• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /********************************************************************
4  * Copyright (c) 1997-2016, International Business Machines
5  * Corporation and others. All Rights Reserved.
6  ********************************************************************/
7 /*****************************************************************************
8 *
9 * File CAPITEST.C
10 *
11 * Modification History:
12 *        Name                     Description
13 *     Madhu Katragadda             Ported for C API
14 *     Brian Rower                  Added TestOpenVsOpenRules
15 ******************************************************************************
16 *//* C API TEST For COLLATOR */
17 
18 #include "unicode/utypes.h"
19 
20 #if !UCONFIG_NO_COLLATION
21 
22 #include <stdio.h>
23 #include <stdlib.h>
24 #include <string.h>
25 #include "unicode/uloc.h"
26 #include "unicode/ulocdata.h"
27 #include "unicode/ustring.h"
28 #include "unicode/ures.h"
29 #include "unicode/ucoleitr.h"
30 #include "cintltst.h"
31 #include "capitst.h"
32 #include "ccolltst.h"
33 #include "putilimp.h"
34 #include "cmemory.h"
35 #include "cstring.h"
36 #include "ucol_imp.h"
37 
38 static void TestAttribute(void);
39 static void TestDefault(void);
40 static void TestDefaultKeyword(void);
41 static void TestBengaliSortKey(void);
42 
43 
ucol_sortKeyToString(const UCollator * coll,const uint8_t * sortkey,char * buffer,uint32_t len)44 static char* U_EXPORT2 ucol_sortKeyToString(const UCollator *coll, const uint8_t *sortkey, char *buffer, uint32_t len) {
45     (void)coll; // suppress compiler warnings about unused variable
46     uint32_t position = 0;
47     uint8_t b;
48 
49     if (position + 1 < len)
50         position += sprintf(buffer + position, "[");
51     while ((b = *sortkey++) != 0) {
52         if (b == 1 && position + 5 < len) {
53             position += sprintf(buffer + position, "%02X . ", b);
54         } else if (b != 1 && position + 3 < len) {
55             position += sprintf(buffer + position, "%02X ", b);
56         }
57     }
58     if (position + 3 < len)
59         position += sprintf(buffer + position, "%02X]", b);
60     return buffer;
61 }
62 
addCollAPITest(TestNode ** root)63 void addCollAPITest(TestNode** root)
64 {
65     /* WEIVTODO: return tests here */
66     addTest(root, &TestProperty,      "tscoll/capitst/TestProperty");
67     addTest(root, &TestRuleBasedColl, "tscoll/capitst/TestRuleBasedColl");
68     addTest(root, &TestCompare,       "tscoll/capitst/TestCompare");
69     addTest(root, &TestSortKey,       "tscoll/capitst/TestSortKey");
70     addTest(root, &TestHashCode,      "tscoll/capitst/TestHashCode");
71     addTest(root, &TestElemIter,      "tscoll/capitst/TestElemIter");
72     addTest(root, &TestGetAll,        "tscoll/capitst/TestGetAll");
73     /*addTest(root, &TestGetDefaultRules, "tscoll/capitst/TestGetDefaultRules");*/
74     addTest(root, &TestDecomposition, "tscoll/capitst/TestDecomposition");
75     addTest(root, &TestSafeClone, "tscoll/capitst/TestSafeClone");
76     addTest(root, &TestClone, "tscoll/capitst/TestClone");
77     addTest(root, &TestCloneBinary, "tscoll/capitst/TestCloneBinary");
78     addTest(root, &TestGetSetAttr, "tscoll/capitst/TestGetSetAttr");
79     addTest(root, &TestBounds, "tscoll/capitst/TestBounds");
80     addTest(root, &TestGetLocale, "tscoll/capitst/TestGetLocale");
81     addTest(root, &TestSortKeyBufferOverrun, "tscoll/capitst/TestSortKeyBufferOverrun");
82     addTest(root, &TestAttribute, "tscoll/capitst/TestAttribute");
83     addTest(root, &TestGetTailoredSet, "tscoll/capitst/TestGetTailoredSet");
84     addTest(root, &TestMergeSortKeys, "tscoll/capitst/TestMergeSortKeys");
85     addTest(root, &TestShortString, "tscoll/capitst/TestShortString");
86     addTest(root, &TestGetContractionsAndUnsafes, "tscoll/capitst/TestGetContractionsAndUnsafes");
87     addTest(root, &TestOpenBinary, "tscoll/capitst/TestOpenBinary");
88     addTest(root, &TestDefault, "tscoll/capitst/TestDefault");
89     addTest(root, &TestDefaultKeyword, "tscoll/capitst/TestDefaultKeyword");
90     addTest(root, &TestOpenVsOpenRules, "tscoll/capitst/TestOpenVsOpenRules");
91     addTest(root, &TestBengaliSortKey, "tscoll/capitst/TestBengaliSortKey");
92     addTest(root, &TestGetKeywordValuesForLocale, "tscoll/capitst/TestGetKeywordValuesForLocale");
93     addTest(root, &TestStrcollNull, "tscoll/capitst/TestStrcollNull");
94     addTest(root, &TestLocaleIDWithUnderscoreAndExtension, "tscoll/capitst/TestLocaleIDWithUnderscoreAndExtension");
95 }
96 
TestGetSetAttr(void)97 void TestGetSetAttr(void) {
98   UErrorCode status = U_ZERO_ERROR;
99   UCollator *coll = ucol_open(NULL, &status);
100   struct attrTest {
101     UColAttribute att;
102     UColAttributeValue val[5];
103     uint32_t valueSize;
104     UColAttributeValue nonValue;
105   } attrs[] = {
106     {UCOL_FRENCH_COLLATION, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},
107     {UCOL_ALTERNATE_HANDLING, {UCOL_NON_IGNORABLE, UCOL_SHIFTED}, 2, UCOL_OFF},/* attribute for handling variable elements*/
108     {UCOL_CASE_FIRST, {UCOL_OFF, UCOL_LOWER_FIRST, UCOL_UPPER_FIRST}, 3, UCOL_SHIFTED},/* who goes first, lower case or uppercase */
109     {UCOL_CASE_LEVEL, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* do we have an extra case level */
110     {UCOL_NORMALIZATION_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* attribute for normalization */
111     {UCOL_DECOMPOSITION_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},
112     {UCOL_STRENGTH,         {UCOL_PRIMARY, UCOL_SECONDARY, UCOL_TERTIARY, UCOL_QUATERNARY, UCOL_IDENTICAL}, 5, UCOL_SHIFTED},/* attribute for strength */
113     {UCOL_HIRAGANA_QUATERNARY_MODE, {UCOL_ON, UCOL_OFF}, 2, UCOL_SHIFTED},/* when turned on, this attribute */
114   };
115   UColAttribute currAttr;
116   UColAttributeValue value;
117   uint32_t i = 0, j = 0;
118 
119   if (coll == NULL) {
120     log_err_status(status, "Unable to open collator. %s\n", u_errorName(status));
121     return;
122   }
123   for(i = 0; i<UPRV_LENGTHOF(attrs); i++) {
124     currAttr = attrs[i].att;
125     ucol_setAttribute(coll, currAttr, UCOL_DEFAULT, &status);
126     if(U_FAILURE(status)) {
127       log_err_status(status, "ucol_setAttribute with the default value returned error: %s\n", u_errorName(status));
128       break;
129     }
130     value = ucol_getAttribute(coll, currAttr, &status);
131     if(U_FAILURE(status)) {
132       log_err("ucol_getAttribute returned error: %s\n", u_errorName(status));
133       break;
134     }
135     for(j = 0; j<attrs[i].valueSize; j++) {
136       ucol_setAttribute(coll, currAttr, attrs[i].val[j], &status);
137       if(U_FAILURE(status)) {
138         log_err("ucol_setAttribute with the value %i returned error: %s\n", attrs[i].val[j], u_errorName(status));
139         break;
140       }
141     }
142     status = U_ZERO_ERROR;
143     ucol_setAttribute(coll, currAttr, attrs[i].nonValue, &status);
144     if(U_SUCCESS(status)) {
145       log_err("ucol_setAttribute with the bad value didn't return an error\n");
146       break;
147     }
148     status = U_ZERO_ERROR;
149 
150     ucol_setAttribute(coll, currAttr, value, &status);
151     if(U_FAILURE(status)) {
152       log_err("ucol_setAttribute with the default valuereturned error: %s\n", u_errorName(status));
153       break;
154     }
155   }
156   status = U_ZERO_ERROR;
157   value = ucol_getAttribute(coll, UCOL_ATTRIBUTE_COUNT, &status);
158   if(U_SUCCESS(status)) {
159     log_err("ucol_getAttribute for UCOL_ATTRIBUTE_COUNT didn't return an error\n");
160   }
161   status = U_ZERO_ERROR;
162   ucol_setAttribute(coll, UCOL_ATTRIBUTE_COUNT, UCOL_DEFAULT, &status);
163   if(U_SUCCESS(status)) {
164     log_err("ucol_setAttribute for UCOL_ATTRIBUTE_COUNT didn't return an error\n");
165   }
166   status = U_ZERO_ERROR;
167   ucol_close(coll);
168 }
169 
170 
doAssert(int condition,const char * message)171 static void doAssert(int condition, const char *message)
172 {
173     if (condition==0) {
174         log_err("ERROR :  %s\n", message);
175     }
176 }
177 
178 #define UTF8_BUF_SIZE 128
179 
doStrcoll(const UCollator * coll,const UChar * src,int32_t srcLen,const UChar * tgt,int32_t tgtLen,UCollationResult expected,const char * message)180 static void doStrcoll(const UCollator* coll, const UChar* src, int32_t srcLen, const UChar* tgt, int32_t tgtLen,
181                     UCollationResult expected, const char *message) {
182     UErrorCode err = U_ZERO_ERROR;
183     char srcU8[UTF8_BUF_SIZE], tgtU8[UTF8_BUF_SIZE];
184     int32_t srcU8Len = -1, tgtU8Len = -1;
185     int32_t len = 0;
186 
187     if (ucol_strcoll(coll, src, srcLen, tgt, tgtLen) != expected) {
188         log_err("ERROR :  %s\n", message);
189     }
190 
191     u_strToUTF8(srcU8, UTF8_BUF_SIZE, &len, src, srcLen, &err);
192     if (U_FAILURE(err) || len >= UTF8_BUF_SIZE) {
193         log_err("ERROR : UTF-8 conversion error\n");
194         return;
195     }
196     if (srcLen >= 0) {
197         srcU8Len = len;
198     }
199     u_strToUTF8(tgtU8, UTF8_BUF_SIZE, &len, tgt, tgtLen, &err);
200     if (U_FAILURE(err) || len >= UTF8_BUF_SIZE) {
201         log_err("ERROR : UTF-8 conversion error\n");
202         return;
203     }
204     if (tgtLen >= 0) {
205         tgtU8Len = len;
206     }
207 
208     if (ucol_strcollUTF8(coll, srcU8, srcU8Len, tgtU8, tgtU8Len, &err) != expected
209         || U_FAILURE(err)) {
210         log_err("ERROR: %s (strcollUTF8)\n", message);
211     }
212 }
213 
214 #if 0
215 /* We don't have default rules, at least not in the previous sense */
216 void TestGetDefaultRules(){
217     uint32_t size=0;
218     UErrorCode status=U_ZERO_ERROR;
219     UCollator *coll=NULL;
220     int32_t len1 = 0, len2=0;
221     uint8_t *binColData = NULL;
222 
223     UResourceBundle *res = NULL;
224     UResourceBundle *binColl = NULL;
225     uint8_t *binResult = NULL;
226 
227 
228     const UChar * defaultRulesArray=ucol_getDefaultRulesArray(&size);
229     log_verbose("Test the function ucol_getDefaultRulesArray()\n");
230 
231     coll = ucol_openRules(defaultRulesArray, size, UCOL_ON, UCOL_PRIMARY, &status);
232     if(U_SUCCESS(status) && coll !=NULL) {
233         binColData = (uint8_t*)ucol_cloneRuleData(coll, &len1, &status);
234 
235     }
236 
237 
238     status=U_ZERO_ERROR;
239     res=ures_open(NULL, "root", &status);
240     if(U_FAILURE(status)){
241         log_err("ERROR: Failed to get resource for \"root Locale\" with %s", myErrorName(status));
242         return;
243     }
244     binColl=ures_getByKey(res, "%%Collation", binColl, &status);
245     if(U_SUCCESS(status)){
246         binResult=(uint8_t*)ures_getBinary(binColl,  &len2, &status);
247         if(U_FAILURE(status)){
248             log_err("ERROR: ures_getBinary() failed\n");
249         }
250     }else{
251         log_err("ERROR: ures_getByKey(locale(default), %%Collation) failed");
252     }
253 
254 
255     if(len1 != len2){
256         log_err("Error: ucol_getDefaultRulesArray() failed to return the correct length.\n");
257     }
258     if(memcmp(binColData, binResult, len1) != 0){
259         log_err("Error: ucol_getDefaultRulesArray() failed\n");
260     }
261 
262     free(binColData);
263     ures_close(binColl);
264     ures_close(res);
265     ucol_close(coll);
266 
267 }
268 #endif
269 
270 /* Collator Properties
271  ucol_open, ucol_strcoll,  getStrength/setStrength
272  getDecomposition/setDecomposition, getDisplayName*/
TestProperty()273 void TestProperty()
274 {
275     UCollator *col, *ruled;
276     const UChar *rules;
277     UChar *disName;
278     int32_t len = 0;
279     UChar source[12], target[12];
280     int32_t tempLength;
281     UErrorCode status = U_ZERO_ERROR;
282     /*
283      * Expected version of the English collator.
284      * Currently, the major/minor version numbers change when the builder code
285      * changes,
286      * number 2 is from the tailoring data version and
287      * number 3 is the UCA version.
288      * This changes with every UCA version change, and the expected value
289      * needs to be adjusted.
290      * Same in intltest/apicoll.cpp.
291      */
292     UVersionInfo currVersionArray = {0x31, 0xC0, 0x05, 0x2A};  /* from ICU 4.4/UCA 5.2 */
293     UVersionInfo versionArray = {0, 0, 0, 0};
294     UVersionInfo versionUCAArray = {0, 0, 0, 0};
295     UVersionInfo versionUCDArray = {0, 0, 0, 0};
296 
297     log_verbose("The property tests begin : \n");
298     log_verbose("Test ucol_strcoll : \n");
299     col = ucol_open("en_US", &status);
300     if (U_FAILURE(status)) {
301         log_err_status(status, "Default Collator creation failed.: %s\n", myErrorName(status));
302         return;
303     }
304 
305     ucol_getVersion(col, versionArray);
306     /* Check for a version greater than some value rather than equality
307      * so that we need not update the expected version each time. */
308     if (uprv_memcmp(versionArray, currVersionArray, 4)<0) {
309       log_err("Testing ucol_getVersion() - unexpected result: %02x.%02x.%02x.%02x\n",
310               versionArray[0], versionArray[1], versionArray[2], versionArray[3]);
311     } else {
312       log_verbose("ucol_getVersion() result: %02x.%02x.%02x.%02x\n",
313                   versionArray[0], versionArray[1], versionArray[2], versionArray[3]);
314     }
315 
316     /* Assume that the UCD and UCA versions are the same,
317      * rather than hardcoding (and updating each time) a particular UCA version. */
318     u_getUnicodeVersion(versionUCDArray);
319     ucol_getUCAVersion(col, versionUCAArray);
320     if (0!=uprv_memcmp(versionUCAArray, versionUCDArray, 4)) {
321       log_err("Testing ucol_getUCAVersion() - unexpected result: %hu.%hu.%hu.%hu\n",
322               versionUCAArray[0], versionUCAArray[1], versionUCAArray[2], versionUCAArray[3]);
323     }
324 
325     u_uastrcpy(source, "ab");
326     u_uastrcpy(target, "abc");
327 
328     doStrcoll(col, source, u_strlen(source), target, u_strlen(target), UCOL_LESS, "ab < abc comparison failed");
329 
330     u_uastrcpy(source, "ab");
331     u_uastrcpy(target, "AB");
332 
333     doStrcoll(col, source, u_strlen(source), target, u_strlen(target), UCOL_LESS, "ab < AB comparison failed");
334 
335     u_uastrcpy(source, "blackbird");
336     u_uastrcpy(target, "black-bird");
337 
338     doStrcoll(col, source, u_strlen(source), target, u_strlen(target), UCOL_GREATER, "black-bird > blackbird comparison failed");
339 
340     u_uastrcpy(source, "black bird");
341     u_uastrcpy(target, "black-bird");
342 
343     doStrcoll(col, source, u_strlen(source), target, u_strlen(target), UCOL_LESS, "black bird < black-bird comparison failed");
344 
345     u_uastrcpy(source, "Hello");
346     u_uastrcpy(target, "hello");
347 
348     doStrcoll(col, source, u_strlen(source), target, u_strlen(target), UCOL_GREATER, "Hello > hello comparison failed");
349 
350     log_verbose("Test ucol_strcoll ends.\n");
351 
352     log_verbose("testing ucol_getStrength() method ...\n");
353     doAssert( (ucol_getStrength(col) == UCOL_TERTIARY), "collation object has the wrong strength");
354     doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
355 
356     log_verbose("testing ucol_setStrength() method ...\n");
357     ucol_setStrength(col, UCOL_SECONDARY);
358     doAssert( (ucol_getStrength(col) != UCOL_TERTIARY), "collation object's strength is secondary difference");
359     doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
360     doAssert( (ucol_getStrength(col) == UCOL_SECONDARY), "collation object has the wrong strength");
361 
362 
363     log_verbose("Get display name for the default collation in German : \n");
364 
365     len=ucol_getDisplayName("en_US", "de_DE", NULL, 0,  &status);
366     if(status==U_BUFFER_OVERFLOW_ERROR){
367         status=U_ZERO_ERROR;
368         disName=(UChar*)malloc(sizeof(UChar) * (len+1));
369         ucol_getDisplayName("en_US", "de_DE", disName, len+1,  &status);
370         log_verbose("the display name for default collation in german: %s\n", austrdup(disName) );
371         free(disName);
372     }
373     if(U_FAILURE(status)){
374         log_err("ERROR: in getDisplayName: %s\n", myErrorName(status));
375         return;
376     }
377     log_verbose("Default collation getDisplayName ended.\n");
378 
379     ruled = ucol_open("da_DK", &status);
380     if(U_FAILURE(status)) {
381         log_data_err("ucol_open(\"da_DK\") failed - %s\n", u_errorName(status));
382         ucol_close(col);
383         return;
384     }
385     log_verbose("ucol_getRules() testing ...\n");
386     rules = ucol_getRules(ruled, &tempLength);
387     if(tempLength == 0) {
388         log_data_err("missing da_DK tailoring rule string\n");
389     } else {
390         UChar aa[2] = { 0x61, 0x61 };
391         doAssert(u_strFindFirst(rules, tempLength, aa, 2) != NULL,
392                  "da_DK rules do not contain 'aa'");
393     }
394     log_verbose("getRules tests end.\n");
395     {
396         UChar *buffer = (UChar *)malloc(200000*sizeof(UChar));
397         int32_t bufLen = 200000;
398         buffer[0] = '\0';
399         log_verbose("ucol_getRulesEx() testing ...\n");
400         tempLength = ucol_getRulesEx(col,UCOL_TAILORING_ONLY,buffer,bufLen );
401         doAssert( tempLength == 0x00, "getRulesEx() result incorrect" );
402         log_verbose("getRules tests end.\n");
403 
404         log_verbose("ucol_getRulesEx() testing ...\n");
405         tempLength=ucol_getRulesEx(col,UCOL_FULL_RULES,buffer,bufLen );
406         if(tempLength == 0) {
407             log_data_err("missing *full* rule string\n");
408         }
409         log_verbose("getRulesEx tests end.\n");
410         free(buffer);
411     }
412     ucol_close(ruled);
413     ucol_close(col);
414 
415     log_verbose("open an collator for french locale");
416     col = ucol_open("fr_FR", &status);
417     if (U_FAILURE(status)) {
418        log_err("ERROR: Creating French collation failed.: %s\n", myErrorName(status));
419         return;
420     }
421     ucol_setStrength(col, UCOL_PRIMARY);
422     log_verbose("testing ucol_getStrength() method again ...\n");
423     doAssert( (ucol_getStrength(col) != UCOL_TERTIARY), "collation object has the wrong strength");
424     doAssert( (ucol_getStrength(col) == UCOL_PRIMARY), "collation object's strength is not primary difference");
425 
426     log_verbose("testing French ucol_setStrength() method ...\n");
427     ucol_setStrength(col, UCOL_TERTIARY);
428     doAssert( (ucol_getStrength(col) == UCOL_TERTIARY), "collation object's strength is not tertiary difference");
429     doAssert( (ucol_getStrength(col) != UCOL_PRIMARY), "collation object's strength is primary difference");
430     doAssert( (ucol_getStrength(col) != UCOL_SECONDARY), "collation object's strength is secondary difference");
431     ucol_close(col);
432 
433     log_verbose("Get display name for the french collation in english : \n");
434     len=ucol_getDisplayName("fr_FR", "en_US", NULL, 0,  &status);
435     if(status==U_BUFFER_OVERFLOW_ERROR){
436         status=U_ZERO_ERROR;
437         disName=(UChar*)malloc(sizeof(UChar) * (len+1));
438         ucol_getDisplayName("fr_FR", "en_US", disName, len+1,  &status);
439         log_verbose("the display name for french collation in english: %s\n", austrdup(disName) );
440         free(disName);
441     }
442     if(U_FAILURE(status)){
443         log_err("ERROR: in getDisplayName: %s\n", myErrorName(status));
444         return;
445     }
446     log_verbose("Default collation getDisplayName ended.\n");
447 
448 }
449 
450 /* Test RuleBasedCollator and getRules*/
TestRuleBasedColl()451 void TestRuleBasedColl()
452 {
453     UCollator *col1, *col2, *col3, *col4;
454     UCollationElements *iter1, *iter2;
455     UChar ruleset1[60];
456     UChar ruleset2[50];
457     UChar teststr[10];
458     const UChar *rule1, *rule2, *rule3, *rule4;
459     int32_t tempLength;
460     UErrorCode status = U_ZERO_ERROR;
461     u_uastrcpy(ruleset1, "&9 < a, A < b, B < c, C; ch, cH, Ch, CH < d, D, e, E");
462     u_uastrcpy(ruleset2, "&9 < a, A < b, B < c, C < d, D, e, E");
463 
464 
465     col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL,&status);
466     if (U_FAILURE(status)) {
467         log_err_status(status, "RuleBased Collator creation failed.: %s\n", myErrorName(status));
468         return;
469     }
470     else
471         log_verbose("PASS: RuleBased Collator creation passed\n");
472 
473     status = U_ZERO_ERROR;
474     col2 = ucol_openRules(ruleset2, u_strlen(ruleset2),  UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
475     if (U_FAILURE(status)) {
476         log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
477         return;
478     }
479     else
480         log_verbose("PASS: RuleBased Collator creation passed\n");
481 
482 
483     status = U_ZERO_ERROR;
484     col3= ucol_open(NULL, &status);
485     if (U_FAILURE(status)) {
486         log_err("Default Collator creation failed.: %s\n", myErrorName(status));
487         return;
488     }
489     else
490         log_verbose("PASS: Default Collator creation passed\n");
491 
492     rule1 = ucol_getRules(col1, &tempLength);
493     rule2 = ucol_getRules(col2, &tempLength);
494     rule3 = ucol_getRules(col3, &tempLength);
495 
496     doAssert((u_strcmp(rule1, rule2) != 0), "Default collator getRules failed");
497     doAssert((u_strcmp(rule2, rule3) != 0), "Default collator getRules failed");
498     doAssert((u_strcmp(rule1, rule3) != 0), "Default collator getRules failed");
499 
500     col4=ucol_openRules(rule2, u_strlen(rule2), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
501     if (U_FAILURE(status)) {
502         log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
503         return;
504     }
505     rule4= ucol_getRules(col4, &tempLength);
506     doAssert((u_strcmp(rule2, rule4) == 0), "Default collator getRules failed");
507 
508     ucol_close(col1);
509     ucol_close(col2);
510     ucol_close(col3);
511     ucol_close(col4);
512 
513     /* tests that modifier ! is always ignored */
514     u_uastrcpy(ruleset1, "!&a<b");
515     teststr[0] = 0x0e40;
516     teststr[1] = 0x0e01;
517     teststr[2] = 0x0e2d;
518     col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
519     if (U_FAILURE(status)) {
520         log_err("RuleBased Collator creation failed.: %s\n", myErrorName(status));
521         return;
522     }
523     col2 = ucol_open("en_US", &status);
524     if (U_FAILURE(status)) {
525         log_err("en_US Collator creation failed.: %s\n", myErrorName(status));
526         return;
527     }
528     iter1 = ucol_openElements(col1, teststr, 3, &status);
529     iter2 = ucol_openElements(col2, teststr, 3, &status);
530     if(U_FAILURE(status)) {
531         log_err("ERROR: CollationElement iterator creation failed.: %s\n", myErrorName(status));
532         return;
533     }
534     while (TRUE) {
535         /* testing with en since thai has its own tailoring */
536         int32_t ce = ucol_next(iter1, &status);
537         int32_t ce2 = ucol_next(iter2, &status);
538         if(U_FAILURE(status)) {
539             log_err("ERROR: CollationElement iterator creation failed.: %s\n", myErrorName(status));
540             return;
541         }
542         if (ce2 != ce) {
543              log_err("! modifier test failed");
544         }
545         if (ce == UCOL_NULLORDER) {
546             break;
547         }
548     }
549     ucol_closeElements(iter1);
550     ucol_closeElements(iter2);
551     ucol_close(col1);
552     ucol_close(col2);
553     /* CLDR 24+ requires a reset before the first relation */
554     u_uastrcpy(ruleset1, "< z < a");
555     col1 = ucol_openRules(ruleset1, u_strlen(ruleset1), UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
556     if (status != U_PARSE_ERROR && status != U_INVALID_FORMAT_ERROR) {
557         log_err("ucol_openRules(without initial reset: '< z < a') "
558                 "should fail with U_PARSE_ERROR or U_INVALID_FORMAT_ERROR but yielded %s\n",
559                 myErrorName(status));
560     }
561     ucol_close(col1);
562 }
563 
TestCompare()564 void TestCompare()
565 {
566     UErrorCode status = U_ZERO_ERROR;
567     UCollator *col;
568     UChar* test1;
569     UChar* test2;
570 
571     log_verbose("The compare tests begin : \n");
572     status=U_ZERO_ERROR;
573     col = ucol_open("en_US", &status);
574     if(U_FAILURE(status)) {
575         log_err_status(status, "ucal_open() collation creation failed.: %s\n", myErrorName(status));
576         return;
577     }
578     test1=(UChar*)malloc(sizeof(UChar) * 6);
579     test2=(UChar*)malloc(sizeof(UChar) * 6);
580     u_uastrcpy(test1, "Abcda");
581     u_uastrcpy(test2, "abcda");
582 
583     log_verbose("Use tertiary comparison level testing ....\n");
584 
585     doAssert( (!ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" != \"abcda\" ");
586     doAssert( (ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" >>> \"abcda\" ");
587     doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" >>> \"abcda\"");
588 
589     ucol_setStrength(col, UCOL_SECONDARY);
590     log_verbose("Use secondary comparison level testing ....\n");
591 
592     doAssert( (ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2) )), "Result should be \"Abcda\" == \"abcda\"");
593     doAssert( (!ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
594     doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2) )), "Result should be \"Abcda\" == \"abcda\"");
595 
596     ucol_setStrength(col, UCOL_PRIMARY);
597     log_verbose("Use primary comparison level testing ....\n");
598 
599     doAssert( (ucol_equal(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
600     doAssert( (!ucol_greater(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
601     doAssert( (ucol_greaterOrEqual(col, test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"Abcda\" == \"abcda\"");
602 
603 
604     log_verbose("The compare tests end.\n");
605     ucol_close(col);
606     free(test1);
607     free(test2);
608 
609 }
610 /*
611 ---------------------------------------------
612  tests decomposition setting
613 */
TestDecomposition()614 void TestDecomposition() {
615     UErrorCode status = U_ZERO_ERROR;
616     UCollator *en_US, *el_GR, *vi_VN;
617     en_US = ucol_open("en_US", &status);
618     el_GR = ucol_open("el_GR", &status);
619     vi_VN = ucol_open("vi_VN", &status);
620 
621     if (U_FAILURE(status)) {
622         log_err_status(status, "ERROR: collation creation failed.: %s\n", myErrorName(status));
623         return;
624     }
625 
626     if (ucol_getAttribute(vi_VN, UCOL_NORMALIZATION_MODE, &status) != UCOL_ON ||
627         U_FAILURE(status))
628     {
629         log_err("ERROR: vi_VN collation did not have canonical decomposition for normalization!\n");
630     }
631 
632     status = U_ZERO_ERROR;
633     if (ucol_getAttribute(el_GR, UCOL_NORMALIZATION_MODE, &status) != UCOL_ON ||
634         U_FAILURE(status))
635     {
636         log_err("ERROR: el_GR collation did not have canonical decomposition for normalization!\n");
637     }
638 
639     status = U_ZERO_ERROR;
640     if (ucol_getAttribute(en_US, UCOL_NORMALIZATION_MODE, &status) != UCOL_OFF ||
641         U_FAILURE(status))
642     {
643         log_err("ERROR: en_US collation had canonical decomposition for normalization!\n");
644     }
645 
646     ucol_close(en_US);
647     ucol_close(el_GR);
648     ucol_close(vi_VN);
649 }
650 
651 #define CLONETEST_COLLATOR_COUNT 4
652 
TestSafeClone()653 void TestSafeClone() {
654     UChar test1[6];
655     UChar test2[6];
656     static const UChar umlautUStr[] = {0x00DC, 0};
657     static const UChar oeStr[] = {0x0055, 0x0045, 0};
658     UCollator * someCollators [CLONETEST_COLLATOR_COUNT];
659     UCollator * someClonedCollators [CLONETEST_COLLATOR_COUNT];
660     UCollator * col;
661     UErrorCode err = U_ZERO_ERROR;
662     int8_t idx = 6;    /* Leave this here to test buffer alignment in memory*/
663     uint8_t buffer [CLONETEST_COLLATOR_COUNT] [U_COL_SAFECLONE_BUFFERSIZE];
664     int32_t bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
665     const char sampleRuleChars[] = "&Z < CH";
666     UChar sampleRule[sizeof(sampleRuleChars)];
667 
668     u_uastrcpy(test1, "abCda");
669     u_uastrcpy(test2, "abcda");
670     u_uastrcpy(sampleRule, sampleRuleChars);
671 
672     /* one default collator & two complex ones */
673     someCollators[0] = ucol_open("en_US", &err);
674     someCollators[1] = ucol_open("ko", &err);
675     someCollators[2] = ucol_open("ja_JP", &err);
676     someCollators[3] = ucol_openRules(sampleRule, -1, UCOL_ON, UCOL_TERTIARY, NULL, &err);
677     if(U_FAILURE(err)) {
678         for (idx = 0; idx < CLONETEST_COLLATOR_COUNT; idx++) {
679             ucol_close(someCollators[idx]);
680         }
681         log_data_err("Couldn't open one or more collators\n");
682         return;
683     }
684 
685     /* Check the various error & informational states: */
686 
687     /* Null status - just returns NULL */
688     if (NULL != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, NULL))
689     {
690         log_err("FAIL: Cloned Collator failed to deal correctly with null status\n");
691     }
692     /* error status - should return 0 & keep error the same */
693     err = U_MEMORY_ALLOCATION_ERROR;
694     if (NULL != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err) || err != U_MEMORY_ALLOCATION_ERROR)
695     {
696         log_err("FAIL: Cloned Collator failed to deal correctly with incoming error status\n");
697     }
698     err = U_ZERO_ERROR;
699 
700     /* Null buffer size pointer is ok */
701     if (NULL == (col = ucol_safeClone(someCollators[0], buffer[0], NULL, &err)) || U_FAILURE(err))
702     {
703         log_err("FAIL: Cloned Collator failed to deal correctly with null bufferSize pointer\n");
704     }
705     ucol_close(col);
706     err = U_ZERO_ERROR;
707 
708     /* buffer size pointer is 0 - fill in pbufferSize with a size */
709     bufferSize = 0;
710     if (NULL != ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err) ||
711             U_FAILURE(err) || bufferSize <= 0)
712     {
713         log_err("FAIL: Cloned Collator failed a sizing request ('preflighting')\n");
714     }
715     /* Verify our define is large enough  */
716     if (U_COL_SAFECLONE_BUFFERSIZE < bufferSize)
717     {
718         log_err("FAIL: Pre-calculated buffer size is too small\n");
719     }
720     /* Verify we can use this run-time calculated size */
721     if (NULL == (col = ucol_safeClone(someCollators[0], buffer[0], &bufferSize, &err)) || U_FAILURE(err))
722     {
723         log_err("FAIL: Collator can't be cloned with run-time size\n");
724     }
725     if (col) ucol_close(col);
726     /* size one byte too small - should allocate & let us know */
727     if (bufferSize > 1) {
728         --bufferSize;
729     }
730     if (NULL == (col = ucol_safeClone(someCollators[0], 0, &bufferSize, &err)) || err != U_SAFECLONE_ALLOCATED_WARNING)
731     {
732         log_err("FAIL: Cloned Collator failed to deal correctly with too-small buffer size\n");
733     }
734     if (col) ucol_close(col);
735     err = U_ZERO_ERROR;
736     bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
737 
738 
739     /* Null buffer pointer - return Collator & set error to U_SAFECLONE_ALLOCATED_ERROR */
740     if (NULL == (col = ucol_safeClone(someCollators[0], 0, &bufferSize, &err)) || err != U_SAFECLONE_ALLOCATED_WARNING)
741     {
742         log_err("FAIL: Cloned Collator failed to deal correctly with null buffer pointer\n");
743     }
744     if (col) ucol_close(col);
745     err = U_ZERO_ERROR;
746 
747     /* Null Collator - return NULL & set U_ILLEGAL_ARGUMENT_ERROR */
748     if (NULL != ucol_safeClone(NULL, buffer[0], &bufferSize, &err) || err != U_ILLEGAL_ARGUMENT_ERROR)
749     {
750         log_err("FAIL: Cloned Collator failed to deal correctly with null Collator pointer\n");
751     }
752 
753     err = U_ZERO_ERROR;
754 
755     /* Test that a cloned collator doesn't accidentally use UCA. */
756     col=ucol_open("de@collation=phonebook", &err);
757     bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
758     someClonedCollators[0] = ucol_safeClone(col, buffer[0], &bufferSize, &err);
759     doAssert( (ucol_greater(col, umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Original German phonebook collation sorts differently than expected");
760     doAssert( (ucol_greater(someClonedCollators[0], umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Cloned German phonebook collation sorts differently than expected");
761     if (!ucol_equals(someClonedCollators[0], col)) {
762         log_err("FAIL: Cloned German phonebook collator is not equal to original.\n");
763     }
764     ucol_close(col);
765     ucol_close(someClonedCollators[0]);
766 
767     err = U_ZERO_ERROR;
768 
769     /* change orig & clone & make sure they are independent */
770 
771     for (idx = 0; idx < CLONETEST_COLLATOR_COUNT; idx++)
772     {
773         ucol_setStrength(someCollators[idx], UCOL_IDENTICAL);
774         bufferSize = 1;
775         err = U_ZERO_ERROR;
776         ucol_close(ucol_safeClone(someCollators[idx], buffer[idx], &bufferSize, &err));
777         if (err != U_SAFECLONE_ALLOCATED_WARNING) {
778             log_err("FAIL: collator number %d was not allocated.\n", idx);
779             log_err("FAIL: status of Collator[%d] is %d  (hex: %x).\n", idx, err, err);
780         }
781 
782         bufferSize = U_COL_SAFECLONE_BUFFERSIZE;
783         err = U_ZERO_ERROR;
784         someClonedCollators[idx] = ucol_safeClone(someCollators[idx], buffer[idx], &bufferSize, &err);
785         if (U_FAILURE(err)) {
786             log_err("FAIL: Unable to clone collator %d - %s\n", idx, u_errorName(err));
787             continue;
788         }
789         if (!ucol_equals(someClonedCollators[idx], someCollators[idx])) {
790             log_err("FAIL: Cloned collator is not equal to original at index = %d.\n", idx);
791         }
792 
793         /* Check the usability */
794         ucol_setStrength(someCollators[idx], UCOL_PRIMARY);
795         ucol_setAttribute(someCollators[idx], UCOL_CASE_LEVEL, UCOL_OFF, &err);
796 
797         doAssert( (ucol_equal(someCollators[idx], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abcda\" == \"abCda\"");
798 
799         /* Close the original to make sure that the clone is usable. */
800         ucol_close(someCollators[idx]);
801 
802         ucol_setStrength(someClonedCollators[idx], UCOL_TERTIARY);
803         ucol_setAttribute(someClonedCollators[idx], UCOL_CASE_LEVEL, UCOL_OFF, &err);
804         doAssert( (ucol_greater(someClonedCollators[idx], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abCda\" >>> \"abcda\" ");
805 
806         ucol_close(someClonedCollators[idx]);
807     }
808 }
809 
TestClone()810 void TestClone() {
811     UChar test1[6];
812     UChar test2[6];
813     static const UChar umlautUStr[] = {0x00DC, 0};
814     static const UChar oeStr[] = {0x0055, 0x0045, 0};
815     UCollator * someCollators [CLONETEST_COLLATOR_COUNT];
816     UCollator * someClonedCollators [CLONETEST_COLLATOR_COUNT];
817     UCollator * col = NULL;
818     UErrorCode err = U_ZERO_ERROR;
819     int8_t idx = 6;    /* Leave this here to test buffer alignment in memory*/
820     const char sampleRuleChars[] = "&Z < CH";
821     UChar sampleRule[sizeof(sampleRuleChars)];
822 
823     u_uastrcpy(test1, "abCda");
824     u_uastrcpy(test2, "abcda");
825     u_uastrcpy(sampleRule, sampleRuleChars);
826 
827     /* one default collator & two complex ones */
828     someCollators[0] = ucol_open("en_US", &err);
829     someCollators[1] = ucol_open("ko", &err);
830     someCollators[2] = ucol_open("ja_JP", &err);
831     someCollators[3] = ucol_openRules(sampleRule, -1, UCOL_ON, UCOL_TERTIARY, NULL, &err);
832     if(U_FAILURE(err)) {
833         for (idx = 0; idx < CLONETEST_COLLATOR_COUNT; idx++) {
834             ucol_close(someCollators[idx]);
835         }
836         log_data_err("Couldn't open one or more collators\n");
837         return;
838     }
839 
840     /* Check the various error & informational states: */
841 
842     /* Null status - just returns NULL */
843     if (NULL != ucol_clone(someCollators[0], NULL))
844     {
845         log_err("FAIL: Cloned Collator failed to deal correctly with null status\n");
846     }
847     /* error status - should return 0 & keep error the same */
848     err = U_MEMORY_ALLOCATION_ERROR;
849     if (NULL != ucol_clone(someCollators[0], &err) || err != U_MEMORY_ALLOCATION_ERROR)
850     {
851         log_err("FAIL: Cloned Collator failed to deal correctly with incoming error status\n");
852     }
853     err = U_ZERO_ERROR;
854 
855     /* Verify we can use this run-time calculated size */
856     if (NULL == (col = ucol_clone(someCollators[0], &err)) || U_FAILURE(err))
857     {
858         log_err("FAIL: Collator can't be cloned.\n");
859     }
860     if (col) ucol_close(col);
861 
862     if (NULL == (col = ucol_clone(someCollators[0], &err)) || err != U_ZERO_ERROR)
863     {
864         log_err("FAIL: Cloned Collator failed to deal correctly\n");
865     }
866     if (col) ucol_close(col);
867     err = U_ZERO_ERROR;
868 
869     /* Null Collator - return NULL & set U_ILLEGAL_ARGUMENT_ERROR */
870     if (NULL != ucol_clone(NULL, &err) || err != U_ILLEGAL_ARGUMENT_ERROR)
871     {
872         log_err("FAIL: Cloned Collator failed to deal correctly with null Collator pointer\n");
873     }
874     err = U_ZERO_ERROR;
875 
876     /* Test that a cloned collator doesn't accidentally use UCA. */
877     col=ucol_open("de@collation=phonebook", &err);
878     someClonedCollators[0] = ucol_clone(col, &err);
879     doAssert( (ucol_greater(col, umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Original German phonebook collation sorts differently than expected");
880     doAssert( (ucol_greater(someClonedCollators[0], umlautUStr, u_strlen(umlautUStr), oeStr, u_strlen(oeStr))), "Cloned German phonebook collation sorts differently than expected");
881     if (!ucol_equals(someClonedCollators[0], col)) {
882         log_err("FAIL: Cloned German phonebook collator is not equal to original.\n");
883     }
884     ucol_close(col);
885     ucol_close(someClonedCollators[0]);
886 
887     err = U_ZERO_ERROR;
888 
889     /* change orig & clone & make sure they are independent */
890 
891     for (idx = 0; idx < CLONETEST_COLLATOR_COUNT; idx++)
892     {
893         ucol_setStrength(someCollators[idx], UCOL_IDENTICAL);
894         err = U_ZERO_ERROR;
895         ucol_close(ucol_clone(someCollators[idx], &err));
896         if (err != U_ZERO_ERROR) {
897             log_err("FAIL: collator number %d was not allocated.\n", idx);
898             log_err("FAIL: status of Collator[%d] is %d  (hex: %x).\n", idx, err, err);
899         }
900 
901         err = U_ZERO_ERROR;
902         someClonedCollators[idx] = ucol_clone(someCollators[idx], &err);
903         if (U_FAILURE(err)) {
904             log_err("FAIL: Unable to clone collator %d - %s\n", idx, u_errorName(err));
905             continue;
906         }
907         if (!ucol_equals(someClonedCollators[idx], someCollators[idx])) {
908             log_err("FAIL: Cloned collator is not equal to original at index = %d.\n", idx);
909         }
910 
911         /* Check the usability */
912         ucol_setStrength(someCollators[idx], UCOL_PRIMARY);
913         ucol_setAttribute(someCollators[idx], UCOL_CASE_LEVEL, UCOL_OFF, &err);
914 
915         doAssert( (ucol_equal(someCollators[idx], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abcda\" == \"abCda\"");
916 
917         /* Close the original to make sure that the clone is usable. */
918         ucol_close(someCollators[idx]);
919 
920         ucol_setStrength(someClonedCollators[idx], UCOL_TERTIARY);
921         ucol_setAttribute(someClonedCollators[idx], UCOL_CASE_LEVEL, UCOL_OFF, &err);
922         doAssert( (ucol_greater(someClonedCollators[idx], test1, u_strlen(test1), test2, u_strlen(test2))), "Result should be \"abCda\" >>> \"abcda\" ");
923 
924         ucol_close(someClonedCollators[idx]);
925     }
926 }
927 
TestCloneBinary()928 void TestCloneBinary(){
929     UErrorCode err = U_ZERO_ERROR;
930     UCollator * col = ucol_open("en_US", &err);
931     UCollator * c;
932     int32_t size;
933     uint8_t * buffer;
934 
935     if (U_FAILURE(err)) {
936         log_data_err("Couldn't open collator. Error: %s\n", u_errorName(err));
937         return;
938     }
939 
940     size = ucol_cloneBinary(col, NULL, 0, &err);
941     if(size==0 || err!=U_BUFFER_OVERFLOW_ERROR) {
942         log_err("ucol_cloneBinary - couldn't check size. Error: %s\n", u_errorName(err));
943         return;
944     }
945     err = U_ZERO_ERROR;
946 
947     buffer = (uint8_t *) malloc(size);
948     ucol_cloneBinary(col, buffer, size, &err);
949     if(U_FAILURE(err)) {
950         log_err("ucol_cloneBinary - couldn't clone.. Error: %s\n", u_errorName(err));
951         free(buffer);
952         return;
953     }
954 
955     /* how to check binary result ? */
956 
957     c = ucol_openBinary(buffer, size, col, &err);
958     if(U_FAILURE(err)) {
959         log_err("ucol_openBinary failed. Error: %s\n", u_errorName(err));
960     } else {
961         UChar t[] = {0x41, 0x42, 0x43, 0};  /* ABC */
962         uint8_t  *k1, *k2;
963         int l1, l2;
964         l1 = ucol_getSortKey(col, t, -1, NULL,0);
965         l2 = ucol_getSortKey(c, t, -1, NULL,0);
966         k1 = (uint8_t *) malloc(sizeof(uint8_t) * l1);
967         k2 = (uint8_t *) malloc(sizeof(uint8_t) * l2);
968         ucol_getSortKey(col, t, -1, k1, l1);
969         ucol_getSortKey(col, t, -1, k2, l2);
970         if (strcmp((char *)k1,(char *)k2) != 0){
971             log_err("ucol_openBinary - new collator should equal to old one\n");
972         }
973         free(k1);
974         free(k2);
975     }
976     free(buffer);
977     ucol_close(c);
978     ucol_close(col);
979 }
980 
981 
TestBengaliSortKey(void)982 static void TestBengaliSortKey(void)
983 {
984   const char *curLoc = "bn";
985   UChar str1[] = { 0x09BE, 0 };
986   UChar str2[] = { 0x0B70, 0 };
987   UCollator *c2 = NULL;
988   const UChar *rules;
989   int32_t rulesLength=-1;
990   uint8_t *sortKey1;
991   int32_t sortKeyLen1 = 0;
992   uint8_t *sortKey2;
993   int32_t sortKeyLen2 = 0;
994   UErrorCode status = U_ZERO_ERROR;
995   char sortKeyStr1[2048];
996   uint32_t sortKeyStrLen1 = UPRV_LENGTHOF(sortKeyStr1);
997   char sortKeyStr2[2048];
998   uint32_t sortKeyStrLen2 = UPRV_LENGTHOF(sortKeyStr2);
999   UCollationResult result;
1000 
1001   static UChar preRules[41] = { 0x26, 0x9fa, 0x3c, 0x98c, 0x3c, 0x9e1, 0x3c, 0x98f, 0x3c, 0x990, 0x3c, 0x993, 0x3c, 0x994, 0x3c, 0x9bc, 0x3c, 0x982, 0x3c, 0x983, 0x3c, 0x981, 0x3c, 0x9b0, 0x3c, 0x9b8, 0x3c, 0x9b9, 0x3c, 0x9bd, 0x3c, 0x9be, 0x3c, 0x9bf, 0x3c, 0x9c8, 0x3c, 0x9cb, 0x3d, 0x9cb , 0};
1002 
1003   rules = preRules;
1004 
1005   log_verbose("Rules: %s\n", aescstrdup(rules, rulesLength));
1006 
1007   c2 = ucol_openRules(rules, rulesLength, UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &status);
1008   if (U_FAILURE(status)) {
1009     log_data_err("ERROR: Creating collator from rules failed with locale: %s : %s\n", curLoc, myErrorName(status));
1010     return;
1011   }
1012 
1013   sortKeyLen1 = ucol_getSortKey(c2, str1, -1, NULL, 0);
1014   sortKey1 = (uint8_t*)malloc(sortKeyLen1+1);
1015   ucol_getSortKey(c2,str1,-1,sortKey1, sortKeyLen1+1);
1016   ucol_sortKeyToString(c2, sortKey1, sortKeyStr1, sortKeyStrLen1);
1017 
1018 
1019   sortKeyLen2 = ucol_getSortKey(c2, str2, -1, NULL, 0);
1020   sortKey2 = (uint8_t*)malloc(sortKeyLen2+1);
1021   ucol_getSortKey(c2,str2,-1,sortKey2, sortKeyLen2+1);
1022 
1023   ucol_sortKeyToString(c2, sortKey2, sortKeyStr2, sortKeyStrLen2);
1024 
1025 
1026 
1027   result=ucol_strcoll(c2, str1, -1, str2, -1);
1028   if(result!=UCOL_LESS) {
1029     log_err("Error: %s was not less than %s: result=%d.\n", aescstrdup(str1,-1), aescstrdup(str2,-1), result);
1030     log_info("[%s] -> %s (%d, from rule)\n", aescstrdup(str1,-1), sortKeyStr1, sortKeyLen1);
1031     log_info("[%s] -> %s (%d, from rule)\n", aescstrdup(str2,-1), sortKeyStr2, sortKeyLen2);
1032   } else {
1033     log_verbose("OK: %s was  less than %s: result=%d.\n", aescstrdup(str1,-1), aescstrdup(str2,-1), result);
1034     log_verbose("[%s] -> %s (%d, from rule)\n", aescstrdup(str1,-1), sortKeyStr1, sortKeyLen1);
1035     log_verbose("[%s] -> %s (%d, from rule)\n", aescstrdup(str2,-1), sortKeyStr2, sortKeyLen2);
1036   }
1037 
1038   free(sortKey1);
1039   free(sortKey2);
1040   ucol_close(c2);
1041 
1042 }
1043 
1044 /*
1045     TestOpenVsOpenRules ensures that collators from ucol_open and ucol_openRules
1046     will generate identical sort keys
1047 */
TestOpenVsOpenRules()1048 void TestOpenVsOpenRules(){
1049 
1050     /* create an array of all the locales */
1051     int32_t numLocales = uloc_countAvailable();
1052     int32_t sizeOfStdSet;
1053     uint32_t adder;
1054     UChar str[41]; /* create an array of UChar of size maximum strSize + 1 */
1055     USet *stdSet;
1056     char* curLoc;
1057     UCollator * c1;
1058     UCollator * c2;
1059     const UChar* rules;
1060     int32_t rulesLength;
1061     int32_t sortKeyLen1, sortKeyLen2;
1062     uint8_t *sortKey1 = NULL, *sortKey2 = NULL;
1063     char sortKeyStr1[512], sortKeyStr2[512];
1064     uint32_t sortKeyStrLen1 = UPRV_LENGTHOF(sortKeyStr1),
1065              sortKeyStrLen2 = UPRV_LENGTHOF(sortKeyStr2);
1066     ULocaleData *uld;
1067     int32_t x, y, z;
1068     USet *eSet;
1069     int32_t eSize;
1070     int strSize;
1071 
1072     UErrorCode err = U_ZERO_ERROR;
1073 
1074     /* create a set of standard characters that aren't very interesting...
1075     and then we can find some interesting ones later */
1076 
1077     stdSet = uset_open(0x61, 0x7A);
1078     uset_addRange(stdSet, 0x41, 0x5A);
1079     uset_addRange(stdSet, 0x30, 0x39);
1080     sizeOfStdSet = uset_size(stdSet);
1081     (void)sizeOfStdSet;   /* Suppress set but not used warning. */
1082 
1083     adder = 1;
1084     if(getTestOption(QUICK_OPTION))
1085     {
1086         adder = 10;
1087     }
1088 
1089     for(x = 0; x < numLocales; x+=adder){
1090         curLoc = (char *)uloc_getAvailable(x);
1091         log_verbose("Processing %s\n", curLoc);
1092 
1093         /* create a collator the normal API way */
1094         c1 = ucol_open(curLoc, &err);
1095         if (U_FAILURE(err)) {
1096             log_err("ERROR: Normal collation creation failed with locale: %s : %s\n", curLoc, myErrorName(err));
1097             return;
1098         }
1099 
1100         /* grab the rules */
1101         rules = ucol_getRules(c1, &rulesLength);
1102         if (rulesLength == 0) {
1103             /* The optional tailoring rule string is either empty (boring) or missing. */
1104             ucol_close(c1);
1105             continue;
1106         }
1107 
1108         /* use those rules to create a collator from rules */
1109         c2 = ucol_openRules(rules, rulesLength, UCOL_DEFAULT, UCOL_DEFAULT_STRENGTH, NULL, &err);
1110         if (U_FAILURE(err)) {
1111             log_err("ERROR: Creating collator from rules failed with locale: %s : %s\n", curLoc, myErrorName(err));
1112             ucol_close(c1);
1113             continue;
1114         }
1115 
1116         uld = ulocdata_open(curLoc, &err);
1117 
1118         /*now that we have some collators, we get several strings */
1119 
1120         for(y = 0; y < 5; y++){
1121 
1122             /* get a set of ALL the characters in this locale */
1123             eSet =  ulocdata_getExemplarSet(uld, NULL, 0, ULOCDATA_ES_STANDARD, &err);
1124             eSize = uset_size(eSet);
1125 
1126             /* make a string with these characters in it */
1127             strSize = (rand()%40) + 1;
1128 
1129             for(z = 0; z < strSize; z++){
1130                 str[z] = uset_charAt(eSet, rand()%eSize);
1131             }
1132 
1133             /* change the set to only include 'abnormal' characters (not A-Z, a-z, 0-9 */
1134             uset_removeAll(eSet, stdSet);
1135             eSize = uset_size(eSet);
1136 
1137             /* if there are some non-normal characters left, put a few into the string, just to make sure we have some */
1138             if(eSize > 0){
1139                 str[2%strSize] = uset_charAt(eSet, rand()%eSize);
1140                 str[3%strSize] = uset_charAt(eSet, rand()%eSize);
1141                 str[5%strSize] = uset_charAt(eSet, rand()%eSize);
1142                 str[10%strSize] = uset_charAt(eSet, rand()%eSize);
1143                 str[13%strSize] = uset_charAt(eSet, rand()%eSize);
1144             }
1145             /* terminate the string */
1146             str[strSize-1] = '\0';
1147             log_verbose("String used: %S\n", str);
1148 
1149             /* get sort keys for both of them, and check that the keys are identicle */
1150             sortKeyLen1 = ucol_getSortKey(c1, str, u_strlen(str),  NULL, 0);
1151             sortKey1 = (uint8_t*)malloc(sizeof(uint8_t) * (sortKeyLen1 + 1));
1152             /*memset(sortKey1, 0xFE, sortKeyLen1);*/
1153             ucol_getSortKey(c1, str, u_strlen(str), sortKey1, sortKeyLen1 + 1);
1154             ucol_sortKeyToString(c1, sortKey1, sortKeyStr1, sortKeyStrLen1);
1155 
1156             sortKeyLen2 = ucol_getSortKey(c2, str, u_strlen(str),  NULL, 0);
1157             sortKey2 = (uint8_t*)malloc(sizeof(uint8_t) * (sortKeyLen2 + 1));
1158             /*memset(sortKey2, 0xFE, sortKeyLen2);*/
1159             ucol_getSortKey(c2, str, u_strlen(str), sortKey2, sortKeyLen2 + 1);
1160             ucol_sortKeyToString(c2, sortKey2, sortKeyStr2, sortKeyStrLen2);
1161 
1162             /* Check that the lengths are the same */
1163             if (sortKeyLen1 != sortKeyLen2) {
1164                 log_err("ERROR : Sort key lengths %d and %d for text '%s' in locale '%s' do not match.\n",
1165                     sortKeyLen1, sortKeyLen2, str, curLoc);
1166             }
1167 
1168             /* check that the keys are the same */
1169             if (memcmp(sortKey1, sortKey2, sortKeyLen1) != 0) {
1170                 log_err("ERROR : Sort keys '%s' and '%s' for text '%s' in locale '%s' are not equivalent.\n",
1171                     sortKeyStr1, sortKeyStr2, str, curLoc);
1172             }
1173 
1174             /* clean up after each string */
1175             free(sortKey1);
1176             free(sortKey2);
1177             uset_close(eSet);
1178         }
1179         /* clean up after each locale */
1180         ulocdata_close(uld);
1181         ucol_close(c1);
1182         ucol_close(c2);
1183     }
1184     /* final clean up */
1185     uset_close(stdSet);
1186 }
1187 /*
1188 ----------------------------------------------------------------------------
1189  ctor -- Tests the getSortKey
1190 */
TestSortKey()1191 void TestSortKey()
1192 {
1193     uint8_t *sortk1 = NULL, *sortk2 = NULL, *sortk3 = NULL, *sortkEmpty = NULL;
1194     int32_t sortklen, osortklen;
1195     UCollator *col;
1196     UChar *test1, *test2, *test3;
1197     UErrorCode status = U_ZERO_ERROR;
1198     char toStringBuffer[256], *resultP;
1199     uint32_t toStringLen=UPRV_LENGTHOF(toStringBuffer);
1200 
1201 
1202     uint8_t s1[] = { 0x9f, 0x00 };
1203     uint8_t s2[] = { 0x61, 0x00 };
1204     int  strcmpResult;
1205 
1206     strcmpResult = strcmp((const char *)s1, (const char *)s2);
1207     log_verbose("strcmp(0x9f..., 0x61...) = %d\n", strcmpResult);
1208 
1209     if(strcmpResult <= 0) {
1210       log_err("ERR: expected strcmp(\"9f 00\", \"61 00\") to be >=0 (GREATER).. got %d. Calling strcmp() for sortkeys may not work! \n",
1211               strcmpResult);
1212     }
1213 
1214 
1215     log_verbose("testing SortKey begins...\n");
1216     /* this is supposed to open default date format, but later on it treats it like it is "en_US"
1217        - very bad if you try to run the tests on machine where default locale is NOT "en_US" */
1218     /* col = ucol_open(NULL, &status); */
1219     col = ucol_open("en_US", &status);
1220     if (U_FAILURE(status)) {
1221         log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1222         return;
1223     }
1224 
1225 
1226     if(ucol_getStrength(col) != UCOL_DEFAULT_STRENGTH)
1227     {
1228         log_err("ERROR: default collation did not have UCOL_DEFAULT_STRENGTH !\n");
1229     }
1230     /* Need to use identical strength */
1231     ucol_setAttribute(col, UCOL_STRENGTH, UCOL_IDENTICAL, &status);
1232 
1233     test1=(UChar*)malloc(sizeof(UChar) * 6);
1234     test2=(UChar*)malloc(sizeof(UChar) * 6);
1235     test3=(UChar*)malloc(sizeof(UChar) * 6);
1236 
1237     memset(test1,0xFE, sizeof(UChar)*6);
1238     memset(test2,0xFE, sizeof(UChar)*6);
1239     memset(test3,0xFE, sizeof(UChar)*6);
1240 
1241 
1242     u_uastrcpy(test1, "Abcda");
1243     u_uastrcpy(test2, "abcda");
1244     u_uastrcpy(test3, "abcda");
1245 
1246     log_verbose("Use tertiary comparison level testing ....\n");
1247 
1248     sortklen=ucol_getSortKey(col, test1, u_strlen(test1),  NULL, 0);
1249     sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1250     memset(sortk1,0xFE, sortklen);
1251     ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortklen+1);
1252 
1253     sortklen=ucol_getSortKey(col, test2, u_strlen(test2),  NULL, 0);
1254     sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1255     memset(sortk2,0xFE, sortklen);
1256     ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortklen+1);
1257 
1258     osortklen = sortklen;
1259     sortklen=ucol_getSortKey(col, test2, u_strlen(test3),  NULL, 0);
1260     sortk3=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1261     memset(sortk3,0xFE, sortklen);
1262     ucol_getSortKey(col, test2, u_strlen(test2), sortk3, sortklen+1);
1263 
1264     doAssert( (sortklen == osortklen), "Sortkey length should be the same (abcda, abcda)");
1265 
1266     doAssert( (memcmp(sortk1, sortk2, sortklen) > 0), "Result should be \"Abcda\" > \"abcda\"");
1267     doAssert( (memcmp(sortk2, sortk1, sortklen) < 0), "Result should be \"abcda\" < \"Abcda\"");
1268     doAssert( (memcmp(sortk2, sortk3, sortklen) == 0), "Result should be \"abcda\" ==  \"abcda\"");
1269 
1270     resultP = ucol_sortKeyToString(col, sortk3, toStringBuffer, toStringLen);
1271     doAssert( (resultP != 0), "sortKeyToString failed!");
1272 
1273 #if 1 /* verobse log of sortkeys */
1274     {
1275       char junk2[1000];
1276       char junk3[1000];
1277       int i;
1278 
1279       strcpy(junk2, "abcda[2] ");
1280       strcpy(junk3, " abcda[3] ");
1281 
1282       for(i=0;i<sortklen;i++)
1283         {
1284           sprintf(junk2+strlen(junk2), "%02X ",(int)( 0xFF & sortk2[i]));
1285           sprintf(junk3+strlen(junk3), "%02X ",(int)( 0xFF & sortk3[i]));
1286         }
1287 
1288       log_verbose("%s\n", junk2);
1289       log_verbose("%s\n", junk3);
1290     }
1291 #endif
1292 
1293     free(sortk1);
1294     free(sortk2);
1295     free(sortk3);
1296 
1297     log_verbose("Use secondary comparison level testing ...\n");
1298     ucol_setStrength(col, UCOL_SECONDARY);
1299     sortklen=ucol_getSortKey(col, test1, u_strlen(test1),  NULL, 0);
1300     sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1301     ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortklen+1);
1302     sortklen=ucol_getSortKey(col, test2, u_strlen(test2),  NULL, 0);
1303     sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortklen+1));
1304     ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortklen+1);
1305 
1306     doAssert( !(memcmp(sortk1, sortk2, sortklen) > 0), "Result should be \"Abcda\" == \"abcda\"");
1307     doAssert( !(memcmp(sortk2, sortk1, sortklen) < 0), "Result should be \"abcda\" == \"Abcda\"");
1308     doAssert( (memcmp(sortk1, sortk2, sortklen) == 0), "Result should be \"abcda\" ==  \"abcda\"");
1309 
1310     log_verbose("getting sortkey for an empty string\n");
1311     ucol_setAttribute(col, UCOL_STRENGTH, UCOL_TERTIARY, &status);
1312     sortklen = ucol_getSortKey(col, test1, 0, NULL, 0);
1313     sortkEmpty = (uint8_t*)malloc(sizeof(uint8_t) * sortklen+1);
1314     sortklen = ucol_getSortKey(col, test1, 0, sortkEmpty, sortklen+1);
1315     if(sortklen != 3 || sortkEmpty[0] != 1 || sortkEmpty[0] != 1 || sortkEmpty[2] != 0) {
1316       log_err("Empty string generated wrong sortkey!\n");
1317     }
1318     free(sortkEmpty);
1319 
1320     log_verbose("testing passing invalid string\n");
1321     sortklen = ucol_getSortKey(col, NULL, 10, NULL, 0);
1322     if(sortklen != 0) {
1323       log_err("Invalid string didn't return sortkey size of 0\n");
1324     }
1325 
1326 
1327     log_verbose("testing sortkey ends...\n");
1328     ucol_close(col);
1329     free(test1);
1330     free(test2);
1331     free(test3);
1332     free(sortk1);
1333     free(sortk2);
1334 
1335 }
TestHashCode()1336 void TestHashCode()
1337 {
1338     uint8_t *sortk1, *sortk2, *sortk3;
1339     int32_t sortk1len, sortk2len, sortk3len;
1340     UCollator *col;
1341     UChar *test1, *test2, *test3;
1342     UErrorCode status = U_ZERO_ERROR;
1343     log_verbose("testing getHashCode begins...\n");
1344     col = ucol_open("en_US", &status);
1345     if (U_FAILURE(status)) {
1346         log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1347         return;
1348     }
1349     test1=(UChar*)malloc(sizeof(UChar) * 6);
1350     test2=(UChar*)malloc(sizeof(UChar) * 6);
1351     test3=(UChar*)malloc(sizeof(UChar) * 6);
1352     u_uastrcpy(test1, "Abcda");
1353     u_uastrcpy(test2, "abcda");
1354     u_uastrcpy(test3, "abcda");
1355 
1356     log_verbose("Use tertiary comparison level testing ....\n");
1357     sortk1len=ucol_getSortKey(col, test1, u_strlen(test1),  NULL, 0);
1358     sortk1=(uint8_t*)malloc(sizeof(uint8_t) * (sortk1len+1));
1359     ucol_getSortKey(col, test1, u_strlen(test1), sortk1, sortk1len+1);
1360     sortk2len=ucol_getSortKey(col, test2, u_strlen(test2),  NULL, 0);
1361     sortk2=(uint8_t*)malloc(sizeof(uint8_t) * (sortk2len+1));
1362     ucol_getSortKey(col, test2, u_strlen(test2), sortk2, sortk2len+1);
1363     sortk3len=ucol_getSortKey(col, test2, u_strlen(test3),  NULL, 0);
1364     sortk3=(uint8_t*)malloc(sizeof(uint8_t) * (sortk3len+1));
1365     ucol_getSortKey(col, test2, u_strlen(test2), sortk3, sortk3len+1);
1366 
1367 
1368     log_verbose("ucol_hashCode() testing ...\n");
1369 
1370     doAssert( ucol_keyHashCode(sortk1, sortk1len) != ucol_keyHashCode(sortk2, sortk2len), "Hash test1 result incorrect" );
1371     doAssert( !(ucol_keyHashCode(sortk1, sortk1len) == ucol_keyHashCode(sortk2, sortk2len)), "Hash test2 result incorrect" );
1372     doAssert( ucol_keyHashCode(sortk2, sortk2len) == ucol_keyHashCode(sortk3, sortk3len), "Hash result not equal" );
1373 
1374     log_verbose("hashCode tests end.\n");
1375     ucol_close(col);
1376     free(sortk1);
1377     free(sortk2);
1378     free(sortk3);
1379     free(test1);
1380     free(test2);
1381     free(test3);
1382 
1383 
1384 }
1385 /*
1386  *----------------------------------------------------------------------------
1387  * Tests the UCollatorElements API.
1388  *
1389  */
TestElemIter()1390 void TestElemIter()
1391 {
1392     int32_t offset;
1393     int32_t order1, order2, order3;
1394     UChar *testString1, *testString2;
1395     UCollator *col;
1396     UCollationElements *iterator1, *iterator2, *iterator3;
1397     UErrorCode status = U_ZERO_ERROR;
1398     log_verbose("testing UCollatorElements begins...\n");
1399     col = ucol_open("en_US", &status);
1400     ucol_setAttribute(col, UCOL_NORMALIZATION_MODE, UCOL_OFF, &status);
1401     if (U_FAILURE(status)) {
1402         log_err_status(status, "ERROR: Default collation creation failed.: %s\n", myErrorName(status));
1403         return;
1404     }
1405 
1406     testString1=(UChar*)malloc(sizeof(UChar) * 150);
1407     testString2=(UChar*)malloc(sizeof(UChar) * 150);
1408     u_uastrcpy(testString1, "XFILE What subset of all possible test cases has the highest probability of detecting the most errors?");
1409     u_uastrcpy(testString2, "Xf_ile What subset of all possible test cases has the lowest probability of detecting the least errors?");
1410 
1411     log_verbose("Constructors and comparison testing....\n");
1412 
1413     iterator1 = ucol_openElements(col, testString1, u_strlen(testString1), &status);
1414     if(U_FAILURE(status)) {
1415         log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1416         ucol_close(col);
1417         return;
1418     }
1419     else{ log_verbose("PASS: Default collationElement iterator1 creation passed\n");}
1420 
1421     iterator2 = ucol_openElements(col, testString1, u_strlen(testString1), &status);
1422     if(U_FAILURE(status)) {
1423         log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1424         ucol_close(col);
1425         return;
1426     }
1427     else{ log_verbose("PASS: Default collationElement iterator2 creation passed\n");}
1428 
1429     iterator3 = ucol_openElements(col, testString2, u_strlen(testString2), &status);
1430     if(U_FAILURE(status)) {
1431         log_err("ERROR: Default collationElement iterator creation failed.: %s\n", myErrorName(status));
1432         ucol_close(col);
1433         return;
1434     }
1435     else{ log_verbose("PASS: Default collationElement iterator3 creation passed\n");}
1436 
1437     offset=ucol_getOffset(iterator1);
1438     (void)offset;   /* Suppress set but not used warning. */
1439     ucol_setOffset(iterator1, 6, &status);
1440     if (U_FAILURE(status)) {
1441         log_err("Error in setOffset for UCollatorElements iterator.: %s\n", myErrorName(status));
1442         return;
1443     }
1444     if(ucol_getOffset(iterator1)==6)
1445         log_verbose("setOffset and getOffset working fine\n");
1446     else{
1447         log_err("error in set and get Offset got %d instead of 6\n", ucol_getOffset(iterator1));
1448     }
1449 
1450     ucol_setOffset(iterator1, 0, &status);
1451     order1 = ucol_next(iterator1, &status);
1452     if (U_FAILURE(status)) {
1453         log_err("Somehow ran out of memory stepping through the iterator1.: %s\n", myErrorName(status));
1454         return;
1455     }
1456     order2=ucol_getOffset(iterator2);
1457     doAssert((order1 != order2), "The first iterator advance failed");
1458     order2 = ucol_next(iterator2, &status);
1459     if (U_FAILURE(status)) {
1460         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1461         return;
1462     }
1463     order3 = ucol_next(iterator3, &status);
1464     if (U_FAILURE(status)) {
1465         log_err("Somehow ran out of memory stepping through the iterator3.: %s\n", myErrorName(status));
1466         return;
1467     }
1468 
1469     doAssert((order1 == order2), "The second iterator advance failed should be the same as first one");
1470 
1471 doAssert( (ucol_primaryOrder(order1) == ucol_primaryOrder(order3)), "The primary orders should be identical");
1472 doAssert( (ucol_secondaryOrder(order1) == ucol_secondaryOrder(order3)), "The secondary orders should be identical");
1473 doAssert( (ucol_tertiaryOrder(order1) == ucol_tertiaryOrder(order3)), "The tertiary orders should be identical");
1474 
1475     order1=ucol_next(iterator1, &status);
1476     if (U_FAILURE(status)) {
1477         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1478         return;
1479     }
1480     order3=ucol_next(iterator3, &status);
1481     if (U_FAILURE(status)) {
1482         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1483         return;
1484     }
1485 doAssert( (ucol_primaryOrder(order1) == ucol_primaryOrder(order3)), "The primary orders should be identical");
1486 doAssert( (ucol_tertiaryOrder(order1) != ucol_tertiaryOrder(order3)), "The tertiary orders should be different");
1487 
1488     order1=ucol_next(iterator1, &status);
1489     if (U_FAILURE(status)) {
1490         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1491         return;
1492     }
1493     order3=ucol_next(iterator3, &status);
1494     if (U_FAILURE(status)) {
1495         log_err("Somehow ran out of memory stepping through the iterator2.: %s\n", myErrorName(status));
1496         return;
1497     }
1498     /* this here, my friends, is either pure lunacy or something so obsolete that even it's mother
1499      * doesn't care about it. Essentially, this test complains if secondary values for 'I' and '_'
1500      * are the same. According to the UCA, this is not true. Therefore, remove the test.
1501      * Besides, if primary strengths for two code points are different, it doesn't matter one bit
1502      * what is the relation between secondary or any other strengths.
1503      * killed by weiv 06/11/2002.
1504      */
1505     /*
1506     doAssert( ((order1 & UCOL_SECONDARYMASK) != (order3 & UCOL_SECONDARYMASK)), "The secondary orders should be different");
1507     */
1508     doAssert( (order1 != UCOL_NULLORDER), "Unexpected end of iterator reached");
1509 
1510     free(testString1);
1511     free(testString2);
1512     ucol_closeElements(iterator1);
1513     ucol_closeElements(iterator2);
1514     ucol_closeElements(iterator3);
1515     ucol_close(col);
1516 
1517     log_verbose("testing CollationElementIterator ends...\n");
1518 }
1519 
TestGetLocale()1520 void TestGetLocale() {
1521   UErrorCode status = U_ZERO_ERROR;
1522   const char *rules = "&a<x<y<z";
1523   UChar rlz[256] = {0};
1524   uint32_t rlzLen = u_unescape(rules, rlz, 256);
1525 
1526   UCollator *coll = NULL;
1527   const char *locale = NULL;
1528 
1529   int32_t i = 0;
1530 
1531   static const struct {
1532     const char* requestedLocale;
1533     const char* validLocale;
1534     const char* actualLocale;
1535   } testStruct[] = {
1536     { "sr_RS", "sr_Cyrl_RS", "sr" },
1537     { "sh_YU", "sr_Latn_RS", "sr_Latn" }, /* was sh, then aliased to hr, now sr_Latn via import per cldrbug 5647: */
1538     { "en_BE_FOO", "en", "root" },
1539     { "sv_SE_NONEXISTANT", "sv", "sv" }
1540   };
1541 
1542   /* test opening collators for different locales */
1543   for(i = 0; i<UPRV_LENGTHOF(testStruct); i++) {
1544     status = U_ZERO_ERROR;
1545     coll = ucol_open(testStruct[i].requestedLocale, &status);
1546     if(U_FAILURE(status)) {
1547       log_err_status(status, "Failed to open collator for %s with %s\n", testStruct[i].requestedLocale, u_errorName(status));
1548       ucol_close(coll);
1549       continue;
1550     }
1551     /*
1552      * The requested locale may be the same as the valid locale,
1553      * or may not be supported at all. See ticket #10477.
1554      */
1555     locale = ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status);
1556     if(U_SUCCESS(status) &&
1557           strcmp(locale, testStruct[i].requestedLocale) != 0 && strcmp(locale, testStruct[i].validLocale) != 0) {
1558       log_err("[Coll %s]: Error in requested locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].requestedLocale, locale);
1559     }
1560     status = U_ZERO_ERROR;
1561     locale = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status);
1562     if(strcmp(locale, testStruct[i].validLocale) != 0) {
1563       log_err("[Coll %s]: Error in valid locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].validLocale, locale);
1564     }
1565     locale = ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status);
1566     if(strcmp(locale, testStruct[i].actualLocale) != 0) {
1567       log_err("[Coll %s]: Error in actual locale, expected %s, got %s\n", testStruct[i].requestedLocale, testStruct[i].actualLocale, locale);
1568     }
1569     ucol_close(coll);
1570   }
1571 
1572   /* completely non-existent locale for collator should get a root collator */
1573   {
1574     UCollator *defaultColl = ucol_open(NULL, &status);
1575     coll = ucol_open("blahaha", &status);
1576     if(U_SUCCESS(status)) {
1577       /* See comment above about ticket #10477.
1578       if(strcmp(ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status), "blahaha")) {
1579         log_err("Nonexisting locale didn't preserve the requested locale\n");
1580       } */
1581       const char *name = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status);
1582       if(*name != 0 && strcmp(name, "root") != 0) {
1583         log_err("Valid locale for nonexisting-locale collator is \"%s\" not root\n", name);
1584       }
1585       name = ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status);
1586       if(*name != 0 && strcmp(name, "root") != 0) {
1587         log_err("Actual locale for nonexisting-locale collator is \"%s\" not root\n", name);
1588       }
1589       ucol_close(coll);
1590       ucol_close(defaultColl);
1591     } else {
1592       log_data_err("Couldn't open collators\n");
1593     }
1594   }
1595 
1596 
1597 
1598   /* collator instantiated from rules should have all three locales NULL */
1599   coll = ucol_openRules(rlz, rlzLen, UCOL_DEFAULT, UCOL_DEFAULT, NULL, &status);
1600   if (coll != NULL) {
1601     locale = ucol_getLocaleByType(coll, ULOC_REQUESTED_LOCALE, &status);
1602     if(U_SUCCESS(status) && locale != NULL) {
1603       log_err("For collator instantiated from rules, requested locale returned %s instead of NULL\n", locale);
1604     }
1605     status = U_ZERO_ERROR;
1606     locale = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, &status);
1607     if(locale != NULL) {
1608       log_err("For collator instantiated from rules,  valid locale returned %s instead of NULL\n", locale);
1609     }
1610     locale = ucol_getLocaleByType(coll, ULOC_ACTUAL_LOCALE, &status);
1611     if(locale != NULL) {
1612       log_err("For collator instantiated from rules, actual locale returned %s instead of NULL\n", locale);
1613     }
1614     ucol_close(coll);
1615   } else {
1616     log_data_err("Couldn't get collator from ucol_openRules() - %s\n", u_errorName(status));
1617   }
1618 }
1619 
1620 
TestGetAll()1621 void TestGetAll()
1622 {
1623     int32_t i, count;
1624     count=ucol_countAvailable();
1625     /* use something sensible w/o hardcoding the count */
1626     if(count < 0){
1627         log_err("Error in countAvailable(), it returned %d\n", count);
1628     }
1629     else{
1630         log_verbose("PASS: countAvailable() successful, it returned %d\n", count);
1631     }
1632     for(i=0;i<count;i++)
1633         log_verbose("%s\n", ucol_getAvailable(i));
1634 
1635 
1636 }
1637 
1638 
1639 struct teststruct {
1640     const char *original;
1641     uint8_t key[256];
1642 } ;
1643 
compare_teststruct(const void * string1,const void * string2)1644 static int compare_teststruct(const void *string1, const void *string2) {
1645     return(strcmp((const char *)((struct teststruct *)string1)->key, (const char *)((struct teststruct *)string2)->key));
1646 }
1647 
TestBounds()1648 void TestBounds() {
1649     UErrorCode status = U_ZERO_ERROR;
1650 
1651     UCollator *coll = ucol_open("sh", &status);
1652 
1653     uint8_t sortkey[512], lower[512], upper[512];
1654     UChar buffer[512];
1655 
1656     static const char * const test[] = {
1657         "John Smith",
1658         "JOHN SMITH",
1659         "john SMITH",
1660         "j\\u00F6hn sm\\u00EFth",
1661         "J\\u00F6hn Sm\\u00EFth",
1662         "J\\u00D6HN SM\\u00CFTH",
1663         "john smithsonian",
1664         "John Smithsonian",
1665     };
1666 
1667     struct teststruct tests[] = {
1668         {"\\u010CAKI MIHALJ", {0}},
1669         {"\\u010CAKI MIHALJ", {0}},
1670         {"\\u010CAKI PIRO\\u0160KA", {0}},
1671         {"\\u010CABAI ANDRIJA", {0}},
1672         {"\\u010CABAI LAJO\\u0160", {0}},
1673         {"\\u010CABAI MARIJA", {0}},
1674         {"\\u010CABAI STEVAN", {0}},
1675         {"\\u010CABAI STEVAN", {0}},
1676         {"\\u010CABARKAPA BRANKO", {0}},
1677         {"\\u010CABARKAPA MILENKO", {0}},
1678         {"\\u010CABARKAPA MIROSLAV", {0}},
1679         {"\\u010CABARKAPA SIMO", {0}},
1680         {"\\u010CABARKAPA STANKO", {0}},
1681         {"\\u010CABARKAPA TAMARA", {0}},
1682         {"\\u010CABARKAPA TOMA\\u0160", {0}},
1683         {"\\u010CABDARI\\u0106 NIKOLA", {0}},
1684         {"\\u010CABDARI\\u0106 ZORICA", {0}},
1685         {"\\u010CABI NANDOR", {0}},
1686         {"\\u010CABOVI\\u0106 MILAN", {0}},
1687         {"\\u010CABRADI AGNEZIJA", {0}},
1688         {"\\u010CABRADI IVAN", {0}},
1689         {"\\u010CABRADI JELENA", {0}},
1690         {"\\u010CABRADI LJUBICA", {0}},
1691         {"\\u010CABRADI STEVAN", {0}},
1692         {"\\u010CABRDA MARTIN", {0}},
1693         {"\\u010CABRILO BOGDAN", {0}},
1694         {"\\u010CABRILO BRANISLAV", {0}},
1695         {"\\u010CABRILO LAZAR", {0}},
1696         {"\\u010CABRILO LJUBICA", {0}},
1697         {"\\u010CABRILO SPASOJA", {0}},
1698         {"\\u010CADE\\u0160 ZDENKA", {0}},
1699         {"\\u010CADESKI BLAGOJE", {0}},
1700         {"\\u010CADOVSKI VLADIMIR", {0}},
1701         {"\\u010CAGLJEVI\\u0106 TOMA", {0}},
1702         {"\\u010CAGOROVI\\u0106 VLADIMIR", {0}},
1703         {"\\u010CAJA VANKA", {0}},
1704         {"\\u010CAJI\\u0106 BOGOLJUB", {0}},
1705         {"\\u010CAJI\\u0106 BORISLAV", {0}},
1706         {"\\u010CAJI\\u0106 RADOSLAV", {0}},
1707         {"\\u010CAK\\u0160IRAN MILADIN", {0}},
1708         {"\\u010CAKAN EUGEN", {0}},
1709         {"\\u010CAKAN EVGENIJE", {0}},
1710         {"\\u010CAKAN IVAN", {0}},
1711         {"\\u010CAKAN JULIJAN", {0}},
1712         {"\\u010CAKAN MIHAJLO", {0}},
1713         {"\\u010CAKAN STEVAN", {0}},
1714         {"\\u010CAKAN VLADIMIR", {0}},
1715         {"\\u010CAKAN VLADIMIR", {0}},
1716         {"\\u010CAKAN VLADIMIR", {0}},
1717         {"\\u010CAKARA ANA", {0}},
1718         {"\\u010CAKAREVI\\u0106 MOMIR", {0}},
1719         {"\\u010CAKAREVI\\u0106 NEDELJKO", {0}},
1720         {"\\u010CAKI \\u0160ANDOR", {0}},
1721         {"\\u010CAKI AMALIJA", {0}},
1722         {"\\u010CAKI ANDRA\\u0160", {0}},
1723         {"\\u010CAKI LADISLAV", {0}},
1724         {"\\u010CAKI LAJO\\u0160", {0}},
1725         {"\\u010CAKI LASLO", {0}},
1726     };
1727 
1728 
1729 
1730     int32_t i = 0, j = 0, k = 0, buffSize = 0, skSize = 0, lowerSize = 0, upperSize = 0;
1731     int32_t arraySize = UPRV_LENGTHOF(tests);
1732 
1733     if(U_SUCCESS(status) && coll) {
1734         for(i = 0; i<arraySize; i++) {
1735             buffSize = u_unescape(tests[i].original, buffer, 512);
1736             skSize = ucol_getSortKey(coll, buffer, buffSize, tests[i].key, 512);
1737         }
1738 
1739         qsort(tests, arraySize, sizeof(struct teststruct), compare_teststruct);
1740 
1741         for(i = 0; i < arraySize-1; i++) {
1742             for(j = i+1; j < arraySize; j++) {
1743                 lowerSize = ucol_getBound(tests[i].key, -1, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1744                 upperSize = ucol_getBound(tests[j].key, -1, UCOL_BOUND_UPPER, 1, upper, 512, &status);
1745                 (void)lowerSize;    /* Suppress set but not used warning. */
1746                 (void)upperSize;
1747                 for(k = i; k <= j; k++) {
1748                     if(strcmp((const char *)lower, (const char *)tests[k].key) > 0) {
1749                         log_err("Problem with lower! j = %i (%s vs %s)\n", k, tests[k].original, tests[i].original);
1750                     }
1751                     if(strcmp((const char *)upper, (const char *)tests[k].key) <= 0) {
1752                         log_err("Problem with upper! j = %i (%s vs %s)\n", k, tests[k].original, tests[j].original);
1753                     }
1754                 }
1755             }
1756         }
1757 
1758 
1759 #if 0
1760         for(i = 0; i < 1000; i++) {
1761             lowerRND = (rand()/(RAND_MAX/arraySize));
1762             upperRND = lowerRND + (rand()/(RAND_MAX/(arraySize-lowerRND)));
1763 
1764             lowerSize = ucol_getBound(tests[lowerRND].key, -1, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1765             upperSize = ucol_getBound(tests[upperRND].key, -1, UCOL_BOUND_UPPER_LONG, 1, upper, 512, &status);
1766 
1767             for(j = lowerRND; j<=upperRND; j++) {
1768                 if(strcmp(lower, tests[j].key) > 0) {
1769                     log_err("Problem with lower! j = %i (%s vs %s)\n", j, tests[j].original, tests[lowerRND].original);
1770                 }
1771                 if(strcmp(upper, tests[j].key) <= 0) {
1772                     log_err("Problem with upper! j = %i (%s vs %s)\n", j, tests[j].original, tests[upperRND].original);
1773                 }
1774             }
1775         }
1776 #endif
1777 
1778 
1779 
1780 
1781 
1782         for(i = 0; i<UPRV_LENGTHOF(test); i++) {
1783             buffSize = u_unescape(test[i], buffer, 512);
1784             skSize = ucol_getSortKey(coll, buffer, buffSize, sortkey, 512);
1785             lowerSize = ucol_getBound(sortkey, skSize, UCOL_BOUND_LOWER, 1, lower, 512, &status);
1786             upperSize = ucol_getBound(sortkey, skSize, UCOL_BOUND_UPPER_LONG, 1, upper, 512, &status);
1787             for(j = i+1; j<UPRV_LENGTHOF(test); j++) {
1788                 buffSize = u_unescape(test[j], buffer, 512);
1789                 skSize = ucol_getSortKey(coll, buffer, buffSize, sortkey, 512);
1790                 if(strcmp((const char *)lower, (const char *)sortkey) > 0) {
1791                     log_err("Problem with lower! i = %i, j = %i (%s vs %s)\n", i, j, test[i], test[j]);
1792                 }
1793                 if(strcmp((const char *)upper, (const char *)sortkey) <= 0) {
1794                     log_err("Problem with upper! i = %i, j = %i (%s vs %s)\n", i, j, test[i], test[j]);
1795                 }
1796             }
1797         }
1798         ucol_close(coll);
1799     } else {
1800         log_data_err("Couldn't open collator\n");
1801     }
1802 
1803 }
1804 
doOverrunTest(UCollator * coll,const UChar * uString,int32_t strLen)1805 static void doOverrunTest(UCollator *coll, const UChar *uString, int32_t strLen) {
1806     int32_t skLen = 0, skLen2 = 0;
1807     uint8_t sortKey[256];
1808     int32_t i, j;
1809     uint8_t filler = 0xFF;
1810 
1811     skLen = ucol_getSortKey(coll, uString, strLen, NULL, 0);
1812 
1813     for(i = 0; i < skLen; i++) {
1814         memset(sortKey, filler, 256);
1815         skLen2 = ucol_getSortKey(coll, uString, strLen, sortKey, i);
1816         if(skLen != skLen2) {
1817             log_err("For buffer size %i, got different sortkey length. Expected %i got %i\n", i, skLen, skLen2);
1818         }
1819         for(j = i; j < 256; j++) {
1820             if(sortKey[j] != filler) {
1821                 log_err("Something run over index %i\n", j);
1822                 break;
1823             }
1824         }
1825     }
1826 }
1827 
1828 /* j1865 reports that if a shorter buffer is passed to
1829 * to get sort key, a buffer overrun happens in some
1830 * cases. This test tries to check this.
1831 */
TestSortKeyBufferOverrun(void)1832 void TestSortKeyBufferOverrun(void) {
1833     UErrorCode status = U_ZERO_ERROR;
1834     const char* cString = "A very Merry liTTle-lamB..";
1835     UChar uString[256];
1836     int32_t strLen = 0;
1837     UCollator *coll = ucol_open("root", &status);
1838     strLen = u_unescape(cString, uString, 256);
1839 
1840     if(U_SUCCESS(status)) {
1841         log_verbose("testing non ignorable\n");
1842         ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &status);
1843         doOverrunTest(coll, uString, strLen);
1844 
1845         log_verbose("testing shifted\n");
1846         ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_SHIFTED, &status);
1847         doOverrunTest(coll, uString, strLen);
1848 
1849         log_verbose("testing shifted quaternary\n");
1850         ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_QUATERNARY, &status);
1851         doOverrunTest(coll, uString, strLen);
1852 
1853         log_verbose("testing with french secondaries\n");
1854         ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_ON, &status);
1855         ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_TERTIARY, &status);
1856         ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &status);
1857         doOverrunTest(coll, uString, strLen);
1858 
1859     }
1860     ucol_close(coll);
1861 }
1862 
TestAttribute()1863 static void TestAttribute()
1864 {
1865     UErrorCode error = U_ZERO_ERROR;
1866     UCollator *coll = ucol_open(NULL, &error);
1867 
1868     if (U_FAILURE(error)) {
1869         log_err_status(error, "Creation of default collator failed\n");
1870         return;
1871     }
1872 
1873     ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_OFF, &error);
1874     if (ucol_getAttribute(coll, UCOL_FRENCH_COLLATION, &error) != UCOL_OFF ||
1875         U_FAILURE(error)) {
1876         log_err_status(error, "Setting and retrieving of the french collation failed\n");
1877     }
1878 
1879     ucol_setAttribute(coll, UCOL_FRENCH_COLLATION, UCOL_ON, &error);
1880     if (ucol_getAttribute(coll, UCOL_FRENCH_COLLATION, &error) != UCOL_ON ||
1881         U_FAILURE(error)) {
1882         log_err_status(error, "Setting and retrieving of the french collation failed\n");
1883     }
1884 
1885     ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_SHIFTED, &error);
1886     if (ucol_getAttribute(coll, UCOL_ALTERNATE_HANDLING, &error) != UCOL_SHIFTED ||
1887         U_FAILURE(error)) {
1888         log_err_status(error, "Setting and retrieving of the alternate handling failed\n");
1889     }
1890 
1891     ucol_setAttribute(coll, UCOL_ALTERNATE_HANDLING, UCOL_NON_IGNORABLE, &error);
1892     if (ucol_getAttribute(coll, UCOL_ALTERNATE_HANDLING, &error) != UCOL_NON_IGNORABLE ||
1893         U_FAILURE(error)) {
1894         log_err_status(error, "Setting and retrieving of the alternate handling failed\n");
1895     }
1896 
1897     ucol_setAttribute(coll, UCOL_CASE_FIRST, UCOL_LOWER_FIRST, &error);
1898     if (ucol_getAttribute(coll, UCOL_CASE_FIRST, &error) != UCOL_LOWER_FIRST ||
1899         U_FAILURE(error)) {
1900         log_err_status(error, "Setting and retrieving of the case first attribute failed\n");
1901     }
1902 
1903     ucol_setAttribute(coll, UCOL_CASE_FIRST, UCOL_UPPER_FIRST, &error);
1904     if (ucol_getAttribute(coll, UCOL_CASE_FIRST, &error) != UCOL_UPPER_FIRST ||
1905         U_FAILURE(error)) {
1906         log_err_status(error, "Setting and retrieving of the case first attribute failed\n");
1907     }
1908 
1909     ucol_setAttribute(coll, UCOL_CASE_LEVEL, UCOL_ON, &error);
1910     if (ucol_getAttribute(coll, UCOL_CASE_LEVEL, &error) != UCOL_ON ||
1911         U_FAILURE(error)) {
1912         log_err_status(error, "Setting and retrieving of the case level attribute failed\n");
1913     }
1914 
1915     ucol_setAttribute(coll, UCOL_CASE_LEVEL, UCOL_OFF, &error);
1916     if (ucol_getAttribute(coll, UCOL_CASE_LEVEL, &error) != UCOL_OFF ||
1917         U_FAILURE(error)) {
1918         log_err_status(error, "Setting and retrieving of the case level attribute failed\n");
1919     }
1920 
1921     ucol_setAttribute(coll, UCOL_NORMALIZATION_MODE, UCOL_ON, &error);
1922     if (ucol_getAttribute(coll, UCOL_NORMALIZATION_MODE, &error) != UCOL_ON ||
1923         U_FAILURE(error)) {
1924         log_err_status(error, "Setting and retrieving of the normalization on/off attribute failed\n");
1925     }
1926 
1927     ucol_setAttribute(coll, UCOL_NORMALIZATION_MODE, UCOL_OFF, &error);
1928     if (ucol_getAttribute(coll, UCOL_NORMALIZATION_MODE, &error) != UCOL_OFF ||
1929         U_FAILURE(error)) {
1930         log_err_status(error, "Setting and retrieving of the normalization on/off attribute failed\n");
1931     }
1932 
1933     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_PRIMARY, &error);
1934     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_PRIMARY ||
1935         U_FAILURE(error)) {
1936         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1937     }
1938 
1939     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_SECONDARY, &error);
1940     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_SECONDARY ||
1941         U_FAILURE(error)) {
1942         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1943     }
1944 
1945     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_TERTIARY, &error);
1946     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_TERTIARY ||
1947         U_FAILURE(error)) {
1948         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1949     }
1950 
1951     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_QUATERNARY, &error);
1952     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_QUATERNARY ||
1953         U_FAILURE(error)) {
1954         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1955     }
1956 
1957     ucol_setAttribute(coll, UCOL_STRENGTH, UCOL_IDENTICAL, &error);
1958     if (ucol_getAttribute(coll, UCOL_STRENGTH, &error) != UCOL_IDENTICAL ||
1959         U_FAILURE(error)) {
1960         log_err_status(error, "Setting and retrieving of the collation strength failed\n");
1961     }
1962 
1963     ucol_close(coll);
1964 }
1965 
TestGetTailoredSet()1966 void TestGetTailoredSet() {
1967   struct {
1968     const char *rules;
1969     const char *tests[20];
1970     int32_t testsize;
1971   } setTest[] = {
1972     { "&a < \\u212b", { "\\u212b", "A\\u030a", "\\u00c5" }, 3},
1973     { "& S < \\u0161 <<< \\u0160", { "\\u0161", "s\\u030C", "\\u0160", "S\\u030C" }, 4}
1974   };
1975 
1976   int32_t i = 0, j = 0;
1977   UErrorCode status = U_ZERO_ERROR;
1978   UParseError pError;
1979 
1980   UCollator *coll = NULL;
1981   UChar buff[1024];
1982   int32_t buffLen = 0;
1983   USet *set = NULL;
1984 
1985   for(i = 0; i < UPRV_LENGTHOF(setTest); i++) {
1986     buffLen = u_unescape(setTest[i].rules, buff, 1024);
1987     coll = ucol_openRules(buff, buffLen, UCOL_DEFAULT, UCOL_DEFAULT, &pError, &status);
1988     if(U_SUCCESS(status)) {
1989       set = ucol_getTailoredSet(coll, &status);
1990       if(uset_size(set) < setTest[i].testsize) {
1991         log_err("Tailored set size smaller (%d) than expected (%d)\n", uset_size(set), setTest[i].testsize);
1992       }
1993       for(j = 0; j < setTest[i].testsize; j++) {
1994         buffLen = u_unescape(setTest[i].tests[j], buff, 1024);
1995         if(!uset_containsString(set, buff, buffLen)) {
1996           log_err("Tailored set doesn't contain %s... It should\n", setTest[i].tests[j]);
1997         }
1998       }
1999       uset_close(set);
2000     } else {
2001       log_err_status(status, "Couldn't open collator with rules %s\n", setTest[i].rules);
2002     }
2003     ucol_close(coll);
2004   }
2005 }
2006 
tMemCmp(const uint8_t * first,const uint8_t * second)2007 static int tMemCmp(const uint8_t *first, const uint8_t *second) {
2008    int32_t firstLen = (int32_t)strlen((const char *)first);
2009    int32_t secondLen = (int32_t)strlen((const char *)second);
2010    return memcmp(first, second, uprv_min(firstLen, secondLen));
2011 }
2012 static const char * strengthsC[] = {
2013      "UCOL_PRIMARY",
2014      "UCOL_SECONDARY",
2015      "UCOL_TERTIARY",
2016      "UCOL_QUATERNARY",
2017      "UCOL_IDENTICAL"
2018 };
2019 
TestMergeSortKeys(void)2020 void TestMergeSortKeys(void) {
2021    UErrorCode status = U_ZERO_ERROR;
2022    UCollator *coll = ucol_open("en", &status);
2023    if(U_SUCCESS(status)) {
2024 
2025      const char* cases[] = {
2026        "abc",
2027          "abcd",
2028          "abcde"
2029      };
2030      uint32_t casesSize = UPRV_LENGTHOF(cases);
2031      const char* prefix = "foo";
2032      const char* suffix = "egg";
2033      char outBuff1[256], outBuff2[256];
2034 
2035      uint8_t **sortkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
2036      uint8_t **mergedPrefixkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
2037      uint8_t **mergedSuffixkeys = (uint8_t **)malloc(casesSize*sizeof(uint8_t *));
2038      uint32_t *sortKeysLen = (uint32_t *)malloc(casesSize*sizeof(uint32_t));
2039      uint8_t prefixKey[256], suffixKey[256];
2040      uint32_t prefixKeyLen = 0, suffixKeyLen = 0, i = 0;
2041      UChar buffer[256];
2042      uint32_t unescapedLen = 0, l1 = 0, l2 = 0;
2043      UColAttributeValue strength;
2044 
2045      log_verbose("ucol_mergeSortkeys test\n");
2046      log_verbose("Testing order of the test cases\n");
2047      genericLocaleStarter("en", cases, casesSize);
2048 
2049      for(i = 0; i<casesSize; i++) {
2050        sortkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
2051        mergedPrefixkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
2052        mergedSuffixkeys[i] = (uint8_t *)malloc(256*sizeof(uint8_t));
2053      }
2054 
2055      unescapedLen = u_unescape(prefix, buffer, 256);
2056      prefixKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, prefixKey, 256);
2057 
2058      unescapedLen = u_unescape(suffix, buffer, 256);
2059      suffixKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, suffixKey, 256);
2060 
2061      log_verbose("Massaging data with prefixes and different strengths\n");
2062      strength = UCOL_PRIMARY;
2063      while(strength <= UCOL_IDENTICAL) {
2064        log_verbose("Strength %s\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
2065        ucol_setAttribute(coll, UCOL_STRENGTH, strength, &status);
2066        for(i = 0; i<casesSize; i++) {
2067          unescapedLen = u_unescape(cases[i], buffer, 256);
2068          sortKeysLen[i] = ucol_getSortKey(coll, buffer, unescapedLen, sortkeys[i], 256);
2069          ucol_mergeSortkeys(prefixKey, prefixKeyLen, sortkeys[i], sortKeysLen[i], mergedPrefixkeys[i], 256);
2070          ucol_mergeSortkeys(sortkeys[i], sortKeysLen[i], suffixKey, suffixKeyLen, mergedSuffixkeys[i], 256);
2071          if(i>0) {
2072            if(tMemCmp(mergedPrefixkeys[i-1], mergedPrefixkeys[i]) >= 0) {
2073              log_err("Error while comparing prefixed keys @ strength %s:\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
2074              log_err("%s\n%s\n",
2075                          ucol_sortKeyToString(coll, mergedPrefixkeys[i-1], outBuff1, l1),
2076                          ucol_sortKeyToString(coll, mergedPrefixkeys[i], outBuff2, l2));
2077            }
2078            if(tMemCmp(mergedSuffixkeys[i-1], mergedSuffixkeys[i]) >= 0) {
2079              log_err("Error while comparing suffixed keys @ strength %s:\n", strengthsC[strength<=UCOL_QUATERNARY?strength:4]);
2080              log_err("%s\n%s\n",
2081                          ucol_sortKeyToString(coll, mergedSuffixkeys[i-1], outBuff1, l1),
2082                          ucol_sortKeyToString(coll, mergedSuffixkeys[i], outBuff2, l2));
2083            }
2084          }
2085        }
2086        if(strength == UCOL_QUATERNARY) {
2087          strength = UCOL_IDENTICAL;
2088        } else {
2089          strength++;
2090        }
2091      }
2092 
2093      {
2094        uint8_t smallBuf[3];
2095        uint32_t reqLen = 0;
2096        log_verbose("testing buffer overflow\n");
2097        reqLen = ucol_mergeSortkeys(prefixKey, prefixKeyLen, suffixKey, suffixKeyLen, smallBuf, 3);
2098        if(reqLen != (prefixKeyLen+suffixKeyLen)) {
2099          log_err("Wrong preflight size for merged sortkey\n");
2100        }
2101      }
2102 
2103      {
2104        UChar empty = 0;
2105        uint8_t emptyKey[20], abcKey[50], mergedKey[100];
2106        int32_t emptyKeyLen = 0, abcKeyLen = 0, mergedKeyLen = 0;
2107 
2108        log_verbose("testing merging with sortkeys generated for empty strings\n");
2109        emptyKeyLen = ucol_getSortKey(coll, &empty, 0, emptyKey, 20);
2110        unescapedLen = u_unescape(cases[0], buffer, 256);
2111        abcKeyLen = ucol_getSortKey(coll, buffer, unescapedLen, abcKey, 50);
2112        mergedKeyLen = ucol_mergeSortkeys(emptyKey, emptyKeyLen, abcKey, abcKeyLen, mergedKey, 100);
2113        if(mergedKey[0] != 2) {
2114          log_err("Empty sortkey didn't produce a level separator\n");
2115        }
2116        /* try with zeros */
2117        mergedKeyLen = ucol_mergeSortkeys(emptyKey, 0, abcKey, abcKeyLen, mergedKey, 100);
2118        if(mergedKeyLen != 0 || mergedKey[0] != 0) {
2119          log_err("Empty key didn't produce null mergedKey\n");
2120        }
2121        mergedKeyLen = ucol_mergeSortkeys(abcKey, abcKeyLen, emptyKey, 0, mergedKey, 100);
2122        if(mergedKeyLen != 0 || mergedKey[0] != 0) {
2123          log_err("Empty key didn't produce null mergedKey\n");
2124        }
2125 
2126      }
2127 
2128      for(i = 0; i<casesSize; i++) {
2129        free(sortkeys[i]);
2130        free(mergedPrefixkeys[i]);
2131        free(mergedSuffixkeys[i]);
2132      }
2133      free(sortkeys);
2134      free(mergedPrefixkeys);
2135      free(mergedSuffixkeys);
2136      free(sortKeysLen);
2137      ucol_close(coll);
2138      /* need to finish this up */
2139    } else {
2140      log_data_err("Couldn't open collator");
2141    }
2142 }
TestShortString(void)2143 static void TestShortString(void)
2144 {
2145     struct {
2146         const char *input;
2147         const char *expectedOutput;
2148         const char *locale;
2149         UErrorCode expectedStatus;
2150         int32_t    expectedOffset;
2151         uint32_t   expectedIdentifier;
2152     } testCases[] = {
2153         /*
2154          * Note: The first test case sets variableTop to the dollar sign '$'.
2155          * We have agreed to drop support for variableTop in ucol_getShortDefinitionString(),
2156          * related to ticket #10372 "deprecate collation APIs for short definition strings",
2157          * and because it did not work for most spaces/punctuation/symbols,
2158          * as documented in ticket #10386 "collation short definition strings issues":
2159          * The old code wrote only 3 hex digits for primary weights below 0x0FFF,
2160          * which is a syntax error, and then failed to normalize the result.
2161          *
2162          * The "B2700" was removed from the expected result ("B2700_KPHONEBOOK_LDE").
2163          *
2164          * Previously, this test had to be adjusted for root collator changes because the
2165          * primary weight of the variable top character naturally changed
2166          * but was baked into the expected result.
2167          */
2168         {"LDE_RDE_KPHONEBOOK_T0024_ZLATN","KPHONEBOOK_LDE", "de@collation=phonebook", U_USING_FALLBACK_WARNING, 0, 0 },
2169 
2170         {"LEN_RUS_NO_AS_S4","AS_LROOT_NO_S4", NULL, U_USING_DEFAULT_WARNING, 0, 0 },
2171         // uloc_canonicalize("de__PHONEBOOK") used to return "de@collation=phonebook"
2172         // and we got U_ZERO_ERROR.
2173         // Since ICU-20187 "drop support for long-obsolete locale ID variants..."
2174         // we actually load the "de__PHONEBOOK" bundle and fall back to "de".
2175         {"LDE_VPHONEBOOK_EO_SI","EO_KPHONEBOOK_LDE_SI", "de@collation=phonebook", U_USING_FALLBACK_WARNING, 0, 0 },
2176         {"LDE_Kphonebook","KPHONEBOOK_LDE", "de@collation=phonebook", U_ZERO_ERROR, 0, 0 },
2177         {"Xqde_DE@collation=phonebookq_S3_EX","KPHONEBOOK_LDE", "de@collation=phonebook", U_USING_FALLBACK_WARNING, 0, 0 },
2178         {"LFR_FO", "FO_LROOT", NULL, U_USING_DEFAULT_WARNING, 0, 0 },
2179         {"SO_LX_AS", "", NULL, U_ILLEGAL_ARGUMENT_ERROR, 8, 0 },
2180         {"S3_ASS_MMM", "", NULL, U_ILLEGAL_ARGUMENT_ERROR, 5, 0 }
2181     };
2182 
2183     int32_t i = 0;
2184     UCollator *coll = NULL, *fromNormalized = NULL;
2185     UParseError parseError;
2186     UErrorCode status = U_ZERO_ERROR;
2187     char fromShortBuffer[256], normalizedBuffer[256], fromNormalizedBuffer[256];
2188     const char* locale = NULL;
2189 
2190 
2191     for(i = 0; i < UPRV_LENGTHOF(testCases); i++) {
2192         status = U_ZERO_ERROR;
2193         if(testCases[i].locale) {
2194             locale = testCases[i].locale;
2195         } else {
2196             locale = NULL;
2197         }
2198 
2199         coll = ucol_openFromShortString(testCases[i].input, FALSE, &parseError, &status);
2200         if(status != testCases[i].expectedStatus) {
2201             log_err_status(status, "Got status '%s' that is different from expected '%s' for '%s'\n",
2202                 u_errorName(status), u_errorName(testCases[i].expectedStatus), testCases[i].input);
2203             continue;
2204         }
2205 
2206         if(U_SUCCESS(status)) {
2207             ucol_getShortDefinitionString(coll, locale, fromShortBuffer, 256, &status);
2208 
2209             if(strcmp(fromShortBuffer, testCases[i].expectedOutput)) {
2210                 log_err("Got short string '%s' from the collator. Expected '%s' for input '%s'\n",
2211                     fromShortBuffer, testCases[i].expectedOutput, testCases[i].input);
2212             }
2213 
2214             ucol_normalizeShortDefinitionString(testCases[i].input, normalizedBuffer, 256, &parseError, &status);
2215             fromNormalized = ucol_openFromShortString(normalizedBuffer, FALSE, &parseError, &status);
2216             ucol_getShortDefinitionString(fromNormalized, locale, fromNormalizedBuffer, 256, &status);
2217 
2218             if(strcmp(fromShortBuffer, fromNormalizedBuffer)) {
2219                 log_err("Strings obtained from collators instantiated by short string ('%s') and from normalized string ('%s') differ\n",
2220                     fromShortBuffer, fromNormalizedBuffer);
2221             }
2222 
2223 
2224             if(!ucol_equals(coll, fromNormalized)) {
2225                 log_err("Collator from short string ('%s') differs from one obtained through a normalized version ('%s')\n",
2226                     testCases[i].input, normalizedBuffer);
2227             }
2228 
2229             ucol_close(fromNormalized);
2230             ucol_close(coll);
2231 
2232         } else {
2233             if(parseError.offset != testCases[i].expectedOffset) {
2234                 log_err("Got parse error offset %i, but expected %i instead for '%s'\n",
2235                     parseError.offset, testCases[i].expectedOffset, testCases[i].input);
2236             }
2237         }
2238     }
2239 
2240 }
2241 
2242 static void
doSetsTest(const char * locale,const USet * ref,USet * set,const char * inSet,const char * outSet,UErrorCode * status)2243 doSetsTest(const char *locale, const USet *ref, USet *set, const char* inSet, const char* outSet, UErrorCode *status) {
2244     UChar buffer[65536];
2245     int32_t bufLen;
2246 
2247     uset_clear(set);
2248     bufLen = u_unescape(inSet, buffer, 512);
2249     uset_applyPattern(set, buffer, bufLen, 0, status);
2250     if(U_FAILURE(*status)) {
2251         log_err("%s: Failure setting pattern %s\n", locale, u_errorName(*status));
2252     }
2253 
2254     if(!uset_containsAll(ref, set)) {
2255         log_err("%s: Some stuff from %s is not present in the set\n", locale, inSet);
2256         uset_removeAll(set, ref);
2257         bufLen = uset_toPattern(set, buffer, UPRV_LENGTHOF(buffer), TRUE, status);
2258         log_info("    missing: %s\n", aescstrdup(buffer, bufLen));
2259         bufLen = uset_toPattern(ref, buffer, UPRV_LENGTHOF(buffer), TRUE, status);
2260         log_info("    total: size=%i  %s\n", uset_getItemCount(ref), aescstrdup(buffer, bufLen));
2261     }
2262 
2263     uset_clear(set);
2264     bufLen = u_unescape(outSet, buffer, 512);
2265     uset_applyPattern(set, buffer, bufLen, 0, status);
2266     if(U_FAILURE(*status)) {
2267         log_err("%s: Failure setting pattern %s\n", locale, u_errorName(*status));
2268     }
2269 
2270     if(!uset_containsNone(ref, set)) {
2271         log_err("%s: Some stuff from %s is present in the set\n", locale, outSet);
2272     }
2273 }
2274 
2275 
2276 
2277 
2278 static void
TestGetContractionsAndUnsafes(void)2279 TestGetContractionsAndUnsafes(void)
2280 {
2281     static struct {
2282         const char* locale;
2283         const char* inConts;
2284         const char* outConts;
2285         const char* inExp;
2286         const char* outExp;
2287         const char* unsafeCodeUnits;
2288         const char* safeCodeUnits;
2289     } tests[] = {
2290         { "ru",
2291             "[{\\u0418\\u0306}{\\u0438\\u0306}]",
2292             "[\\u0439\\u0457]",
2293             "[\\u00e6]",
2294             "[ae]",
2295             "[\\u0418\\u0438]",
2296             "[aAbB\\u0430\\u0410\\u0433\\u0413]"
2297         },
2298         { "uk",
2299             "[{\\u0406\\u0308}{\\u0456\\u0308}{\\u0418\\u0306}{\\u0438\\u0306}]",
2300             "[\\u0407\\u0419\\u0439\\u0457]",
2301             "[\\u00e6]",
2302             "[ae]",
2303             "[\\u0406\\u0456\\u0418\\u0438]",
2304             "[aAbBxv]",
2305         },
2306         { "sh",
2307             "[{C\\u0301}{C\\u030C}{C\\u0341}{DZ\\u030C}{Dz\\u030C}{D\\u017D}{D\\u017E}{lj}{nj}]",
2308             "[{\\u309d\\u3099}{\\u30fd\\u3099}]",
2309             "[\\u00e6]",
2310             "[a]",
2311             "[nlcdzNLCDZ]",
2312             "[jabv]"
2313         },
2314         { "ja",
2315           /*
2316            * The "collv2" builder omits mappings if the collator maps their
2317            * character sequences to the same CEs.
2318            * For example, it omits Japanese contractions for NFD forms
2319            * of the voiced iteration mark (U+309E = U+309D + U+3099), such as
2320            * {\\u3053\\u3099\\u309D\\u3099}{\\u3053\\u309D\\u3099}
2321            * {\\u30B3\\u3099\\u30FD\\u3099}{\\u30B3\\u30FD\\u3099}.
2322            * It does add mappings for the precomposed forms.
2323            */
2324           "[{\\u3053\\u3099\\u309D}{\\u3053\\u3099\\u309E}{\\u3053\\u3099\\u30FC}"
2325            "{\\u3053\\u309D}{\\u3053\\u309E}{\\u3053\\u30FC}"
2326            "{\\u30B3\\u3099\\u30FC}{\\u30B3\\u3099\\u30FD}{\\u30B3\\u3099\\u30FE}"
2327            "{\\u30B3\\u30FC}{\\u30B3\\u30FD}{\\u30B3\\u30FE}]",
2328           "[{\\u30FD\\u3099}{\\u309D\\u3099}{\\u3053\\u3099}{\\u30B3\\u3099}{lj}{nj}]",
2329             "[\\u30FE\\u00e6]",
2330             "[a]",
2331             "[\\u3099]",
2332             "[]"
2333         }
2334     };
2335 
2336     UErrorCode status = U_ZERO_ERROR;
2337     UCollator *coll = NULL;
2338     int32_t i = 0;
2339     int32_t noConts = 0;
2340     USet *conts = uset_open(0,0);
2341     USet *exp = uset_open(0, 0);
2342     USet *set  = uset_open(0,0);
2343     int32_t setBufferLen = 65536;
2344     UChar buffer[65536];
2345     int32_t setLen = 0;
2346 
2347     for(i = 0; i < UPRV_LENGTHOF(tests); i++) {
2348         log_verbose("Testing locale: %s\n", tests[i].locale);
2349         coll = ucol_open(tests[i].locale, &status);
2350         if (coll == NULL || U_FAILURE(status)) {
2351             log_err_status(status, "Unable to open collator for locale %s ==> %s\n", tests[i].locale, u_errorName(status));
2352             continue;
2353         }
2354         ucol_getContractionsAndExpansions(coll, conts, exp, TRUE, &status);
2355         doSetsTest(tests[i].locale, conts, set, tests[i].inConts, tests[i].outConts, &status);
2356         setLen = uset_toPattern(conts, buffer, setBufferLen, TRUE, &status);
2357         if(U_SUCCESS(status)) {
2358             /*log_verbose("Contractions %i: %s\n", uset_getItemCount(conts), aescstrdup(buffer, setLen));*/
2359         } else {
2360             log_err("error %s. %i\n", u_errorName(status), setLen);
2361             status = U_ZERO_ERROR;
2362         }
2363         doSetsTest(tests[i].locale, exp, set, tests[i].inExp, tests[i].outExp, &status);
2364         setLen = uset_toPattern(exp, buffer, setBufferLen, TRUE, &status);
2365         if(U_SUCCESS(status)) {
2366             /*log_verbose("Expansions %i: %s\n", uset_getItemCount(exp), aescstrdup(buffer, setLen));*/
2367         } else {
2368             log_err("error %s. %i\n", u_errorName(status), setLen);
2369             status = U_ZERO_ERROR;
2370         }
2371 
2372         noConts = ucol_getUnsafeSet(coll, conts, &status);
2373         (void)noConts;   /* Suppress set but not used warning */
2374         doSetsTest(tests[i].locale, conts, set, tests[i].unsafeCodeUnits, tests[i].safeCodeUnits, &status);
2375         setLen = uset_toPattern(conts, buffer, setBufferLen, TRUE, &status);
2376         if(U_SUCCESS(status)) {
2377             log_verbose("Unsafe %i: %s\n", uset_getItemCount(exp), aescstrdup(buffer, setLen));
2378         } else {
2379             log_err("error %s. %i\n", u_errorName(status), setLen);
2380             status = U_ZERO_ERROR;
2381         }
2382 
2383         ucol_close(coll);
2384     }
2385 
2386 
2387     uset_close(conts);
2388     uset_close(exp);
2389     uset_close(set);
2390 }
2391 
2392 static void
TestOpenBinary(void)2393 TestOpenBinary(void)
2394 {
2395     /*
2396      * ucol_openBinary() documents:
2397      * "The API also takes a base collator which usually should be UCA."
2398      * and
2399      * "Currently it cannot be NULL."
2400      *
2401      * However, the check for NULL was commented out in ICU 3.4 (r18149).
2402      * Ticket #4355 requested "Make collation work with minimal data.
2403      * Optionally without UCA, with relevant parts of UCA copied into the tailoring table."
2404      *
2405      * The ICU team agreed with ticket #10517 "require base collator in ucol_openBinary() etc."
2406      * to require base!=NULL again.
2407      */
2408 #define OPEN_BINARY_ACCEPTS_NULL_BASE 0
2409     UErrorCode status = U_ZERO_ERROR;
2410     /*
2411     char rule[] = "&h < d < c < b";
2412     char *wUCA[] = { "a", "h", "d", "c", "b", "i" };
2413     char *noUCA[] = {"d", "c", "b", "a", "h", "i" };
2414     */
2415     /* we have to use Cyrillic letters because latin-1 always gets copied */
2416     const char rule[] = "&\\u0452 < \\u0434 < \\u0433 < \\u0432"; /* &dje < d < g < v */
2417     const char *wUCA[] = { "\\u0430", "\\u0452", "\\u0434", "\\u0433", "\\u0432", "\\u0435" }; /* a, dje, d, g, v, e */
2418 #if OPEN_BINARY_ACCEPTS_NULL_BASE
2419     const char *noUCA[] = {"\\u0434", "\\u0433", "\\u0432", "\\u0430", "\\u0435", "\\u0452" }; /* d, g, v, a, e, dje */
2420 #endif
2421 
2422     UChar uRules[256];
2423     int32_t uRulesLen = u_unescape(rule, uRules, 256);
2424 
2425     UCollator *coll = ucol_openRules(uRules, uRulesLen, UCOL_DEFAULT, UCOL_DEFAULT, NULL, &status);
2426     UCollator *UCA = NULL;
2427     UCollator *cloneNOUCA = NULL, *cloneWUCA = NULL;
2428 
2429     uint8_t imageBuffer[32768];
2430     uint8_t *image = imageBuffer;
2431     int32_t imageBufferCapacity = 32768;
2432 
2433     int32_t imageSize;
2434 
2435     if((coll==NULL)||(U_FAILURE(status))) {
2436         log_data_err("could not load collators or error occurred: %s\n",
2437             u_errorName(status));
2438         return;
2439     }
2440     UCA = ucol_open("root", &status);
2441     if((UCA==NULL)||(U_FAILURE(status))) {
2442         log_data_err("could not load UCA collator or error occurred: %s\n",
2443             u_errorName(status));
2444         return;
2445     }
2446     imageSize = ucol_cloneBinary(coll, image, imageBufferCapacity, &status);
2447     if(U_FAILURE(status)) {
2448         image = (uint8_t *)malloc(imageSize*sizeof(uint8_t));
2449         status = U_ZERO_ERROR;
2450         imageSize = ucol_cloneBinary(coll, imageBuffer, imageSize, &status);
2451     }
2452 
2453 
2454     cloneWUCA = ucol_openBinary(image, imageSize, UCA, &status);
2455     cloneNOUCA = ucol_openBinary(image, imageSize, NULL, &status);
2456 #if !OPEN_BINARY_ACCEPTS_NULL_BASE
2457     if(status != U_ILLEGAL_ARGUMENT_ERROR) {
2458         log_err("ucol_openBinary(base=NULL) unexpectedly did not fail - %s\n", u_errorName(status));
2459     }
2460 #endif
2461 
2462     genericOrderingTest(coll, wUCA, UPRV_LENGTHOF(wUCA));
2463 
2464     genericOrderingTest(cloneWUCA, wUCA, UPRV_LENGTHOF(wUCA));
2465 #if OPEN_BINARY_ACCEPTS_NULL_BASE
2466     genericOrderingTest(cloneNOUCA, noUCA, UPRV_LENGTHOF(noUCA));
2467 #endif
2468 
2469     if(image != imageBuffer) {
2470         free(image);
2471     }
2472     ucol_close(coll);
2473     ucol_close(cloneNOUCA);
2474     ucol_close(cloneWUCA);
2475     ucol_close(UCA);
2476 }
2477 
TestDefault(void)2478 static void TestDefault(void) {
2479     /* Tests for code coverage. */
2480     UErrorCode status = U_ZERO_ERROR;
2481     UCollator *coll = ucol_open("es@collation=pinyin", &status);
2482     if (coll == NULL || status == U_FILE_ACCESS_ERROR) {
2483         log_data_err("Unable to open collator es@collation=pinyin\n");
2484         return;
2485     }
2486     if (status != U_USING_DEFAULT_WARNING) {
2487         /* What do you mean that you know about using pinyin collation in Spanish!? This should be in the zh locale. */
2488         log_err("es@collation=pinyin should return U_USING_DEFAULT_WARNING, but returned %s\n", u_errorName(status));
2489     }
2490     ucol_close(coll);
2491     if (ucol_getKeywordValues("funky", &status) != NULL) {
2492         log_err("Collators should not know about the funky keyword.\n");
2493     }
2494     if (status != U_ILLEGAL_ARGUMENT_ERROR) {
2495         log_err("funky keyword didn't fail as expected %s\n", u_errorName(status));
2496     }
2497     if (ucol_getKeywordValues("collation", &status) != NULL) {
2498         log_err("ucol_getKeywordValues should not work when given a bad status.\n");
2499     }
2500 }
2501 
TestDefaultKeyword(void)2502 static void TestDefaultKeyword(void) {
2503     /* Tests for code coverage. */
2504     UErrorCode status = U_ZERO_ERROR;
2505     const char *loc = "zh_TW@collation=default";
2506     UCollator *coll = ucol_open(loc, &status);
2507     if(U_FAILURE(status)) {
2508         log_info("Warning: ucol_open(%s, ...) returned %s, at least it didn't crash.\n", loc, u_errorName(status));
2509     } else if (status != U_USING_FALLBACK_WARNING) {
2510         /* Hmm, skip the following test for CLDR 1.9 data and/or ICU 4.6, no longer seems to apply */
2511         #if 0
2512         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));
2513         #endif
2514     }
2515     ucol_close(coll);
2516 }
2517 
uenum_contains(UEnumeration * e,const char * s,UErrorCode * status)2518 static UBool uenum_contains(UEnumeration *e, const char *s, UErrorCode *status) {
2519     const char *t;
2520     uenum_reset(e, status);
2521     while(((t = uenum_next(e, NULL, status)) != NULL) && U_SUCCESS(*status)) {
2522         if(uprv_strcmp(s, t) == 0) {
2523             return TRUE;
2524         }
2525     }
2526     return FALSE;
2527 }
2528 
TestGetKeywordValuesForLocale(void)2529 static void TestGetKeywordValuesForLocale(void) {
2530 #define MAX_NUMBER_OF_KEYWORDS 9
2531     const char *PREFERRED[][MAX_NUMBER_OF_KEYWORDS+1] = {
2532             { "und",            "standard", "eor", "search", NULL, NULL, NULL, NULL, NULL, NULL },
2533             { "en_US",          "standard", "eor", "search", NULL, NULL, NULL, NULL, NULL, NULL },
2534             { "en_029",         "standard", "eor", "search", NULL, NULL, NULL, NULL, NULL, NULL },
2535             { "de_DE",          "standard", "phonebook", "search", "eor", NULL, NULL, NULL, NULL, NULL },
2536             { "de_Latn_DE",     "standard", "phonebook", "search", "eor", NULL, NULL, NULL, NULL, NULL },
2537             { "zh",             "pinyin", "stroke", "eor", "search", "standard", NULL },
2538             { "zh_Hans",        "pinyin", "stroke", "eor", "search", "standard", NULL },
2539             { "zh_CN",          "pinyin", "stroke", "eor", "search", "standard", NULL },
2540             { "zh_Hant",        "stroke", "pinyin", "eor", "search", "standard", NULL },
2541             { "zh_TW",          "stroke", "pinyin", "eor", "search", "standard", NULL },
2542             { "zh__PINYIN",     "pinyin", "stroke", "eor", "search", "standard", NULL },
2543             { "es_ES",          "standard", "search", "traditional", "eor", NULL, NULL, NULL, NULL, NULL },
2544             { "es__TRADITIONAL","traditional", "search", "standard", "eor", NULL, NULL, NULL, NULL, NULL },
2545             { "und@collation=phonebook",    "standard", "eor", "search", NULL, NULL, NULL, NULL, NULL, NULL },
2546             { "de_DE@collation=pinyin",     "standard", "phonebook", "search", "eor", NULL, NULL, NULL, NULL, NULL },
2547             { "zzz@collation=xxx",          "standard", "eor", "search", NULL, NULL, NULL, NULL, NULL, NULL }
2548     };
2549 
2550     UErrorCode status = U_ZERO_ERROR;
2551     UEnumeration *keywordValues = NULL;
2552     int32_t i, n, size;
2553     const char *locale = NULL, *value = NULL;
2554     UBool errorOccurred = FALSE;
2555 
2556     for (i = 0; i < UPRV_LENGTHOF(PREFERRED) && !errorOccurred; i++) {
2557         locale = PREFERRED[i][0];
2558         value = NULL;
2559         size = 0;
2560 
2561         keywordValues = ucol_getKeywordValuesForLocale("collation", locale, TRUE, &status);
2562         if (keywordValues == NULL || U_FAILURE(status)) {
2563             log_err_status(status, "Error getting keyword values: %s\n", u_errorName(status));
2564             break;
2565         }
2566         size = uenum_count(keywordValues, &status);
2567         (void)size;
2568 
2569         for (n = 0; (value = PREFERRED[i][n+1]) != NULL; n++) {
2570             if (!uenum_contains(keywordValues, value, &status)) {
2571                 if (U_SUCCESS(status)) {
2572                     log_err("Keyword value \"%s\" missing for locale: %s\n", value, locale);
2573                 } else {
2574                     log_err("While getting keyword value from locale: %s got this error: %s\n", locale, u_errorName(status));
2575                     errorOccurred = TRUE;
2576                     break;
2577                 }
2578             }
2579         }
2580         uenum_close(keywordValues);
2581         keywordValues = NULL;
2582     }
2583     uenum_close(keywordValues);
2584 }
2585 
TestStrcollNull(void)2586 static void TestStrcollNull(void) {
2587     UErrorCode status = U_ZERO_ERROR;
2588     UCollator *coll;
2589 
2590     const UChar u16asc[] = {0x0049, 0x0042, 0x004D, 0};
2591     const int32_t u16ascLen = 3;
2592 
2593     const UChar u16han[] = {0x5c71, 0x5ddd, 0};
2594     const int32_t u16hanLen = 2;
2595 
2596     const char *u8asc = "\x49\x42\x4D";
2597     const int32_t u8ascLen = 3;
2598 
2599     const char *u8han = "\xE5\xB1\xB1\xE5\xB7\x9D";
2600     const int32_t u8hanLen = 6;
2601 
2602     coll = ucol_open(NULL, &status);
2603     if (U_FAILURE(status)) {
2604         log_err_status(status, "Default Collator creation failed.: %s\n", myErrorName(status));
2605         return;
2606     }
2607 
2608     /* UChar API */
2609     if (ucol_strcoll(coll, NULL, 0, NULL, 0) != 0) {
2610         log_err("ERROR : ucol_strcoll NULL/0 and NULL/0");
2611     }
2612 
2613     if (ucol_strcoll(coll, NULL, -1, NULL, 0) != 0) {
2614         /* No error arg, should return equal without crash */
2615         log_err("ERROR : ucol_strcoll NULL/-1 and NULL/0");
2616     }
2617 
2618     if (ucol_strcoll(coll, u16asc, -1, NULL, 10) != 0) {
2619         /* No error arg, should return equal without crash */
2620         log_err("ERROR : ucol_strcoll u16asc/u16ascLen and NULL/10");
2621     }
2622 
2623     if (ucol_strcoll(coll, u16asc, -1, NULL, 0) <= 0) {
2624         log_err("ERROR : ucol_strcoll u16asc/-1 and NULL/0");
2625     }
2626     if (ucol_strcoll(coll, NULL, 0, u16asc, -1) >= 0) {
2627         log_err("ERROR : ucol_strcoll NULL/0 and u16asc/-1");
2628     }
2629     if (ucol_strcoll(coll, u16asc, u16ascLen, NULL, 0) <= 0) {
2630         log_err("ERROR : ucol_strcoll u16asc/u16ascLen and NULL/0");
2631     }
2632 
2633     if (ucol_strcoll(coll, u16han, -1, NULL, 0) <= 0) {
2634         log_err("ERROR : ucol_strcoll u16han/-1 and NULL/0");
2635     }
2636     if (ucol_strcoll(coll, NULL, 0, u16han, -1) >= 0) {
2637         log_err("ERROR : ucol_strcoll NULL/0 and u16han/-1");
2638     }
2639     if (ucol_strcoll(coll, NULL, 0, u16han, u16hanLen) >= 0) {
2640         log_err("ERROR : ucol_strcoll NULL/0 and u16han/u16hanLen");
2641     }
2642 
2643     /* UTF-8 API */
2644     status = U_ZERO_ERROR;
2645     if (ucol_strcollUTF8(coll, NULL, 0, NULL, 0, &status) != 0 || U_FAILURE(status)) {
2646         log_err("ERROR : ucol_strcollUTF8 NULL/0 and NULL/0");
2647     }
2648     status = U_ZERO_ERROR;
2649     ucol_strcollUTF8(coll, NULL, -1, NULL, 0, &status);
2650     if (status != U_ILLEGAL_ARGUMENT_ERROR) {
2651         log_err("ERROR: ucol_strcollUTF8 NULL/-1 and NULL/0, should return U_ILLEGAL_ARGUMENT_ERROR");
2652     }
2653     status = U_ZERO_ERROR;
2654     ucol_strcollUTF8(coll, u8asc, u8ascLen, NULL, 10, &status);
2655     if (status != U_ILLEGAL_ARGUMENT_ERROR) {
2656         log_err("ERROR: ucol_strcollUTF8 u8asc/u8ascLen and NULL/10, should return U_ILLEGAL_ARGUMENT_ERROR");
2657     }
2658 
2659     status = U_ZERO_ERROR;
2660     if (ucol_strcollUTF8(coll, u8asc, -1, NULL, 0, &status) <= 0  || U_FAILURE(status)) {
2661         log_err("ERROR : ucol_strcollUTF8 u8asc/-1 and NULL/0");
2662     }
2663     status = U_ZERO_ERROR;
2664     if (ucol_strcollUTF8(coll, NULL, 0, u8asc, -1, &status) >= 0  || U_FAILURE(status)) {
2665         log_err("ERROR : ucol_strcollUTF8 NULL/0 and u8asc/-1");
2666     }
2667     status = U_ZERO_ERROR;
2668     if (ucol_strcollUTF8(coll, u8asc, u8ascLen, NULL, 0, &status) <= 0 || U_FAILURE(status)) {
2669         log_err("ERROR : ucol_strcollUTF8 u8asc/u8ascLen and NULL/0");
2670     }
2671 
2672     status = U_ZERO_ERROR;
2673     if (ucol_strcollUTF8(coll, u8han, -1, NULL, 0, &status) <= 0 || U_FAILURE(status)) {
2674         log_err("ERROR : ucol_strcollUTF8 u8han/-1 and NULL/0");
2675     }
2676     status = U_ZERO_ERROR;
2677     if (ucol_strcollUTF8(coll, NULL, 0, u8han, -1, &status) >= 0 || U_FAILURE(status)) {
2678         log_err("ERROR : ucol_strcollUTF8 NULL/0 and u8han/-1");
2679     }
2680     status = U_ZERO_ERROR;
2681     if (ucol_strcollUTF8(coll, NULL, 0, u8han, u8hanLen, &status) >= 0 || U_FAILURE(status)) {
2682         log_err("ERROR : ucol_strcollUTF8 NULL/0 and u8han/u8hanLen");
2683     }
2684 
2685     ucol_close(coll);
2686 }
2687 
TestLocaleIDWithUnderscoreAndExtension(void)2688 static void TestLocaleIDWithUnderscoreAndExtension(void) {
2689     UErrorCode err = U_ZERO_ERROR;
2690     UCollator* c1 = ucol_open("en-US-u-kn-true", &err);
2691     UCollator* c2 = ucol_open("en_US-u-kn-true", &err);
2692 
2693     if (assertSuccess("Failed to create collators", &err)) {
2694         assertTrue("Comparison using \"normal\" collator failed", !ucol_greater(c1, u"2", -1, u"10", -1));
2695         assertTrue("Comparison using \"bad\" collator failed", !ucol_greater(c2, u"2", -1, u"10", -1));
2696     }
2697 
2698     ucol_close(c1);
2699     ucol_close(c2);
2700 }
2701 
2702 #endif /* #if !UCONFIG_NO_COLLATION */
2703