1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4 *******************************************************************************
5 * Copyright (C) 2004-2016, International Business Machines
6 * Corporation and others. All Rights Reserved.
7 *******************************************************************************
8 * file name: ucol_sit.cpp
9 * encoding: UTF-8
10 * tab size: 8 (not used)
11 * indentation:4
12 *
13 * Modification history
14 * Date Name Comments
15 * 03/12/2004 weiv Creation
16 */
17
18 #include "unicode/ustring.h"
19 #include "unicode/udata.h"
20 #include "unicode/utf16.h"
21 #include "utracimp.h"
22 #include "ucol_imp.h"
23 #include "ulocimp.h"
24 #include "bytesinkutil.h"
25 #include "cmemory.h"
26 #include "cstring.h"
27 #include "uresimp.h"
28 #include "unicode/coll.h"
29 #include "unicode/stringpiece.h"
30 #include "charstr.h"
31
32 U_NAMESPACE_USE
33
34 #ifdef UCOL_TRACE_SIT
35 # include <stdio.h>
36 #endif
37
38 #if !UCONFIG_NO_COLLATION
39
40 #include "unicode/tblcoll.h"
41
42 enum OptionsList {
43 UCOL_SIT_LANGUAGE = 0,
44 UCOL_SIT_SCRIPT = 1,
45 UCOL_SIT_REGION = 2,
46 UCOL_SIT_VARIANT = 3,
47 UCOL_SIT_KEYWORD = 4,
48 UCOL_SIT_PROVIDER = 5,
49 UCOL_SIT_LOCELEMENT_MAX = UCOL_SIT_PROVIDER, /* the last element that's part of LocElements */
50
51 UCOL_SIT_BCP47,
52 UCOL_SIT_STRENGTH,
53 UCOL_SIT_CASE_LEVEL,
54 UCOL_SIT_CASE_FIRST,
55 UCOL_SIT_NUMERIC_COLLATION,
56 UCOL_SIT_ALTERNATE_HANDLING,
57 UCOL_SIT_NORMALIZATION_MODE,
58 UCOL_SIT_FRENCH_COLLATION,
59 UCOL_SIT_HIRAGANA_QUATERNARY,
60 UCOL_SIT_VARIABLE_TOP,
61 UCOL_SIT_VARIABLE_TOP_VALUE,
62 UCOL_SIT_ITEMS_COUNT
63 };
64
65 /* option starters chars. */
66 static const char alternateHArg = 'A';
67 static const char variableTopValArg = 'B';
68 static const char caseFirstArg = 'C';
69 static const char numericCollArg = 'D';
70 static const char caseLevelArg = 'E';
71 static const char frenchCollArg = 'F';
72 static const char hiraganaQArg = 'H';
73 static const char keywordArg = 'K';
74 static const char languageArg = 'L';
75 static const char normArg = 'N';
76 static const char providerArg = 'P';
77 static const char regionArg = 'R';
78 static const char strengthArg = 'S';
79 static const char variableTopArg = 'T';
80 static const char variantArg = 'V';
81 static const char RFC3066Arg = 'X';
82 static const char scriptArg = 'Z';
83
84 static const char collationKeyword[] = "@collation=";
85 static const char providerKeyword[] = "@sp=";
86
87
88 static const int32_t locElementCount = UCOL_SIT_LOCELEMENT_MAX+1;
89 static const int32_t locElementCapacity = 32;
90 static const int32_t loc3066Capacity = 256;
91
92 /* structure containing specification of a collator. Initialized
93 * from a short string. Also used to construct a short string from a
94 * collator instance
95 */
96 struct CollatorSpec {
97 inline CollatorSpec();
98
99 CharString locElements[locElementCount];
100 CharString locale;
101 UColAttributeValue options[UCOL_ATTRIBUTE_COUNT];
102 uint32_t variableTopValue;
103 char16_t variableTopString[locElementCapacity];
104 int32_t variableTopStringLen;
105 UBool variableTopSet;
106 CharString entries[UCOL_SIT_ITEMS_COUNT];
107 };
108
CollatorSpec()109 CollatorSpec::CollatorSpec() :
110 locale(),
111 variableTopValue(0),
112 variableTopString(),
113 variableTopSet(false)
114 {
115 // set collation options to default
116 for(int32_t i = 0; i < UCOL_ATTRIBUTE_COUNT; i++) {
117 options[i] = UCOL_DEFAULT;
118 }
119 }
120
121
122 /* structure for converting between character attribute
123 * representation and real collation attribute value.
124 */
125 struct AttributeConversion {
126 char letter;
127 UColAttributeValue value;
128 };
129
130 static const AttributeConversion conversions[12] = {
131 { '1', UCOL_PRIMARY },
132 { '2', UCOL_SECONDARY },
133 { '3', UCOL_TERTIARY },
134 { '4', UCOL_QUATERNARY },
135 { 'D', UCOL_DEFAULT },
136 { 'I', UCOL_IDENTICAL },
137 { 'L', UCOL_LOWER_FIRST },
138 { 'N', UCOL_NON_IGNORABLE },
139 { 'O', UCOL_ON },
140 { 'S', UCOL_SHIFTED },
141 { 'U', UCOL_UPPER_FIRST },
142 { 'X', UCOL_OFF }
143 };
144
145
146 static UColAttributeValue
ucol_sit_letterToAttributeValue(char letter,UErrorCode * status)147 ucol_sit_letterToAttributeValue(char letter, UErrorCode *status) {
148 uint32_t i = 0;
149 for(i = 0; i < UPRV_LENGTHOF(conversions); i++) {
150 if(conversions[i].letter == letter) {
151 return conversions[i].value;
152 }
153 }
154 *status = U_ILLEGAL_ARGUMENT_ERROR;
155 #ifdef UCOL_TRACE_SIT
156 fprintf(stderr, "%s:%d: unknown letter %c: %s\n", __FILE__, __LINE__, letter, u_errorName(*status));
157 #endif
158 return UCOL_DEFAULT;
159 }
160
161 /* function prototype for functions used to parse a short string */
162 U_CDECL_BEGIN
163 typedef const char* U_CALLCONV
164 ActionFunction(CollatorSpec *spec, uint32_t value1, const char* string,
165 UErrorCode *status);
166 U_CDECL_END
167
168 U_CDECL_BEGIN
169 static const char* U_CALLCONV
_processLocaleElement(CollatorSpec * spec,uint32_t value,const char * string,UErrorCode * status)170 _processLocaleElement(CollatorSpec *spec, uint32_t value, const char* string,
171 UErrorCode *status)
172 {
173 do {
174 if(value == UCOL_SIT_LANGUAGE || value == UCOL_SIT_KEYWORD || value == UCOL_SIT_PROVIDER) {
175 spec->locElements[value].append(uprv_tolower(*string), *status);
176 } else {
177 spec->locElements[value].append(*string, *status);
178 }
179 } while(*(++string) != '_' && *string && U_SUCCESS(*status));
180 // don't skip the underscore at the end
181 return string;
182 }
183 U_CDECL_END
184
185 U_CDECL_BEGIN
186 static const char* U_CALLCONV
_processRFC3066Locale(CollatorSpec * spec,uint32_t,const char * string,UErrorCode * status)187 _processRFC3066Locale(CollatorSpec *spec, uint32_t, const char* string,
188 UErrorCode *status)
189 {
190 char terminator = *string;
191 string++;
192 const char *end = uprv_strchr(string+1, terminator);
193 if(end == nullptr || end - string >= loc3066Capacity) {
194 *status = U_BUFFER_OVERFLOW_ERROR;
195 return string;
196 } else {
197 spec->locale.copyFrom(CharString(string, static_cast<int32_t>(end-string), *status), *status);
198 return end+1;
199 }
200 }
201
202 U_CDECL_END
203
204 U_CDECL_BEGIN
205 static const char* U_CALLCONV
_processCollatorOption(CollatorSpec * spec,uint32_t option,const char * string,UErrorCode * status)206 _processCollatorOption(CollatorSpec *spec, uint32_t option, const char* string,
207 UErrorCode *status)
208 {
209 spec->options[option] = ucol_sit_letterToAttributeValue(*string, status);
210 if((*(++string) != '_' && *string) || U_FAILURE(*status)) {
211 #ifdef UCOL_TRACE_SIT
212 fprintf(stderr, "%s:%d: unknown collator option at '%s': %s\n", __FILE__, __LINE__, string, u_errorName(*status));
213 #endif
214 *status = U_ILLEGAL_ARGUMENT_ERROR;
215 }
216 return string;
217 }
218 U_CDECL_END
219
220
221 static char16_t
readHexCodeUnit(const char ** string,UErrorCode * status)222 readHexCodeUnit(const char **string, UErrorCode *status)
223 {
224 char16_t result = 0;
225 int32_t value = 0;
226 char c;
227 int32_t noDigits = 0;
228 while((c = **string) != 0 && noDigits < 4) {
229 if( c >= '0' && c <= '9') {
230 value = c - '0';
231 } else if ( c >= 'a' && c <= 'f') {
232 value = c - 'a' + 10;
233 } else if ( c >= 'A' && c <= 'F') {
234 value = c - 'A' + 10;
235 } else {
236 *status = U_ILLEGAL_ARGUMENT_ERROR;
237 #ifdef UCOL_TRACE_SIT
238 fprintf(stderr, "%s:%d: Bad hex char at '%s': %s\n", __FILE__, __LINE__, *string, u_errorName(*status));
239 #endif
240 return 0;
241 }
242 result = (result << 4) | (char16_t)value;
243 noDigits++;
244 (*string)++;
245 }
246 // if the string was terminated before we read 4 digits, set an error
247 if(noDigits < 4) {
248 *status = U_ILLEGAL_ARGUMENT_ERROR;
249 #ifdef UCOL_TRACE_SIT
250 fprintf(stderr, "%s:%d: Short (only %d digits, wanted 4) at '%s': %s\n", __FILE__, __LINE__, noDigits,*string, u_errorName(*status));
251 #endif
252 }
253 return result;
254 }
255
256 U_CDECL_BEGIN
257 static const char* U_CALLCONV
_processVariableTop(CollatorSpec * spec,uint32_t value1,const char * string,UErrorCode * status)258 _processVariableTop(CollatorSpec *spec, uint32_t value1, const char* string, UErrorCode *status)
259 {
260 // get four digits
261 int32_t i = 0;
262 if(!value1) {
263 while(U_SUCCESS(*status) && i < locElementCapacity && *string != 0 && *string != '_') {
264 spec->variableTopString[i++] = readHexCodeUnit(&string, status);
265 }
266 spec->variableTopStringLen = i;
267 if(i == locElementCapacity && *string != 0 && *string != '_') {
268 *status = U_BUFFER_OVERFLOW_ERROR;
269 }
270 } else {
271 spec->variableTopValue = readHexCodeUnit(&string, status);
272 }
273 if(U_SUCCESS(*status)) {
274 spec->variableTopSet = true;
275 }
276 return string;
277 }
278 U_CDECL_END
279
280
281 /* Table for parsing short strings */
282 struct ShortStringOptions {
283 char optionStart;
284 ActionFunction *action;
285 uint32_t attr;
286 };
287
288 static const ShortStringOptions options[UCOL_SIT_ITEMS_COUNT] =
289 {
290 /* 10 ALTERNATE_HANDLING */ {alternateHArg, _processCollatorOption, UCOL_ALTERNATE_HANDLING }, // alternate N, S, D
291 /* 15 VARIABLE_TOP_VALUE */ {variableTopValArg, _processVariableTop, 1 },
292 /* 08 CASE_FIRST */ {caseFirstArg, _processCollatorOption, UCOL_CASE_FIRST }, // case first L, U, X, D
293 /* 09 NUMERIC_COLLATION */ {numericCollArg, _processCollatorOption, UCOL_NUMERIC_COLLATION }, // codan O, X, D
294 /* 07 CASE_LEVEL */ {caseLevelArg, _processCollatorOption, UCOL_CASE_LEVEL }, // case level O, X, D
295 /* 12 FRENCH_COLLATION */ {frenchCollArg, _processCollatorOption, UCOL_FRENCH_COLLATION }, // french O, X, D
296 /* 13 HIRAGANA_QUATERNARY] */ {hiraganaQArg, _processCollatorOption, UCOL_HIRAGANA_QUATERNARY_MODE }, // hiragana O, X, D
297 /* 04 KEYWORD */ {keywordArg, _processLocaleElement, UCOL_SIT_KEYWORD }, // keyword
298 /* 00 LANGUAGE */ {languageArg, _processLocaleElement, UCOL_SIT_LANGUAGE }, // language
299 /* 11 NORMALIZATION_MODE */ {normArg, _processCollatorOption, UCOL_NORMALIZATION_MODE }, // norm O, X, D
300 /* 02 REGION */ {regionArg, _processLocaleElement, UCOL_SIT_REGION }, // region
301 /* 06 STRENGTH */ {strengthArg, _processCollatorOption, UCOL_STRENGTH }, // strength 1, 2, 3, 4, I, D
302 /* 14 VARIABLE_TOP */ {variableTopArg, _processVariableTop, 0 },
303 /* 03 VARIANT */ {variantArg, _processLocaleElement, UCOL_SIT_VARIANT }, // variant
304 /* 05 RFC3066BIS */ {RFC3066Arg, _processRFC3066Locale, 0 }, // rfc3066bis locale name
305 /* 01 SCRIPT */ {scriptArg, _processLocaleElement, UCOL_SIT_SCRIPT }, // script
306 /* PROVIDER */ {providerArg, _processLocaleElement, UCOL_SIT_PROVIDER }
307 };
308
309
310 static
ucol_sit_readOption(const char * start,CollatorSpec * spec,UErrorCode * status)311 const char* ucol_sit_readOption(const char *start, CollatorSpec *spec,
312 UErrorCode *status)
313 {
314 int32_t i = 0;
315
316 for(i = 0; i < UCOL_SIT_ITEMS_COUNT; i++) {
317 if(*start == options[i].optionStart) {
318 const char* end = options[i].action(spec, options[i].attr, start+1, status);
319 #ifdef UCOL_TRACE_SIT
320 fprintf(stderr, "***Set %d to %s...\n", i, start);
321 #endif
322 // assume 'start' does not go away through all this
323 spec->entries[i].copyFrom(CharString(start, (int32_t)(end - start), *status), *status);
324 return end;
325 }
326 }
327 *status = U_ILLEGAL_ARGUMENT_ERROR;
328 #ifdef UCOL_TRACE_SIT
329 fprintf(stderr, "%s:%d: Unknown option at '%s': %s\n", __FILE__, __LINE__, start, u_errorName(*status));
330 #endif
331 return start;
332 }
333
334 static const char*
ucol_sit_readSpecs(CollatorSpec * s,const char * string,UParseError * parseError,UErrorCode * status)335 ucol_sit_readSpecs(CollatorSpec *s, const char *string,
336 UParseError *parseError, UErrorCode *status)
337 {
338 const char *definition = string;
339 while(U_SUCCESS(*status) && *string) {
340 string = ucol_sit_readOption(string, s, status);
341 // advance over '_'
342 while(*string && *string == '_') {
343 string++;
344 }
345 }
346 if(U_FAILURE(*status)) {
347 parseError->offset = (int32_t)(string - definition);
348 }
349 return string;
350 }
351
352 static
ucol_sit_dumpSpecs(CollatorSpec * s,char * destination,int32_t capacity,UErrorCode * status)353 int32_t ucol_sit_dumpSpecs(CollatorSpec *s, char *destination, int32_t capacity, UErrorCode *status)
354 {
355 int32_t i = 0, j = 0;
356 int32_t len = 0;
357 char optName;
358 if(U_SUCCESS(*status)) {
359 for(i = 0; i < UCOL_SIT_ITEMS_COUNT; i++) {
360 if(!s->entries[i].isEmpty()) {
361 if(len) {
362 if(len < capacity) {
363 uprv_strcat(destination, "_");
364 }
365 len++;
366 }
367 optName = s->entries[i][0];
368 if(optName == languageArg || optName == regionArg || optName == variantArg || optName == keywordArg) {
369 for(j = 0; j < s->entries[i].length(); j++) {
370 if(len + j < capacity) {
371 destination[len+j] = uprv_toupper(s->entries[i][j]);
372 }
373 }
374 len += s->entries[i].length();
375 } else {
376 len += s->entries[i].extract(destination + len, capacity - len, *status);
377 }
378 }
379 }
380 return len;
381 } else {
382 return 0;
383 }
384 }
385
386 static void
ucol_sit_calculateWholeLocale(CollatorSpec * s,UErrorCode & status)387 ucol_sit_calculateWholeLocale(CollatorSpec *s, UErrorCode &status) {
388 // put the locale together, unless we have a done
389 // locale
390 if(s->locale.isEmpty()) {
391 // first the language
392 s->locale.append(s->locElements[UCOL_SIT_LANGUAGE], status);
393 // then the script, if present
394 if(!s->locElements[UCOL_SIT_SCRIPT].isEmpty()) {
395 s->locale.append("_", status);
396 s->locale.append(s->locElements[UCOL_SIT_SCRIPT], status);
397 }
398 // then the region, if present
399 if(!s->locElements[UCOL_SIT_REGION].isEmpty()) {
400 s->locale.append("_", status);
401 s->locale.append(s->locElements[UCOL_SIT_REGION], status);
402 } else if(!s->locElements[UCOL_SIT_VARIANT].isEmpty()) { // if there is a variant, we need an underscore
403 s->locale.append("_", status);
404 }
405 // add variant, if there
406 if(!s->locElements[UCOL_SIT_VARIANT].isEmpty()) {
407 s->locale.append("_", status);
408 s->locale.append(s->locElements[UCOL_SIT_VARIANT], status);
409 }
410
411 // if there is a collation keyword, add that too
412 if(!s->locElements[UCOL_SIT_KEYWORD].isEmpty()) {
413 s->locale.append(collationKeyword, status);
414 s->locale.append(s->locElements[UCOL_SIT_KEYWORD], status);
415 }
416
417 // if there is a provider keyword, add that too
418 if(!s->locElements[UCOL_SIT_PROVIDER].isEmpty()) {
419 s->locale.append(providerKeyword, status);
420 s->locale.append(s->locElements[UCOL_SIT_PROVIDER], status);
421 }
422 }
423 }
424
425
426 U_CAPI void U_EXPORT2
ucol_prepareShortStringOpen(const char * definition,UBool,UParseError * parseError,UErrorCode * status)427 ucol_prepareShortStringOpen( const char *definition,
428 UBool,
429 UParseError *parseError,
430 UErrorCode *status)
431 {
432 if(U_FAILURE(*status)) return;
433
434 UParseError internalParseError;
435
436 if(!parseError) {
437 parseError = &internalParseError;
438 }
439 parseError->line = 0;
440 parseError->offset = 0;
441 parseError->preContext[0] = 0;
442 parseError->postContext[0] = 0;
443
444
445 // first we want to pick stuff out of short string.
446 // we'll end up with an UCA version, locale and a bunch of
447 // settings
448
449 // analyse the string in order to get everything we need.
450 CollatorSpec s;
451 ucol_sit_readSpecs(&s, definition, parseError, status);
452 ucol_sit_calculateWholeLocale(&s, *status);
453
454 CharString buffer;
455 {
456 CharStringByteSink sink(&buffer);
457 ulocimp_canonicalize(s.locale.data(), sink, status);
458 }
459
460 UResourceBundle *b = ures_open(U_ICUDATA_COLL, buffer.data(), status);
461 /* we try to find stuff from keyword */
462 UResourceBundle *collations = ures_getByKey(b, "collations", nullptr, status);
463 UResourceBundle *collElem = nullptr;
464 CharString keyBuffer;
465 {
466 // if there is a keyword, we pick it up and try to get elements
467 CharStringByteSink sink(&keyBuffer);
468 ulocimp_getKeywordValue(buffer.data(), "collation", sink, status);
469 }
470 if(keyBuffer.isEmpty()) {
471 // no keyword
472 // we try to find the default setting, which will give us the keyword value
473 UResourceBundle *defaultColl = ures_getByKeyWithFallback(collations, "default", nullptr, status);
474 if(U_SUCCESS(*status)) {
475 int32_t defaultKeyLen = 0;
476 const char16_t *defaultKey = ures_getString(defaultColl, &defaultKeyLen, status);
477 keyBuffer.appendInvariantChars(defaultKey, defaultKeyLen, *status);
478 } else {
479 *status = U_INTERNAL_PROGRAM_ERROR;
480 return;
481 }
482 ures_close(defaultColl);
483 }
484 collElem = ures_getByKeyWithFallback(collations, keyBuffer.data(), collElem, status);
485 ures_close(collElem);
486 ures_close(collations);
487 ures_close(b);
488 }
489
490
491 U_CAPI UCollator* U_EXPORT2
ucol_openFromShortString(const char * definition,UBool forceDefaults,UParseError * parseError,UErrorCode * status)492 ucol_openFromShortString( const char *definition,
493 UBool forceDefaults,
494 UParseError *parseError,
495 UErrorCode *status)
496 {
497 UTRACE_ENTRY_OC(UTRACE_UCOL_OPEN_FROM_SHORT_STRING);
498 UTRACE_DATA1(UTRACE_INFO, "short string = \"%s\"", definition);
499
500 if(U_FAILURE(*status)) return 0;
501
502 UParseError internalParseError;
503
504 if(!parseError) {
505 parseError = &internalParseError;
506 }
507 parseError->line = 0;
508 parseError->offset = 0;
509 parseError->preContext[0] = 0;
510 parseError->postContext[0] = 0;
511
512
513 // first we want to pick stuff out of short string.
514 // we'll end up with an UCA version, locale and a bunch of
515 // settings
516
517 // analyse the string in order to get everything we need.
518 const char *string = definition;
519 CollatorSpec s;
520 string = ucol_sit_readSpecs(&s, definition, parseError, status);
521 ucol_sit_calculateWholeLocale(&s, *status);
522
523 #ifdef UCOL_TRACE_SIT
524 fprintf(stderr, "DEF %s, DATA %s, ERR %s\n", definition, s.locale.data(), u_errorName(*status));
525 #endif
526 CharString buffer;
527 {
528 CharStringByteSink sink(&buffer);
529 ulocimp_canonicalize(s.locale.data(), sink, status);
530 }
531
532 UCollator *result = ucol_open(buffer.data(), status);
533 int32_t i = 0;
534
535 for(i = 0; i < UCOL_ATTRIBUTE_COUNT; i++) {
536 if(s.options[i] != UCOL_DEFAULT) {
537 if(forceDefaults || ucol_getAttribute(result, (UColAttribute)i, status) != s.options[i]) {
538 ucol_setAttribute(result, (UColAttribute)i, s.options[i], status);
539 }
540
541 if(U_FAILURE(*status)) {
542 parseError->offset = (int32_t)(string - definition);
543 ucol_close(result);
544 return nullptr;
545 }
546
547 }
548 }
549 if(s.variableTopSet) {
550 if(s.variableTopString[0]) {
551 ucol_setVariableTop(result, s.variableTopString, s.variableTopStringLen, status);
552 } else { // we set by value, using 'B'
553 ucol_restoreVariableTop(result, s.variableTopValue, status);
554 }
555 }
556
557
558 if(U_FAILURE(*status)) { // here it can only be a bogus value
559 ucol_close(result);
560 result = nullptr;
561 }
562
563 UTRACE_EXIT_PTR_STATUS(result, *status);
564 return result;
565 }
566
567
568 U_CAPI int32_t U_EXPORT2
ucol_getShortDefinitionString(const UCollator * coll,const char * locale,char * dst,int32_t capacity,UErrorCode * status)569 ucol_getShortDefinitionString(const UCollator *coll,
570 const char *locale,
571 char *dst,
572 int32_t capacity,
573 UErrorCode *status)
574 {
575 if(U_FAILURE(*status)) return 0;
576 if(coll == nullptr) {
577 *status = U_ILLEGAL_ARGUMENT_ERROR;
578 return 0;
579 }
580 return ((icu::Collator*)coll)->internalGetShortDefinitionString(locale,dst,capacity,*status);
581 }
582
583 U_CAPI int32_t U_EXPORT2
ucol_normalizeShortDefinitionString(const char * definition,char * destination,int32_t capacity,UParseError * parseError,UErrorCode * status)584 ucol_normalizeShortDefinitionString(const char *definition,
585 char *destination,
586 int32_t capacity,
587 UParseError *parseError,
588 UErrorCode *status)
589 {
590
591 if(U_FAILURE(*status)) {
592 return 0;
593 }
594
595 if(destination) {
596 uprv_memset(destination, 0, capacity*sizeof(char));
597 }
598
599 UParseError pe;
600 if(!parseError) {
601 parseError = &pe;
602 }
603
604 // validate
605 CollatorSpec s;
606 ucol_sit_readSpecs(&s, definition, parseError, status);
607 return ucol_sit_dumpSpecs(&s, destination, capacity, status);
608 }
609
610 /**
611 * Get a set containing the contractions defined by the collator. The set includes
612 * both the UCA contractions and the contractions defined by the collator
613 * @param coll collator
614 * @param conts the set to hold the result
615 * @param status to hold the error code
616 * @return the size of the contraction set
617 */
618 U_CAPI int32_t U_EXPORT2
ucol_getContractions(const UCollator * coll,USet * contractions,UErrorCode * status)619 ucol_getContractions( const UCollator *coll,
620 USet *contractions,
621 UErrorCode *status)
622 {
623 ucol_getContractionsAndExpansions(coll, contractions, nullptr, false, status);
624 return uset_getItemCount(contractions);
625 }
626
627 /**
628 * Get a set containing the expansions defined by the collator. The set includes
629 * both the UCA expansions and the expansions defined by the tailoring
630 * @param coll collator
631 * @param conts the set to hold the result
632 * @param addPrefixes add the prefix contextual elements to contractions
633 * @param status to hold the error code
634 *
635 * @draft ICU 3.4
636 */
637 U_CAPI void U_EXPORT2
ucol_getContractionsAndExpansions(const UCollator * coll,USet * contractions,USet * expansions,UBool addPrefixes,UErrorCode * status)638 ucol_getContractionsAndExpansions( const UCollator *coll,
639 USet *contractions,
640 USet *expansions,
641 UBool addPrefixes,
642 UErrorCode *status)
643 {
644 if(U_FAILURE(*status)) {
645 return;
646 }
647 if(coll == nullptr) {
648 *status = U_ILLEGAL_ARGUMENT_ERROR;
649 return;
650 }
651 const icu::RuleBasedCollator *rbc = icu::RuleBasedCollator::rbcFromUCollator(coll);
652 if(rbc == nullptr) {
653 *status = U_UNSUPPORTED_ERROR;
654 return;
655 }
656 rbc->internalGetContractionsAndExpansions(
657 icu::UnicodeSet::fromUSet(contractions),
658 icu::UnicodeSet::fromUSet(expansions),
659 addPrefixes, *status);
660 }
661 #endif
662