• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /********************************************************************
2  * Copyright (c) 1997-2010 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         int TestBufferSize();    /* defined in "colutil.c" */
39 
40 
41 
42 
43 /* next two function is modified from "i18n/ucol.cpp" to avoid include "ucol_imp.h" */
uprv_appendByteToHexString(char * dst,uint8_t val)44 static void uprv_appendByteToHexString(char *dst, uint8_t val) {
45   uint32_t len = (uint32_t)strlen(dst);
46   sprintf(dst+len, "%02X", val);
47 }
48 
ucol_sortKeyToString(const UCollator * coll,const uint8_t * sortkey,char * buffer,uint32_t * len)49 static char* U_EXPORT2 ucol_sortKeyToString(const UCollator *coll, const uint8_t *sortkey, char *buffer, uint32_t *len) {
50   int32_t strength = UCOL_PRIMARY;
51   uint32_t res_size = 0;
52   UBool doneCase = FALSE;
53 
54   char *current = buffer;
55   const uint8_t *currentSk = sortkey;
56 
57   UErrorCode error_code = U_ZERO_ERROR;
58 
59   strcpy(current, "[");
60 
61   while(strength <= UCOL_QUATERNARY && strength <= ucol_getAttribute(coll,UCOL_STRENGTH, &error_code)) {
62     if(U_FAILURE(error_code)) {
63       log_err("ucol_getAttribute returned error: %s\n", u_errorName(error_code));
64     }
65     if(strength > UCOL_PRIMARY) {
66       strcat(current, " . ");
67     }
68     while(*currentSk != 0x01 && *currentSk != 0x00) { /* print a level */
69       uprv_appendByteToHexString(current, *currentSk++);
70       strcat(current, " ");
71     }
72     if(ucol_getAttribute(coll,UCOL_CASE_LEVEL, &error_code) == UCOL_ON && strength == UCOL_SECONDARY && doneCase == FALSE) {
73         doneCase = TRUE;
74     } else if(ucol_getAttribute(coll,UCOL_CASE_LEVEL, &error_code) == UCOL_OFF || doneCase == TRUE || strength != UCOL_SECONDARY) {
75       strength ++;
76     }
77     if(U_FAILURE(error_code)) {
78       log_err("ucol_getAttribute returned error: %s\n", u_errorName(error_code));
79     }
80     uprv_appendByteToHexString(current, *currentSk++); /* This should print '01' */
81     if(strength == UCOL_QUATERNARY && ucol_getAttribute(coll,UCOL_ALTERNATE_HANDLING, &error_code) == UCOL_NON_IGNORABLE) {
82       break;
83     }
84   }
85 
86   if(ucol_getAttribute(coll,UCOL_STRENGTH, &error_code) == UCOL_IDENTICAL) {
87     strcat(current, " . ");
88     while(*currentSk != 0) {
89       uprv_appendByteToHexString(current, *currentSk++);
90       strcat(current, " ");
91     }
92 
93     uprv_appendByteToHexString(current, *currentSk++);
94   }
95   if(U_FAILURE(error_code)) {
96     log_err("ucol_getAttribute returned error: %s\n", u_errorName(error_code));
97   }
98   strcat(current, "]");
99 
100   if(res_size > *len) {
101     return NULL;
102   }
103 
104   return buffer;
105 }
106 /* end of  avoid include "ucol_imp.h" */
107 
108 
addCollAPITest(TestNode ** root)109 void addCollAPITest(TestNode** root)
110 {
111     /* WEIVTODO: return tests here */
112     addTest(root, &TestProperty,      "tscoll/capitst/TestProperty");
113     addTest(root, &TestRuleBasedColl, "tscoll/capitst/TestRuleBasedColl");
114     addTest(root, &TestCompare,       "tscoll/capitst/TestCompare");
115     addTest(root, &TestSortKey,       "tscoll/capitst/TestSortKey");
116     addTest(root, &TestHashCode,      "tscoll/capitst/TestHashCode");
117     addTest(root, &TestElemIter,      "tscoll/capitst/TestElemIter");
118     addTest(root, &TestGetAll,        "tscoll/capitst/TestGetAll");
119     /*addTest(root, &TestGetDefaultRules, "tscoll/capitst/TestGetDefaultRules");*/
120     addTest(root, &TestDecomposition, "tscoll/capitst/TestDecomposition");
121     addTest(root, &TestSafeClone, "tscoll/capitst/TestSafeClone");
122     addTest(root, &TestCloneBinary, "tscoll/capitst/TestCloneBinary");
123     addTest(root, &TestGetSetAttr, "tscoll/capitst/TestGetSetAttr");
124     addTest(root, &TestBounds, "tscoll/capitst/TestBounds");
125     addTest(root, &TestGetLocale, "tscoll/capitst/TestGetLocale");
126     addTest(root, &TestSortKeyBufferOverrun, "tscoll/capitst/TestSortKeyBufferOverrun");
127     addTest(root, &TestAttribute, "tscoll/capitst/TestAttribute");
128     addTest(root, &TestGetTailoredSet, "tscoll/capitst/TestGetTailoredSet");
129     addTest(root, &TestMergeSortKeys, "tscoll/capitst/TestMergeSortKeys");
130     addTest(root, &TestShortString, "tscoll/capitst/TestShortString");
131     /* BEGIN android-removed
132        To save space, Android does not include the collation tailoring rules.
133        We skip the tailing tests for collations. */
134     /* addTest(root, &TestGetContractionsAndUnsafes, "tscoll/capitst/TestGetContractionsAndUnsafes"); */
135     /* END android-removed */
136     addTest(root, &TestOpenBinary, "tscoll/capitst/TestOpenBinary");
137     addTest(root, &TestDefault, "tscoll/capitst/TestDefault");
138     addTest(root, &TestDefaultKeyword, "tscoll/capitst/TestDefaultKeyword");
139     /* BEGIN android-removed
140        To save space, Android does not build full collation tables and tailing rules.
141        We skip the related tests. */
142     /* addTest(root, &TestOpenVsOpenRules, "tscoll/capitst/TestOpenVsOpenRules"); */
143     /* addTest(root, &TestGetKeywordValuesForLocale, "tscoll/capitst/TestGetKeywordValuesForLocale"); */
144     /* END android-removed */
145 }
146 
TestGetSetAttr(void)147 void TestGetSetAttr(void) {
148   UErrorCode status = U_ZERO_ERROR;
149   UCollator *coll = ucol_open(NULL, &status);
150   struct attrTest {
151     UColAttribute att;
152     UColAttributeValue val[5];
153     uint32_t valueSize;
154     UColAttributeValue nonValue;
155   } attrs[] = {
156     {UCOL_FRENCH_COLLATION, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},
157     {UCOL_ALTERNATE_HANDLING, {UCOL_NON_IGNORABLE, UCOL_SHIFTED}, 2, UCOL_OFF},/* attribute for handling variable elements*/
158     {UCOL_CASE_FIRST, {UCOL_OFF, UCOL_LOWER_FIRST, UCOL_UPPER_FIRST}, 3, UCOL_SHIFTED},/* who goes first, lower case or uppercase */
159     {UCOL_CASE_LEVEL, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* do we have an extra case level */
160     {UCOL_NORMALIZATION_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* attribute for normalization */
161     {UCOL_DECOMPOSITION_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},
162     {UCOL_STRENGTH,         {UCOL_PRIMARY, UCOL_SECONDARY, UCOL_TERTIARY, UCOL_QUATERNARY, UCOL_IDENTICAL}, 5, UCOL_SHIFTED},/* attribute for strength */
163     {UCOL_HIRAGANA_QUATERNARY_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* when turned on, this attribute */
164   };
165   UColAttribute currAttr;
166   UColAttributeValue value;
167   uint32_t i = 0, j = 0;
168 
169   if (coll == NULL) {
170     log_err_status(status, "Unable to open collator. %s\n", u_errorName(status));
171     return;
172   }
173   for(i = 0; i<sizeof(attrs)/sizeof(attrs[0]); i++) {
174     currAttr = attrs[i].att;
175     ucol_setAttribute(coll, currAttr, UCOL_DEFAULT, &status);
176     if(U_FAILURE(status)) {
177       log_err_status(status, "ucol_setAttribute with the default value returned error: %s\n", u_errorName(status));
178       break;
179     }
180     value = ucol_getAttribute(coll, currAttr, &status);
181     if(U_FAILURE(status)) {
182       log_err("ucol_getAttribute returned error: %s\n", u_errorName(status));
183       break;
184     }
185     for(j = 0; j<attrs[i].valueSize; j++) {
186       ucol_setAttribute(coll, currAttr, attrs[i].val[j], &status);
187       if(U_FAILURE(status)) {
188         log_err("ucol_setAttribute with the value %i returned error: %s\n", attrs[i].val[j], u_errorName(status));
189         break;
190       }
191     }
192     status = U_ZERO_ERROR;
193     ucol_setAttribute(coll, currAttr, attrs[i].nonValue, &status);
194     if(U_SUCCESS(status)) {
195       log_err("ucol_setAttribute with the bad value didn't return an error\n");
196       break;
197     }
198     status = U_ZERO_ERROR;
199 
200     ucol_setAttribute(coll, currAttr, value, &status);
201     if(U_FAILURE(status)) {
202       log_err("ucol_setAttribute with the default valuereturned error: %s\n", u_errorName(status));
203       break;
204     }
205   }
206   status = U_ZERO_ERROR;
207   value = ucol_getAttribute(coll, UCOL_ATTRIBUTE_COUNT, &status);
208   if(U_SUCCESS(status)) {
209     log_err("ucol_getAttribute for UCOL_ATTRIBUTE_COUNT didn't return an error\n");
210   }
211   status = U_ZERO_ERROR;
212   ucol_setAttribute(coll, UCOL_ATTRIBUTE_COUNT, UCOL_DEFAULT, &status);
213   if(U_SUCCESS(status)) {
214     log_err("ucol_setAttribute for UCOL_ATTRIBUTE_COUNT didn't return an error\n");
215   }
216   status = U_ZERO_ERROR;
217   ucol_close(coll);
218 }
219 
220 
doAssert(int condition,const char * message)221 static void doAssert(int condition, const char *message)
222 {
223     if (condition==0) {
224         log_err("ERROR :  %s\n", message);
225     }
226 }
227 
228 #if 0
229 /* We don't have default rules, at least not in the previous sense */
230 void TestGetDefaultRules(){
231     uint32_t size=0;
232     UErrorCode status=U_ZERO_ERROR;
233     UCollator *coll=NULL;
234     int32_t len1 = 0, len2=0;
235     uint8_t *binColData = NULL;
236 
237     UResourceBundle *res = NULL;
238     UResourceBundle *binColl = NULL;
239     uint8_t *binResult = NULL;
240 
241 
242     const UChar * defaultRulesArray=ucol_getDefaultRulesArray(&size);
243     log_verbose("Test the function ucol_getDefaultRulesArray()\n");
244 
245     coll = ucol_openRules(defaultRulesArray, size, UCOL_ON, UCOL_PRIMARY, &status);
246     if(U_SUCCESS(status) && coll !=NULL) {
247         binColData = (uint8_t*)ucol_cloneRuleData(coll, &len1, &status);
248 
249     }
250 
251 
252     status=U_ZERO_ERROR;
253     res=ures_open(NULL, "root", &status);
254     if(U_FAILURE(status)){
255         log_err("ERROR: Failed to get resource for \"root Locale\" with %s", myErrorName(status));
256         return;
257     }
258     binColl=ures_getByKey(res, "%%Collation", binColl, &status);
259     if(U_SUCCESS(status)){
260         binResult=(uint8_t*)ures_getBinary(binColl,  &len2, &status);
261         if(U_FAILURE(status)){
262             log_err("ERROR: ures_getBinary() failed\n");
263         }
264     }else{
265         log_err("ERROR: ures_getByKey(locale(default), %%Collation) failed");
266     }
267 
268 
269     if(len1 != len2){
270         log_err("Error: ucol_getDefaultRulesArray() failed to return the correct length.\n");
271     }
272     if(memcmp(binColData, binResult, len1) != 0){
273         log_err("Error: ucol_getDefaultRulesArray() failed\n");
274     }
275 
276     free(binColData);
277     ures_close(binColl);
278     ures_close(res);
279     ucol_close(coll);
280 
281 }
282 #endif
283 
284 /* Collator Properties
285  ucol_open, ucol_strcoll,  getStrength/setStrength
286  getDecomposition/setDecomposition, getDisplayName*/
TestProperty()287 void TestProperty()
288 {
289     UCollator *col, *ruled;
290     UChar *disName;
291     int32_t len = 0;
292     UChar *source, *target;
293     int32_t tempLength;
294     UErrorCode status = U_ZERO_ERROR;
295     /*
296      * Expected version of the English collator.
297      * Currently, the major/minor version numbers change when the builder code
298      * changes,
299      * number 2 is from the tailoring data version and
300      * number 3 is the UCA version.
301      * This changes with every UCA version change, and the expected value
302      * needs to be adjusted.
303      * Same in intltest/apicoll.cpp.
304      */
305     UVersionInfo currVersionArray = {0x31, 0xC0, 0x05, 0x2A};  /* from ICU 4.4/UCA 5.2 */
306     UVersionInfo versionArray = {0, 0, 0, 0};
307     UVersionInfo versionUCAArray = {0, 0, 0, 0};
308     UVersionInfo versionUCDArray = {0, 0, 0, 0};
309 
310     log_verbose("The property tests begin : \n");
311     log_verbose("Test ucol_strcoll : \n");
312     col = ucol_open("en_US", &status);
313     if (U_FAILURE(status)) {
314         log_err_status(status, "Default Collator creation failed.: %s\n", myErrorName(status));
315         return;
316     }
317 
318     ucol_getVersion(col, versionArray);
319     /* Check for a version greater than some value rather than equality
320      * so that we need not update the expected version each time. */
321     if (uprv_memcmp(versionArray, currVersionArray, 4)<0) {
322       log_err("Testing ucol_getVersion() - unexpected result: %02x.%02x.%02x.%02x\n",
323               versionArray[0], versionArray[1], versionArray[2], versionArray[3]);
324     } else {
325       log_verbose("ucol_getVersion() result: %02x.%02x.%02x.%02x\n",
326                   versionArray[0], versionArray[1], versionArray[2], versionArray[3]);
327     }
328 
329     /* Assume that the UCD and UCA versions are the same,
330      * rather than hardcoding (and updating each time) a particular UCA version. */
331     u_getUnicodeVersion(versionUCDArray);
332     ucol_getUCAVersion(col, versionUCAArray);
333     if (0!=uprv_memcmp(versionUCAArray, versionUCDArray, 4)) {
334       log_err("Testing ucol_getUCAVersion() - unexpected result: %hu.%hu.%hu.%hu\n",
335               versionUCAArray[0], versionUCAArray[1], versionUCAArray[2], versionUCAArray[3]);
336     }
337 
338     source=(UChar*)malloc(sizeof(UChar) * 12);
339     target=(UChar*)malloc(sizeof(UChar) * 12);
340 
341 
342     u_uastrcpy(source, "ab");
343     u_uastrcpy(target, "abc");
344 
345     doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_LESS), "ab < abc comparison failed");
346 
347     u_uastrcpy(source, "ab");
348     u_uastrcpy(target, "AB");
349 
350     doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_LESS), "ab < AB comparison failed");
351 /*    u_uastrcpy(source, "black-bird");
352     u_uastrcpy(target, "blackbird"); */
353     u_uastrcpy(target, "black-bird");
354     u_uastrcpy(source, "blackbird");
355 
356     doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_GREATER),
357         "black-bird > blackbird comparison failed");
358     u_uastrcpy(source, "black bird");
359     u_uastrcpy(target, "black-bird");
360     doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_LESS),
361         "black bird < black-bird comparison failed");
362     u_uastrcpy(source, "Hello");
363     u_uastrcpy(target, "hello");
364 
365     doAssert((ucol_strcoll(col, source, u_strlen(source), target, u_strlen(target)) == UCOL_GREATER),
366         "Hello > hello comparison failed");
367     free(source);
368     free(target);
369     log_verbose("Test ucol_strcoll ends.\n");
370 
371     log_verbose("testing ucol_getStrength() method ...\n");
372     doAssert( (ucol_getStrength(col) == UCOL_TERTIARY), "collation object has the wrong strength");
373     doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
374 
375     log_verbose("testing ucol_setStrength() method ...\n");
376     ucol_setStrength(col, UCOL_SECONDARY);
377     doAssert( (ucol_getStrength(col) != UCOL_TERTIARY), "collation object's strength is secondary difference");
378     doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
379     doAssert( (ucol_getStrength(col) == UCOL_SECONDARY), "collation object has the wrong strength");
380 
381 
382     log_verbose("Get display name for the default collation in German : \n");
383 
384     len=ucol_getDisplayName("en_US", "de_DE", NULL, 0,  &status);
385     if(status==U_BUFFER_OVERFLOW_ERROR){
386         status=U_ZERO_ERROR;
387         disName=(UChar*)malloc(sizeof(UChar) * (len+1));
388         ucol_getDisplayName("en_US", "de_DE", disName, len+1,  &status);
389         log_verbose("the display name for default collation in german: %s\n", austrdup(disName) );
390         free(disName);
391     }
392     if(U_FAILURE(status)){
393         log_err("ERROR: in getDisplayName: %s\n", myErrorName(status));
394         return;
395     }
396     log_verbose("Default collation getDisplayName ended.\n");
397 
398 
399     /* BEGIN android-removed
400        To save space, Android does not include the collation tailoring rules.
401        Skip the related tests.
402     ruled = ucol_open("da_DK", &status);
403     log_verbose("ucol_getRules() testing ...\n");
404     ucol_getRules(ruled, &tempLength);
405     doAssert( tempLength != 0, "getRules() result incorrect" );
406     log_verbose("getRules tests end.\n");
407     {
408         UChar *buffer = (UChar *)malloc(200000*sizeof(UChar));
409         int32_t bufLen = 200000;
410         buffer[0] = '\0';
411         log_verbose("ucol_getRulesEx() testing ...\n");
412         tempLength = ucol_getRulesEx(col,UCOL_TAILORING_ONLY,buffer,bufLen );
413         doAssert( tempLength == 0x0a, "getRulesEx() result incorrect" );
414         log_verbose("getRules tests end.\n");
415 
416         log_verbose("ucol_getRulesEx() testing ...\n");
417         tempLength=ucol_getRulesEx(col,UCOL_FULL_RULES,buffer,bufLen );
418         doAssert( tempLength != 0, "getRulesEx() result incorrect" );
419         log_verbose("getRules tests end.\n");
420         free(buffer);
421     }
422     ucol_close(ruled);
423     ucol_close(col);
424 
425     END android-removed */
426 
427     log_verbose("open an collator for french locale");
428     col = ucol_open("fr_FR", &status);
429     if (U_FAILURE(status)) {
430        log_err("ERROR: Creating French collation failed.: %s\n", myErrorName(status));
431         return;
432     }
433     ucol_setStrength(col, UCOL_PRIMARY);
434     log_verbose("testing ucol_getStrength() method again ...\n");
435     doAssert( (ucol_getStrength(col) != UCOL_TERTIARY), "collation object has the wrong strength");
436     doAssert( (ucol_getStrength(col) == UCOL_PRIMARY), "collation object's strength is not primary difference");
437 
438     log_verbose("testing French ucol_setStrength() method ...\n");
439     ucol_setStrength(col, UCOL_TERTIARY);
440     doAssert( (ucol_getStrength(col) == UCOL_TERTIARY), "collation object's strength is not tertiary difference");
441     doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
442     doAssert( (ucol_getStrength(col) != UCOL_SECONDARY), "collation object's strength is secondary difference");
443     ucol_close(col);
444 
445     log_verbose("Get display name for the french collation in english : \n");
446     len=ucol_getDisplayName("fr_FR", "en_US", NULL, 0,  &status);
447     if(status==U_BUFFER_OVERFLOW_ERROR){
448         status=U_ZERO_ERROR;
449         disName=(UChar*)malloc(sizeof(UChar) * (len+1));
450         ucol_getDisplayName("fr_FR", "en_US", disName, len+1,  &status);
451         log_verbose("the display name for french collation in english: %s\n", austrdup(disName) );
452         free(disName);
453     }
454     if(U_FAILURE(status)){
455         log_err("ERROR: in getDisplayName: %s\n", myErrorName(status));
456         return;
457     }
458     log_verbose("Default collation getDisplayName ended.\n");
459 
460 }
461 
462 /* Test RuleBasedCollator and getRules*/
TestRuleBasedColl()463 void TestRuleBasedColl()
464 {
465     UCollator *col1, *col2, *col3, *col4;
466     UCollationElements *iter1, *iter2;
467     UChar ruleset1[60];
468     UChar ruleset2[50];
469     UChar teststr[10];
470     UChar teststr2[10];
471     const UChar *rule1, *rule2, *rule3, *rule4;
472     int32_t tempLength;
473     UErrorCode status = U_ZERO_ERROR;
474     u_uastrcpy(ruleset1, "&9 < a, A < b, B < c, C; ch, cH, Ch, CH < d, D, e, E");
475     u_uastrcpy(ruleset2, "&9 < a, A < b, B < c, C < d, D, e, E");
476 
477 
478     col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL,&status);
479     if (U_FAILURE(status)) {
480         log_err_status(status, "RuleBased Collator creation failed.: %s\n", myErrorName(status));
481         return;
482     }
483     else
484         log_verbose("PASS: RuleBased Collator creation passed\n");
485 
486     status = U_ZERO_ERROR;
487     col2 = ucol_openRules(ruleset2, u_strlen(ruleset2),  UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
488     if (U_FAILURE(status)) {
489         log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
490         return;
491     }
492     else
493         log_verbose("PASS: RuleBased Collator creation passed\n");
494 
495 
496     status = U_ZERO_ERROR;
497     col3= ucol_open(NULL, &status);
498     if (U_FAILURE(status)) {
499         log_err("Default Collator creation failed.: %s\n", myErrorName(status));
500         return;
501     }
502     else
503         log_verbose("PASS: Default Collator creation passed\n");
504 
505     rule1 = ucol_getRules(col1, &tempLength);
506     rule2 = ucol_getRules(col2, &tempLength);
507     rule3 = ucol_getRules(col3, &tempLength);
508 
509     doAssert((u_strcmp(rule1, rule2) != 0), "Default collator getRules failed");
510     doAssert((u_strcmp(rule2, rule3) != 0), "Default collator getRules failed");
511     doAssert((u_strcmp(rule1, rule3) != 0), "Default collator getRules failed");
512 
513     col4=ucol_openRules(rule2, u_strlen(rule2), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
514     if (U_FAILURE(status)) {
515         log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
516         return;
517     }
518     rule4= ucol_getRules(col4, &tempLength);
519     doAssert((u_strcmp(rule2, rule4) == 0), "Default collator getRules failed");
520 
521     ucol_close(col1);
522     ucol_close(col2);
523     ucol_close(col3);
524     ucol_close(col4);
525 
526     /* tests that modifier ! is always ignored */
527     u_uastrcpy(ruleset1, "!&a<b");
528     teststr[0] = 0x0e40;
529     teststr[1] = 0x0e01;
530     teststr[2] = 0x0e2d;
531     col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
532     if (U_FAILURE(status)) {
533         log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
534         return;
535     }
536     col2 = ucol_open("en_US", &status);
537     if (U_FAILURE(status)) {
538         log_err("en_US Collator creation failed.: %s\n", myErrorName(status));
539         return;
540     }
541     iter1 = ucol_openElements(col1, teststr, 3, &status);
542     iter2 = ucol_openElements(col2, teststr, 3, &status);
543     if(U_FAILURE(status)) {
544         log_err("ERROR: CollationElement iterator creation failed.: %s\n", myErrorName(status));
545         return;
546     }
547     while (TRUE) {
548         /* testing with en since thai has its own tailoring */
549         uint32_t ce = ucol_next(iter1, &status);
550         uint32_t ce2 = ucol_next(iter2, &status);
551         if(U_FAILURE(status)) {
552             log_err("ERROR: CollationElement iterator creation failed.: %s\n", myErrorName(status));
553             return;
554         }
555         if (ce2 != ce) {
556              log_err("! modifier test failed");
557         }
558         if (ce == UCOL_NULLORDER) {
559             break;
560         }
561     }
562     ucol_closeElements(iter1);
563     ucol_closeElements(iter2);
564     ucol_close(col1);
565     ucol_close(col2);
566     /* test that we can start a rule without a & or < */
567     u_uastrcpy(ruleset1, "< z < a");
568     col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
569     if (U_FAILURE(status)) {
570         log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
571         return;
572     }
573     u_uastrcpy(teststr, "z");
574     u_uastrcpy(teststr2, "a");
575     if (ucol_greaterOrEqual(col1, teststr, 1, teststr2, 1)) {
576         log_err("Rule \"z < a\" fails");
577     }
578     ucol_close(col1);
579 }
580 
TestCompare()581 void TestCompare()
582 {
583     UErrorCode status = U_ZERO_ERROR;
584     UCollator *col;
585     UChar* test1;
586     UChar* test2;
587 
588     log_verbose("The compare tests begin : \n");
589     status=U_ZERO_ERROR;
590     col = ucol_open("en_US", &status);
591     if(U_FAILURE(status)) {
592         log_err_status(status, "ucal_open() collation creation failed.: %s\n", myErrorName(status));
593         return;
594     }
595     test1=(UChar*)malloc(sizeof(UChar) * 6);
596     test2=(UChar*)malloc(sizeof(UChar) * 6);
597     u_uastrcpy(test1, "Abcda");
598     u_uastrcpy(test2, "abcda");
599 
600     log_verbose("Use tertiary comparison level testing ....\n");
601 
602     doAssert( (!ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" != \"abcda\" ");
603     doAssert( (ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" >>> \"abcda\" ");
604     doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" >>> \"abcda\"");
605 
606     ucol_setStrength(col, UCOL_SECONDARY);
607     log_verbose("Use secondary comparison level testing ....\n");
608 
609     doAssert( (ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2) )), "Result should be \"Abcda\" == \"abcda\"");
610     doAssert( (!ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
611     doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2) )), "Result should be \"Abcda\" == \"abcda\"");
612 
613     ucol_setStrength(col, UCOL_PRIMARY);
614     log_verbose("Use primary comparison level testing ....\n");
615 
616     doAssert( (ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
617     doAssert( (!ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
618     doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
619 
620 
621     log_verbose("The compare tests end.\n");
622     ucol_close(col);
623     free(test1);
624     free(test2);
625 
626 }
627 /*
628 ---------------------------------------------
629  tests decomposition setting
630 */
TestDecomposition()631 void TestDecomposition() {
632     UErrorCode status = U_ZERO_ERROR;
633     UCollator *en_US, *el_GR, *vi_VN;
634     en_US = ucol_open("en_US", &status);
635     el_GR = ucol_open("el_GR", &status);
636     vi_VN = ucol_open("vi_VN", &status);
637 
638     if (U_FAILURE(status)) {
639         log_err_status(status, "ERROR: collation creation failed.: %s\n", myErrorName(status));
640         return;
641     }
642 
643     if (ucol_getAttribute(vi_VN, UCOL_NORMALIZATION_MODE, &status) != UCOL_ON ||
644         U_FAILURE(status))
645     {
646         log_err("ERROR: vi_VN collation did not have cannonical decomposition for normalization!\n");
647     }
648 
649     status = U_ZERO_ERROR;
650     if (ucol_getAttribute(el_GR, UCOL_NORMALIZATION_MODE, &status) != UCOL_ON ||
651         U_FAILURE(status))
652     {
653         log_err("ERROR: el_GR collation did not have cannonical decomposition for normalization!\n");
654     }
655 
656     status = U_ZERO_ERROR;
657     if (ucol_getAttribute(en_US, UCOL_NORMALIZATION_MODE, &status) != UCOL_OFF ||
658         U_FAILURE(status))
659     {
660         log_err("ERROR: en_US collation had cannonical decomposition for normalization!\n");
661     }
662 
663     ucol_close(en_US);
664     ucol_close(el_GR);
665     ucol_close(vi_VN);
666 }
667 
668 #define CLONETEST_COLLATOR_COUNT 4
669 
TestSafeClone()670 void TestSafeClone() {
671     UChar test1[6];
672     UChar test2[6];
673     static const UChar umlautUStr[] = {0x00DC, 0};
674     static const UChar oeStr[] = {0x0055, 0x0045, 0};
675     UCollator * someCollators [CLONETEST_COLLATOR_COUNT];
676     UCollator * someClonedCollators [CLONETEST_COLLATOR_COUNT];
677     UCollator * col;
678     UErrorCode err = U_ZERO_ERROR;
679     int8_t index = 6;    /* Leave this here to test buffer alingment in memory*/
680     uint8_t buffer [CLONETEST_COLLATOR_COUNT] [U_COL_SAFECLONE_BUFFERSIZE];
681     int32_t bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
682     const char sampleRuleChars[] = "&Z < CH";
683     UChar sampleRule[sizeof(sampleRuleChars)];
684 
685     if (TestBufferSize()) {
686         log_err("U_COL_SAFECLONE_BUFFERSIZE should be larger than sizeof(UCollator)\n");
687         return;
688     }
689 
690     u_uastrcpy(test1, "abCda");
691     u_uastrcpy(test2, "abcda");
692     u_uastrcpy(sampleRule, sampleRuleChars);
693 
694     /* one default collator & two complex ones */
695     someCollators[0] = ucol_open("en_US", &err);
696     someCollators[1] = ucol_open("ko", &err);
697     someCollators[2] = ucol_open("ja_JP", &err);
698     someCollators[3] = ucol_openRules(sampleRule, -1, UCOL_ON, UCOL_TERTIARY, NULL, &err);
699     if(U_FAILURE(err)) {
700         for (index = 0; index < CLONETEST_COLLATOR_COUNT; index++) {
701             ucol_close(someCollators[index]);
702         }
703         log_data_err("Couldn't open one or more collators\n");
704         return;
705     }
706 
707     /* Check the various error & informational states: */
708 
709     /* Null status - just returns NULL */
710     if (0 != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, 0))
711     {
712         log_err("FAIL: Cloned Collator failed to deal correctly with null status\n");
713     }
714     /* error status - should return 0 & keep error the same */
715     err = U_MEMORY_ALLOCATION_ERROR;
716     if (0 != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err) || err != U_MEMORY_ALLOCATION_ERROR)
717     {
718         log_err("FAIL: Cloned Collator failed to deal correctly with incoming error status\n");
719     }
720     err = U_ZERO_ERROR;
721 
722     /* Null buffer size pointer - just returns NULL & set error to U_ILLEGAL_ARGUMENT_ERROR*/
723     if (0 != ucol_safeClone(someCollators[0], buffer[0], 0, &err) || err != U_ILLEGAL_ARGUMENT_ERROR)
724     {
725         log_err("FAIL: Cloned Collator failed to deal correctly with null bufferSize pointer\n");
726     }
727     err = U_ZERO_ERROR;
728 
729     /* buffer size pointer is 0 - fill in pbufferSize with a size */
730     bufferSize = 0;
731     if (0 != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err) || U_FAILURE(err) || bufferSize <= 0)
732     {
733         log_err("FAIL: Cloned Collator failed a sizing request ('preflighting')\n");
734     }
735     /* Verify our define is large enough  */
736     if (U_COL_SAFECLONE_BUFFERSIZE < bufferSize)
737     {
738         log_err("FAIL: Pre-calculated buffer size is too small\n");
739     }
740     /* Verify we can use this run-time calculated size */
741     if (0 == (col = ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err)) || U_FAILURE(err))
742     {
743         log_err("FAIL: Collator can't be cloned with run-time size\n");
744     }
745     if (col) ucol_close(col);
746     /* size one byte too small - should allocate & let us know */
747     --bufferSize;
748     if (0 == (col = ucol_safeClone(someCollators[0], 0, &bufferSize, &err)) || err != U_SAFECLONE_ALLOCATED_WARNING)
749     {
750         log_err("FAIL: Cloned Collator failed to deal correctly with too-small buffer size\n");
751     }
752     if (col) ucol_close(col);
753     err = U_ZERO_ERROR;
754     bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
755 
756 
757     /* Null buffer pointer - return Collator & set error to U_SAFECLONE_ALLOCATED_ERROR */
758     if (0 == (col = ucol_safeClone(someCollators[0], 0, &bufferSize, &err)) || err != U_SAFECLONE_ALLOCATED_WARNING)
759     {
760         log_err("FAIL: Cloned Collator failed to deal correctly with null buffer pointer\n");
761     }
762     if (col) ucol_close(col);
763     err = U_ZERO_ERROR;
764 
765     /* Null Collator - return NULL & set U_ILLEGAL_ARGUMENT_ERROR */
766     if (0 != ucol_safeClone(0, buffer[0], &bufferSize, &err) || err != U_ILLEGAL_ARGUMENT_ERROR)
767     {
768         log_err("FAIL: Cloned Collator failed to deal correctly with null Collator pointer\n");
769     }
770 
771     err = U_ZERO_ERROR;
772 
773     /* Test that a cloned collator doesn't accidentally use UCA. */
774     col=ucol_open("de@collation=phonebook", &err);
775     bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
776     someClonedCollators[0] = ucol_safeClone(col, buffer[0], &bufferSize, &err);
777     doAssert( (ucol_greater(col, umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Original German phonebook collation sorts differently than expected");
778     doAssert( (ucol_greater(someClonedCollators[0], umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Cloned German phonebook collation sorts differently than expected");
779     if (!ucol_equals(someClonedCollators[0], col)) {
780         log_err("FAIL: Cloned German phonebook collator is not equal to original.\n");
781     }
782     ucol_close(col);
783     ucol_close(someClonedCollators[0]);
784 
785     err = U_ZERO_ERROR;
786 
787     /* change orig & clone & make sure they are independent */
788 
789     for (index = 0; index < CLONETEST_COLLATOR_COUNT; index++)
790     {
791         ucol_setStrength(someCollators[index], UCOL_IDENTICAL);
792         bufferSize = 1;
793         err = U_ZERO_ERROR;
794         ucol_close(ucol_safeClone(someCollators[index], buffer[index], &bufferSize, &err));
795         if (err != U_SAFECLONE_ALLOCATED_WARNING) {
796             log_err("FAIL: collator number %d was not allocated.\n", index);
797             log_err("FAIL: status of Collator[%d] is %d  (hex: %x).\n", index, err, err);
798         }
799 
800         bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
801         err = U_ZERO_ERROR;
802         someClonedCollators[index] = ucol_safeClone(someCollators[index], buffer[index], &bufferSize, &err);
803         if (someClonedCollators[index] == NULL
804             || someClonedCollators[index] < (UCollator *)buffer[index]
805             || someClonedCollators[index] > (UCollator *)(buffer[index]+(U_COL_SAFECLONE_BUFFERSIZE-1)))
806         {
807             log_err("FAIL: Cloned collator didn't use provided buffer.\n");
808             return;
809         }
810         if (!ucol_equals(someClonedCollators[index], someCollators[index])) {
811             log_err("FAIL: Cloned collator is not equal to original at index = %d.\n", index);
812         }
813 
814         /* Check the usability */
815         ucol_setStrength(someCollators[index], UCOL_PRIMARY);
816         ucol_setAttribute(someCollators[index], UCOL_CASE_LEVEL, UCOL_OFF, &err);
817 
818         doAssert( (ucol_equal(someCollators[index], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abcda\" == \"abCda\"");
819 
820         /* Close the original to make sure that the clone is usable. */
821         ucol_close(someCollators[index]);
822 
823         ucol_setStrength(someClonedCollators[index], UCOL_TERTIARY);
824         ucol_setAttribute(someClonedCollators[index], UCOL_CASE_LEVEL, UCOL_OFF, &err);
825         doAssert( (ucol_greater(someClonedCollators[index], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abCda\" >>> \"abcda\" ");
826 
827         ucol_close(someClonedCollators[index]);
828     }
829 }
830 
TestCloneBinary()831 void TestCloneBinary(){
832     UErrorCode err = U_ZERO_ERROR;
833     UCollator * col = ucol_open("en_US", &err);
834     UCollator * c;
835     int32_t size;
836     uint8_t * buffer;
837 
838     if (U_FAILURE(err)) {
839         log_data_err("Couldn't open collator. Error: %s\n", u_errorName(err));
840         return;
841     }
842 
843     size = ucol_cloneBinary(col, NULL, 0, &err);
844     if(size==0 || err!=U_BUFFER_OVERFLOW_ERROR) {
845         log_err("ucol_cloneBinary - couldn't check size. Error: %s\n", u_errorName(err));
846         return;
847     }
848     err = U_ZERO_ERROR;
849 
850     buffer = (uint8_t *) malloc(size);
851     ucol_cloneBinary(col, buffer, size, &err);
852     if(U_FAILURE(err)) {
853         log_err("ucol_cloneBinary - couldn't clone.. Error: %s\n", u_errorName(err));
854         free(buffer);
855         return;
856     }
857 
858     /* how to check binary result ? */
859 
860     c = ucol_openBinary(buffer, size, col, &err);
861     if(U_FAILURE(err)) {
862         log_err("ucol_openBinary failed. Error: %s\n", u_errorName(err));
863     } else {
864         UChar t[] = {0x41, 0x42, 0x43, 0};  /* ABC */
865         uint8_t  *k1, *k2;
866         int l1, l2;
867         l1 = ucol_getSortKey(col, t, -1, NULL,0);
868         l2 = ucol_getSortKey(c, t, -1, NULL,0);
869         k1 = (uint8_t *) malloc(sizeof(uint8_t) * l1);
870         k2 = (uint8_t *) malloc(sizeof(uint8_t) * l2);
871         ucol_getSortKey(col, t, -1, k1, l1);
872         ucol_getSortKey(col, t, -1, k2, l2);
873         if (strcmp((char *)k1,(char *)k2) != 0){
874             log_err("ucol_openBinary - new collator should equal to old one\n");
875         };
876         free(k1);
877         free(k2);
878     }
879     free(buffer);
880     ucol_close(c);
881     ucol_close(col);
882 }
883 /*
884     TestOpenVsOpenRules ensures that collators from ucol_open and ucol_openRules
885     will generate identical sort keys
886 */
TestOpenVsOpenRules()887 void TestOpenVsOpenRules(){
888 
889     /* create an array of all the locales */
890     int32_t numLocales = uloc_countAvailable();
891     int32_t sizeOfStdSet;
892     uint32_t adder;
893     UChar str[41]; /* create an array of UChar of size maximum strSize + 1 */
894     USet *stdSet;
895     char* curLoc;
896     UCollator * c1;
897     UCollator * c2;
898     const UChar* rules;
899     int32_t rulesLength;
900     int32_t sortKeyLen1, sortKeyLen2;
901     uint8_t *sortKey1 = NULL, *sortKey2 = NULL;
902     ULocaleData *uld;
903     int32_t x, y, z;
904     USet *eSet;
905     int32_t eSize;
906     int strSize;
907 
908     UErrorCode err = U_ZERO_ERROR;
909 
910     /* create a set of standard characters that aren't very interesting...
911     and then we can find some interesting ones later */
912 
913     stdSet = uset_open(0x61, 0x7A);
914     uset_addRange(stdSet, 0x41, 0x5A);
915     uset_addRange(stdSet, 0x30, 0x39);
916     sizeOfStdSet = uset_size(stdSet);
917 
918     adder = 1;
919     if(getTestOption(QUICK_OPTION))
920     {
921         adder = 10;
922     }
923 
924     for(x = 0; x < numLocales; x+=adder){
925         curLoc = (char *)uloc_getAvailable(x);
926         log_verbose("Processing %s\n", curLoc);
927 
928         /* create a collator the normal API way */
929         c1 = ucol_open(curLoc, &err);
930         if (U_FAILURE(err)) {
931             log_err("ERROR: Normal collation creation failed with locale: %s : %s\n", curLoc, myErrorName(err));
932             return;
933         }
934 
935         /* grab the rules */
936         rules = ucol_getRules(c1, &rulesLength);
937 
938         /* use those rules to create a collator from rules */
939         c2 = ucol_openRules(rules, rulesLength, UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &err);
940         if (U_FAILURE(err)) {
941             log_err("ERROR: Creating collator from rules failed with locale: %s : %s\n", curLoc, myErrorName(err));
942             return;
943         }
944 
945         uld = ulocdata_open(curLoc, &err);
946 
947         /*now that we have some collators, we get several strings */
948 
949         for(y = 0; y < 5; y++){
950 
951             /* get a set of ALL the characters in this locale */
952             eSet =  ulocdata_getExemplarSet(uld, NULL, 0, ULOCDATA_ES_STANDARD, &err);
953             eSize = uset_size(eSet);
954 
955             /* make a string with these characters in it */
956             strSize = (rand()%40) + 1;
957 
958             for(z = 0; z < strSize; z++){
959                 str[z] = uset_charAt(eSet, rand()%eSize);
960             }
961 
962             /* change the set to only include 'abnormal' characters (not A-Z, a-z, 0-9 */
963             uset_removeAll(eSet, stdSet);
964             eSize = uset_size(eSet);
965 
966             /* if there are some non-normal characters left, put a few into the string, just to make sure we have some */
967             if(eSize > 0){
968                 str[2%strSize] = uset_charAt(eSet, rand()%eSize);
969                 str[3%strSize] = uset_charAt(eSet, rand()%eSize);
970                 str[5%strSize] = uset_charAt(eSet, rand()%eSize);
971                 str[10%strSize] = uset_charAt(eSet, rand()%eSize);
972                 str[13%strSize] = uset_charAt(eSet, rand()%eSize);
973             }
974             /* terminate the string */
975             str[strSize-1] = '\0';
976             log_verbose("String used: %S\n", str);
977 
978             /* get sort keys for both of them, and check that the keys are identicle */
979             sortKeyLen1 = ucol_getSortKey(c1, str, u_strlen(str),  NULL, 0);
980             sortKey1 = (uint8_t*)malloc(sizeof(uint8_t) * (sortKeyLen1 + 1));
981             /*memset(sortKey1, 0xFE, sortKeyLen1);*/
982             ucol_getSortKey(c1, str, u_strlen(str), sortKey1, sortKeyLen1 + 1);
983 
984             sortKeyLen2 = ucol_getSortKey(c2, str, u_strlen(str),  NULL, 0);
985             sortKey2 = (uint8_t*)malloc(sizeof(uint8_t) * (sortKeyLen2 + 1));
986             /*memset(sortKey2, 0xFE, sortKeyLen2);*/
987             ucol_getSortKey(c2, str, u_strlen(str), sortKey2, sortKeyLen2 + 1);
988 
989             /* Check that the lengths are the same */
990             doAssert((sortKeyLen1 == sortKeyLen2), "Sort key lengths do not match.");
991 
992             /* check that the keys are the same */
993             doAssert((memcmp(sortKey1, sortKey2, sortKeyLen1) == 0), "Keys are not equivalent");
994 
995             /* clean up after each string */
996             free(sortKey1);
997             free(sortKey2);
998             uset_close(eSet);
999         }
1000         /* clean up after each locale */
1001         ulocdata_close(uld);
1002         ucol_close(c1);
1003         ucol_close(c2);
1004     }
1005     /* final clean up */
1006     uset_close(stdSet);
1007 }
1008 /*
1009 ----------------------------------------------------------------------------
1010  ctor -- Tests the getSortKey
1011 */
TestSortKey()1012 void TestSortKey()
1013 {
1014     uint8_t *sortk1 = NULL, *sortk2 = NULL, *sortk3 = NULL, *sortkEmpty = NULL;
1015     int32_t sortklen, osortklen;
1016     uint32_t toStringLen=0;
1017     UCollator *col;
1018     UChar *test1, *test2, *test3;
1019     UErrorCode status = U_ZERO_ERROR;
1020     char toStringBuffer[256], *resultP;
1021 
1022 
1023     uint8_t s1[] = { 0x9f, 0x00 };
1024     uint8_t s2[] = { 0x61, 0x00 };
1025     int  strcmpResult;
1026 
1027     strcmpResult = strcmp((const char *)s1, (const char *)s2);
1028     log_verbose("strcmp(0x9f..., 0x61...) = %d\n", strcmpResult);
1029 
1030     if(strcmpResult <= 0) {
1031       log_err("ERR: expected strcmp(\"9f 00\", \"61 00\") to be >=0 (GREATER).. got %d. Calling strcmp() for sortkeys may not work! \n",
1032               strcmpResult);
1033     }
1034 
1035 
1036     log_verbose("testing SortKey begins...\n");
1037     /* this is supposed to open default date format, but later on it treats it like it is "en_US"
1038        - very bad if you try to run the tests on machine where default locale is NOT "en_US" */
1039     /* col = ucol_open(NULL, &status); */
1040     col = ucol_open("en_US", &status);
1041     if (U_FAILURE(status)) {
1042         log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1043         return;
1044     }
1045 
1046 
1047     if(ucol_getStrength(col) != UCOL_DEFAULT_STRENGTH)
1048     {
1049         log_err("ERROR: default collation did not have UCOL_DEFAULT_STRENGTH !\n");
1050     }
1051     /* Need to use identical strength */
1052     ucol_setAttribute(col, UCOL_STRENGTH, UCOL_IDENTICAL, &status);
1053 
1054     test1=(UChar*)malloc(sizeof(UChar) * 6);
1055     test2=(UChar*)malloc(sizeof(UChar) * 6);
1056     test3=(UChar*)malloc(sizeof(UChar) * 6);
1057 
1058     memset(test1,0xFE, sizeof(UChar)*6);
1059     memset(test2,0xFE, sizeof(UChar)*6);
1060     memset(test3,0xFE, sizeof(UChar)*6);
1061 
1062 
1063     u_uastrcpy(test1, "Abcda");
1064     u_uastrcpy(test2, "abcda");
1065     u_uastrcpy(test3, "abcda");
1066 
1067     log_verbose("Use tertiary comparison level testing ....\n");
1068 
1069     sortklen=ucol_getSortKey(col, test1, u_strlen(test1),  NULL, 0);
1070     sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1071     memset(sortk1,0xFE, sortklen);
1072     ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortklen+1);
1073 
1074     sortklen=ucol_getSortKey(col, test2, u_strlen(test2),  NULL, 0);
1075     sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1076     memset(sortk2,0xFE, sortklen);
1077     ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortklen+1);
1078 
1079     osortklen = sortklen;
1080     sortklen=ucol_getSortKey(col, test2, u_strlen(test3),  NULL, 0);
1081     sortk3=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1082     memset(sortk3,0xFE, sortklen);
1083     ucol_getSortKey(col, test2, u_strlen(test2), sortk3, sortklen+1);
1084 
1085     doAssert( (sortklen == osortklen), "Sortkey length should be the same (abcda, abcda)");
1086 
1087     doAssert( (memcmp(sortk1, sortk2, sortklen) > 0), "Result should be \"Abcda\" > \"abcda\"");
1088     doAssert( (memcmp(sortk2, sortk1, sortklen) < 0), "Result should be \"abcda\" < \"Abcda\"");
1089     doAssert( (memcmp(sortk2, sortk3, sortklen) == 0), "Result should be \"abcda\" ==  \"abcda\"");
1090 
1091     resultP = ucol_sortKeyToString(col, sortk3, toStringBuffer, &toStringLen);
1092     doAssert( (resultP != 0), "sortKeyToString failed!");
1093 
1094 #if 1 /* verobse log of sortkeys */
1095     {
1096       char junk2[1000];
1097       char junk3[1000];
1098       int i;
1099 
1100       strcpy(junk2, "abcda[2] ");
1101       strcpy(junk3, " abcda[3] ");
1102 
1103       for(i=0;i<sortklen;i++)
1104         {
1105           sprintf(junk2+strlen(junk2), "%02X ",(int)( 0xFF & sortk2[i]));
1106           sprintf(junk3+strlen(junk3), "%02X ",(int)( 0xFF & sortk3[i]));
1107         }
1108 
1109       log_verbose("%s\n", junk2);
1110       log_verbose("%s\n", junk3);
1111     }
1112 #endif
1113 
1114     free(sortk1);
1115     free(sortk2);
1116     free(sortk3);
1117 
1118     log_verbose("Use secondary comparision level testing ...\n");
1119     ucol_setStrength(col, UCOL_SECONDARY);
1120     sortklen=ucol_getSortKey(col, test1, u_strlen(test1),  NULL, 0);
1121     sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1122     ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortklen+1);
1123     sortklen=ucol_getSortKey(col, test2, u_strlen(test2),  NULL, 0);
1124     sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1125     ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortklen+1);
1126 
1127     doAssert( !(memcmp(sortk1, sortk2, sortklen) > 0), "Result should be \"Abcda\" == \"abcda\"");
1128     doAssert( !(memcmp(sortk2, sortk1, sortklen) < 0), "Result should be \"abcda\" == \"Abcda\"");
1129     doAssert( (memcmp(sortk1, sortk2, sortklen) == 0), "Result should be \"abcda\" ==  \"abcda\"");
1130 
1131     log_verbose("getting sortkey for an empty string\n");
1132     ucol_setAttribute(col, UCOL_STRENGTH, UCOL_TERTIARY, &status);
1133     sortklen = ucol_getSortKey(col, test1, 0, NULL, 0);
1134     sortkEmpty = (uint8_t*)malloc(sizeof(uint8_t) * sortklen+1);
1135     sortklen = ucol_getSortKey(col, test1, 0, sortkEmpty, sortklen+1);
1136     if(sortklen != 3 || sortkEmpty[0] != 1 || sortkEmpty[0] != 1 || sortkEmpty[2] != 0) {
1137       log_err("Empty string generated wrong sortkey!\n");
1138     }
1139     free(sortkEmpty);
1140 
1141     log_verbose("testing passing invalid string\n");
1142     sortklen = ucol_getSortKey(col, NULL, 0, NULL, 0);
1143     if(sortklen != 0) {
1144       log_err("Invalid string didn't return sortkey size of 0\n");
1145     }
1146 
1147 
1148     log_verbose("testing sortkey ends...\n");
1149     ucol_close(col);
1150     free(test1);
1151     free(test2);
1152     free(test3);
1153     free(sortk1);
1154     free(sortk2);
1155 
1156 }
TestHashCode()1157 void TestHashCode()
1158 {
1159     uint8_t *sortk1, *sortk2, *sortk3;
1160     int32_t sortk1len, sortk2len, sortk3len;
1161     UCollator *col;
1162     UChar *test1, *test2, *test3;
1163     UErrorCode status = U_ZERO_ERROR;
1164     log_verbose("testing getHashCode begins...\n");
1165     col = ucol_open("en_US", &status);
1166     if (U_FAILURE(status)) {
1167         log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1168         return;
1169     }
1170     test1=(UChar*)malloc(sizeof(UChar) * 6);
1171     test2=(UChar*)malloc(sizeof(UChar) * 6);
1172     test3=(UChar*)malloc(sizeof(UChar) * 6);
1173     u_uastrcpy(test1, "Abcda");
1174     u_uastrcpy(test2, "abcda");
1175     u_uastrcpy(test3, "abcda");
1176 
1177     log_verbose("Use tertiary comparison level testing ....\n");
1178     sortk1len=ucol_getSortKey(col, test1, u_strlen(test1),  NULL, 0);
1179     sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortk1len+1));
1180     ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortk1len+1);
1181     sortk2len=ucol_getSortKey(col, test2, u_strlen(test2),  NULL, 0);
1182     sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortk2len+1));
1183     ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortk2len+1);
1184     sortk3len=ucol_getSortKey(col, test2, u_strlen(test3),  NULL, 0);
1185     sortk3=(uint8_t*)malloc(sizeof(uint8_t) * (sortk3len+1));
1186     ucol_getSortKey(col, test2, u_strlen(test2), sortk3, sortk3len+1);
1187 
1188 
1189     log_verbose("ucol_hashCode() testing ...\n");
1190 
1191     doAssert( ucol_keyHashCode(sortk1, sortk1len) != ucol_keyHashCode(sortk2, sortk2len), "Hash test1 result incorrect" );
1192     doAssert( !(ucol_keyHashCode(sortk1, sortk1len) == ucol_keyHashCode(sortk2, sortk2len)), "Hash test2 result incorrect" );
1193     doAssert( ucol_keyHashCode(sortk2, sortk2len) == ucol_keyHashCode(sortk3, sortk3len), "Hash result not equal" );
1194 
1195     log_verbose("hashCode tests end.\n");
1196     ucol_close(col);
1197     free(sortk1);
1198     free(sortk2);
1199     free(sortk3);
1200     free(test1);
1201     free(test2);
1202     free(test3);
1203 
1204 
1205 }
1206 /*
1207  *----------------------------------------------------------------------------
1208  * Tests the UCollatorElements API.
1209  *
1210  */
TestElemIter()1211 void TestElemIter()
1212 {
1213     int32_t offset;
1214     int32_t order1, order2, order3;
1215     UChar *testString1, *testString2;
1216     UCollator *col;
1217     UCollationElements *iterator1, *iterator2, *iterator3;
1218     UErrorCode status = U_ZERO_ERROR;
1219     log_verbose("testing UCollatorElements begins...\n");
1220     col = ucol_open("en_US", &status);
1221     ucol_setAttribute(col, UCOL_NORMALIZATION_MODE, UCOL_OFF, &status);
1222     if (U_FAILURE(status)) {
1223         log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1224         return;
1225     }
1226 
1227     testString1=(UChar*)malloc(sizeof(UChar) * 150);
1228     testString2=(UChar*)malloc(sizeof(UChar) * 150);
1229     u_uastrcpy(testString1, "XFILE What subset of all possible test cases has the highest probability of detecting the most errors?");
1230     u_uastrcpy(testString2, "Xf_ile What subset of all possible test cases has the lowest probability of detecting the least errors?");
1231 
1232     log_verbose("Constructors and comparison testing....\n");
1233 
1234     iterator1 = ucol_openElements(col, testString1, u_strlen(testString1), &status);
1235     if(U_FAILURE(status)) {
1236         log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1237         ucol_close(col);
1238         return;
1239     }
1240     else{ log_verbose("PASS: Default collationElement iterator1 creation passed\n");}
1241 
1242     iterator2 = ucol_openElements(col, testString1, u_strlen(testString1), &status);
1243     if(U_FAILURE(status)) {
1244         log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1245         ucol_close(col);
1246         return;
1247     }
1248     else{ log_verbose("PASS: Default collationElement iterator2 creation passed\n");}
1249 
1250     iterator3 = ucol_openElements(col, testString2, u_strlen(testString2), &status);
1251     if(U_FAILURE(status)) {
1252         log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1253         ucol_close(col);
1254         return;
1255     }
1256     else{ log_verbose("PASS: Default collationElement iterator3 creation passed\n");}
1257 
1258     offset=ucol_getOffset(iterator1);
1259     ucol_setOffset(iterator1, 6, &status);
1260     if (U_FAILURE(status)) {
1261         log_err("Error in setOffset for UCollatorElements iterator.: %s\n", myErrorName(status));
1262         return;
1263     }
1264     if(ucol_getOffset(iterator1)==6)
1265         log_verbose("setOffset and getOffset working fine\n");
1266     else{
1267         log_err("error in set and get Offset got %d instead of 6\n", ucol_getOffset(iterator1));
1268     }
1269 
1270     ucol_setOffset(iterator1, 0, &status);
1271     order1 = ucol_next(iterator1, &status);
1272     if (U_FAILURE(status)) {
1273         log_err("Somehow ran out of memory stepping through the iterator1.: %s\n", myErrorName(status));
1274         return;
1275     }
1276     order2=ucol_getOffset(iterator2);
1277     doAssert((order1 != order2), "The first iterator advance failed");
1278     order2 = ucol_next(iterator2, &status);
1279     if (U_FAILURE(status)) {
1280         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1281         return;
1282     }
1283     order3 = ucol_next(iterator3, &status);
1284     if (U_FAILURE(status)) {
1285         log_err("Somehow ran out of memory stepping through the iterator3.: %s\n", myErrorName(status));
1286         return;
1287     }
1288 
1289     doAssert((order1 == order2), "The second iterator advance failed should be the same as first one");
1290 
1291 doAssert( (ucol_primaryOrder(order1) == ucol_primaryOrder(order3)), "The primary orders should be identical");
1292 doAssert( (ucol_secondaryOrder(order1) == ucol_secondaryOrder(order3)), "The secondary orders should be identical");
1293 doAssert( (ucol_tertiaryOrder(order1) == ucol_tertiaryOrder(order3)), "The tertiary orders should be identical");
1294 
1295     order1=ucol_next(iterator1, &status);
1296     if (U_FAILURE(status)) {
1297         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1298         return;
1299     }
1300     order3=ucol_next(iterator3, &status);
1301     if (U_FAILURE(status)) {
1302         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1303         return;
1304     }
1305 doAssert( (ucol_primaryOrder(order1) == ucol_primaryOrder(order3)), "The primary orders should be identical");
1306 doAssert( (ucol_tertiaryOrder(order1) != ucol_tertiaryOrder(order3)), "The tertiary orders should be different");
1307 
1308     order1=ucol_next(iterator1, &status);
1309     if (U_FAILURE(status)) {
1310         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1311         return;
1312     }
1313     order3=ucol_next(iterator3, &status);
1314     if (U_FAILURE(status)) {
1315         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1316         return;
1317     }
1318     /* this here, my friends, is either pure lunacy or something so obsolete that even it's mother
1319      * doesn't care about it. Essentialy, this test complains if secondary values for 'I' and '_'
1320      * are the same. According to the UCA, this is not true. Therefore, remove the test.
1321      * Besides, if primary strengths for two code points are different, it doesn't matter one bit
1322      * what is the relation between secondary or any other strengths.
1323      * killed by weiv 06/11/2002.
1324      */
1325     /*
1326     doAssert( ((order1 & UCOL_SECONDARYMASK) != (order3 & UCOL_SECONDARYMASK)), "The secondary orders should be different");
1327     */
1328     doAssert( (order1 != UCOL_NULLORDER), "Unexpected end of iterator reached");
1329 
1330     free(testString1);
1331     free(testString2);
1332     ucol_closeElements(iterator1);
1333     ucol_closeElements(iterator2);
1334     ucol_closeElements(iterator3);
1335     ucol_close(col);
1336 
1337     log_verbose("testing CollationElementIterator ends...\n");
1338 }
1339 
TestGetLocale()1340 void TestGetLocale() {
1341   UErrorCode status = U_ZERO_ERROR;
1342   const char *rules = "&a<x<y<z";
1343   UChar rlz[256] = {0};
1344   uint32_t rlzLen = u_unescape(rules, rlz, 256);
1345 
1346   UCollator *coll = NULL;
1347   const char *locale = NULL;
1348 
1349   int32_t i = 0;
1350 
1351   /* Now that the collation tree is separate, actual==valid at all times. [alan] */
1352   static const struct {
1353     const char* requestedLocale;
1354     const char* validLocale;
1355     const char* actualLocale;
1356   } testStruct[] = {
1357     { "sr_RS", "sr_Cyrl_RS", "sr" },
1358     { "sh_YU", "sr_Latn_RS", "hr" }, /* this used to be sh, but now sh collation aliases hr */
1359     { "en_BE_FOO", "en_BE", "root" },
1360     { "de_DE_NONEXISTANT", "de_DE", "de" }
1361   };
1362 
1363   /* test opening collators for different locales */
1364   for(i = 0; i<sizeof(testStruct)/sizeof(testStruct[0]); i++) {
1365     status = U_ZERO_ERROR;
1366     coll = ucol_open(testStruct[i].requestedLocale, &status);
1367     if(U_FAILURE(status)) {
1368       log_err_status(status, "Failed to open collator for %s with %s\n", testStruct[i].requestedLocale, u_errorName(status));
1369       ucol_close(coll);
1370       continue;
1371     }
1372    locale = ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status);
1373     if(strcmp(locale, testStruct[i].requestedLocale) != 0) {
1374       log_err("[Coll %s]: Error in requested locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].requestedLocale, locale);
1375     }
1376     locale = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status);
1377     if(strcmp(locale, testStruct[i].validLocale) != 0) {
1378       log_err("[Coll %s]: Error in valid locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].validLocale, locale);
1379     }
1380     locale = ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status);
1381     if(strcmp(locale, testStruct[i].actualLocale) != 0) {
1382       log_err("[Coll %s]: Error in actual locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].actualLocale, locale);
1383     }
1384     ucol_close(coll);
1385   }
1386 
1387   /* completely non-existant locale for collator should get a default collator */
1388   {
1389     UCollator *defaultColl = ucol_open(NULL, &status);
1390     coll = ucol_open("blahaha", &status);
1391     if(U_SUCCESS(status)) {
1392       if(strcmp(ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status), "blahaha")) {
1393         log_err("Nonexisting locale didn't preserve the requested locale\n");
1394       }
1395       if(strcmp(ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status),
1396         ucol_getLocaleByType(defaultColl, ULOC_VALID_LOCALE, &status))) {
1397         log_err("Valid locale for nonexisting locale locale collator differs "
1398           "from valid locale for default collator\n");
1399       }
1400       if(strcmp(ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status),
1401         ucol_getLocaleByType(defaultColl, ULOC_ACTUAL_LOCALE, &status))) {
1402         log_err("Actual locale for nonexisting locale locale collator differs "
1403           "from actual locale for default collator\n");
1404       }
1405       ucol_close(coll);
1406       ucol_close(defaultColl);
1407     } else {
1408       log_data_err("Couldn't open collators\n");
1409     }
1410   }
1411 
1412 
1413 
1414   /* collator instantiated from rules should have all three locales NULL */
1415   coll = ucol_openRules(rlz, rlzLen, UCOL_DEFAULT, UCOL_DEFAULT, NULL, &status);
1416   locale = ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status);
1417   if(locale != NULL) {
1418     log_err("For collator instantiated from rules, requested locale returned %s instead of NULL\n", locale);
1419   }
1420   locale = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status);
1421   if(locale != NULL) {
1422     log_err("For collator instantiated from rules,  valid locale returned %s instead of NULL\n", locale);
1423   }
1424   locale = ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status);
1425   if(locale != NULL) {
1426     log_err("For collator instantiated from rules, actual locale returned %s instead of NULL\n", locale);
1427   }
1428   ucol_close(coll);
1429 
1430 }
1431 
1432 
TestGetAll()1433 void TestGetAll()
1434 {
1435     int32_t i, count;
1436     count=ucol_countAvailable();
1437     /* use something sensible w/o hardcoding the count */
1438     if(count < 0){
1439         log_err("Error in countAvailable(), it returned %d\n", count);
1440     }
1441     else{
1442         log_verbose("PASS: countAvailable() successful, it returned %d\n", count);
1443     }
1444     for(i=0;i<count;i++)
1445         log_verbose("%s\n", ucol_getAvailable(i));
1446 
1447 
1448 }
1449 
1450 
1451 struct teststruct {
1452     const char *original;
1453     uint8_t key[256];
1454 } ;
1455 
compare_teststruct(const void * string1,const void * string2)1456 static int compare_teststruct(const void *string1, const void *string2) {
1457     return(strcmp((const char *)((struct teststruct *)string1)->key, (const char *)((struct teststruct *)string2)->key));
1458 }
1459 
TestBounds()1460 void TestBounds() {
1461     UErrorCode status = U_ZERO_ERROR;
1462 
1463     UCollator *coll = ucol_open("sh", &status);
1464 
1465     uint8_t sortkey[512], lower[512], upper[512];
1466     UChar buffer[512];
1467 
1468     static const char * const test[] = {
1469         "John Smith",
1470         "JOHN SMITH",
1471         "john SMITH",
1472         "j\\u00F6hn sm\\u00EFth",
1473         "J\\u00F6hn Sm\\u00EFth",
1474         "J\\u00D6HN SM\\u00CFTH",
1475         "john smithsonian",
1476         "John Smithsonian",
1477     };
1478 
1479     struct teststruct tests[] = {
1480         {"\\u010CAKI MIHALJ" } ,
1481         {"\\u010CAKI MIHALJ" } ,
1482         {"\\u010CAKI PIRO\\u0160KA" },
1483         {"\\u010CABAI ANDRIJA" } ,
1484         {"\\u010CABAI LAJO\\u0160" } ,
1485         {"\\u010CABAI MARIJA" } ,
1486         {"\\u010CABAI STEVAN" } ,
1487         {"\\u010CABAI STEVAN" } ,
1488         {"\\u010CABARKAPA BRANKO" } ,
1489         {"\\u010CABARKAPA MILENKO" } ,
1490         {"\\u010CABARKAPA MIROSLAV" } ,
1491         {"\\u010CABARKAPA SIMO" } ,
1492         {"\\u010CABARKAPA STANKO" } ,
1493         {"\\u010CABARKAPA TAMARA" } ,
1494         {"\\u010CABARKAPA TOMA\\u0160" } ,
1495         {"\\u010CABDARI\\u0106 NIKOLA" } ,
1496         {"\\u010CABDARI\\u0106 ZORICA" } ,
1497         {"\\u010CABI NANDOR" } ,
1498         {"\\u010CABOVI\\u0106 MILAN" } ,
1499         {"\\u010CABRADI AGNEZIJA" } ,
1500         {"\\u010CABRADI IVAN" } ,
1501         {"\\u010CABRADI JELENA" } ,
1502         {"\\u010CABRADI LJUBICA" } ,
1503         {"\\u010CABRADI STEVAN" } ,
1504         {"\\u010CABRDA MARTIN" } ,
1505         {"\\u010CABRILO BOGDAN" } ,
1506         {"\\u010CABRILO BRANISLAV" } ,
1507         {"\\u010CABRILO LAZAR" } ,
1508         {"\\u010CABRILO LJUBICA" } ,
1509         {"\\u010CABRILO SPASOJA" } ,
1510         {"\\u010CADE\\u0160 ZDENKA" } ,
1511         {"\\u010CADESKI BLAGOJE" } ,
1512         {"\\u010CADOVSKI VLADIMIR" } ,
1513         {"\\u010CAGLJEVI\\u0106 TOMA" } ,
1514         {"\\u010CAGOROVI\\u0106 VLADIMIR" } ,
1515         {"\\u010CAJA VANKA" } ,
1516         {"\\u010CAJI\\u0106 BOGOLJUB" } ,
1517         {"\\u010CAJI\\u0106 BORISLAV" } ,
1518         {"\\u010CAJI\\u0106 RADOSLAV" } ,
1519         {"\\u010CAK\\u0160IRAN MILADIN" } ,
1520         {"\\u010CAKAN EUGEN" } ,
1521         {"\\u010CAKAN EVGENIJE" } ,
1522         {"\\u010CAKAN IVAN" } ,
1523         {"\\u010CAKAN JULIJAN" } ,
1524         {"\\u010CAKAN MIHAJLO" } ,
1525         {"\\u010CAKAN STEVAN" } ,
1526         {"\\u010CAKAN VLADIMIR" } ,
1527         {"\\u010CAKAN VLADIMIR" } ,
1528         {"\\u010CAKAN VLADIMIR" } ,
1529         {"\\u010CAKARA ANA" } ,
1530         {"\\u010CAKAREVI\\u0106 MOMIR" } ,
1531         {"\\u010CAKAREVI\\u0106 NEDELJKO" } ,
1532         {"\\u010CAKI \\u0160ANDOR" } ,
1533         {"\\u010CAKI AMALIJA" } ,
1534         {"\\u010CAKI ANDRA\\u0160" } ,
1535         {"\\u010CAKI LADISLAV" } ,
1536         {"\\u010CAKI LAJO\\u0160" } ,
1537         {"\\u010CAKI LASLO" } ,
1538     };
1539 
1540 
1541 
1542     int32_t i = 0, j = 0, k = 0, buffSize = 0, skSize = 0, lowerSize = 0, upperSize = 0;
1543     int32_t arraySize = sizeof(tests)/sizeof(tests[0]);
1544 
1545     if(U_SUCCESS(status) && coll) {
1546         for(i = 0; i<arraySize; i++) {
1547             buffSize = u_unescape(tests[i].original, buffer, 512);
1548             skSize = ucol_getSortKey(coll, buffer, buffSize, tests[i].key, 512);
1549         }
1550 
1551         qsort(tests, arraySize, sizeof(struct teststruct), compare_teststruct);
1552 
1553         for(i = 0; i < arraySize-1; i++) {
1554             for(j = i+1; j < arraySize; j++) {
1555                 lowerSize = ucol_getBound(tests[i].key, -1, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1556                 upperSize = ucol_getBound(tests[j].key, -1, UCOL_BOUND_UPPER, 1, upper, 512, &status);
1557                 for(k = i; k <= j; k++) {
1558                     if(strcmp((const char *)lower, (const char *)tests[k].key) > 0) {
1559                         log_err("Problem with lower! j = %i (%s vs %s)\n", k, tests[k].original, tests[i].original);
1560                     }
1561                     if(strcmp((const char *)upper, (const char *)tests[k].key) <= 0) {
1562                         log_err("Problem with upper! j = %i (%s vs %s)\n", k, tests[k].original, tests[j].original);
1563                     }
1564                 }
1565             }
1566         }
1567 
1568 
1569 #if 0
1570         for(i = 0; i < 1000; i++) {
1571             lowerRND = (rand()/(RAND_MAX/arraySize));
1572             upperRND = lowerRND + (rand()/(RAND_MAX/(arraySize-lowerRND)));
1573 
1574             lowerSize = ucol_getBound(tests[lowerRND].key, -1, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1575             upperSize = ucol_getBound(tests[upperRND].key, -1, UCOL_BOUND_UPPER_LONG, 1, upper, 512, &status);
1576 
1577             for(j = lowerRND; j<=upperRND; j++) {
1578                 if(strcmp(lower, tests[j].key) > 0) {
1579                     log_err("Problem with lower! j = %i (%s vs %s)\n", j, tests[j].original, tests[lowerRND].original);
1580                 }
1581                 if(strcmp(upper, tests[j].key) <= 0) {
1582                     log_err("Problem with upper! j = %i (%s vs %s)\n", j, tests[j].original, tests[upperRND].original);
1583                 }
1584             }
1585         }
1586 #endif
1587 
1588 
1589 
1590 
1591 
1592         for(i = 0; i<sizeof(test)/sizeof(test[0]); i++) {
1593             buffSize = u_unescape(test[i], buffer, 512);
1594             skSize = ucol_getSortKey(coll, buffer, buffSize, sortkey, 512);
1595             lowerSize = ucol_getBound(sortkey, skSize, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1596             upperSize = ucol_getBound(sortkey, skSize, UCOL_BOUND_UPPER_LONG, 1, upper, 512, &status);
1597             for(j = i+1; j<sizeof(test)/sizeof(test[0]); j++) {
1598                 buffSize = u_unescape(test[j], buffer, 512);
1599                 skSize = ucol_getSortKey(coll, buffer, buffSize, sortkey, 512);
1600                 if(strcmp((const char *)lower, (const char *)sortkey) > 0) {
1601                     log_err("Problem with lower! i = %i, j = %i (%s vs %s)\n", i, j, test[i], test[j]);
1602                 }
1603                 if(strcmp((const char *)upper, (const char *)sortkey) <= 0) {
1604                     log_err("Problem with upper! i = %i, j = %i (%s vs %s)\n", i, j, test[i], test[j]);
1605                 }
1606             }
1607         }
1608         ucol_close(coll);
1609     } else {
1610         log_data_err("Couldn't open collator\n");
1611     }
1612 
1613 }
1614 
doOverrunTest(UCollator * coll,const UChar * uString,int32_t strLen)1615 static void doOverrunTest(UCollator *coll, const UChar *uString, int32_t strLen) {
1616     int32_t skLen = 0, skLen2 = 0;
1617     uint8_t sortKey[256];
1618     int32_t i, j;
1619     uint8_t filler = 0xFF;
1620 
1621     skLen = ucol_getSortKey(coll, uString, strLen, NULL, 0);
1622 
1623     for(i = 0; i < skLen; i++) {
1624         memset(sortKey, filler, 256);
1625         skLen2 = ucol_getSortKey(coll, uString, strLen, sortKey, i);
1626         if(skLen != skLen2) {
1627             log_err("For buffer size %i, got different sortkey length. Expected %i got %i\n", i, skLen, skLen2);
1628         }
1629         for(j = i; j < 256; j++) {
1630             if(sortKey[j] != filler) {
1631                 log_err("Something run over index %i\n", j);
1632                 break;
1633             }
1634         }
1635     }
1636 }
1637 
1638 /* j1865 reports that if a shorter buffer is passed to
1639 * to get sort key, a buffer overrun happens in some
1640 * cases. This test tries to check this.
1641 */
TestSortKeyBufferOverrun(void)1642 void TestSortKeyBufferOverrun(void) {
1643     UErrorCode status = U_ZERO_ERROR;
1644     const char* cString = "A very Merry liTTle-lamB..";
1645     UChar uString[256];
1646     int32_t strLen = 0;
1647     UCollator *coll = ucol_open("root", &status);
1648     strLen = u_unescape(cString, uString, 256);
1649 
1650     if(U_SUCCESS(status)) {
1651         log_verbose("testing non ignorable\n");
1652         ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &status);
1653         doOverrunTest(coll, uString, strLen);
1654 
1655         log_verbose("testing shifted\n");
1656         ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_SHIFTED, &status);
1657         doOverrunTest(coll, uString, strLen);
1658 
1659         log_verbose("testing shifted quaternary\n");
1660         ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_QUATERNARY, &status);
1661         doOverrunTest(coll, uString, strLen);
1662 
1663         log_verbose("testing with french secondaries\n");
1664         ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_ON, &status);
1665         ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_TERTIARY, &status);
1666         ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &status);
1667         doOverrunTest(coll, uString, strLen);
1668 
1669     }
1670     ucol_close(coll);
1671 }
1672 
TestAttribute()1673 static void TestAttribute()
1674 {
1675     UErrorCode error = U_ZERO_ERROR;
1676     UCollator *coll = ucol_open(NULL, &error);
1677 
1678     if (U_FAILURE(error)) {
1679         log_err_status(error, "Creation of default collator failed\n");
1680         return;
1681     }
1682 
1683     ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_OFF, &error);
1684     if (ucol_getAttribute(coll, UCOL_FRENCH_COLLATION, &error) != UCOL_OFF ||
1685         U_FAILURE(error)) {
1686         log_err_status(error, "Setting and retrieving of the french collation failed\n");
1687     }
1688 
1689     ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_ON, &error);
1690     if (ucol_getAttribute(coll, UCOL_FRENCH_COLLATION, &error) != UCOL_ON ||
1691         U_FAILURE(error)) {
1692         log_err_status(error, "Setting and retrieving of the french collation failed\n");
1693     }
1694 
1695     ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_SHIFTED, &error);
1696     if (ucol_getAttribute(coll, UCOL_ALTERNATE_HANDLING, &error) != UCOL_SHIFTED ||
1697         U_FAILURE(error)) {
1698         log_err_status(error, "Setting and retrieving of the alternate handling failed\n");
1699     }
1700 
1701     ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &error);
1702     if (ucol_getAttribute(coll, UCOL_ALTERNATE_HANDLING, &error) != UCOL_NON_IGNORABLE ||
1703         U_FAILURE(error)) {
1704         log_err_status(error, "Setting and retrieving of the alternate handling failed\n");
1705     }
1706 
1707     ucol_setAttribute(coll, UCOL_CASE_FIRST, UCOL_LOWER_FIRST, &error);
1708     if (ucol_getAttribute(coll, UCOL_CASE_FIRST, &error) != UCOL_LOWER_FIRST ||
1709         U_FAILURE(error)) {
1710         log_err_status(error, "Setting and retrieving of the case first attribute failed\n");
1711     }
1712 
1713     ucol_setAttribute(coll, UCOL_CASE_FIRST, UCOL_UPPER_FIRST, &error);
1714     if (ucol_getAttribute(coll, UCOL_CASE_FIRST, &error) != UCOL_UPPER_FIRST ||
1715         U_FAILURE(error)) {
1716         log_err_status(error, "Setting and retrieving of the case first attribute failed\n");
1717     }
1718 
1719     ucol_setAttribute(coll, UCOL_CASE_LEVEL, UCOL_ON, &error);
1720     if (ucol_getAttribute(coll, UCOL_CASE_LEVEL, &error) != UCOL_ON ||
1721         U_FAILURE(error)) {
1722         log_err_status(error, "Setting and retrieving of the case level attribute failed\n");
1723     }
1724 
1725     ucol_setAttribute(coll, UCOL_CASE_LEVEL, UCOL_OFF, &error);
1726     if (ucol_getAttribute(coll, UCOL_CASE_LEVEL, &error) != UCOL_OFF ||
1727         U_FAILURE(error)) {
1728         log_err_status(error, "Setting and retrieving of the case level attribute failed\n");
1729     }
1730 
1731     ucol_setAttribute(coll, UCOL_NORMALIZATION_MODE, UCOL_ON, &error);
1732     if (ucol_getAttribute(coll, UCOL_NORMALIZATION_MODE, &error) != UCOL_ON ||
1733         U_FAILURE(error)) {
1734         log_err_status(error, "Setting and retrieving of the normalization on/off attribute failed\n");
1735     }
1736 
1737     ucol_setAttribute(coll, UCOL_NORMALIZATION_MODE, UCOL_OFF, &error);
1738     if (ucol_getAttribute(coll, UCOL_NORMALIZATION_MODE, &error) != UCOL_OFF ||
1739         U_FAILURE(error)) {
1740         log_err_status(error, "Setting and retrieving of the normalization on/off attribute failed\n");
1741     }
1742 
1743     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_PRIMARY, &error);
1744     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_PRIMARY ||
1745         U_FAILURE(error)) {
1746         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1747     }
1748 
1749     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_SECONDARY, &error);
1750     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_SECONDARY ||
1751         U_FAILURE(error)) {
1752         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1753     }
1754 
1755     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_TERTIARY, &error);
1756     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_TERTIARY ||
1757         U_FAILURE(error)) {
1758         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1759     }
1760 
1761     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_QUATERNARY, &error);
1762     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_QUATERNARY ||
1763         U_FAILURE(error)) {
1764         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1765     }
1766 
1767     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_IDENTICAL, &error);
1768     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_IDENTICAL ||
1769         U_FAILURE(error)) {
1770         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1771     }
1772 
1773     ucol_close(coll);
1774 }
1775 
TestGetTailoredSet()1776 void TestGetTailoredSet() {
1777   struct {
1778     const char *rules;
1779     const char *tests[20];
1780     int32_t testsize;
1781   } setTest[] = {
1782     { "&a < \\u212b", { "\\u212b", "A\\u030a", "\\u00c5" }, 3},
1783     { "& S < \\u0161 <<< \\u0160", { "\\u0161", "s\\u030C", "\\u0160", "S\\u030C" }, 4}
1784   };
1785 
1786   int32_t i = 0, j = 0;
1787   UErrorCode status = U_ZERO_ERROR;
1788   UParseError pError;
1789 
1790   UCollator *coll = NULL;
1791   UChar buff[1024];
1792   int32_t buffLen = 0;
1793   USet *set = NULL;
1794 
1795   for(i = 0; i < sizeof(setTest)/sizeof(setTest[0]); i++) {
1796     buffLen = u_unescape(setTest[i].rules, buff, 1024);
1797     coll = ucol_openRules(buff, buffLen, UCOL_DEFAULT, UCOL_DEFAULT, &pError, &status);
1798     if(U_SUCCESS(status)) {
1799       set = ucol_getTailoredSet(coll, &status);
1800       if(uset_size(set) != setTest[i].testsize) {
1801         log_err("Tailored set size different (%d) than expected (%d)\n", uset_size(set), setTest[i].testsize);
1802       }
1803       for(j = 0; j < setTest[i].testsize; j++) {
1804         buffLen = u_unescape(setTest[i].tests[j], buff, 1024);
1805         if(!uset_containsString(set, buff, buffLen)) {
1806           log_err("Tailored set doesn't contain %s... It should\n", setTest[i].tests[j]);
1807         }
1808       }
1809       uset_close(set);
1810     } else {
1811       log_err_status(status, "Couldn't open collator with rules %s\n", setTest[i].rules);
1812     }
1813     ucol_close(coll);
1814   }
1815 }
1816 
tMemCmp(const uint8_t * first,const uint8_t * second)1817 static int tMemCmp(const uint8_t *first, const uint8_t *second) {
1818    int32_t firstLen = (int32_t)strlen((const char *)first);
1819    int32_t secondLen = (int32_t)strlen((const char *)second);
1820    return memcmp(first, second, uprv_min(firstLen, secondLen));
1821 }
1822 static const char * strengthsC[] = {
1823      "UCOL_PRIMARY",
1824      "UCOL_SECONDARY",
1825      "UCOL_TERTIARY",
1826      "UCOL_QUATERNARY",
1827      "UCOL_IDENTICAL"
1828 };
1829 
TestMergeSortKeys(void)1830 void TestMergeSortKeys(void) {
1831    UErrorCode status = U_ZERO_ERROR;
1832    UCollator *coll = ucol_open("en", &status);
1833    if(U_SUCCESS(status)) {
1834 
1835      const char* cases[] = {
1836        "abc",
1837          "abcd",
1838          "abcde"
1839      };
1840      uint32_t casesSize = sizeof(cases)/sizeof(cases[0]);
1841      const char* prefix = "foo";
1842      const char* suffix = "egg";
1843      char outBuff1[256], outBuff2[256];
1844 
1845      uint8_t **sortkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
1846      uint8_t **mergedPrefixkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
1847      uint8_t **mergedSuffixkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
1848      uint32_t *sortKeysLen = (uint32_t *)malloc(casesSize*sizeof(uint32_t));
1849      uint8_t prefixKey[256], suffixKey[256];
1850      uint32_t prefixKeyLen = 0, suffixKeyLen = 0, i = 0;
1851      UChar buffer[256];
1852      uint32_t unescapedLen = 0, l1 = 0, l2 = 0;
1853      UColAttributeValue strength;
1854 
1855      log_verbose("ucol_mergeSortkeys test\n");
1856      log_verbose("Testing order of the test cases\n");
1857      genericLocaleStarter("en", cases, casesSize);
1858 
1859      for(i = 0; i<casesSize; i++) {
1860        sortkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
1861        mergedPrefixkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
1862        mergedSuffixkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
1863      }
1864 
1865      unescapedLen = u_unescape(prefix, buffer, 256);
1866      prefixKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, prefixKey, 256);
1867 
1868      unescapedLen = u_unescape(suffix, buffer, 256);
1869      suffixKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, suffixKey, 256);
1870 
1871      log_verbose("Massaging data with prefixes and different strengths\n");
1872      strength = UCOL_PRIMARY;
1873      while(strength <= UCOL_IDENTICAL) {
1874        log_verbose("Strength %s\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
1875        ucol_setAttribute(coll, UCOL_STRENGTH, strength, &status);
1876        for(i = 0; i<casesSize; i++) {
1877          unescapedLen = u_unescape(cases[i], buffer, 256);
1878          sortKeysLen[i] = ucol_getSortKey(coll, buffer, unescapedLen, sortkeys[i], 256);
1879          ucol_mergeSortkeys(prefixKey, prefixKeyLen, sortkeys[i], sortKeysLen[i], mergedPrefixkeys[i], 256);
1880          ucol_mergeSortkeys(sortkeys[i], sortKeysLen[i], suffixKey, suffixKeyLen, mergedSuffixkeys[i], 256);
1881          if(i>0) {
1882            if(tMemCmp(mergedPrefixkeys[i-1], mergedPrefixkeys[i]) >= 0) {
1883              log_err("Error while comparing prefixed keys @ strength %s:\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
1884              log_err("%s\n%s\n",
1885                          ucol_sortKeyToString(coll, mergedPrefixkeys[i-1], outBuff1, &l1),
1886                          ucol_sortKeyToString(coll, mergedPrefixkeys[i], outBuff2, &l2));
1887            }
1888            if(tMemCmp(mergedSuffixkeys[i-1], mergedSuffixkeys[i]) >= 0) {
1889              log_err("Error while comparing suffixed keys @ strength %s:\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
1890              log_err("%s\n%s\n",
1891                          ucol_sortKeyToString(coll, mergedSuffixkeys[i-1], outBuff1, &l1),
1892                          ucol_sortKeyToString(coll, mergedSuffixkeys[i], outBuff2, &l2));
1893            }
1894          }
1895        }
1896        if(strength == UCOL_QUATERNARY) {
1897          strength = UCOL_IDENTICAL;
1898        } else {
1899          strength++;
1900        }
1901      }
1902 
1903      {
1904        uint8_t smallBuf[3];
1905        uint32_t reqLen = 0;
1906        log_verbose("testing buffer overflow\n");
1907        reqLen = ucol_mergeSortkeys(prefixKey, prefixKeyLen, suffixKey, suffixKeyLen, smallBuf, 3);
1908        if(reqLen != (prefixKeyLen+suffixKeyLen-1)) {
1909          log_err("Wrong preflight size for merged sortkey\n");
1910        }
1911      }
1912 
1913      {
1914        UChar empty = 0;
1915        uint8_t emptyKey[20], abcKey[50], mergedKey[100];
1916        int32_t emptyKeyLen = 0, abcKeyLen = 0, mergedKeyLen = 0;
1917 
1918        log_verbose("testing merging with sortkeys generated for empty strings\n");
1919        emptyKeyLen = ucol_getSortKey(coll, &empty, 0, emptyKey, 20);
1920        unescapedLen = u_unescape(cases[0], buffer, 256);
1921        abcKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, abcKey, 50);
1922        mergedKeyLen = ucol_mergeSortkeys(emptyKey, emptyKeyLen, abcKey, abcKeyLen, mergedKey, 100);
1923        if(mergedKey[0] != 2) {
1924          log_err("Empty sortkey didn't produce a level separator\n");
1925        }
1926        /* try with zeros */
1927        mergedKeyLen = ucol_mergeSortkeys(emptyKey, 0, abcKey, abcKeyLen, mergedKey, 100);
1928        if(mergedKeyLen != 0 || mergedKey[0] != 0) {
1929          log_err("Empty key didn't produce null mergedKey\n");
1930        }
1931        mergedKeyLen = ucol_mergeSortkeys(abcKey, abcKeyLen, emptyKey, 0, mergedKey, 100);
1932        if(mergedKeyLen != 0 || mergedKey[0] != 0) {
1933          log_err("Empty key didn't produce null mergedKey\n");
1934        }
1935 
1936      }
1937 
1938      for(i = 0; i<casesSize; i++) {
1939        free(sortkeys[i]);
1940        free(mergedPrefixkeys[i]);
1941        free(mergedSuffixkeys[i]);
1942      }
1943      free(sortkeys);
1944      free(mergedPrefixkeys);
1945      free(mergedSuffixkeys);
1946      free(sortKeysLen);
1947      ucol_close(coll);
1948      /* need to finish this up */
1949    } else {
1950      log_data_err("Couldn't open collator");
1951    }
1952 }
TestShortString(void)1953 static void TestShortString(void)
1954 {
1955     struct {
1956         const char *input;
1957         const char *expectedOutput;
1958         const char *locale;
1959         UErrorCode expectedStatus;
1960         int32_t    expectedOffset;
1961         uint32_t   expectedIdentifier;
1962     } testCases[] = {
1963         /*
1964          * The following expectedOutput contains a collation weight (2700 from UCA 6.0)
1965          * which is the primary weight for the T character (U+0041) in the input.
1966          * When that character gets a different weight in FractionalUCA.txt,
1967          * the expectedOutput needs to be adjusted.
1968          * That is, when we upgrade to a new UCA version or change collation
1969          * in such a way that the absolute weight for 'A' changes,
1970          * we will get a test failure here and need to adjust the test case.
1971          */
1972         {"LDE_RDE_KPHONEBOOK_T0041_ZLATN","B2700_KPHONEBOOK_LDE", "de@collation=phonebook", U_USING_FALLBACK_WARNING, 0, 0 },
1973 
1974         {"LEN_RUS_NO_AS_S4","AS_LROOT_NO_S4", NULL, U_USING_DEFAULT_WARNING, 0, 0 },
1975         {"LDE_VPHONEBOOK_EO_SI","EO_KPHONEBOOK_LDE_SI", "de@collation=phonebook", U_ZERO_ERROR, 0, 0 },
1976         {"LDE_Kphonebook","KPHONEBOOK_LDE", "de@collation=phonebook", U_ZERO_ERROR, 0, 0 },
1977         {"Xqde_DE@collation=phonebookq_S3_EX","KPHONEBOOK_LDE", "de@collation=phonebook", U_USING_FALLBACK_WARNING, 0, 0 },
1978         {"LFR_FO", "FO_LROOT", NULL, U_USING_DEFAULT_WARNING, 0, 0 },
1979         {"SO_LX_AS", "", NULL, U_ILLEGAL_ARGUMENT_ERROR, 8, 0 },
1980         {"S3_ASS_MMM", "", NULL, U_ILLEGAL_ARGUMENT_ERROR, 5, 0 }
1981     };
1982 
1983     int32_t i = 0;
1984     UCollator *coll = NULL, *fromNormalized = NULL;
1985     UParseError parseError;
1986     UErrorCode status = U_ZERO_ERROR;
1987     char fromShortBuffer[256], normalizedBuffer[256], fromNormalizedBuffer[256];
1988     const char* locale = NULL;
1989 
1990 
1991     for(i = 0; i < sizeof(testCases)/sizeof(testCases[0]); i++) {
1992         status = U_ZERO_ERROR;
1993         if(testCases[i].locale) {
1994             locale = testCases[i].locale;
1995         } else {
1996             locale = NULL;
1997         }
1998 
1999         coll = ucol_openFromShortString(testCases[i].input, FALSE, &parseError, &status);
2000         if(status != testCases[i].expectedStatus) {
2001             log_err_status(status, "Got status '%s' that is different from expected '%s' for '%s'\n",
2002                 u_errorName(status), u_errorName(testCases[i].expectedStatus), testCases[i].input);
2003             continue;
2004         }
2005 
2006         if(U_SUCCESS(status)) {
2007             ucol_getShortDefinitionString(coll, locale, fromShortBuffer, 256, &status);
2008 
2009             if(strcmp(fromShortBuffer, testCases[i].expectedOutput)) {
2010                 log_err("Got short string '%s' from the collator. Expected '%s' for input '%s'\n",
2011                     fromShortBuffer, testCases[i].expectedOutput, testCases[i].input);
2012             }
2013 
2014             ucol_normalizeShortDefinitionString(testCases[i].input, normalizedBuffer, 256, &parseError, &status);
2015             fromNormalized = ucol_openFromShortString(normalizedBuffer, FALSE, &parseError, &status);
2016             ucol_getShortDefinitionString(fromNormalized, locale, fromNormalizedBuffer, 256, &status);
2017 
2018             if(strcmp(fromShortBuffer, fromNormalizedBuffer)) {
2019                 log_err("Strings obtained from collators instantiated by short string ('%s') and from normalized string ('%s') differ\n",
2020                     fromShortBuffer, fromNormalizedBuffer);
2021             }
2022 
2023 
2024             if(!ucol_equals(coll, fromNormalized)) {
2025                 log_err("Collator from short string ('%s') differs from one obtained through a normalized version ('%s')\n",
2026                     testCases[i].input, normalizedBuffer);
2027             }
2028 
2029             ucol_close(fromNormalized);
2030             ucol_close(coll);
2031 
2032         } else {
2033             if(parseError.offset != testCases[i].expectedOffset) {
2034                 log_err("Got parse error offset %i, but expected %i instead for '%s'\n",
2035                     parseError.offset, testCases[i].expectedOffset, testCases[i].input);
2036             }
2037         }
2038     }
2039 
2040 }
2041 
2042 static void
doSetsTest(const char * locale,const USet * ref,USet * set,const char * inSet,const char * outSet,UErrorCode * status)2043 doSetsTest(const char *locale, const USet *ref, USet *set, const char* inSet, const char* outSet, UErrorCode *status) {
2044     UChar buffer[512];
2045     int32_t bufLen;
2046 
2047     uset_clear(set);
2048     bufLen = u_unescape(inSet, buffer, 512);
2049     uset_applyPattern(set, buffer, bufLen, 0, status);
2050     if(U_FAILURE(*status)) {
2051         log_err("%s: Failure setting pattern %s\n", locale, u_errorName(*status));
2052     }
2053 
2054     if(!uset_containsAll(ref, set)) {
2055         log_err("%s: Some stuff from %s is not present in the set\n", locale, inSet);
2056     }
2057 
2058     uset_clear(set);
2059     bufLen = u_unescape(outSet, buffer, 512);
2060     uset_applyPattern(set, buffer, bufLen, 0, status);
2061     if(U_FAILURE(*status)) {
2062         log_err("%s: Failure setting pattern %s\n", locale, u_errorName(*status));
2063     }
2064 
2065     if(!uset_containsNone(ref, set)) {
2066         log_err("%s: Some stuff from %s is present in the set\n", locale, outSet);
2067     }
2068 }
2069 
2070 
2071 
2072 
2073 static void
TestGetContractionsAndUnsafes(void)2074 TestGetContractionsAndUnsafes(void)
2075 {
2076     static struct {
2077         const char* locale;
2078         const char* inConts;
2079         const char* outConts;
2080         const char* inExp;
2081         const char* outExp;
2082         const char* unsafeCodeUnits;
2083         const char* safeCodeUnits;
2084     } tests[] = {
2085         { "ru",
2086             "[{\\u0418\\u0306}{\\u0438\\u0306}]",
2087             "[\\u0439\\u0457]",
2088             "[\\u00e6]",
2089             "[ae]",
2090             "[\\u0418\\u0438]",
2091             "[aAbB\\u0430\\u0410\\u0433\\u0413]"
2092         },
2093         { "uk",
2094             "[{\\u0406\\u0308}{\\u0456\\u0308}{\\u0418\\u0306}{\\u0438\\u0306}]",
2095             "[\\u0407\\u0419\\u0439\\u0457]",
2096             "[\\u00e6]",
2097             "[ae]",
2098             "[\\u0406\\u0456\\u0418\\u0438]",
2099             "[aAbBxv]",
2100         },
2101         { "sh",
2102             "[{C\\u0301}{C\\u030C}{C\\u0341}{DZ\\u030C}{Dz\\u030C}{D\\u017D}{D\\u017E}{lj}{nj}]",
2103             "[{\\u309d\\u3099}{\\u30fd\\u3099}]",
2104             "[\\u00e6]",
2105             "[a]",
2106             "[nlcdzNLCDZ]",
2107             "[jabv]"
2108         },
2109         { "ja",
2110           "[{\\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}]",
2111           "[{\\u30FD\\u3099}{\\u309D\\u3099}{\\u3053\\u3099}{\\u30B3\\u3099}{lj}{nj}]",
2112             "[\\u30FE\\u00e6]",
2113             "[a]",
2114             "[\\u3099]",
2115             "[]"
2116         }
2117     };
2118 
2119 
2120 
2121 
2122     UErrorCode status = U_ZERO_ERROR;
2123     UCollator *coll = NULL;
2124     int32_t i = 0;
2125     int32_t noConts = 0;
2126     USet *conts = uset_open(0,0);
2127     USet *exp = uset_open(0, 0);
2128     USet *set  = uset_open(0,0);
2129     int32_t setBufferLen = 65536;
2130     UChar buffer[65536];
2131     int32_t setLen = 0;
2132 
2133     for(i = 0; i < sizeof(tests)/sizeof(tests[0]); i++) {
2134         log_verbose("Testing locale: %s\n", tests[i].locale);
2135         coll = ucol_open(tests[i].locale, &status);
2136         if (coll == NULL || U_FAILURE(status)) {
2137             log_err_status(status, "Unable to open collator for locale %s ==> %s\n", tests[i].locale, u_errorName(status));
2138             continue;
2139         }
2140         ucol_getContractionsAndExpansions(coll, conts, exp, TRUE, &status);
2141         doSetsTest(tests[i].locale, conts, set, tests[i].inConts, tests[i].outConts, &status);
2142         setLen = uset_toPattern(conts, buffer, setBufferLen, TRUE, &status);
2143         if(U_SUCCESS(status)) {
2144             /*log_verbose("Contractions %i: %s\n", uset_getItemCount(conts), aescstrdup(buffer, setLen));*/
2145         } else {
2146             log_err("error %s. %i\n", u_errorName(status), setLen);
2147             status = U_ZERO_ERROR;
2148         }
2149         doSetsTest(tests[i].locale, exp, set, tests[i].inExp, tests[i].outExp, &status);
2150         setLen = uset_toPattern(exp, buffer, setBufferLen, TRUE, &status);
2151         if(U_SUCCESS(status)) {
2152             /*log_verbose("Expansions %i: %s\n", uset_getItemCount(exp), aescstrdup(buffer, setLen));*/
2153         } else {
2154             log_err("error %s. %i\n", u_errorName(status), setLen);
2155             status = U_ZERO_ERROR;
2156         }
2157 
2158         noConts = ucol_getUnsafeSet(coll, conts, &status);
2159         doSetsTest(tests[i].locale, conts, set, tests[i].unsafeCodeUnits, tests[i].safeCodeUnits, &status);
2160         setLen = uset_toPattern(conts, buffer, setBufferLen, TRUE, &status);
2161         if(U_SUCCESS(status)) {
2162             log_verbose("Unsafe %i: %s\n", uset_getItemCount(exp), aescstrdup(buffer, setLen));
2163         } else {
2164             log_err("error %s. %i\n", u_errorName(status), setLen);
2165             status = U_ZERO_ERROR;
2166         }
2167 
2168         ucol_close(coll);
2169     }
2170 
2171 
2172     uset_close(conts);
2173     uset_close(exp);
2174     uset_close(set);
2175 }
2176 
2177 static void
TestOpenBinary(void)2178 TestOpenBinary(void)
2179 {
2180     UErrorCode status = U_ZERO_ERROR;
2181     /*
2182     char rule[] = "&h < d < c < b";
2183     char *wUCA[] = { "a", "h", "d", "c", "b", "i" };
2184     char *noUCA[] = {"d", "c", "b", "a", "h", "i" };
2185     */
2186     /* we have to use Cyrillic letters because latin-1 always gets copied */
2187     const char rule[] = "&\\u0452 < \\u0434 < \\u0433 < \\u0432"; /* &dje < d < g < v */
2188     const char *wUCA[] = { "\\u0430", "\\u0452", "\\u0434", "\\u0433", "\\u0432", "\\u0435" }; /* a, dje, d, g, v, e */
2189     const char *noUCA[] = {"\\u0434", "\\u0433", "\\u0432", "\\u0430", "\\u0435", "\\u0452" }; /* d, g, v, a, e, dje */
2190 
2191     UChar uRules[256];
2192     int32_t uRulesLen = u_unescape(rule, uRules, 256);
2193 
2194     UCollator *coll = ucol_openRules(uRules, uRulesLen, UCOL_DEFAULT, UCOL_DEFAULT, NULL, &status);
2195     UCollator *UCA = NULL;
2196     UCollator *cloneNOUCA = NULL, *cloneWUCA = NULL;
2197 
2198     uint8_t imageBuffer[32768];
2199     uint8_t *image = imageBuffer;
2200     int32_t imageBufferCapacity = 32768;
2201 
2202     int32_t imageSize;
2203 
2204     if((coll==NULL)||(U_FAILURE(status))) {
2205         log_data_err("could not load collators or error occured: %s\n",
2206             u_errorName(status));
2207         return;
2208     }
2209     UCA = ucol_open("root", &status);
2210     if((UCA==NULL)||(U_FAILURE(status))) {
2211         log_data_err("could not load UCA collator or error occured: %s\n",
2212             u_errorName(status));
2213         return;
2214     }
2215     imageSize = ucol_cloneBinary(coll, image, imageBufferCapacity, &status);
2216     if(U_FAILURE(status)) {
2217         image = (uint8_t *)malloc(imageSize*sizeof(uint8_t));
2218         status = U_ZERO_ERROR;
2219         imageSize = ucol_cloneBinary(coll, imageBuffer, imageSize, &status);
2220     }
2221 
2222 
2223     cloneWUCA = ucol_openBinary(image, imageSize, UCA, &status);
2224     cloneNOUCA = ucol_openBinary(image, imageSize, NULL, &status);
2225 
2226     genericOrderingTest(coll, wUCA, sizeof(wUCA)/sizeof(wUCA[0]));
2227 
2228     genericOrderingTest(cloneWUCA, wUCA, sizeof(wUCA)/sizeof(wUCA[0]));
2229     genericOrderingTest(cloneNOUCA, noUCA, sizeof(noUCA)/sizeof(noUCA[0]));
2230 
2231     if(image != imageBuffer) {
2232         free(image);
2233     }
2234     ucol_close(coll);
2235     ucol_close(cloneNOUCA);
2236     ucol_close(cloneWUCA);
2237     ucol_close(UCA);
2238 }
2239 
TestDefault(void)2240 static void TestDefault(void) {
2241     /* Tests for code coverage. */
2242     UErrorCode status = U_ZERO_ERROR;
2243     UCollator *coll = ucol_open("es@collation=pinyin", &status);
2244     if (coll == NULL || status == U_FILE_ACCESS_ERROR) {
2245         log_data_err("Unable to open collator es@collation=pinyin\n");
2246         return;
2247     }
2248     if (status != U_USING_DEFAULT_WARNING) {
2249         /* What do you mean that you know about using pinyin collation in Spanish!? This should be in the zh locale. */
2250         log_err("es@collation=pinyin should return U_USING_DEFAULT_WARNING, but returned %s\n", u_errorName(status));
2251     }
2252     ucol_close(coll);
2253     if (ucol_getKeywordValues("funky", &status) != NULL) {
2254         log_err("Collators should not know about the funky keyword.\n");
2255     }
2256     if (status != U_ILLEGAL_ARGUMENT_ERROR) {
2257         log_err("funky keyword didn't fail as expected %s\n", u_errorName(status));
2258     }
2259     if (ucol_getKeywordValues("collation", &status) != NULL) {
2260         log_err("ucol_getKeywordValues should not work when given a bad status.\n");
2261     }
2262 }
2263 
TestDefaultKeyword(void)2264 static void TestDefaultKeyword(void) {
2265     /* Tests for code coverage. */
2266     UErrorCode status = U_ZERO_ERROR;
2267     const char *loc = "zh_TW@collation=default";
2268     UCollator *coll = ucol_open(loc, &status);
2269     if(U_FAILURE(status)) {
2270         log_info("Warning: ucol_open(%s, ...) returned %s, at least it didn't crash.\n", loc, u_errorName(status));
2271     } else if (status != U_USING_FALLBACK_WARNING) {
2272         /* Hmm, skip the following test for CLDR 1.9 data and/or ICU 4.6, no longer seems to apply */
2273         #if 0
2274         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));
2275         #endif
2276     }
2277     ucol_close(coll);
2278 }
2279 
TestGetKeywordValuesForLocale(void)2280 static void TestGetKeywordValuesForLocale(void) {
2281 #define INCLUDE_UNIHAN_COLLATION 0
2282 #define PREFERRED_SIZE 16
2283 #define MAX_NUMBER_OF_KEYWORDS 8
2284     const char *PREFERRED[PREFERRED_SIZE][MAX_NUMBER_OF_KEYWORDS+1] = {
2285             { "und",            "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2286             { "en_US",          "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2287             { "en_029",         "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2288             { "de_DE",          "standard", "phonebook", "search", "ducet", NULL, NULL, NULL, NULL },
2289             { "de_Latn_DE",     "standard", "phonebook", "search", "ducet", NULL, NULL, NULL, NULL },
2290 #if INCLUDE_UNIHAN_COLLATION
2291             { "zh",             "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2292             { "zh_Hans",        "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2293             { "zh_CN",          "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2294             { "zh_Hant",        "stroke", "big5han", "gb2312han", "pinyin", "standard", "unihan", "ducet", "search" },
2295             { "zh_TW",          "stroke", "big5han", "gb2312han", "pinyin", "standard", "unihan", "ducet", "search" },
2296             { "zh__PINYIN",     "pinyin", "big5han", "gb2312han", "standard", "stroke", "unihan", "ducet", "search" },
2297 #else
2298             { "zh",             "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2299             { "zh_Hans",        "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2300             { "zh_CN",          "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2301             { "zh_Hant",        "stroke", "big5han", "gb2312han", "pinyin", "standard", "ducet", "search", NULL },
2302             { "zh_TW",          "stroke", "big5han", "gb2312han", "pinyin", "standard", "ducet", "search", NULL },
2303             { "zh__PINYIN",     "pinyin", "big5han", "gb2312han", "standard", "stroke", "ducet", "search", NULL },
2304 #endif
2305             { "es_ES",          "standard", "search", "traditional", "ducet", NULL, NULL, NULL, NULL },
2306             { "es__TRADITIONAL","traditional", "search", "standard", "ducet", NULL, NULL, NULL, NULL },
2307             { "und@collation=phonebook",    "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL },
2308             { "de_DE@collation=big5han",    "standard", "phonebook", "search", "ducet", NULL, NULL, NULL, NULL },
2309             { "zzz@collation=xxx",          "standard", "ducet", "search", NULL, NULL, NULL, NULL, NULL }
2310     };
2311 #if INCLUDE_UNIHAN_COLLATION
2312     const int32_t expectedLength[PREFERRED_SIZE] = { 3, 3, 3, 4, 4, 8, 8, 8, 8, 8, 8, 4, 4, 3, 4, 3 };
2313 #else
2314     const int32_t expectedLength[PREFERRED_SIZE] = { 3, 3, 3, 4, 4, 7, 7, 7, 7, 7, 7, 4, 4, 3, 4, 3 };
2315 #endif
2316 
2317     UErrorCode status = U_ZERO_ERROR;
2318     UEnumeration *keywordValues = NULL;
2319     int32_t i, n, size, valueLength;
2320     const char *locale = NULL, *value = NULL;
2321     UBool errorOccurred = FALSE;
2322 
2323     for (i = 0; i < PREFERRED_SIZE; i++) {
2324         locale = PREFERRED[i][0];
2325         value = NULL;
2326         valueLength = 0;
2327         size = 0;
2328 
2329         keywordValues = ucol_getKeywordValuesForLocale("collation", locale, TRUE, &status);
2330         if (keywordValues == NULL || U_FAILURE(status)) {
2331             log_err_status(status, "Error getting keyword values: %s\n", u_errorName(status));
2332             break;
2333         }
2334         size = uenum_count(keywordValues, &status);
2335 
2336         if (size == expectedLength[i]) {
2337             for (n = 0; n < expectedLength[i]; n++) {
2338                 if ((value = uenum_next(keywordValues, &valueLength, &status)) != NULL && U_SUCCESS(status)) {
2339                     if (uprv_strcmp(value, PREFERRED[i][n+1]) != 0) {
2340                         log_err("Keyword values differ: Got [%s] Expected [%s] for locale: %s\n", value, PREFERRED[i][n+1], locale);
2341                         errorOccurred = TRUE;
2342                         break;
2343                     }
2344 
2345                 } else {
2346                     log_err("While getting keyword value from locale: %s got this error: %s\n", locale, u_errorName(status));
2347                     errorOccurred = TRUE;
2348                     break;
2349                 }
2350             }
2351             if (errorOccurred) {
2352                 break;
2353             }
2354         } else {
2355             log_err("Number of keywords (%d) does not match expected size (%d) for locale: %s\n", size, expectedLength[i], locale);
2356             break;
2357         }
2358         uenum_close(keywordValues);
2359         keywordValues = NULL;
2360     }
2361     if (keywordValues != NULL) {
2362         uenum_close(keywordValues);
2363     }
2364 }
2365 
2366 
2367 #endif /* #if !UCONFIG_NO_COLLATION */
2368