1, &status);
552 uint32_t ce2 = ucol_next(iter2, &status);
553 if(U_FAILURE(status)) {
554 log_err("ERROR: CollationElement iterator creation failed.: %s\n", myErrorName(status));
555 return;
556 }
557 if (ce2 != ce) {
558 log_err("! modifier test failed");
559 }
560 if (ce == UCOL_NULLORDER) {
561 break;
562 }
563 }
564 ucol_closeElements(iter1);
565 ucol_closeElements(iter2);
566 ucol_close(col1);
567 ucol_close(col2);
568 /* test that we can start a rule without a & or < */
569 u_uastrcpy(ruleset1, "< z < a");
570 col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
571 if (U_FAILURE(status)) {
572 log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
573 return;
574 }
575 u_uastrcpy(teststr, "z");
576 u_uastrcpy(teststr2, "a");
577 if (ucol_greaterOrEqual(col1, teststr, 1, teststr2, 1)) {
578 log_err("Rule \"z < a\" fails");
579 }
580 ucol_close(col1);
581 }
582
TestCompare()583 void TestCompare()
584 {
585 UErrorCode status = U_ZERO_ERROR;
586 UCollator *col;
587 UChar* test1;
588 UChar* test2;
589
590 log_verbose("The compare tests begin : \n");
591 status=U_ZERO_ERROR;
592 col = ucol_open("en_US", &status);
593 if(U_FAILURE(status)) {
594 log_err_status(status, "ucal_open() collation creation failed.: %s\n", myErrorName(status));
595 return;
596 }
597 test1=(UChar*)malloc(sizeof(UChar) * 6);
598 test2=(UChar*)malloc(sizeof(UChar) * 6);
599 u_uastrcpy(test1, "Abcda");
600 u_uastrcpy(test2, "abcda");
601
602 log_verbose("Use tertiary comparison level testing ....\n");
603
604 doAssert( (!ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" != \"abcda\" ");
605 doAssert( (ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" >>> \"abcda\" ");
606 doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" >>> \"abcda\"");
607
608 ucol_setStrength(col, UCOL_SECONDARY);
609 log_verbose("Use secondary comparison level testing ....\n");
610
611 doAssert( (ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2) )), "Result should be \"Abcda\" == \"abcda\"");
612 doAssert( (!ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
613 doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2) )), "Result should be \"Abcda\" == \"abcda\"");
614
615 ucol_setStrength(col, UCOL_PRIMARY);
616 log_verbose("Use primary comparison level testing ....\n");
617
618 doAssert( (ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
619 doAssert( (!ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
620 doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
621
622
623 log_verbose("The compare tests end.\n");
624 ucol_close(col);
625 free(test1);
626 free(test2);
627
628 }
629 /*
630 ---------------------------------------------
631 tests decomposition setting
632 */
TestDecomposition()633 void TestDecomposition() {
634 UErrorCode status = U_ZERO_ERROR;
635 UCollator *en_US, *el_GR, *vi_VN;
636 en_US = ucol_open("en_US", &status);
637 el_GR = ucol_open("el_GR", &status);
638 vi_VN = ucol_open("vi_VN", &status);
639
640 if (U_FAILURE(status)) {
641 log_err_status(status, "ERROR: collation creation failed.: %s\n", myErrorName(status));
642 return;
643 }
644
645 if (ucol_getAttribute(vi_VN, UCOL_NORMALIZATION_MODE, &status) != UCOL_ON ||
646 U_FAILURE(status))
647 {
648 log_err("ERROR: vi_VN collation did not have cannonical decomposition for normalization!\n");
649 }
650
651 status = U_ZERO_ERROR;
652 if (ucol_getAttribute(el_GR, UCOL_NORMALIZATION_MODE, &status) != UCOL_ON ||
653 U_FAILURE(status))
654 {
655 log_err("ERROR: el_GR collation did not have cannonical decomposition for normalization!\n");
656 }
657
658 status = U_ZERO_ERROR;
659 if (ucol_getAttribute(en_US, UCOL_NORMALIZATION_MODE, &status) != UCOL_OFF ||
660 U_FAILURE(status))
661 {
662 log_err("ERROR: en_US collation had cannonical decomposition for normalization!\n");
663 }
664
665 ucol_close(en_US);
666 ucol_close(el_GR);
667 ucol_close(vi_VN);
668 }
669
670 #define CLONETEST_COLLATOR_COUNT 4
671
TestSafeClone()672 void TestSafeClone() {
673 UChar test1[6];
674 UChar test2[6];
675 static const UChar umlautUStr[] = {0x00DC, 0};
676 static const UChar oeStr[] = {0x0055, 0x0045, 0};
677 UCollator * someCollators [CLONETEST_COLLATOR_COUNT];
678 UCollator * someClonedCollators [CLONETEST_COLLATOR_COUNT];
679 UCollator * col;
680 UErrorCode err = U_ZERO_ERROR;
681 int8_t index = 6; /* Leave this here to test buffer alingment in memory*/
682 uint8_t buffer [CLONETEST_COLLATOR_COUNT] [U_COL_SAFECLONE_BUFFERSIZE];
683 int32_t bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
684 const char sampleRuleChars[] = "&Z < CH";
685 UChar sampleRule[sizeof(sampleRuleChars)];
686
687 if (TestBufferSize()) {
688 log_err("U_COL_SAFECLONE_BUFFERSIZE should be larger than sizeof(UCollator)\n");
689 return;
690 }
691
692 u_uastrcpy(test1, "abCda");
693 u_uastrcpy(test2, "abcda");
694 u_uastrcpy(sampleRule, sampleRuleChars);
695
696 /* one default collator & two complex ones */
697 someCollators[0] = ucol_open("en_US", &err);
698 someCollators[1] = ucol_open("ko", &err);
699 someCollators[2] = ucol_open("ja_JP", &err);
700 someCollators[3] = ucol_openRules(sampleRule, -1, UCOL_ON, UCOL_TERTIARY, NULL, &err);
701 if(U_FAILURE(err)) {
702 for (index = 0; index < CLONETEST_COLLATOR_COUNT; index++) {
703 ucol_close(someCollators[index]);
704 }
705 log_data_err("Couldn't open one or more collators\n");
706 return;
707 }
708
709 /* Check the various error & informational states: */
710
711 /* Null status - just returns NULL */
712 if (0 != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, 0))
713 {
714 log_err("FAIL: Cloned Collator failed to deal correctly with null status\n");
715 }
716 /* error status - should return 0 & keep error the same */
717 err = U_MEMORY_ALLOCATION_ERROR;
718 if (0 != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err) || err != U_MEMORY_ALLOCATION_ERROR)
719 {
720 log_err("FAIL: Cloned Collator failed to deal correctly with incoming error status\n");
721 }
722 err = U_ZERO_ERROR;
723
724 /* Null buffer size pointer - just returns NULL & set error to U_ILLEGAL_ARGUMENT_ERROR*/
725 if (0 != ucol_safeClone(someCollators[0], buffer[0], 0, &err) || err != U_ILLEGAL_ARGUMENT_ERROR)
726 {
727 log_err("FAIL: Cloned Collator failed to deal correctly with null bufferSize pointer\n");
728 }
729 err = U_ZERO_ERROR;
730
731 /* buffer size pointer is 0 - fill in pbufferSize with a size */
732 bufferSize = 0;
733 if (0 != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err) || U_FAILURE(err) || bufferSize <= 0)
734 {
735 log_err("FAIL: Cloned Collator failed a sizing request ('preflighting')\n");
736 }
737 /* Verify our define is large enough */
738 if (U_COL_SAFECLONE_BUFFERSIZE < bufferSize)
739 {
740 log_err("FAIL: Pre-calculated buffer size is too small\n");
741 }
742 /* Verify we can use this run-time calculated size */
743 if (0 == (col = ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err)) || U_FAILURE(err))
744 {
745 log_err("FAIL: Collator can't be cloned with run-time size\n");
746 }
747 if (col) ucol_close(col);
748 /* size one byte too small - should allocate & let us know */
749 --bufferSize;
750 if (0 == (col = ucol_safeClone(someCollators[0], 0, &bufferSize, &err)) || err != U_SAFECLONE_ALLOCATED_WARNING)
751 {
752 log_err("FAIL: Cloned Collator failed to deal correctly with too-small buffer size\n");
753 }
754 if (col) ucol_close(col);
755 err = U_ZERO_ERROR;
756 bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
757
758
759 /* Null buffer pointer - return Collator & set error to U_SAFECLONE_ALLOCATED_ERROR */
760 if (0 == (col = ucol_safeClone(someCollators[0], 0, &bufferSize, &err)) || err != U_SAFECLONE_ALLOCATED_WARNING)
761 {
762 log_err("FAIL: Cloned Collator failed to deal correctly with null buffer pointer\n");
763 }
764 if (col) ucol_close(col);
765 err = U_ZERO_ERROR;
766
767 /* Null Collator - return NULL & set U_ILLEGAL_ARGUMENT_ERROR */
768 if (0 != ucol_safeClone(0, buffer[0], &bufferSize, &err) || err != U_ILLEGAL_ARGUMENT_ERROR)
769 {
770 log_err("FAIL: Cloned Collator failed to deal correctly with null Collator pointer\n");
771 }
772
773 err = U_ZERO_ERROR;
774
775 /* Test that a cloned collator doesn't accidentally use UCA. */
776 col=ucol_open("de@collation=phonebook", &err);
777 bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
778 someClonedCollators[0] = ucol_safeClone(col, buffer[0], &bufferSize, &err);
779 doAssert( (ucol_greater(col, umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Original German phonebook collation sorts differently than expected");
780 doAssert( (ucol_greater(someClonedCollators[0], umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Cloned German phonebook collation sorts differently than expected");
781 if (!ucol_equals(someClonedCollators[0], col)) {
782 log_err("FAIL: Cloned German phonebook collator is not equal to original.\n");
783 }
784 ucol_close(col);
785 ucol_close(someClonedCollators[0]);
786
787 err = U_ZERO_ERROR;
788
789 /* change orig & clone & make sure they are independent */
790
791 for (index = 0; index < CLONETEST_COLLATOR_COUNT; index++)
792 {
793 ucol_setStrength(someCollators[index], UCOL_IDENTICAL);
794 bufferSize = 1;
795 err = U_ZERO_ERROR;
796 ucol_close(ucol_safeClone(someCollators[index], buffer[index], &bufferSize, &err));
797 if (err != U_SAFECLONE_ALLOCATED_WARNING) {
798 log_err("FAIL: collator number %d was not allocated.\n", index);
799 log_err("FAIL: status of Collator[%d] is %d (hex: %x).\n", index, err, err);
800 }
801
802 bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
803 err = U_ZERO_ERROR;
804 someClonedCollators[index] = ucol_safeClone(someCollators[index], buffer[index], &bufferSize, &err);
805 if (someClonedCollators[index] == NULL
806 || someClonedCollators[index] < (UCollator *)buffer[index]
807 || someClonedCollators[index] > (UCollator *)(buffer[index]+(U_COL_SAFECLONE_BUFFERSIZE-1)))
808 {
809 log_err("FAIL: Cloned collator didn't use provided buffer.\n");
810 return;
811 }
812 if (!ucol_equals(someClonedCollators[index], someCollators[index])) {
813 log_err("FAIL: Cloned collator is not equal to original at index = %d.\n", index);
814 }
815
816 /* Check the usability */
817 ucol_setStrength(someCollators[index], UCOL_PRIMARY);
818 ucol_setAttribute(someCollators[index], UCOL_CASE_LEVEL, UCOL_OFF, &err);
819
820 doAssert( (ucol_equal(someCollators[index], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abcda\" == \"abCda\"");
821
822 /* Close the original to make sure that the clone is usable. */
823 ucol_close(someCollators[index]);
824
825 ucol_setStrength(someClonedCollators[index], UCOL_TERTIARY);
826 ucol_setAttribute(someClonedCollators[index], UCOL_CASE_LEVEL, UCOL_OFF, &err);
827 doAssert( (ucol_greater(someClonedCollators[index], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abCda\" >>> \"abcda\" ");
828
829 ucol_close(someClonedCollators[index]);
830 }
831 }
832
TestCloneBinary()833 void TestCloneBinary(){
834 UErrorCode err = U_ZERO_ERROR;
835 UCollator * col = ucol_open("en_US", &err);
836 UCollator * c;
837 int32_t size;
838 uint8_t * buffer;
839
840 if (U_FAILURE(err)) {
841 log_data_err("Couldn't open collator. Error: %s\n", u_errorName(err));
842 return;
843 }
844
845 size = ucol_cloneBinary(col, NULL, 0, &err);
846 if(size==0 || err!=U_BUFFER_OVERFLOW_ERROR) {
847 log_err("ucol_cloneBinary - couldn't check size. Error: %s\n", u_errorName(err));
848 return;
849 }
850 err = U_ZERO_ERROR;
851
852 buffer = (uint8_t *) malloc(size);
853 ucol_cloneBinary(col, buffer, size, &err);
854 if(U_FAILURE(err)) {
855 log_err("ucol_cloneBinary - couldn't clone.. Error: %s\n", u_errorName(err));
856 free(buffer);
857 return;
858 }
859
860 /* how to check binary result ? */
861
862 c = ucol_openBinary(buffer, size, col, &err);
863 if(U_FAILURE(err)) {
864 log_err("ucol_openBinary failed. Error: %s\n", u_errorName(err));
865 } else {
866 UChar t[] = {0x41, 0x42, 0x43, 0}; /* ABC */
867 uint8_t *k1, *k2;
868 int l1, l2;
869 l1 = ucol_getSortKey(col, t, -1, NULL,0);
870 l2 = ucol_getSortKey(c, t, -1, NULL,0);
871 k1 = (uint8_t *) malloc(sizeof(uint8_t) * l1);
872 k2 = (uint8_t *) malloc(sizeof(uint8_t) * l2);
873 ucol_getSortKey(col, t, -1, k1, l1);
874 ucol_getSortKey(col, t, -1, k2, l2);
875 if (strcmp((char *)k1,(char *)k2) != 0){
876 log_err("ucol_openBinary - new collator should equal to old one\n");
877 };
878 free(k1);
879 free(k2);
880 }
881 free(buffer);
882 ucol_close(c);
883 ucol_close(col);
884 }
885
886
TestBengaliSortKey(void)887 static void TestBengaliSortKey(void)
888 {
889 const char *curLoc = "bn";
890 UChar str1[] = { 0x09BE, 0 };
891 UChar str2[] = { 0x0B70, 0 };
892 UCollator *c2 = NULL;
893 const UChar *rules;
894 int32_t rulesLength=-1;
895 uint8_t *sortKey1;
896 int32_t sortKeyLen1 = 0;
897 uint8_t *sortKey2;
898 int32_t sortKeyLen2 = 0;
899 UErrorCode status = U_ZERO_ERROR;
900 char sortKeyStr1[2048];
901 uint32_t sortKeyStrLen1 = sizeof(sortKeyStr1)/sizeof(sortKeyStr1[0]);
902 char sortKeyStr2[2048];
903 uint32_t sortKeyStrLen2 = sizeof(sortKeyStr2)/sizeof(sortKeyStr2[0]);
904 UCollationResult result;
905
906 static UChar preRules[41] = { 0x26, 0x9fa, 0x3c, 0x98c, 0x3c, 0x9e1, 0x3c, 0x98f, 0x3c, 0x990, 0x3c, 0x993, 0x3c, 0x994, 0x3c, 0x9bc, 0x3c, 0x982, 0x3c, 0x983, 0x3c, 0x981, 0x3c, 0x9b0, 0x3c, 0x9b8, 0x3c, 0x9b9, 0x3c, 0x9bd, 0x3c, 0x9be, 0x3c, 0x9bf, 0x3c, 0x9c8, 0x3c, 0x9cb, 0x3d, 0x9cb , 0};
907
908 rules = preRules;
909
910 log_verbose("Rules: %s\n", aescstrdup(rules, rulesLength));
911
912 c2 = ucol_openRules(rules, rulesLength, UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
913 if (U_FAILURE(status)) {
914 log_data_err("ERROR: Creating collator from rules failed with locale: %s : %s\n", curLoc, myErrorName(status));
915 return;
916 }
917
918 sortKeyLen1 = ucol_getSortKey(c2, str1, -1, NULL, 0);
919 sortKey1 = (uint8_t*)malloc(sortKeyLen1+1);
920 ucol_getSortKey(c2,str1,-1,sortKey1, sortKeyLen1+1);
921 ucol_sortKeyToString(c2, sortKey1, sortKeyStr1, &sortKeyStrLen1);
922
923
924 sortKeyLen2 = ucol_getSortKey(c2, str2, -1, NULL, 0);
925 sortKey2 = (uint8_t*)malloc(sortKeyLen2+1);
926 ucol_getSortKey(c2,str2,-1,sortKey2, sortKeyLen2+1);
927
928 ucol_sortKeyToString(c2, sortKey2, sortKeyStr2, &sortKeyStrLen2);
929
930
931
932 result=ucol_strcoll(c2, str1, -1, str2, -1);
933 if(result!=UCOL_LESS) {
934 log_err("Error: %s was not less than %s: result=%d.\n", aescstrdup(str1,-1), aescstrdup(str2,-1), result);
935 log_info("[%s] -> %s (%d, from rule)\n", aescstrdup(str1,-1), sortKeyStr1, sortKeyLen1);
936 log_info("[%s] -> %s (%d, from rule)\n", aescstrdup(str2,-1), sortKeyStr2, sortKeyLen2);
937 } else {
938 log_verbose("OK: %s was less than %s: result=%d.\n", aescstrdup(str1,-1), aescstrdup(str2,-1), result);
939 log_verbose("[%s] -> %s (%d, from rule)\n", aescstrdup(str1,-1), sortKeyStr1, sortKeyLen1);
940 log_verbose("[%s] -> %s (%d, from rule)\n", aescstrdup(str2,-1), sortKeyStr2, sortKeyLen2);
941 }
942
943 free(sortKey1);
944 free(sortKey2);
945 ucol_close(c2);
946
947 }
948
949 /*
950 TestOpenVsOpenRules ensures that collators from ucol_open and ucol_openRules
951 will generate identical sort keys
952 */
TestOpenVsOpenRules()953 void TestOpenVsOpenRules(){
954
955 /* create an array of all the locales */
956 int32_t numLocales = uloc_countAvailable();
957 int32_t sizeOfStdSet;
958 uint32_t adder;
959 UChar str[41]; /* create an array of UChar of size maximum strSize + 1 */
960 USet *stdSet;
961 char* curLoc;
962 UCollator * c1;
963 UCollator * c2;
964 const UChar* rules;
965 int32_t rulesLength;
966 int32_t sortKeyLen1, sortKeyLen2;
967 uint8_t *sortKey1 = NULL, *sortKey2 = NULL;
968 ULocaleData *uld;
969 int32_t x, y, z;
970 USet *eSet;
971 int32_t eSize;
972 int strSize;
973
974 UErrorCode err = U_ZERO_ERROR;
975
976 /* create a set of standard characters that aren't very interesting...
977 and then we can find some interesting ones later */
978
979 stdSet = uset_open(0x61, 0x7A);
980 uset_addRange(stdSet, 0x41, 0x5A);
981 uset_addRange(stdSet, 0x30, 0x39);
982 sizeOfStdSet = uset_size(stdSet);
983
984 adder = 1;
985 if(getTestOption(QUICK_OPTION))
986 {
987 adder = 10;
988 }
989
990 for(x = 0; x < numLocales; x+=adder){
991 curLoc = (char *)uloc_getAvailable(x);
992 log_verbose("Processing %s\n", curLoc);
993
994 /* create a collator the normal API way */
995 c1 = ucol_open(curLoc, &err);
996 if (U_FAILURE(err)) {
997 log_err("ERROR: Normal collation creation failed with locale: %s : %s\n", curLoc, myErrorName(err));
998 return;
999 }
1000
1001 /* grab the rules */
1002 rules = ucol_getRules(c1, &rulesLength);
1003
1004 /* use those rules to create a collator from rules */
1005 c2 = ucol_openRules(rules, rulesLength, UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &err);
1006 if (U_FAILURE(err)) {
1007 log_err("ERROR: Creating collator from rules failed with locale: %s : %s\n", curLoc, myErrorName(err));
1008 return;
1009 }
1010
1011 uld = ulocdata_open(curLoc, &err);
1012
1013 /*now that we have some collators, we get several strings */
1014
1015 for(y = 0; y < 5; y++){
1016
1017 /* get a set of ALL the characters in this locale */
1018 eSet = ulocdata_getExemplarSet(uld, NULL, 0, ULOCDATA_ES_STANDARD, &err);
1019 eSize = uset_size(eSet);
1020
1021 /* make a string with these characters in it */
1022 strSize = (rand()%40) + 1;
1023
1024 for(z = 0; z < strSize; z++){
1025 str[z] = uset_charAt(eSet, rand()%eSize);
1026 }
1027
1028 /* change the set to only include 'abnormal' characters (not A-Z, a-z, 0-9 */
1029 uset_removeAll(eSet, stdSet);
1030 eSize = uset_size(eSet);
1031
1032 /* if there are some non-normal characters left, put a few into the string, just to make sure we have some */
1033 if(eSize > 0){
1034 str[2%strSize] = uset_charAt(eSet, rand()%eSize);
1035 str[3%strSize] = uset_charAt(eSet, rand()%eSize);
1036 str[5%strSize] = uset_charAt(eSet, rand()%eSize);
1037 str[10%strSize] = uset_charAt(eSet, rand()%eSize);
1038 str[13%strSize] = uset_charAt(eSet, rand()%eSize);
1039 }
1040 /* terminate the string */
1041 str[strSize-1] = '\0';
1042 log_verbose("String used: %S\n", str);
1043
1044 /* get sort keys for both of them, and check that the keys are identicle */
1045 sortKeyLen1 = ucol_getSortKey(c1, str, u_strlen(str), NULL, 0);
1046 sortKey1 = (uint8_t*)malloc(sizeof(uint8_t) * (sortKeyLen1 + 1));
1047 /*memset(sortKey1, 0xFE, sortKeyLen1);*/
1048 ucol_getSortKey(c1, str, u_strlen(str), sortKey1, sortKeyLen1 + 1);
1049
1050 sortKeyLen2 = ucol_getSortKey(c2, str, u_strlen(str), NULL, 0);
1051 sortKey2 = (uint8_t*)malloc(sizeof(uint8_t) * (sortKeyLen2 + 1));
1052 /*memset(sortKey2, 0xFE, sortKeyLen2);*/
1053 ucol_getSortKey(c2, str, u_strlen(str), sortKey2, sortKeyLen2 + 1);
1054
1055 /* Check that the lengths are the same */
1056 doAssert((sortKeyLen1 == sortKeyLen2), "Sort key lengths do not match.");
1057
1058 /* check that the keys are the same */
1059 doAssert((memcmp(sortKey1, sortKey2, sortKeyLen1) == 0), "Keys are not equivalent");
1060
1061 /* clean up after each string */
1062 free(sortKey1);
1063 free(sortKey2);
1064 uset_close(eSet);
1065 }
1066 /* clean up after each locale */
1067 ulocdata_close(uld);
1068 ucol_close(c1);
1069 ucol_close(c2);
1070 }
1071 /* final clean up */
1072 uset_close(stdSet);
1073 }
1074 /*
1075 ----------------------------------------------------------------------------
1076 ctor -- Tests the getSortKey
1077 */
TestSortKey()1078 void TestSortKey()
1079 {
1080 uint8_t *sortk1 = NULL, *sortk2 = NULL, *sortk3 = NULL, *sortkEmpty = NULL;
1081 int32_t sortklen, osortklen;
1082 uint32_t toStringLen=0;
1083 UCollator *col;
1084 UChar *test1, *test2, *test3;
1085 UErrorCode status = U_ZERO_ERROR;
1086 char toStringBuffer[256], *resultP;
1087
1088
1089 uint8_t s1[] = { 0x9f, 0x00 };
1090 uint8_t s2[] = { 0x61, 0x00 };
1091 int strcmpResult;
1092
1093 strcmpResult = strcmp((const char *)s1, (const char *)s2);
1094 log_verbose("strcmp(0x9f..., 0x61...) = %d\n", strcmpResult);
1095
1096 if(strcmpResult <= 0) {
1097 log_err("ERR: expected strcmp(\"9f 00\", \"61 00\") to be >=0 (GREATER).. got %d. Calling strcmp() for sortkeys may not work! \n",
1098 strcmpResult);
1099 }
1100
1101
1102 log_verbose("testing SortKey begins...\n");
1103 /* this is supposed to open default date format, but later on it treats it like it is "en_US"
1104 - very bad if you try to run the tests on machine where default locale is NOT "en_US" */
1105 /* col = ucol_open(NULL, &status); */
1106 col = ucol_open("en_US", &status);
1107 if (U_FAILURE(status)) {
1108 log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1109 return;
1110 }
1111
1112
1113 if(ucol_getStrength(col) != UCOL_DEFAULT_STRENGTH)
1114 {
1115 log_err("ERROR: default collation did not have UCOL_DEFAULT_STRENGTH !\n");
1116 }
1117 /* Need to use identical strength */
1118 ucol_setAttribute(col, UCOL_STRENGTH, UCOL_IDENTICAL, &status);
1119
1120 test1=(UChar*)malloc(sizeof(UChar) * 6);
1121 test2=(UChar*)malloc(sizeof(UChar) * 6);
1122 test3=(UChar*)malloc(sizeof(UChar) * 6);
1123
1124 memset(test1,0xFE, sizeof(UChar)*6);
1125 memset(test2,0xFE, sizeof(UChar)*6);
1126 memset(test3,0xFE, sizeof(UChar)*6);
1127
1128
1129 u_uastrcpy(test1, "Abcda");
1130 u_uastrcpy(test2, "abcda");
1131 u_uastrcpy(test3, "abcda");
1132
1133 log_verbose("Use tertiary comparison level testing ....\n");
1134
1135 sortklen=ucol_getSortKey(col, test1, u_strlen(test1), NULL, 0);
1136 sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1137 memset(sortk1,0xFE, sortklen);
1138 ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortklen+1);
1139
1140 sortklen=ucol_getSortKey(col, test2, u_strlen(test2), NULL, 0);
1141 sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1142 memset(sortk2,0xFE, sortklen);
1143 ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortklen+1);
1144
1145 osortklen = sortklen;
1146 sortklen=ucol_getSortKey(col, test2, u_strlen(test3), NULL, 0);
1147 sortk3=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1148 memset(sortk3,0xFE, sortklen);
1149 ucol_getSortKey(col, test2, u_strlen(test2), sortk3, sortklen+1);
1150
1151 doAssert( (sortklen == osortklen), "Sortkey length should be the same (abcda, abcda)");
1152
1153 doAssert( (memcmp(sortk1, sortk2, sortklen) > 0), "Result should be \"Abcda\" > \"abcda\"");
1154 doAssert( (memcmp(sortk2, sortk1, sortklen) < 0), "Result should be \"abcda\" < \"Abcda\"");
1155 doAssert( (memcmp(sortk2, sortk3, sortklen) == 0), "Result should be \"abcda\" == \"abcda\"");
1156
1157 resultP = ucol_sortKeyToString(col, sortk3, toStringBuffer, &toStringLen);
1158 doAssert( (resultP != 0), "sortKeyToString failed!");
1159
1160 #if 1 /* verobse log of sortkeys */
1161 {
1162 char junk2[1000];
1163 char junk3[1000];
1164 int i;
1165
1166 strcpy(junk2, "abcda[2] ");
1167 strcpy(junk3, " abcda[3] ");
1168
1169 for(i=0;i<sortklen;i++)
1170 {
1171 sprintf(junk2+strlen(junk2), "%02X ",(int)( 0xFF & sortk2[i]));
1172 sprintf(junk3+strlen(junk3), "%02X ",(int)( 0xFF & sortk3[i]));
1173 }
1174
1175 log_verbose("%s\n", junk2);
1176 log_verbose("%s\n", junk3);
1177 }
1178 #endif
1179
1180 free(sortk1);
1181 free(sortk2);
1182 free(sortk3);
1183
1184 log_verbose("Use secondary comparision level testing ...\n");
1185 ucol_setStrength(col, UCOL_SECONDARY);
1186 sortklen=ucol_getSortKey(col, test1, u_strlen(test1), NULL, 0);
1187 sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1188 ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortklen+1);
1189 sortklen=ucol_getSortKey(col, test2, u_strlen(test2), NULL, 0);
1190 sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1191 ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortklen+1);
1192
1193 doAssert( !(memcmp(sortk1, sortk2, sortklen) > 0), "Result should be \"Abcda\" == \"abcda\"");
1194 doAssert( !(memcmp(sortk2, sortk1, sortklen) < 0), "Result should be \"abcda\" == \"Abcda\"");
1195 doAssert( (memcmp(sortk1, sortk2, sortklen) == 0), "Result should be \"abcda\" == \"abcda\"");
1196
1197 log_verbose("getting sortkey for an empty string\n");
1198 ucol_setAttribute(col, UCOL_STRENGTH, UCOL_TERTIARY, &status);
1199 sortklen = ucol_getSortKey(col, test1, 0, NULL, 0);
1200 sortkEmpty = (uint8_t*)malloc(sizeof(uint8_t) * sortklen+1);
1201 sortklen = ucol_getSortKey(col, test1, 0, sortkEmpty, sortklen+1);
1202 if(sortklen != 3 || sortkEmpty[0] != 1 || sortkEmpty[0] != 1 || sortkEmpty[2] != 0) {
1203 log_err("Empty string generated wrong sortkey!\n");
1204 }
1205 free(sortkEmpty);
1206
1207 log_verbose("testing passing invalid string\n");
1208 sortklen = ucol_getSortKey(col, NULL, 0, NULL, 0);
1209 if(sortklen != 0) {
1210 log_err("Invalid string didn't return sortkey size of 0\n");
1211 }
1212
1213
1214 log_verbose("testing sortkey ends...\n");
1215 ucol_close(col);
1216 free(test1);
1217 free(test2);
1218 free(test3);
1219 free(sortk1);
1220 free(sortk2);
1221
1222 }
TestHashCode()1223 void TestHashCode()
1224 {
1225 uint8_t *sortk1, *sortk2, *sortk3;
1226 int32_t sortk1len, sortk2len, sortk3len;
1227 UCollator *col;
1228 UChar *test1, *test2, *test3;
1229 UErrorCode status = U_ZERO_ERROR;
1230 log_verbose("testing getHashCode begins...\n");
1231 col = ucol_open("en_US", &status);
1232 if (U_FAILURE(status)) {
1233 log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1234 return;
1235 }
1236 test1=(UChar*)malloc(sizeof(UChar) * 6);
1237 test2=(UChar*)malloc(sizeof(UChar) * 6);
1238 test3=(UChar*)malloc(sizeof(UChar) * 6);
1239 u_uastrcpy(test1, "Abcda");
1240 u_uastrcpy(test2, "abcda");
1241 u_uastrcpy(test3, "abcda");
1242
1243 log_verbose("Use tertiary comparison level testing ....\n");
1244 sortk1len=ucol_getSortKey(col, test1, u_strlen(test1), NULL, 0);
1245 sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortk1len+1));
1246 ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortk1len+1);
1247 sortk2len=ucol_getSortKey(col, test2, u_strlen(test2), NULL, 0);
1248 sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortk2len+1));
1249 ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortk2len+1);
1250 sortk3len=ucol_getSortKey(col, test2, u_strlen(test3), NULL, 0);
1251 sortk3=(uint8_t*)malloc(sizeof(uint8_t) * (sortk3len+1));
1252 ucol_getSortKey(col, test2, u_strlen(test2), sortk3, sortk3len+1);
1253
1254
1255 log_verbose("ucol_hashCode() testing ...\n");
1256
1257 doAssert( ucol_keyHashCode(sortk1, sortk1len) != ucol_keyHashCode(sortk2, sortk2len), "Hash test1 result incorrect" );
1258 doAssert( !(ucol_keyHashCode(sortk1, sortk1len) == ucol_keyHashCode(sortk2, sortk2len)), "Hash test2 result incorrect" );
1259 doAssert( ucol_keyHashCode(sortk2, sortk2len) == ucol_keyHashCode(sortk3, sortk3len), "Hash result not equal" );
1260
1261 log_verbose("hashCode tests end.\n");
1262 ucol_close(col);
1263 free(sortk1);
1264 free(sortk2);
1265 free(sortk3);
1266 free(test1);
1267 free(test2);
1268 free(test3);
1269
1270
1271 }
1272 /*
1273 *----------------------------------------------------------------------------
1274 * Tests the UCollatorElements API.
1275 *
1276 */
TestElemIter()1277 void TestElemIter()
1278 {
1279 int32_t offset;
1280 int32_t order1, order2, order3;
1281 UChar *testString1, *testString2;
1282 UCollator *col;
1283 UCollationElements *iterator1, *iterator2, *iterator3;
1284 UErrorCode status = U_ZERO_ERROR;
1285 log_verbose("testing UCollatorElements begins...\n");
1286 col = ucol_open("en_US", &status);
1287 ucol_setAttribute(col, UCOL_NORMALIZATION_MODE, UCOL_OFF, &status);
1288 if (U_FAILURE(status)) {
1289 log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1290 return;
1291 }
1292
1293 testString1=(UChar*)malloc(sizeof(UChar) * 150);
1294 testString2=(UChar*)malloc(sizeof(UChar) * 150);
1295 u_uastrcpy(testString1, "XFILE What subset of all possible test cases has the highest probability of detecting the most errors?");
1296 u_uastrcpy(testString2, "Xf_ile What subset of all possible test cases has the lowest probability of detecting the least errors?");
1297
1298 log_verbose("Constructors and comparison testing....\n");
1299
1300 iterator1 = ucol_openElements(col, testString1, u_strlen(testString1), &status);
1301 if(U_FAILURE(status)) {
1302 log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1303 ucol_close(col);
1304 return;
1305 }
1306 else{ log_verbose("PASS: Default collationElement iterator1 creation passed\n");}
1307
1308 iterator2 = ucol_openElements(col, testString1, u_strlen(testString1), &status);
1309 if(U_FAILURE(status)) {
1310 log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1311 ucol_close(col);
1312 return;
1313 }
1314 else{ log_verbose("PASS: Default collationElement iterator2 creation passed\n");}
1315
1316 iterator3 = ucol_openElements(col, testString2, u_strlen(testString2), &status);
1317 if(U_FAILURE(status)) {
1318 log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1319 ucol_close(col);
1320 return;
1321 }
1322 else{ log_verbose("PASS: Default collationElement iterator3 creation passed\n");}
1323
1324 offset=ucol_getOffset(iterator1);
1325 ucol_setOffset(iterator1, 6, &status);
1326 if (U_FAILURE(status)) {
1327 log_err("Error in setOffset for UCollatorElements iterator.: %s\n", myErrorName(status));
1328 return;
1329 }
1330 if(ucol_getOffset(iterator1)==6)
1331 log_verbose("setOffset and getOffset working fine\n");
1332 else{
1333 log_err("error in set and get Offset got %d instead of 6\n", ucol_getOffset(iterator1));
1334 }
1335
1336 ucol_setOffset(iterator1, 0, &status);
1337 order1 = ucol_next(iterator1, &status);
1338 if (U_FAILURE(status)) {
1339 log_err("Somehow ran out of memory stepping through the iterator1.: %s\n", myErrorName(status));
1340 return;
1341 }
1342 order2=ucol_getOffset(iterator2);
1343 doAssert((order1 != order2), "The first iterator advance failed");
1344 order2 = ucol_next(iterator2, &status);
1345 if (U_FAILURE(status)) {
1346 log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1347 return;
1348 }
1349 order3 = ucol_next(iterator3, &status);
1350 if (U_FAILURE(status)) {
1351 log_err("Somehow ran out of memory stepping through the iterator3.: %s\n", myErrorName(status));
1352 return;
1353 }
1354
1355 doAssert((order1 == order2), "The second iterator advance failed should be the same as first one");
1356
1357 doAssert( (ucol_primaryOrder(order1) == ucol_primaryOrder(order3)), "The primary orders should be identical");
1358 doAssert( (ucol_secondaryOrder(order1) == ucol_secondaryOrder(order3)), "The secondary orders should be identical");
1359 doAssert( (ucol_tertiaryOrder(order1) == ucol_tertiaryOrder(order3)), "The tertiary orders should be identical");
1360
1361 order1=ucol_next(iterator1, &status);
1362 if (U_FAILURE(status)) {
1363 log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1364 return;
1365 }
1366 order3=ucol_next(iterator3, &status);
1367 if (U_FAILURE(status)) {
1368 log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1369 return;
1370 }
1371 doAssert( (ucol_primaryOrder(order1) == ucol_primaryOrder(order3)), "The primary orders should be identical");
1372 doAssert( (ucol_tertiaryOrder(order1) != ucol_tertiaryOrder(order3)), "The tertiary orders should be different");
1373
1374 order1=ucol_next(iterator1, &status);
1375 if (U_FAILURE(status)) {
1376 log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1377 return;
1378 }
1379 order3=ucol_next(iterator3, &status);
1380 if (U_FAILURE(status)) {
1381 log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1382 return;
1383 }
1384 /* this here, my friends, is either pure lunacy or something so obsolete that even it's mother
1385 * doesn't care about it. Essentialy, this test complains if secondary values for 'I' and '_'
1386 * are the same. According to the UCA, this is not true. Therefore, remove the test.
1387 * Besides, if primary strengths for two code points are different, it doesn't matter one bit
1388 * what is the relation between secondary or any other strengths.
1389 * killed by weiv 06/11/2002.
1390 */
1391 /*
1392 doAssert( ((order1 & UCOL_SECONDARYMASK) != (order3 & UCOL_SECONDARYMASK)), "The secondary orders should be different");
1393 */
1394 doAssert( (order1 != UCOL_NULLORDER), "Unexpected end of iterator reached");
1395
1396 free(testString1);
1397 free(testString2);
1398 ucol_closeElements(iterator1);
1399 ucol_closeElements(iterator2);
1400 ucol_closeElements(iterator3);
1401 ucol_close(col);
1402
1403 log_verbose("testing CollationElementIterator ends...\n");
1404 }
1405
TestGetLocale()1406 void TestGetLocale() {
1407 UErrorCode status = U_ZERO_ERROR;
1408 const char *rules = "&a<x<y<z";
1409 UChar rlz[256] = {0};
1410 uint32_t rlzLen = u_unescape(rules, rlz, 256);
1411
1412 UCollator *coll = NULL;
1413 const char *locale = NULL;
1414
1415 int32_t i = 0;
1416
1417 /* Now that the collation tree is separate, actual==valid at all times. [alan] */
1418 static const struct {
1419 const char* requestedLocale;
1420 const char* validLocale;
1421 const char* actualLocale;
1422 } testStruct[] = {
1423 { "sr_RS", "sr_Cyrl_RS", "sr" },
1424 { "sh_YU", "sr_Latn_RS", "hr" }, /* this used to be sh, but now sh collation aliases hr */
1425 { "en_BE_FOO", "en_BE", "root" },
1426 { "de_DE_NONEXISTANT", "de_DE", "de" }
1427 };
1428
1429 /* test opening collators for different locales */
1430 for(i = 0; i<sizeof(testStruct)/sizeof(testStruct[0]); i++) {
1431 status = U_ZERO_ERROR;
1432 coll = ucol_open(testStruct[i].requestedLocale, &status);
1433 if(U_FAILURE(status)) {
1434 log_err_status(status, "Failed to open collator for %s with %s\n", testStruct[i].requestedLocale, u_errorName(status));
1435 ucol_close(coll);
1436 continue;
1437 }
1438 locale = ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status);
1439 if(strcmp(locale, testStruct[i].requestedLocale) != 0) {
1440 log_err("[Coll %s]: Error in requested locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].requestedLocale, locale);
1441 }
1442 locale = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status);
1443 if(strcmp(locale, testStruct[i].validLocale) != 0) {
1444 log_err("[Coll %s]: Error in valid locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].validLocale, locale);
1445 }
1446 locale = ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status);
1447 if(strcmp(locale, testStruct[i].actualLocale) != 0) {
1448 log_err("[Coll %s]: Error in actual locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].actualLocale, locale);
1449 }
1450 ucol_close(coll);
1451 }
1452
1453 /* completely non-existant locale for collator should get a default collator */
1454 {
1455 UCollator *defaultColl = ucol_open(NULL, &status);
1456 coll = ucol_open("blahaha", &status);
1457 if(U_SUCCESS(status)) {
1458 if(strcmp(ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status), "blahaha")) {
1459 log_err("Nonexisting locale didn't preserve the requested locale\n");
1460 }
1461 if(strcmp(ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status),
1462 ucol_getLocaleByType(defaultColl, ULOC_VALID_LOCALE, &status))) {
1463 log_err("Valid locale for nonexisting locale locale collator differs "
1464 "from valid locale for default collator\n");
1465 }
1466 if(strcmp(ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status),
1467 ucol_getLocaleByType(defaultColl, ULOC_ACTUAL_LOCALE, &status))) {
1468 log_err("Actual locale for nonexisting locale locale collator differs "
1469 "from actual locale for default collator\n");
1470 }
1471 ucol_close(coll);
1472 ucol_close(defaultColl);
1473 } else {
1474 log_data_err("Couldn't open collators\n");
1475 }
1476 }
1477
1478
1479
1480 /* collator instantiated from rules should have all three locales NULL */
1481 coll = ucol_openRules(rlz, rlzLen, UCOL_DEFAULT, UCOL_DEFAULT, NULL, &status);
1482 locale = ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status);
1483 if(locale != NULL) {
1484 log_err("For collator instantiated from rules, requested locale returned %s instead of NULL\n", locale);
1485 }
1486 locale = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status);
1487 if(locale != NULL) {
1488 log_err("For collator instantiated from rules, valid locale returned %s instead of NULL\n", locale);
1489 }
1490 locale = ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status);
1491 if(locale != NULL) {
1492 log_err("For collator instantiated from rules, actual locale returned %s instead of NULL\n", locale);
1493 }
1494 ucol_close(coll);
1495
1496 }
1497
1498
TestGetAll()1499 void TestGetAll()
1500 {
1501 int32_t i, count;
1502 count=ucol_countAvailable();
1503 /* use something sensible w/o hardcoding the count */
1504 if(count < 0){
1505 log_err("Error in countAvailable(), it returned %d\n", count);
1506 }
1507 else{
1508 log_verbose("PASS: countAvailable() successful, it returned %d\n", count);
1509 }
1510 for(i=0;i<count;i++)
1511 log_verbose("%s\n", ucol_getAvailable(i));
1512
1513
1514 }
1515
1516
1517 struct teststruct {
1518 const char *original;
1519 uint8_t key[256];
1520 } ;
1521
compare_teststruct(const void * string1,const void * string2)1522 static int compare_teststruct(const void *string1, const void *string2) {
1523 return(strcmp((const char *)((struct teststruct *)string1)->key, (const char *)((struct teststruct *)string2)->key));
1524 }
1525
TestBounds()1526 void TestBounds() {
1527 UErrorCode status = U_ZERO_ERROR;
1528
1529 UCollator *coll = ucol_open("sh", &status);
1530
1531 uint8_t sortkey[512], lower[512], upper[512];
1532 UChar buffer[512];
1533
1534 static const char * const test[] = {
1535 "John Smith",
1536 "JOHN SMITH",
1537 "john SMITH",
1538 "j\\u00F6hn sm\\u00EFth",
1539 "J\\u00F6hn Sm\\u00EFth",
1540 "J\\u00D6HN SM\\u00CFTH",
1541 "john smithsonian",
1542 "John Smithsonian",
1543 };
1544
1545 struct teststruct tests[] = {
1546 {"\\u010CAKI MIHALJ" } ,
1547 {"\\u010CAKI MIHALJ" } ,
1548 {"\\u010CAKI PIRO\\u0160KA" },
1549 {"\\u010CABAI ANDRIJA" } ,
1550 {"\\u010CABAI LAJO\\u0160" } ,
1551 {"\\u010CABAI MARIJA" } ,
1552 {"\\u010CABAI STEVAN" } ,
1553 {"\\u010CABAI STEVAN" } ,
1554 {"\\u010CABARKAPA BRANKO" } ,
1555 {"\\u010CABARKAPA MILENKO" } ,
1556 {"\\u010CABARKAPA MIROSLAV" } ,
1557 {"\\u010CABARKAPA SIMO" } ,
1558 {"\\u010CABARKAPA STANKO" } ,
1559 {"\\u010CABARKAPA TAMARA" } ,
1560 {"\\u010CABARKAPA TOMA\\u0160" } ,
1561 {"\\u010CABDARI\\u0106 NIKOLA" } ,
1562 {"\\u010CABDARI\\u0106 ZORICA" } ,
1563 {"\\u010CABI NANDOR" } ,
1564 {"\\u010CABOVI\\u0106 MILAN" } ,
1565 {"\\u010CABRADI AGNEZIJA" } ,
1566 {"\\u010CABRADI IVAN" } ,
1567 {"\\u010CABRADI JELENA" } ,
1568 {"\\u010CABRADI LJUBICA" } ,
1569 {"\\u010CABRADI STEVAN" } ,
1570 {"\\u010CABRDA MARTIN" } ,
1571 {"\\u010CABRILO BOGDAN" } ,
1572 {"\\u010CABRILO BRANISLAV" } ,
1573 {"\\u010CABRILO LAZAR" } ,
1574 {"\\u010CABRILO LJUBICA" } ,
1575 {"\\u010CABRILO SPASOJA" } ,
1576 {"\\u010CADE\\u0160 ZDENKA" } ,
1577 {"\\u010CADESKI BLAGOJE" } ,
1578 {"\\u010CADOVSKI VLADIMIR" } ,
1579 {"\\u010CAGLJEVI\\u0106 TOMA" } ,
1580 {"\\u010CAGOROVI\\u0106 VLADIMIR" } ,
1581 {"\\u010CAJA VANKA" } ,
1582 {"\\u010CAJI\\u0106 BOGOLJUB" } ,
1583 {"\\u010CAJI\\u0106 BORISLAV" } ,
1584 {"\\u010CAJI\\u0106 RADOSLAV" } ,
1585 {"\\u010CAK\\u0160IRAN MILADIN" } ,
1586 {"\\u010CAKAN EUGEN" } ,
1587 {"\\u010CAKAN EVGENIJE" } ,
1588 {"\\u010CAKAN IVAN" } ,
1589 {"\\u010CAKAN JULIJAN" } ,
1590 {"\\u010CAKAN MIHAJLO" } ,
1591 {"\\u010CAKAN STEVAN" } ,
1592 {"\\u010CAKAN VLADIMIR" } ,
1593 {"\\u010CAKAN VLADIMIR" } ,
1594 {"\\u010CAKAN VLADIMIR" } ,
1595 {"\\u010CAKARA ANA" } ,
1596 {"\\u010CAKAREVI\\u0106 MOMIR" } ,
1597 {"\\u010CAKAREVI\\u0106 NEDELJKO" } ,
1598 {"\\u010CAKI \\u0160ANDOR" } ,
1599 {"\\u010CAKI AMALIJA" } ,
1600 {"\\u010CAKI ANDRA\\u0160" } ,
1601 {"\\u010CAKI LADISLAV" } ,
1602 {"\\u010CAKI LAJO\\u0160" } ,
1603 {"\\u010CAKI LASLO" } ,
1604 };
1605
1606
1607
1608 int32_t i = 0, j = 0, k = 0, buffSize = 0, skSize = 0, lowerSize = 0, upperSize = 0;
1609 int32_t arraySize = sizeof(tests)/sizeof(tests[0]);
1610
1611 if(U_SUCCESS(status) && coll) {
1612 for(i = 0; i<arraySize; i++) {
1613 buffSize = u_unescape(tests[i].original, buffer, 512);
1614 skSize = ucol_getSortKey(coll, buffer, buffSize, tests[i].key, 512);
1615 }
1616
1617 qsort(tests, arraySize, sizeof(struct teststruct), compare_teststruct);
1618
1619 for(i = 0; i < arraySize-1; i++) {
1620 for(j = i+1; j < arraySize; j++) {
1621 lowerSize = ucol_getBound(tests[i].key, -1, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1622 upperSize = ucol_getBound(tests[j].key, -1, UCOL_BOUND_UPPER, 1, upper, 512, &status);
1623 for(k = i; k <= j; k++) {
1624 if(strcmp((const char *)lower, (const char *)tests[k].key) > 0) {
1625 log_err("Problem with lower! j = %i (%s vs %s)\n", k, tests[k].original, tests[i].original);
1626 }
1627 if(strcmp((const char *)upper, (const char *)tests[k].key) <= 0) {
1628 log_err("Problem with upper! j = %i (%s vs %s)\n", k, tests[k].original, tests[j].original);
1629 }
1630 }
1631 }
1632 }
1633
1634
1635 #if 0
1636 for(i = 0; i < 1000; i++) {
1637 lowerRND = (rand()/(RAND_MAX/arraySize));
1638 upperRND = lowerRND + (rand()/(RAND_MAX/(arraySize-lowerRND)));
1639
1640 lowerSize = ucol_getBound(tests[lowerRND].key, -1, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1641 upperSize = ucol_getBound(tests[upperRND].key, -1, UCOL_BOUND_UPPER_LONG, 1, upper, 512, &status);
1642
1643 for(j = lowerRND; j<=upperRND; j++) {
1644 if(strcmp(lower, tests[j].key) > 0) {
1645 log_err("Problem with lower! j = %i (%s vs %s)\n", j, tests[j].original, tests[lowerRND].original);
1646 }
1647 if(strcmp(upper, tests[j].key) <= 0) {
1648 log_err("Problem with upper! j = %i (%s vs %s)\n", j, tests[j].original, tests[upperRND].original);
1649 }
1650 }
1651 }
1652 #endif
1653
1654
1655
1656
1657
1658 for(i = 0; i<sizeof(test)/sizeof(test[0]); i++) {
1659 buffSize = u_unescape(test[i], buffer, 512);
1660 skSize = ucol_getSortKey(coll, buffer, buffSize, sortkey, 512);
1661 lowerSize = ucol_getBound(sortkey, skSize, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1662 upperSize = ucol_getBound(sortkey, skSize, UCOL_BOUND_UPPER_LONG, 1, upper, 512, &status);
1663 for(j = i+1; j<sizeof(test)/sizeof(test[0]); j++) {
1664 buffSize = u_unescape(test[j], buffer, 512);
1665 skSize = ucol_getSortKey(coll, buffer, buffSize, sortkey, 512);
1666 if(strcmp((const char *)lower, (const char *)sortkey) > 0) {
1667 log_err("Problem with lower! i = %i, j = %i (%s vs %s)\n", i, j, test[i], test[j]);
1668 }
1669 if(strcmp((const char *)upper, (const char *)sortkey) <= 0) {
1670 log_err("Problem with upper! i = %i, j = %i (%s vs %s)\n", i, j, test[i], test[j]);
1671 }
1672 }
1673 }
1674 ucol_close(coll);
1675 } else {
1676 log_data_err("Couldn't open collator\n");
1677 }
1678
1679 }
1680
doOverrunTest(UCollator * coll,const UChar * uString,int32_t strLen)1681 static void doOverrunTest(UCollator *coll, const UChar *uString, int32_t strLen) {
1682 int32_t skLen = 0, skLen2 = 0;
1683 uint8_t sortKey[256];
1684 int32_t i, j;
1685 uint8_t filler = 0xFF;
1686
1687 skLen = ucol_getSortKey(coll, uString, strLen, NULL, 0);
1688
1689 for(i = 0; i < skLen; i++) {
1690 memset(sortKey, filler, 256);
1691 skLen2 = ucol_getSortKey(coll, uString, strLen, sortKey, i);
1692 if(skLen != skLen2) {
1693 log_err("For buffer size %i, got different sortkey length. Expected %i got %i\n", i, skLen, skLen2);
1694 }
1695 for(j = i; j < 256; j++) {
1696 if(sortKey[j] != filler) {
1697 log_err("Something run over index %i\n", j);
1698 break;
1699 }
1700 }
1701 }
1702 }
1703
1704 /* j1865 reports that if a shorter buffer is passed to
1705 * to get sort key, a buffer overrun happens in some
1706 * cases. This test tries to check this.
1707 */
TestSortKeyBufferOverrun(void)1708 void TestSortKeyBufferOverrun(void) {
1709 UErrorCode status = U_ZERO_ERROR;
1710 const char* cString = "A very Merry liTTle-lamB..";
1711 UChar uString[256];
1712 int32_t strLen = 0;
1713 UCollator *coll = ucol_open("root", &status);
1714 strLen = u_unescape(cString, uString, 256);
1715
1716 if(U_SUCCESS(status)) {
1717 log_verbose("testing non ignorable\n");
1718 ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &status);
1719 doOverrunTest(coll, uString, strLen);
1720
1721 log_verbose("testing shifted\n");
1722 ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_SHIFTED, &status);
1723 doOverrunTest(coll, uString, strLen);
1724
1725 log_verbose("testing shifted quaternary\n");
1726 ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_QUATERNARY, &status);
1727 doOverrunTest(coll, uString, strLen);
1728
1729 log_verbose("testing with french secondaries\n");
1730 ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_ON, &status);
1731 ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_TERTIARY, &status);
1732 ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &status);
1733 doOverrunTest(coll, uString, strLen);
1734
1735 }
1736 ucol_close(coll);
1737 }
1738
TestAttribute()1739 static void TestAttribute()
1740 {
1741 UErrorCode error = U_ZERO_ERROR;
1742 UCollator *coll = ucol_open(NULL, &error);
1743
1744 if (U_FAILURE(error)) {
1745 log_err_status(error, "Creation of default collator failed\n");
1746 return;
1747 }
1748
1749 ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_OFF, &error);
1750 if (ucol_getAttribute(coll, UCOL_FRENCH_COLLATION, &error) != UCOL_OFF ||
1751 U_FAILURE(error)) {
1752 log_err_status(error, "Setting and retrieving of the french collation failed\n");
1753 }
1754
1755 ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_ON, &error);
1756 if (ucol_getAttribute(coll, UCOL_FRENCH_COLLATION, &error) != UCOL_ON ||
1757 U_FAILURE(error)) {
1758 log_err_status(error, "Setting and retrieving of the french collation failed\n");
1759 }
1760
1761 ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_SHIFTED, &error);
1762 if (ucol_getAttribute(coll, UCOL_ALTERNATE_HANDLING, &error) != UCOL_SHIFTED ||
1763 U_FAILURE(error)) {
1764 log_err_status(error, "Setting and retrieving of the alternate handling failed\n");
1765 }
1766
1767 ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &error);
1768 if (ucol_getAttribute(coll, UCOL_ALTERNATE_HANDLING, &error) != UCOL_NON_IGNORABLE ||
1769 U_FAILURE(error)) {
1770 log_err_status(error, "Setting and retrieving of the alternate handling failed\n");
1771 }
1772
1773 ucol_setAttribute(coll, UCOL_CASE_FIRST, UCOL_LOWER_FIRST, &error);
1774 if (ucol_getAttribute(coll, UCOL_CASE_FIRST, &error) != UCOL_LOWER_FIRST ||
1775 U_FAILURE(error)) {
1776 log_err_status(error, "Setting and retrieving of the case first attribute failed\n");
1777 }
1778
1779 ucol_setAttribute(coll, UCOL_CASE_FIRST, UCOL_UPPER_FIRST, &error);
1780 if (ucol_getAttribute(coll, UCOL_CASE_FIRST, &error) != UCOL_UPPER_FIRST ||
1781 U_FAILURE(error)) {
1782 log_err_status(error, "Setting and retrieving of the case first attribute failed\n");
1783 }
1784
1785 ucol_setAttribute(coll, UCOL_CASE_LEVEL, UCOL_ON, &error);
1786 if (ucol_getAttribute(coll, UCOL_CASE_LEVEL, &error) != UCOL_ON ||
1787 U_FAILURE(error)) {
1788 log_err_status(error, "Setting and retrieving of the case level attribute failed\n");
1789 }
1790
1791 ucol_setAttribute(coll, UCOL_CASE_LEVEL, UCOL_OFF, &error);
1792 if (ucol_getAttribute(coll, UCOL_CASE_LEVEL, &error) != UCOL_OFF ||
1793 U_FAILURE(error)) {
1794 log_err_status(error, "Setting and retrieving of the case level attribute failed\n");
1795 }
1796
1797 ucol_setAttribute(coll, UCOL_NORMALIZATION_MODE, UCOL_ON, &error);
1798 if (ucol_getAttribute(coll, UCOL_NORMALIZATION_MODE, &error) != UCOL_ON ||
1799 U_FAILURE(error)) {
1800 log_err_status(error, "Setting and retrieving of the normalization on/off attribute failed\n");
1801 }
1802
1803 ucol_setAttribute(coll, UCOL_NORMALIZATION_MODE, UCOL_OFF, &error);
1804 if (ucol_getAttribute(coll, UCOL_NORMALIZATION_MODE, &error) != UCOL_OFF ||
1805 U_FAILURE(error)) {
1806 log_err_status(error, "Setting and retrieving of the normalization on/off attribute failed\n");
1807 }
1808
1809 ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_PRIMARY, &error);
1810 if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_PRIMARY ||
1811 U_FAILURE(error)) {
1812 log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1813 }
1814
1815 ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_SECONDARY, &error);
1816 if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_SECONDARY ||
1817 U_FAILURE(error)) {
1818 log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1819 }
1820
1821 ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_TERTIARY, &error);
1822 if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_TERTIARY ||
1823 U_FAILURE(error)) {
1824 log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1825 }
1826
1827 ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_QUATERNARY, &error);
1828 if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_QUATERNARY ||
1829 U_FAILURE(error)) {
1830 log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1831 }
1832
1833 ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_IDENTICAL, &error);
1834 if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_IDENTICAL ||
1835 U_FAILURE(error)) {
1836 log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1837 }
1838
1839 ucol_close(coll);
1840 }
1841
TestGetTailoredSet()1842 void TestGetTailoredSet() {
1843 struct {
1844 const char *rules;
1845 const char *tests[20];
1846 int32_t testsize;
1847 } setTest[] = {
1848 { "&a < \\u212b", { "\\u212b", "A\\u030a", "\\u00c5" }, 3},
1849 { "& S < \\u0161 <<< \\u0160", { "\\u0161", "s\\u030C", "\\u0160", "S\\u030C" }, 4}
1850 };
1851
1852 int32_t i = 0, j = 0;
1853 UErrorCode status = U_ZERO_ERROR;
1854 UParseError pError;
1855
1856 UCollator *coll = NULL;
1857 UChar buff[1024];
1858 int32_t buffLen = 0;
1859 USet *set = NULL;
1860
1861 for(i = 0; i < sizeof(setTest)/sizeof(setTest[0]); i++) {
1862 buffLen = u_unescape(setTest[i].rules, buff, 1024);
1863 coll = ucol_openRules(buff, buffLen, UCOL_DEFAULT, UCOL_DEFAULT, &pError, &status);
1864 if(U_SUCCESS(status)) {
1865 set = ucol_getTailoredSet(coll, &status);
1866 if(uset_size(set) != setTest[i].testsize) {
1867 log_err("Tailored set size different (%d) than expected (%d)\n", uset_size(set), setTest[i].testsize);
1868 }
1869 for(j = 0; j < setTest[i].testsize; j++) {
1870 buffLen = u_unescape(setTest[i].tests[j], buff, 1024);
1871 if(!uset_containsString(set, buff, buffLen)) {
1872 log_err("Tailored set doesn't contain %s... It should\n", setTest[i].tests[j]);
1873 }
1874 }
1875 uset_close(set);
1876 } else {
1877 log_err_status(status, "Couldn't open collator with rules %s\n", setTest[i].rules);
1878 }
1879 ucol_close(coll);
1880 }
1881 }
1882
tMemCmp(const uint8_t * first,const uint8_t * second)1883 static int tMemCmp(const uint8_t *first, const uint8_t *second) {
1884 int32_t firstLen = (int32_t)strlen((const char *)first);
1885 int32_t secondLen = (int32_t)strlen((const char *)second);
1886 return memcmp(first, second, uprv_min(firstLen, secondLen));
1887 }
1888 static const char * strengthsC[] = {
1889 "UCOL_PRIMARY",
1890 "UCOL_SECONDARY",
1891 "UCOL_TERTIARY",
1892 "UCOL_QUATERNARY",
1893 "UCOL_IDENTICAL"
1894 };
1895
TestMergeSortKeys(void)1896 void TestMergeSortKeys(void) {
1897 UErrorCode status = U_ZERO_ERROR;
1898 UCollator *coll = ucol_open("en", &status);
1899 if(U_SUCCESS(status)) {
1900
1901 const char* cases[] = {
1902 "abc",
1903 "abcd",
1904 "abcde"
1905 };
1906 uint32_t casesSize = sizeof(cases)/sizeof(cases[0]);
1907 const char* prefix = "foo";
1908 const char* suffix = "egg";
1909 char outBuff1[256], outBuff2[256];
1910
1911 uint8_t **sortkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
1912 uint8_t **mergedPrefixkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
1913 uint8_t **mergedSuffixkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
1914 uint32_t *sortKeysLen = (uint32_t *)malloc(casesSize*sizeof(uint32_t));
1915 uint8_t prefixKey[256], suffixKey[256];
1916 uint32_t prefixKeyLen = 0, suffixKeyLen = 0, i = 0;
1917 UChar buffer[256];
1918 uint32_t unescapedLen = 0, l1 = 0, l2 = 0;
1919 UColAttributeValue strength;
1920
1921 log_verbose("ucol_mergeSortkeys test\n");
1922 log_verbose("Testing order of the test cases\n");
1923 genericLocaleStarter("en", cases, casesSize);
1924
1925 for(i = 0; i<casesSize; i++) {
1926 sortkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
1927 mergedPrefixkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
1928 mergedSuffixkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
1929 }
1930
1931 unescapedLen = u_unescape(prefix, buffer, 256);
1932 prefixKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, prefixKey, 256);
1933
1934 unescapedLen = u_unescape(suffix, buffer, 256);
1935 suffixKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, suffixKey, 256);
1936
1937 log_verbose("Massaging data with prefixes and different strengths\n");
1938 strength = UCOL_PRIMARY;
1939 while(strength <= UCOL_IDENTICAL) {
1940 log_verbose("Strength %s\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
1941 ucol_setAttribute(coll, UCOL_STRENGTH, strength, &status);
1942 for(i = 0; i<casesSize; i++) {
1943 unescapedLen = u_unescape(cases[i], buffer, 256);
1944 sortKeysLen[i] = ucol_getSortKey(coll, buffer, unescapedLen, sortkeys[i], 256);
1945 ucol_mergeSortkeys(prefixKey, prefixKeyLen, sortkeys[i], sortKeysLen[i], mergedPrefixkeys[i], 256);
1946 ucol_mergeSortkeys(sortkeys[i], sortKeysLen[i], suffixKey, suffixKeyLen, mergedSuffixkeys[i], 256);
1947 if(i>0) {
1948 if(tMemCmp(mergedPrefixkeys[i-1], mergedPrefixkeys[i]) >= 0) {
1949 log_err("Error while comparing prefixed keys @ strength %s:\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
1950 log_err("%s\n%s\n",
1951 ucol_sortKeyToString(coll, mergedPrefixkeys[i-1], outBuff1, &l1),
1952 ucol_sortKeyToString(coll, mergedPrefixkeys[i], outBuff2, &l2));
1953 }
1954 if(tMemCmp(mergedSuffixkeys[i-1], mergedSuffixkeys[i]) >= 0) {
1955 log_err("Error while comparing suffixed keys @ strength %s:\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
1956 log_err("%s\n%s\n",
1957 ucol_sortKeyToString(coll, mergedSuffixkeys[i-1], outBuff1, &l1),
1958 ucol_sortKeyToString(coll, mergedSuffixkeys[i], outBuff2, &l2));
1959 }
1960 }
1961 }
1962 if(strength == UCOL_QUATERNARY) {
1963 strength = UCOL_IDENTICAL;
1964 } else {
1965 strength++;
1966 }
1967 }
1968
1969 {
1970 uint8_t smallBuf[3];
1971 uint32_t reqLen = 0;
1972 log_verbose("testing buffer overflow\n");
1973 reqLen = ucol_mergeSortkeys(prefixKey, prefixKeyLen, suffixKey, suffixKeyLen, smallBuf, 3);
1974 if(reqLen != (prefixKeyLen+suffixKeyLen-1)) {
1975 log_err("Wrong preflight size for merged sortkey\n");
1976 }
1977 }
1978
1979 {
1980 UChar empty = 0;
1981 uint8_t emptyKey[20], abcKey[50], mergedKey[100];
1982 int32_t emptyKeyLen = 0, abcKeyLen = 0, mergedKeyLen = 0;
1983
1984 log_verbose("testing merging with sortkeys generated for empty strings\n");
1985 emptyKeyLen = ucol_getSortKey(coll, &empty, 0, emptyKey, 20);
1986 unescapedLen = u_unescape(cases[0], buffer, 256);
1987 abcKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, abcKey, 50);
1988 mergedKeyLen = ucol_mergeSortkeys(emptyKey, emptyKeyLen, abcKey, abcKeyLen, mergedKey, 100);
1989 if(mergedKey[0] != 2) {
1990 log_err("Empty sortkey didn't produce a level separator\n");
1991 }
1992 /* try with zeros */
1993 mergedKeyLen = ucol_mergeSortkeys(emptyKey, 0, abcKey, abcKeyLen, mergedKey, 100);
1994 if(mergedKeyLen != 0 || mergedKey[0] != 0) {
1995 log_err("Empty key didn't produce null mergedKey\n");
1996 }
1997 mergedKeyLen = ucol_mergeSortkeys(abcKey, abcKeyLen, emptyKey, 0, mergedKey, 100);
1998 if(mergedKeyLen != 0 || mergedKey[0] != 0) {
1999 log_err("Empty key didn't produce null mergedKey\n");
2000 }
2001
2002 }
2003
2004 for(i = 0; i<casesSize; i++) {
2005 free(sortkeys[i]);
2006 free(mergedPrefixkeys[i]);
2007 free(mergedSuffixkeys[i]);
2008 }
2009 free(sortkeys);
2010 free(mergedPrefixkeys);
2011 free(mergedSuffixkeys);
2012 free(sortKeysLen);
2013 ucol_close(coll);
2014 /* need to finish this up */
2015 } else {
2016 log_data_err("Couldn't open collator");
2017 }
2018 }
TestShortString(void)2019 static void TestShortString(void)
2020 {
2021 struct {
2022 const char *input;
2023 const char *expectedOutput;
2024 const char *locale;
2025 UErrorCode expectedStatus;
2026 int32_t expectedOffset;
2027 uint32_t expectedIdentifier;
2028 } testCases[] = {
2029 /*
2030 * The following expectedOutput contains a collation weight (2700 from UCA 6.0)
2031 * which is the primary weight for the T character (U+0041) in the input.
2032 * When that character gets a different weight in FractionalUCA.txt,
2033 * the expectedOutput needs to be adjusted.
2034 * That is, when we upgrade to a new UCA version or change collation
2035 * in such a way that the absolute weight for 'A' changes,
2036 * we will get a test failure here and need to adjust the test case.
2037 */
2038 {"LDE_RDE_KPHONEBOOK_T0041_ZLATN","B2700_KPHONEBOOK_LDE", "de@collation=phonebook", U_USING_FALLBACK_WARNING, 0, 0 },
2039
2040 {"LEN_RUS_NO_AS_S4","AS_LROOT_NO_S4", NULL, U_USING_DEFAULT_WARNING, 0, 0 },
2041 {"LDE_VPHONEBOOK_EO_SI","EO_KPHONEBOOK_LDE_SI", "de@collation=phonebook", U_ZERO_ERROR, 0, 0 },
2042 {"LDE_Kphonebook","KPHONEBOOK_LDE", "de@collation=phonebook", U_ZERO_ERROR, 0, 0 },
2043 {"Xqde_DE@collation=phonebookq_S3_EX","KPHONEBOOK_LDE", "de@collation=phonebook", U_USING_FALLBACK_WARNING, 0, 0 },
2044 {"LFR_FO", "FO_LROOT", NULL, U_USING_DEFAULT_WARNING, 0, 0 },
2045 {"SO_LX_AS", "", NULL, U_ILLEGAL_ARGUMENT_ERROR, 8, 0 },
2046 {"S3_ASS_MMM", "", NULL, U_ILLEGAL_ARGUMENT_ERROR, 5, 0 }
2047 };
2048
2049 int32_t i = 0;
2050 UCollator *coll = NULL, *fromNormalized = NULL;
2051 UParseError parseError;
2052 UErrorCode status = U_ZERO_ERROR;
2053 char fromShortBuffer[256], normalizedBuffer[256], fromNormalizedBuffer[256];
2054 const char* locale = NULL;
2055
2056
2057 for(i = 0; i < sizeof(testCases)/sizeof(testCases[0]); i++) {
2058 status = U_ZERO_ERROR;
2059 if(testCases[i].locale) {
2060 locale = testCases[i].locale;
2061 } else {
2062 locale = NULL;
2063 }
2064
2065 coll = ucol_openFromShortString(testCases[i].input, FALSE, &parseError, &status);
2066 if(status != testCases[i].expectedStatus) {
2067 log_err_status(status, "Got status '%s' that is different from expected '%s' for '%s'\n",
2068 u_errorName(status), u_errorName(testCases[i].expectedStatus), testCases[i].input);
2069 continue;
2070 }
2071
2072 if(U_SUCCESS(status)) {
2073 ucol_getShortDefinitionString(coll, locale, fromShortBuffer, 256, &status);
2074
2075 if(strcmp(fromShortBuffer, testCases[i].expectedOutput)) {
2076 log_err("Got short string '%s' from the collator. Expected '%s' for input '%s'\n",
2077 fromShortBuffer, testCases[i].expectedOutput, testCases[i].input);
2078 }
2079
2080 ucol_normalizeShortDefinitionString(testCases[i].input, normalizedBuffer, 256, &parseError, &status);
2081 fromNormalized = ucol_openFromShortString(normalizedBuffer, FALSE, &parseError, &status);
2082 ucol_getShortDefinitionString(fromNormalized, locale, fromNormalizedBuffer, 256, &status);
2083
2084 if(strcmp(fromShortBuffer, fromNormalizedBuffer)) {
2085 log_err("Strings obtained from collators instantiated by short string ('%s') and from normalized string ('%s') differ\n",
2086 fromShortBuffer, fromNormalizedBuffer);
2087 }
2088
2089
2090 if(!ucol_equals(coll, fromNormalized)) {
2091 log_err("Collator from short string ('%s') differs from one obtained through a normalized version ('%s')\n",
2092 testCases[i].input, normalizedBuffer);
2093 }
2094
2095 ucol_close(fromNormalized);
2096 ucol_close(coll);
2097
2098 } else {
2099 if(parseError.offset != testCases[i].expectedOffset) {
2100 log_err("Got parse error offset %i, but expected %i instead for '%s'\n",
2101 parseError.offset, testCases[i].expectedOffset, testCases[i].input);
2102 }
2103 }
2104 }
2105
2106 }
2107
2108 static void
doSetsTest(const char * locale,const USet * ref,USet * set,const char * inSet,const char * outSet,UErrorCode * status)2109 doSetsTest(const char *locale, const USet *ref, USet *set, const char* inSet, const char* outSet, UErrorCode *status) {
2110 UChar buffer[512];
2111 int32_t bufLen;
2112
2113 uset_clear(set);
2114 bufLen = u_unescape(inSet, buffer, 512);
2115 uset_applyPattern(set, buffer, bufLen, 0, status);
2116 if(U_FAILURE(*status)) {
2117 log_err("%s: Failure setting pattern %s\n", locale, u_errorName(*status));
2118 }
2119
2120 if(!uset_containsAll(ref, set)) {
2121 log_err("%s: Some stuff from %s is not present in the set\n", locale, inSet);
2122 }
2123
2124 uset_clear(set);
2125 bufLen = u_unescape(outSet, buffer, 512);
2126 uset_applyPattern(set, buffer, bufLen, 0, status);
2127 if(U_FAILURE(*status)) {
2128 log_err("%s: Failure setting pattern %s\n", locale, u_errorName(*status));
2129 }
2130
2131 if(!uset_containsNone(ref, set)) {
2132 log_err("%s: Some stuff from %s is present in the set\n", locale, outSet);
2133 }
2134 }
2135
2136
2137
2138
2139 static void
TestGetContractionsAndUnsafes(void)2140 TestGetContractionsAndUnsafes(void)
2141 {
2142 static struct {
2143 const char* locale;
2144 const char* inConts;
2145 const char* outConts;
2146 const char* inExp;
2147 const char* outExp;
2148 const char* unsafeCodeUnits;
2149 const char* safeCodeUnits;
2150 } tests[] = {
2151 { "ru",
2152 "[{\\u0418\\u0306}{\\u0438\\u0306}]",
2153 "[\\u0439\\u0457]",
2154 "[\\u00e6]",
2155 "[ae]",
2156 "[\\u0418\\u0438]",
2157 "[aAbB\\u0430\\u0410\\u0433\\u0413]"
2158 },
2159 { "uk",
2160 "[{\\u0406\\u0308}{\\u0456\\u0308}{\\u0418\\u0306}{\\u0438\\u0306}]",
2161 "[\\u0407\\u0419\\u0439\\u0457]",
2162 "[\\u00e6]",
2163 "[ae]",
2164 "[\\u0406\\u0456\\u0418\\u0438]",
2165 "[aAbBxv]",
2166 },
2167 { "sh",
2168 "[{C\\u0301}{C\\u030C}{C\\u0341}{DZ\\u030C}{Dz\\u030C}{D\\u017D}{D\\u017E}{lj}{nj}]",
2169 "[{\\u309d\\u3099}{\\u30fd\\u3099}]",
2170 "[\\u00e6]",
2171 "[a]",
2172 "[nlcdzNLCDZ]",
2173 "[jabv]"
2174 },
2175 { "ja",
2176 "[{\\u3053\\u3099\\u309D}{\\u3053\\u3099\\u309D\\u3099}{\\u3053\\u3099\\u309E}{\\u3053\\u3099\\u30FC}{\\u3053\\u309D}{\\u3053\\u309D\\u3099}{\\u3053\\u309E}{\\u3053\\u30FC}{\\u30B3\\u3099\\u30FC}{\\u30B3\\u3099\\u30FD}{\\u30B3\\u3099\\u30FD\\u3099}{\\u30B3\\u3099\\u30FE}{\\u30B3\\u30FC}{\\u30B3\\u30FD}{\\u30B3\\u30FD\\u3099}{\\u30B3\\u30FE}]",
2177 "[{\\u30FD\\u3099}{\\u309D\\u3099}{\\u3053\\u3099}{\\u30B3\\u3099}{lj}{nj}]",
2178 "[\\u30FE\\u00e6]",
2179 "[a]",
2180 "[\\u3099]",
2181 "[]"
2182 }
2183 };
2184
2185
2186
2187
2188 UErrorCode status = U_ZERO_ERROR;
2189 UCollator *coll = NULL;
2190 int32_t i = 0;
2191 int32_t noConts = 0;
2192 USet *conts = uset_open(0,0);
2193 USet *exp = uset_open(0, 0);
2194 USet *set = uset_open(0,0);
2195 int32_t setBufferLen = 65536;
2196 UChar buffer[65536];
2197 int32_t setLen = 0;
2198
2199 for(i = 0; i < sizeof(tests)/sizeof(tests[0]); i++) {
2200 log_verbose("Testing locale: %s\n", tests[i].locale);
2201 coll = ucol_open(tests[i].locale, &status);
2202 if (coll == NULL || U_FAILURE(status)) {
2203 log_err_status(status, "Unable to open collator for locale %s ==> %s\n", tests[i].locale, u_errorName(status));
2204 continue;
2205 }
2206 ucol_getContractionsAndExpansions(coll, conts, exp, TRUE, &status);
2207 doSetsTest(tests[i].locale, conts, set, tests[i].inConts, tests[i].outConts, &status);
2208 setLen = uset_toPattern(conts, buffer, setBufferLen, TRUE, &status);
2209 if(U_SUCCESS(status)) {
2210 /*log_verbose("Contractions %i: %s\n", uset_getItemCount(conts), aescstrdup(buffer, setLen));*/
2211 } else {
2212 log_err("error %s. %i\n", u_errorName(status), setLen);
2213 status = U_ZERO_ERROR;
2214 }
2215 doSetsTest(tests[i].locale, exp, set, tests[i].inExp, tests[i].outExp, &status);
2216 setLen = uset_toPattern(exp, buffer, setBufferLen, TRUE, &status);
2217 if(U_SUCCESS(status)) {
2218 /*log_verbose("Expansions %i: %s\n", uset_getItemCount(exp), aescstrdup(buffer, setLen));*/
2219 } else {
2220 log_err("error %s. %i\n", u_errorName(status), setLen);
2221 status = U_ZERO_ERROR;
2222 }
2223
2224 noConts = ucol_getUnsafeSet(coll, conts, &status);
2225 doSetsTest(tests[i].locale, conts, set, tests[i].unsafeCodeUnits, tests[i].safeCodeUnits, &status);
2226 setLen = uset_toPattern(conts, buffer, setBufferLen, TRUE, &status);
2227 if(U_SUCCESS(status)) {
2228 log_verbose("Unsafe %i: %s\n", uset_getItemCount(exp), aescstrdup(buffer, setLen));
2229 } else {
2230 log_err("error %s. %i\n", u_errorName(status), setLen);
2231 status = U_ZERO_ERROR;
2232 }
2233
2234 ucol_close(coll);
2235 }
2236
2237
2238 uset_close(conts);
2239 uset_close(exp);
2240 uset_close(set);
2241 }
2242
2243 static void
TestOpenBinary(void)2244 TestOpenBinary(void)
2245 {
2246 UErrorCode status = U_ZERO_ERROR;
2247 /*
2248 char rule[] = "&h < d < c < b";
2249 char *wUCA[] = { "a", "h", "d", "c", "b", "i" };
2250 char *noUCA[] = {"d", "c", "b", "a", "h", "i" };
2251 */
2252 /* we have to use Cyrillic letters because latin-1 always gets copied */
2253 const char rule[] = "&\\u0452 < \\u0434 < \\u0433 < \\u0432"; /* &dje < d < g < v */
2254 const char *wUCA[] = { "\\u0430", "\\u0452", "\\u0434", "\\u0433", "\\u0432", "\\u0435" }; /* a, dje, d, g, v, e */
2255 const char *noUCA[] = {"\\u0434", "\\u0433", "\\u0432", "\\u0430", "\\u0435", "\\u0452" }; /* d, g, v, a, e, dje */
2256
2257 UChar uRules[256];
2258 int32_t uRulesLen = u_unescape(rule, uRules, 256);
2259
2260 UCollator *coll = ucol_openRules(uRules, uRulesLen, UCOL_DEFAULT, UCOL_DEFAULT, NULL, &status);
2261 UCollator *UCA = NULL;
2262 UCollator *cloneNOUCA = NULL, *cloneWUCA = NULL;
2263
2264 uint8_t imageBuffer[32768];
2265 uint8_t *image = imageBuffer;
2266 int32_t imageBufferCapacity = 32768;
2267
2268 int32_t imageSize;
2269
2270 if((coll==NULL)||(U_FAILURE(status))) {
2271 log_data_err("could not load collators or error occured: %s\n",
2272 u_errorName(status));
2273 return;
2274 }
2275 UCA = ucol_open("root", &status);
2276 if((UCA==NULL)||(U_FAILURE(status))) {
2277 log_data_err("could not load UCA collator or error occured: %s\n",
2278 u_errorName(status));
2279 return;
2280 }
2281 imageSize = ucol_cloneBinary(coll, image, imageBufferCapacity, &status);
2282 if(U_FAILURE(status)) {
2283 image = (uint8_t *)malloc(imageSize*sizeof(uint8_t));
2284 status = U_ZERO_ERROR;
2285 imageSize = ucol_cloneBinary(coll, imageBuffer, imageSize, &status);
2286 }
2287
2288
2289 cloneWUCA = ucol_openBinary(image, imageSize, UCA, &status);
2290 cloneNOUCA = ucol_openBinary(image, imageSize, NULL, &status);
2291
2292 genericOrderingTest(coll, wUCA, sizeof(wUCA)/sizeof(wUCA[0]));
2293
2294 genericOrderingTest(cloneWUCA, wUCA, sizeof(wUCA)/sizeof(wUCA[0]));
2295 genericOrderingTest(cloneNOUCA, noUCA, sizeof(noUCA)/sizeof(noUCA[0]));
2296
2297 if(image != imageBuffer) {
2298 free(image);
2299 }
2300 ucol_close(coll);
2301 ucol_close(cloneNOUCA);
2302 ucol_close(cloneWUCA);
2303 ucol_close(UCA);
2304 }
2305
TestDefault(void)2306 static void TestDefault(void) {
2307 /* Tests for code coverage. */
2308 UErrorCode status = U_ZERO_ERROR;
2309 UCollator *coll = ucol_open("es@collation=pinyin", &status);
2310 if (coll == NULL || status == U_FILE_ACCESS_ERROR) {
2311 log_data_err("Unable to open collator es@collation=pinyin\n");
2312 return;
2313 }
2314 if (status != U_USING_DEFAULT_WARNING) {
2315 /* What do you mean that you know about using pinyin collation in Spanish!? This should be in the zh locale. */
2316 log_err("es@collation=pinyin should return U_USING_DEFAULT_WARNING, but returned %s\n", u_errorName(status));
2317 }
2318 ucol_close(coll);
2319 if (ucol_getKeywordValues("funky", &status) != NULL) {
2320 log_err("Collators should not know about the funky keyword.\n");
2321 }
2322 if (status != U_ILLEGAL_ARGUMENT_ERROR) {
2323 log_err("funky keyword didn't fail as expected %s\n", u_errorName(status));
2324 }
2325 if (ucol_getKeywordValues("collation", &status) != NULL) {
2326 log_err("ucol_getKeywordValues should not work when given a bad status.\n");
2327 }
2328 }
2329
TestDefaultKeyword(void)2330 static void TestDefaultKeyword(void) {
2331 /* Tests for code coverage. */
2332 UErrorCode status = U_ZERO_ERROR;
2333 const char *loc = "zh_TW@collation=default";
2334 UCollator *coll = ucol_open(loc, &status);
2335 if(U_FAILURE(status)) {
2336 log_info("Warning: ucol_open(%s, ...) returned %s, at least it didn't crash.\n", loc, u_errorName(status));
2337 } else if (status != U_USING_FALLBACK_WARNING) {
2338 /* Hmm, skip the following test for CLDR 1.9 data and/or ICU 4.6, no longer seems to apply */
2339 #if 0
2340 log_err("ucol_open(%s, ...) should return an error or some sort of U_USING_FALLBACK_WARNING, but returned %s\n", loc, u_errorName(status));
2341 #endif
2342 }
2343 ucol_close(coll);
2344 }
2345
TestGetKeywordValuesForLocale(void)2346 static void TestGetKeywordValuesForLocale(void) {
2347 #define INCLUDE_UNIHAN_COLLATION 0
2348 #define PREFERRED_SIZE 16
2349 #define MAX_NUMBER_OF_KEYWORDS 8
2350 const char *PREFERRED[PREFERRED_SIZE][MAX_NUMBER_OF_KEYWORDS+1] = {
2351 { "und", "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2352 { "en_US", "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2353 { "en_029", "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2354 { "de_DE", "standard", "phonebook", "search", "ducet", NULL, NULL, NULL, NULL },
2355 { "de_Latn_DE", "standard", "phonebook", "search", "ducet", NULL, NULL, NULL, NULL },
2356 #if INCLUDE_UNIHAN_COLLATION
2357 { "zh", "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2358 { "zh_Hans", "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2359 { "zh_CN", "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2360 { "zh_Hant", "stroke", "big5han", "gb2312han", "pinyin", "standard", "unihan", "ducet", "search" },
2361 { "zh_TW", "stroke", "big5han", "gb2312han", "pinyin", "standard", "unihan", "ducet", "search" },
2362 { "zh__PINYIN", "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2363 #else
2364 { "zh", "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2365 { "zh_Hans", "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2366 { "zh_CN", "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2367 { "zh_Hant", "stroke", "big5han", "gb2312han", "pinyin", "standard", "ducet", "search", NULL },
2368 { "zh_TW", "stroke", "big5han", "gb2312han", "pinyin", "standard", "ducet", "search", NULL },
2369 { "zh__PINYIN", "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2370 #endif
2371 { "es_ES", "standard", "search", "traditional", "ducet", NULL, NULL, NULL, NULL },
2372 { "es__TRADITIONAL","traditional", "search", "standard", "ducet", NULL, NULL, NULL, NULL },
2373 { "und@collation=phonebook", "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2374 { "de_DE@collation=big5han", "standard", "phonebook", "search", "ducet", NULL, NULL, NULL, NULL },
2375 { "zzz@collation=xxx", "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL }
2376 };
2377 #if INCLUDE_UNIHAN_COLLATION
2378 const int32_t expectedLength[PREFERRED_SIZE] = { 3, 3, 3, 4, 4, 8, 8, 8, 8, 8, 8, 4, 4, 3, 4, 3 };
2379 #else
2380 const int32_t expectedLength[PREFERRED_SIZE] = { 3, 3, 3, 4, 4, 7, 7, 7, 7, 7, 7, 4, 4, 3, 4, 3 };
2381 #endif
2382
2383 UErrorCode status = U_ZERO_ERROR;
2384 UEnumeration *keywordValues = NULL;
2385 int32_t i, n, size, valueLength;
2386 const char *locale = NULL, *value = NULL;
2387 UBool errorOccurred = FALSE;
2388
2389 for (i = 0; i < PREFERRED_SIZE; i++) {
2390 locale = PREFERRED[i][0];
2391 value = NULL;
2392 valueLength = 0;
2393 size = 0;
2394
2395 keywordValues = ucol_getKeywordValuesForLocale("collation", locale, TRUE, &status);
2396 if (keywordValues == NULL || U_FAILURE(status)) {
2397 log_err_status(status, "Error getting keyword values: %s\n", u_errorName(status));
2398 break;
2399 }
2400 size = uenum_count(keywordValues, &status);
2401
2402 if (size == expectedLength[i]) {
2403 for (n = 0; n < expectedLength[i]; n++) {
2404 if ((value = uenum_next(keywordValues, &valueLength, &status)) != NULL && U_SUCCESS(status)) {
2405 if (uprv_strcmp(value, PREFERRED[i][n+1]) != 0) {
2406 log_err("Keyword values differ: Got [%s] Expected [%s] for locale: %s\n", value, PREFERRED[i][n+1], locale);
2407 errorOccurred = TRUE;
2408 break;
2409 }
2410
2411 } else {
2412 log_err("While getting keyword value from locale: %s got this error: %s\n", locale, u_errorName(status));
2413 errorOccurred = TRUE;
2414 break;
2415 }
2416 }
2417 if (errorOccurred) {
2418 break;
2419 }
2420 } else {
2421 log_err("Number of keywords (%d) does not match expected size (%d) for locale: %s\n", size, expectedLength[i], locale);
2422 break;
2423 }
2424 uenum_close(keywordValues);
2425 keywordValues = NULL;
2426 }
2427 if (keywordValues != NULL) {
2428 uenum_close(keywordValues);
2429 }
2430 }
2431
2432
2433 #endif /* #if !UCONFIG_NO_COLLATION */
2434