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