1// © 2016 and later: Unicode, Inc. and others.
2// License & terms of use: http://www.unicode.org/copyright.html
3/*
4*******************************************************************************
5* Copyright (c) 1996-2015, International Business Machines Corporation and others.
6* All Rights Reserved.
7*******************************************************************************
8*/
9
10#ifndef UCOL_H
11#define UCOL_H
12
13#include "unicode/utypes.h"
14
15#if !UCONFIG_NO_COLLATION
16
17#include "unicode/unorm.h"
18#include "unicode/localpointer.h"
19#include "unicode/parseerr.h"
20#include "unicode/uloc.h"
21#include "unicode/uset.h"
22#include "unicode/uscript.h"
23
24/**
25 * \file
26 * \brief C API: Collator
27 *
28 * <h2> Collator C API </h2>
29 *
30 * The C API for Collator performs locale-sensitive
31 * string comparison. You use this service to build
32 * searching and sorting routines for natural language text.
33 * <p>
34 * For more information about the collation service see
35 * <a href="http://userguide.icu-project.org/collation">the User Guide</a>.
36 * <p>
37 * Collation service provides correct sorting orders for most locales supported in ICU.
38 * If specific data for a locale is not available, the orders eventually falls back
39 * to the <a href="http://www.unicode.org/reports/tr35/tr35-collation.html#Root_Collation">CLDR root sort order</a>.
40 * <p>
41 * Sort ordering may be customized by providing your own set of rules. For more on
42 * this subject see the <a href="http://userguide.icu-project.org/collation/customization">
43 * Collation Customization</a> section of the User Guide.
44 * <p>
45 * @see UCollationResult
46 * @see UNormalizationMode
47 * @see UCollationStrength
48 * @see UCollationElements
49 */
50
51/** A collator.
52* For usage in C programs.
53*/
54struct UCollator;
55/** structure representing a collator object instance
56 * @stable ICU 2.0
57 */
58typedef struct UCollator UCollator;
59
60
61/**
62 * UCOL_LESS is returned if source string is compared to be less than target
63 * string in the ucol_strcoll() method.
64 * UCOL_EQUAL is returned if source string is compared to be equal to target
65 * string in the ucol_strcoll() method.
66 * UCOL_GREATER is returned if source string is compared to be greater than
67 * target string in the ucol_strcoll() method.
68 * @see ucol_strcoll()
69 * <p>
70 * Possible values for a comparison result
71 * @stable ICU 2.0
72 */
73typedef enum {
74 /** string a == string b */
75 UCOL_EQUAL = 0,
76 /** string a > string b */
77 UCOL_GREATER = 1,
78 /** string a < string b */
79 UCOL_LESS = -1
80} UCollationResult ;
81
82
83/** Enum containing attribute values for controling collation behavior.
84 * Here are all the allowable values. Not every attribute can take every value. The only
85 * universal value is UCOL_DEFAULT, which resets the attribute value to the predefined
86 * value for that locale
87 * @stable ICU 2.0
88 */
89typedef enum {
90 /** accepted by most attributes */
91 UCOL_DEFAULT = -1,
92
93 /** Primary collation strength */
94 UCOL_PRIMARY = 0,
95 /** Secondary collation strength */
96 UCOL_SECONDARY = 1,
97 /** Tertiary collation strength */
98 UCOL_TERTIARY = 2,
99 /** Default collation strength */
100 UCOL_DEFAULT_STRENGTH = UCOL_TERTIARY,
101 UCOL_CE_STRENGTH_LIMIT,
102 /** Quaternary collation strength */
103 UCOL_QUATERNARY=3,
104 /** Identical collation strength */
105 UCOL_IDENTICAL=15,
106 UCOL_STRENGTH_LIMIT,
107
108 /** Turn the feature off - works for UCOL_FRENCH_COLLATION,
109 UCOL_CASE_LEVEL, UCOL_HIRAGANA_QUATERNARY_MODE
110 & UCOL_DECOMPOSITION_MODE*/
111 UCOL_OFF = 16,
112 /** Turn the feature on - works for UCOL_FRENCH_COLLATION,
113 UCOL_CASE_LEVEL, UCOL_HIRAGANA_QUATERNARY_MODE
114 & UCOL_DECOMPOSITION_MODE*/
115 UCOL_ON = 17,
116
117 /** Valid for UCOL_ALTERNATE_HANDLING. Alternate handling will be shifted */
118 UCOL_SHIFTED = 20,
119 /** Valid for UCOL_ALTERNATE_HANDLING. Alternate handling will be non ignorable */
120 UCOL_NON_IGNORABLE = 21,
121
122 /** Valid for UCOL_CASE_FIRST -
123 lower case sorts before upper case */
124 UCOL_LOWER_FIRST = 24,
125 /** upper case sorts before lower case */
126 UCOL_UPPER_FIRST = 25,
127
128#ifndef U_HIDE_DEPRECATED_API
129 /**
130 * One more than the highest normal UColAttributeValue value.
131 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
132 */
133 UCOL_ATTRIBUTE_VALUE_COUNT
134#endif /* U_HIDE_DEPRECATED_API */
135} UColAttributeValue;
136
137/**
138 * Enum containing the codes for reordering segments of the collation table that are not script
139 * codes. These reordering codes are to be used in conjunction with the script codes.
140 * @see ucol_getReorderCodes
141 * @see ucol_setReorderCodes
142 * @see ucol_getEquivalentReorderCodes
143 * @see UScriptCode
144 * @stable ICU 4.8
145 */
146 typedef enum {
147 /**
148 * A special reordering code that is used to specify the default
149 * reordering codes for a locale.
150 * @stable ICU 4.8
151 */
152 UCOL_REORDER_CODE_DEFAULT = -1,
153 /**
154 * A special reordering code that is used to specify no reordering codes.
155 * @stable ICU 4.8
156 */
157 UCOL_REORDER_CODE_NONE = USCRIPT_UNKNOWN,
158 /**
159 * A special reordering code that is used to specify all other codes used for
160 * reordering except for the codes lised as UColReorderCode values and those
161 * listed explicitly in a reordering.
162 * @stable ICU 4.8
163 */
164 UCOL_REORDER_CODE_OTHERS = USCRIPT_UNKNOWN,
165 /**
166 * Characters with the space property.
167 * This is equivalent to the rule value "space".
168 * @stable ICU 4.8
169 */
170 UCOL_REORDER_CODE_SPACE = 0x1000,
171 /**
172 * The first entry in the enumeration of reordering groups. This is intended for use in
173 * range checking and enumeration of the reorder codes.
174 * @stable ICU 4.8
175 */
176 UCOL_REORDER_CODE_FIRST = UCOL_REORDER_CODE_SPACE,
177 /**
178 * Characters with the punctuation property.
179 * This is equivalent to the rule value "punct".
180 * @stable ICU 4.8
181 */
182 UCOL_REORDER_CODE_PUNCTUATION = 0x1001,
183 /**
184 * Characters with the symbol property.
185 * This is equivalent to the rule value "symbol".
186 * @stable ICU 4.8
187 */
188 UCOL_REORDER_CODE_SYMBOL = 0x1002,
189 /**
190 * Characters with the currency property.
191 * This is equivalent to the rule value "currency".
192 * @stable ICU 4.8
193 */
194 UCOL_REORDER_CODE_CURRENCY = 0x1003,
195 /**
196 * Characters with the digit property.
197 * This is equivalent to the rule value "digit".
198 * @stable ICU 4.8
199 */
200 UCOL_REORDER_CODE_DIGIT = 0x1004,
201#ifndef U_HIDE_DEPRECATED_API
202 /**
203 * One more than the highest normal UColReorderCode value.
204 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
205 */
206 UCOL_REORDER_CODE_LIMIT = 0x1005
207#endif /* U_HIDE_DEPRECATED_API */
208} UColReorderCode;
209
210/**
211 * Base letter represents a primary difference. Set comparison
212 * level to UCOL_PRIMARY to ignore secondary and tertiary differences.
213 * Use this to set the strength of a Collator object.
214 * Example of primary difference, "abc" &lt; "abd"
215 *
216 * Diacritical differences on the same base letter represent a secondary
217 * difference. Set comparison level to UCOL_SECONDARY to ignore tertiary
218 * differences. Use this to set the strength of a Collator object.
219 * Example of secondary difference, "&auml;" >> "a".
220 *
221 * Uppercase and lowercase versions of the same character represents a
222 * tertiary difference. Set comparison level to UCOL_TERTIARY to include
223 * all comparison differences. Use this to set the strength of a Collator
224 * object.
225 * Example of tertiary difference, "abc" &lt;&lt;&lt; "ABC".
226 *
227 * Two characters are considered "identical" when they have the same
228 * unicode spellings. UCOL_IDENTICAL.
229 * For example, "&auml;" == "&auml;".
230 *
231 * UCollationStrength is also used to determine the strength of sort keys
232 * generated from UCollator objects
233 * These values can be now found in the UColAttributeValue enum.
234 * @stable ICU 2.0
235 **/
236typedef UColAttributeValue UCollationStrength;
237
238/** Attributes that collation service understands. All the attributes can take UCOL_DEFAULT
239 * value, as well as the values specific to each one.
240 * @stable ICU 2.0
241 */
242typedef enum {
243 /** Attribute for direction of secondary weights - used in Canadian French.
244 * Acceptable values are UCOL_ON, which results in secondary weights
245 * being considered backwards and UCOL_OFF which treats secondary
246 * weights in the order they appear.
247 * @stable ICU 2.0
248 */
249 UCOL_FRENCH_COLLATION,
250 /** Attribute for handling variable elements.
251 * Acceptable values are UCOL_NON_IGNORABLE (default)
252 * which treats all the codepoints with non-ignorable
253 * primary weights in the same way,
254 * and UCOL_SHIFTED which causes codepoints with primary
255 * weights that are equal or below the variable top value
256 * to be ignored on primary level and moved to the quaternary
257 * level.
258 * @stable ICU 2.0
259 */
260 UCOL_ALTERNATE_HANDLING,
261 /** Controls the ordering of upper and lower case letters.
262 * Acceptable values are UCOL_OFF (default), which orders
263 * upper and lower case letters in accordance to their tertiary
264 * weights, UCOL_UPPER_FIRST which forces upper case letters to
265 * sort before lower case letters, and UCOL_LOWER_FIRST which does
266 * the opposite.
267 * @stable ICU 2.0
268 */
269 UCOL_CASE_FIRST,
270 /** Controls whether an extra case level (positioned before the third
271 * level) is generated or not. Acceptable values are UCOL_OFF (default),
272 * when case level is not generated, and UCOL_ON which causes the case
273 * level to be generated. Contents of the case level are affected by
274 * the value of UCOL_CASE_FIRST attribute. A simple way to ignore
275 * accent differences in a string is to set the strength to UCOL_PRIMARY
276 * and enable case level.
277 * @stable ICU 2.0
278 */
279 UCOL_CASE_LEVEL,
280 /** Controls whether the normalization check and necessary normalizations
281 * are performed. When set to UCOL_OFF (default) no normalization check
282 * is performed. The correctness of the result is guaranteed only if the
283 * input data is in so-called FCD form (see users manual for more info).
284 * When set to UCOL_ON, an incremental check is performed to see whether
285 * the input data is in the FCD form. If the data is not in the FCD form,
286 * incremental NFD normalization is performed.
287 * @stable ICU 2.0
288 */
289 UCOL_NORMALIZATION_MODE,
290 /** An alias for UCOL_NORMALIZATION_MODE attribute.
291 * @stable ICU 2.0
292 */
293 UCOL_DECOMPOSITION_MODE = UCOL_NORMALIZATION_MODE,
294 /** The strength attribute. Can be either UCOL_PRIMARY, UCOL_SECONDARY,
295 * UCOL_TERTIARY, UCOL_QUATERNARY or UCOL_IDENTICAL. The usual strength
296 * for most locales (except Japanese) is tertiary.
297 *
298 * Quaternary strength
299 * is useful when combined with shifted setting for alternate handling
300 * attribute and for JIS X 4061 collation, when it is used to distinguish
301 * between Katakana and Hiragana.
302 * Otherwise, quaternary level
303 * is affected only by the number of non-ignorable code points in
304 * the string.
305 *
306 * Identical strength is rarely useful, as it amounts
307 * to codepoints of the NFD form of the string.
308 * @stable ICU 2.0
309 */
310 UCOL_STRENGTH,
311#ifndef U_HIDE_DEPRECATED_API
312 /** When turned on, this attribute positions Hiragana before all
313 * non-ignorables on quaternary level This is a sneaky way to produce JIS
314 * sort order.
315 *
316 * This attribute was an implementation detail of the CLDR Japanese tailoring.
317 * Since ICU 50, this attribute is not settable any more via API functions.
318 * Since CLDR 25/ICU 53, explicit quaternary relations are used
319 * to achieve the same Japanese sort order.
320 *
321 * @deprecated ICU 50 Implementation detail, cannot be set via API, was removed from implementation.
322 */
323 UCOL_HIRAGANA_QUATERNARY_MODE = UCOL_STRENGTH + 1,
324#endif /* U_HIDE_DEPRECATED_API */
325 /**
326 * When turned on, this attribute makes
327 * substrings of digits sort according to their numeric values.
328 *
329 * This is a way to get '100' to sort AFTER '2'. Note that the longest
330 * digit substring that can be treated as a single unit is
331 * 254 digits (not counting leading zeros). If a digit substring is
332 * longer than that, the digits beyond the limit will be treated as a
333 * separate digit substring.
334 *
335 * A "digit" in this sense is a code point with General_Category=Nd,
336 * which does not include circled numbers, roman numerals, etc.
337 * Only a contiguous digit substring is considered, that is,
338 * non-negative integers without separators.
339 * There is no support for plus/minus signs, decimals, exponents, etc.
340 *
341 * @stable ICU 2.8
342 */
343 UCOL_NUMERIC_COLLATION = UCOL_STRENGTH + 2,
344
345 /* Do not conditionalize the following with #ifndef U_HIDE_DEPRECATED_API,
346 * it is needed for layout of RuleBasedCollator object. */
347#ifndef U_FORCE_HIDE_DEPRECATED_API
348 /**
349 * One more than the highest normal UColAttribute value.
350 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
351 */
352 UCOL_ATTRIBUTE_COUNT
353#endif // U_FORCE_HIDE_DEPRECATED_API
354} UColAttribute;
355
356/** Options for retrieving the rule string
357 * @stable ICU 2.0
358 */
359typedef enum {
360 /**
361 * Retrieves the tailoring rules only.
362 * Same as calling the version of getRules() without UColRuleOption.
363 * @stable ICU 2.0
364 */
365 UCOL_TAILORING_ONLY,
366 /**
367 * Retrieves the "UCA rules" concatenated with the tailoring rules.
368 * The "UCA rules" are an <i>approximation</i> of the root collator's sort order.
369 * They are almost never used or useful at runtime and can be removed from the data.
370 * See http://userguide.icu-project.org/collation/customization#TOC-Building-on-Existing-Locales
371 * @stable ICU 2.0
372 */
373 UCOL_FULL_RULES
374} UColRuleOption ;
375
376/**
377 * Open a UCollator for comparing strings.
378 *
379 * For some languages, multiple collation types are available;
380 * for example, "de@collation=phonebook".
381 * Starting with ICU 54, collation attributes can be specified via locale keywords as well,
382 * in the old locale extension syntax ("el@colCaseFirst=upper")
383 * or in language tag syntax ("el-u-kf-upper").
384 * See <a href="http://userguide.icu-project.org/collation/api">User Guide: Collation API</a>.
385 *
386 * The UCollator pointer is used in all the calls to the Collation
387 * service. After finished, collator must be disposed of by calling
388 * {@link #ucol_close }.
389 * @param loc The locale containing the required collation rules.
390 * Special values for locales can be passed in -
391 * if NULL is passed for the locale, the default locale
392 * collation rules will be used. If empty string ("") or
393 * "root" are passed, the root collator will be returned.
394 * @param status A pointer to a UErrorCode to receive any errors
395 * @return A pointer to a UCollator, or 0 if an error occurred.
396 * @see ucol_openRules
397 * @see ucol_safeClone
398 * @see ucol_close
399 * @stable ICU 2.0
400 */
401U_STABLE UCollator* U_EXPORT2
402ucol_open(const char *loc, UErrorCode *status);
403
404/**
405 * Produce a UCollator instance according to the rules supplied.
406 * The rules are used to change the default ordering, defined in the
407 * UCA in a process called tailoring. The resulting UCollator pointer
408 * can be used in the same way as the one obtained by {@link #ucol_strcoll }.
409 * @param rules A string describing the collation rules. For the syntax
410 * of the rules please see users guide.
411 * @param rulesLength The length of rules, or -1 if null-terminated.
412 * @param normalizationMode The normalization mode: One of
413 * UCOL_OFF (expect the text to not need normalization),
414 * UCOL_ON (normalize), or
415 * UCOL_DEFAULT (set the mode according to the rules)
416 * @param strength The default collation strength; one of UCOL_PRIMARY, UCOL_SECONDARY,
417 * UCOL_TERTIARY, UCOL_IDENTICAL,UCOL_DEFAULT_STRENGTH - can be also set in the rules.
418 * @param parseError A pointer to UParseError to recieve information about errors
419 * occurred during parsing. This argument can currently be set
420 * to NULL, but at users own risk. Please provide a real structure.
421 * @param status A pointer to a UErrorCode to receive any errors
422 * @return A pointer to a UCollator. It is not guaranteed that NULL be returned in case
423 * of error - please use status argument to check for errors.
424 * @see ucol_open
425 * @see ucol_safeClone
426 * @see ucol_close
427 * @stable ICU 2.0
428 */
429U_STABLE UCollator* U_EXPORT2
430ucol_openRules( const UChar *rules,
431 int32_t rulesLength,
432 UColAttributeValue normalizationMode,
433 UCollationStrength strength,
434 UParseError *parseError,
435 UErrorCode *status);
436
437#ifndef U_HIDE_DEPRECATED_API
438/**
439 * Open a collator defined by a short form string.
440 * The structure and the syntax of the string is defined in the "Naming collators"
441 * section of the users guide:
442 * http://userguide.icu-project.org/collation/concepts#TOC-Collator-naming-scheme
443 * Attributes are overriden by the subsequent attributes. So, for "S2_S3", final
444 * strength will be 3. 3066bis locale overrides individual locale parts.
445 * The call to this function is equivalent to a call to ucol_open, followed by a
446 * series of calls to ucol_setAttribute and ucol_setVariableTop.
447 * @param definition A short string containing a locale and a set of attributes.
448 * Attributes not explicitly mentioned are left at the default
449 * state for a locale.
450 * @param parseError if not NULL, structure that will get filled with error's pre
451 * and post context in case of error.
452 * @param forceDefaults if FALSE, the settings that are the same as the collator
453 * default settings will not be applied (for example, setting
454 * French secondary on a French collator would not be executed).
455 * If TRUE, all the settings will be applied regardless of the
456 * collator default value. If the definition
457 * strings are to be cached, should be set to FALSE.
458 * @param status Error code. Apart from regular error conditions connected to
459 * instantiating collators (like out of memory or similar), this
460 * API will return an error if an invalid attribute or attribute/value
461 * combination is specified.
462 * @return A pointer to a UCollator or 0 if an error occured (including an
463 * invalid attribute).
464 * @see ucol_open
465 * @see ucol_setAttribute
466 * @see ucol_setVariableTop
467 * @see ucol_getShortDefinitionString
468 * @see ucol_normalizeShortDefinitionString
469 * @deprecated ICU 54 Use ucol_open() with language tag collation keywords instead.
470 */
471U_DEPRECATED UCollator* U_EXPORT2
472ucol_openFromShortString( const char *definition,
473 UBool forceDefaults,
474 UParseError *parseError,
475 UErrorCode *status);
476#endif /* U_HIDE_DEPRECATED_API */
477
478#ifndef U_HIDE_DEPRECATED_API
479/**
480 * Get a set containing the contractions defined by the collator. The set includes
481 * both the root collator's contractions and the contractions defined by the collator. This set
482 * will contain only strings. If a tailoring explicitly suppresses contractions from
483 * the root collator (like Russian), removed contractions will not be in the resulting set.
484 * @param coll collator
485 * @param conts the set to hold the result. It gets emptied before
486 * contractions are added.
487 * @param status to hold the error code
488 * @return the size of the contraction set
489 *
490 * @deprecated ICU 3.4, use ucol_getContractionsAndExpansions instead
491 */
492U_DEPRECATED int32_t U_EXPORT2
493ucol_getContractions( const UCollator *coll,
494 USet *conts,
495 UErrorCode *status);
496#endif /* U_HIDE_DEPRECATED_API */
497
498/**
499 * Get a set containing the expansions defined by the collator. The set includes
500 * both the root collator's expansions and the expansions defined by the tailoring
501 * @param coll collator
502 * @param contractions if not NULL, the set to hold the contractions
503 * @param expansions if not NULL, the set to hold the expansions
504 * @param addPrefixes add the prefix contextual elements to contractions
505 * @param status to hold the error code
506 *
507 * @stable ICU 3.4
508 */
509U_STABLE void U_EXPORT2
510ucol_getContractionsAndExpansions( const UCollator *coll,
511 USet *contractions, USet *expansions,
512 UBool addPrefixes, UErrorCode *status);
513
514/**
515 * Close a UCollator.
516 * Once closed, a UCollator should not be used. Every open collator should
517 * be closed. Otherwise, a memory leak will result.
518 * @param coll The UCollator to close.
519 * @see ucol_open
520 * @see ucol_openRules
521 * @see ucol_safeClone
522 * @stable ICU 2.0
523 */
524U_STABLE void U_EXPORT2
525ucol_close(UCollator *coll);
526
527#if U_SHOW_CPLUSPLUS_API
528
529U_NAMESPACE_BEGIN
530
531/**
532 * \class LocalUCollatorPointer
533 * "Smart pointer" class, closes a UCollator via ucol_close().
534 * For most methods see the LocalPointerBase base class.
535 *
536 * @see LocalPointerBase
537 * @see LocalPointer
538 * @stable ICU 4.4
539 */
540U_DEFINE_LOCAL_OPEN_POINTER(LocalUCollatorPointer, UCollator, ucol_close);
541
542U_NAMESPACE_END
543
544#endif
545
546/**
547 * Compare two strings.
548 * The strings will be compared using the options already specified.
549 * @param coll The UCollator containing the comparison rules.
550 * @param source The source string.
551 * @param sourceLength The length of source, or -1 if null-terminated.
552 * @param target The target string.
553 * @param targetLength The length of target, or -1 if null-terminated.
554 * @return The result of comparing the strings; one of UCOL_EQUAL,
555 * UCOL_GREATER, UCOL_LESS
556 * @see ucol_greater
557 * @see ucol_greaterOrEqual
558 * @see ucol_equal
559 * @stable ICU 2.0
560 */
561U_STABLE UCollationResult U_EXPORT2
562ucol_strcoll( const UCollator *coll,
563 const UChar *source,
564 int32_t sourceLength,
565 const UChar *target,
566 int32_t targetLength);
567
568/**
569* Compare two strings in UTF-8.
570* The strings will be compared using the options already specified.
571* Note: When input string contains malformed a UTF-8 byte sequence,
572* this function treats these bytes as REPLACEMENT CHARACTER (U+FFFD).
573* @param coll The UCollator containing the comparison rules.
574* @param source The source UTF-8 string.
575* @param sourceLength The length of source, or -1 if null-terminated.
576* @param target The target UTF-8 string.
577* @param targetLength The length of target, or -1 if null-terminated.
578* @param status A pointer to a UErrorCode to receive any errors
579* @return The result of comparing the strings; one of UCOL_EQUAL,
580* UCOL_GREATER, UCOL_LESS
581* @see ucol_greater
582* @see ucol_greaterOrEqual
583* @see ucol_equal
584* @stable ICU 50
585*/
586U_STABLE UCollationResult U_EXPORT2
587ucol_strcollUTF8(
588 const UCollator *coll,
589 const char *source,
590 int32_t sourceLength,
591 const char *target,
592 int32_t targetLength,
593 UErrorCode *status);
594
595/**
596 * Determine if one string is greater than another.
597 * This function is equivalent to {@link #ucol_strcoll } == UCOL_GREATER
598 * @param coll The UCollator containing the comparison rules.
599 * @param source The source string.
600 * @param sourceLength The length of source, or -1 if null-terminated.
601 * @param target The target string.
602 * @param targetLength The length of target, or -1 if null-terminated.
603 * @return TRUE if source is greater than target, FALSE otherwise.
604 * @see ucol_strcoll
605 * @see ucol_greaterOrEqual
606 * @see ucol_equal
607 * @stable ICU 2.0
608 */
609U_STABLE UBool U_EXPORT2
610ucol_greater(const UCollator *coll,
611 const UChar *source, int32_t sourceLength,
612 const UChar *target, int32_t targetLength);
613
614/**
615 * Determine if one string is greater than or equal to another.
616 * This function is equivalent to {@link #ucol_strcoll } != UCOL_LESS
617 * @param coll The UCollator containing the comparison rules.
618 * @param source The source string.
619 * @param sourceLength The length of source, or -1 if null-terminated.
620 * @param target The target string.
621 * @param targetLength The length of target, or -1 if null-terminated.
622 * @return TRUE if source is greater than or equal to target, FALSE otherwise.
623 * @see ucol_strcoll
624 * @see ucol_greater
625 * @see ucol_equal
626 * @stable ICU 2.0
627 */
628U_STABLE UBool U_EXPORT2
629ucol_greaterOrEqual(const UCollator *coll,
630 const UChar *source, int32_t sourceLength,
631 const UChar *target, int32_t targetLength);
632
633/**
634 * Compare two strings for equality.
635 * This function is equivalent to {@link #ucol_strcoll } == UCOL_EQUAL
636 * @param coll The UCollator containing the comparison rules.
637 * @param source The source string.
638 * @param sourceLength The length of source, or -1 if null-terminated.
639 * @param target The target string.
640 * @param targetLength The length of target, or -1 if null-terminated.
641 * @return TRUE if source is equal to target, FALSE otherwise
642 * @see ucol_strcoll
643 * @see ucol_greater
644 * @see ucol_greaterOrEqual
645 * @stable ICU 2.0
646 */
647U_STABLE UBool U_EXPORT2
648ucol_equal(const UCollator *coll,
649 const UChar *source, int32_t sourceLength,
650 const UChar *target, int32_t targetLength);
651
652/**
653 * Compare two UTF-8 encoded trings.
654 * The strings will be compared using the options already specified.
655 * @param coll The UCollator containing the comparison rules.
656 * @param sIter The source string iterator.
657 * @param tIter The target string iterator.
658 * @return The result of comparing the strings; one of UCOL_EQUAL,
659 * UCOL_GREATER, UCOL_LESS
660 * @param status A pointer to a UErrorCode to receive any errors
661 * @see ucol_strcoll
662 * @stable ICU 2.6
663 */
664U_STABLE UCollationResult U_EXPORT2
665ucol_strcollIter( const UCollator *coll,
666 UCharIterator *sIter,
667 UCharIterator *tIter,
668 UErrorCode *status);
669
670/**
671 * Get the collation strength used in a UCollator.
672 * The strength influences how strings are compared.
673 * @param coll The UCollator to query.
674 * @return The collation strength; one of UCOL_PRIMARY, UCOL_SECONDARY,
675 * UCOL_TERTIARY, UCOL_QUATERNARY, UCOL_IDENTICAL
676 * @see ucol_setStrength
677 * @stable ICU 2.0
678 */
679U_STABLE UCollationStrength U_EXPORT2
680ucol_getStrength(const UCollator *coll);
681
682/**
683 * Set the collation strength used in a UCollator.
684 * The strength influences how strings are compared.
685 * @param coll The UCollator to set.
686 * @param strength The desired collation strength; one of UCOL_PRIMARY,
687 * UCOL_SECONDARY, UCOL_TERTIARY, UCOL_QUATERNARY, UCOL_IDENTICAL, UCOL_DEFAULT
688 * @see ucol_getStrength
689 * @stable ICU 2.0
690 */
691U_STABLE void U_EXPORT2
692ucol_setStrength(UCollator *coll,
693 UCollationStrength strength);
694
695/**
696 * Retrieves the reordering codes for this collator.
697 * These reordering codes are a combination of UScript codes and UColReorderCode entries.
698 * @param coll The UCollator to query.
699 * @param dest The array to fill with the script ordering.
700 * @param destCapacity The length of dest. If it is 0, then dest may be NULL and the function
701 * will only return the length of the result without writing any codes (pre-flighting).
702 * @param pErrorCode Must be a valid pointer to an error code value, which must not indicate a
703 * failure before the function call.
704 * @return The number of reordering codes written to the dest array.
705 * @see ucol_setReorderCodes
706 * @see ucol_getEquivalentReorderCodes
707 * @see UScriptCode
708 * @see UColReorderCode
709 * @stable ICU 4.8
710 */
711U_STABLE int32_t U_EXPORT2
712ucol_getReorderCodes(const UCollator* coll,
713 int32_t* dest,
714 int32_t destCapacity,
715 UErrorCode *pErrorCode);
716/**
717 * Sets the reordering codes for this collator.
718 * Collation reordering allows scripts and some other groups of characters
719 * to be moved relative to each other. This reordering is done on top of
720 * the DUCET/CLDR standard collation order. Reordering can specify groups to be placed
721 * at the start and/or the end of the collation order. These groups are specified using
722 * UScript codes and UColReorderCode entries.
723 *
724 * <p>By default, reordering codes specified for the start of the order are placed in the
725 * order given after several special non-script blocks. These special groups of characters
726 * are space, punctuation, symbol, currency, and digit. These special groups are represented with
727 * UColReorderCode entries. Script groups can be intermingled with
728 * these special non-script groups if those special groups are explicitly specified in the reordering.
729 *
730 * <p>The special code OTHERS stands for any script that is not explicitly
731 * mentioned in the list of reordering codes given. Anything that is after OTHERS
732 * will go at the very end of the reordering in the order given.
733 *
734 * <p>The special reorder code DEFAULT will reset the reordering for this collator
735 * to the default for this collator. The default reordering may be the DUCET/CLDR order or may be a reordering that
736 * was specified when this collator was created from resource data or from rules. The
737 * DEFAULT code <b>must</b> be the sole code supplied when it is used.
738 * If not, then U_ILLEGAL_ARGUMENT_ERROR will be set.
739 *
740 * <p>The special reorder code NONE will remove any reordering for this collator.
741 * The result of setting no reordering will be to have the DUCET/CLDR ordering used. The
742 * NONE code <b>must</b> be the sole code supplied when it is used.
743 *
744 * @param coll The UCollator to set.
745 * @param reorderCodes An array of script codes in the new order. This can be NULL if the
746 * length is also set to 0. An empty array will clear any reordering codes on the collator.
747 * @param reorderCodesLength The length of reorderCodes.
748 * @param pErrorCode Must be a valid pointer to an error code value, which must not indicate a
749 * failure before the function call.
750 * @see ucol_getReorderCodes
751 * @see ucol_getEquivalentReorderCodes
752 * @see UScriptCode
753 * @see UColReorderCode
754 * @stable ICU 4.8
755 */
756U_STABLE void U_EXPORT2
757ucol_setReorderCodes(UCollator* coll,
758 const int32_t* reorderCodes,
759 int32_t reorderCodesLength,
760 UErrorCode *pErrorCode);
761
762/**
763 * Retrieves the reorder codes that are grouped with the given reorder code. Some reorder
764 * codes will be grouped and must reorder together.
765 * Beginning with ICU 55, scripts only reorder together if they are primary-equal,
766 * for example Hiragana and Katakana.
767 *
768 * @param reorderCode The reorder code to determine equivalence for.
769 * @param dest The array to fill with the script ordering.
770 * @param destCapacity The length of dest. If it is 0, then dest may be NULL and the function
771 * will only return the length of the result without writing any codes (pre-flighting).
772 * @param pErrorCode Must be a valid pointer to an error code value, which must not indicate
773 * a failure before the function call.
774 * @return The number of reordering codes written to the dest array.
775 * @see ucol_setReorderCodes
776 * @see ucol_getReorderCodes
777 * @see UScriptCode
778 * @see UColReorderCode
779 * @stable ICU 4.8
780 */
781U_STABLE int32_t U_EXPORT2
782ucol_getEquivalentReorderCodes(int32_t reorderCode,
783 int32_t* dest,
784 int32_t destCapacity,
785 UErrorCode *pErrorCode);
786
787/**
788 * Get the display name for a UCollator.
789 * The display name is suitable for presentation to a user.
790 * @param objLoc The locale of the collator in question.
791 * @param dispLoc The locale for display.
792 * @param result A pointer to a buffer to receive the attribute.
793 * @param resultLength The maximum size of result.
794 * @param status A pointer to a UErrorCode to receive any errors
795 * @return The total buffer size needed; if greater than resultLength,
796 * the output was truncated.
797 * @stable ICU 2.0
798 */
799U_STABLE int32_t U_EXPORT2
800ucol_getDisplayName( const char *objLoc,
801 const char *dispLoc,
802 UChar *result,
803 int32_t resultLength,
804 UErrorCode *status);
805
806/**
807 * Get a locale for which collation rules are available.
808 * A UCollator in a locale returned by this function will perform the correct
809 * collation for the locale.
810 * @param localeIndex The index of the desired locale.
811 * @return A locale for which collation rules are available, or 0 if none.
812 * @see ucol_countAvailable
813 * @stable ICU 2.0
814 */
815U_STABLE const char* U_EXPORT2
816ucol_getAvailable(int32_t localeIndex);
817
818/**
819 * Determine how many locales have collation rules available.
820 * This function is most useful as determining the loop ending condition for
821 * calls to {@link #ucol_getAvailable }.
822 * @return The number of locales for which collation rules are available.
823 * @see ucol_getAvailable
824 * @stable ICU 2.0
825 */
826U_STABLE int32_t U_EXPORT2
827ucol_countAvailable(void);
828
829#if !UCONFIG_NO_SERVICE
830/**
831 * Create a string enumerator of all locales for which a valid
832 * collator may be opened.
833 * @param status input-output error code
834 * @return a string enumeration over locale strings. The caller is
835 * responsible for closing the result.
836 * @stable ICU 3.0
837 */
838U_STABLE UEnumeration* U_EXPORT2
839ucol_openAvailableLocales(UErrorCode *status);
840#endif
841
842/**
843 * Create a string enumerator of all possible keywords that are relevant to
844 * collation. At this point, the only recognized keyword for this
845 * service is "collation".
846 * @param status input-output error code
847 * @return a string enumeration over locale strings. The caller is
848 * responsible for closing the result.
849 * @stable ICU 3.0
850 */
851U_STABLE UEnumeration* U_EXPORT2
852ucol_getKeywords(UErrorCode *status);
853
854/**
855 * Given a keyword, create a string enumeration of all values
856 * for that keyword that are currently in use.
857 * @param keyword a particular keyword as enumerated by
858 * ucol_getKeywords. If any other keyword is passed in, *status is set
859 * to U_ILLEGAL_ARGUMENT_ERROR.
860 * @param status input-output error code
861 * @return a string enumeration over collation keyword values, or NULL
862 * upon error. The caller is responsible for closing the result.
863 * @stable ICU 3.0
864 */
865U_STABLE UEnumeration* U_EXPORT2
866ucol_getKeywordValues(const char *keyword, UErrorCode *status);
867
868/**
869 * Given a key and a locale, returns an array of string values in a preferred
870 * order that would make a difference. These are all and only those values where
871 * the open (creation) of the service with the locale formed from the input locale
872 * plus input keyword and that value has different behavior than creation with the
873 * input locale alone.
874 * @param key one of the keys supported by this service. For now, only
875 * "collation" is supported.
876 * @param locale the locale
877 * @param commonlyUsed if set to true it will return only commonly used values
878 * with the given locale in preferred order. Otherwise,
879 * it will return all the available values for the locale.
880 * @param status error status
881 * @return a string enumeration over keyword values for the given key and the locale.
882 * @stable ICU 4.2
883 */
884U_STABLE UEnumeration* U_EXPORT2
885ucol_getKeywordValuesForLocale(const char* key,
886 const char* locale,
887 UBool commonlyUsed,
888 UErrorCode* status);
889
890/**
891 * Return the functionally equivalent locale for the specified
892 * input locale, with respect to given keyword, for the
893 * collation service. If two different input locale + keyword
894 * combinations produce the same result locale, then collators
895 * instantiated for these two different input locales will behave
896 * equivalently. The converse is not always true; two collators
897 * may in fact be equivalent, but return different results, due to
898 * internal details. The return result has no other meaning than
899 * that stated above, and implies nothing as to the relationship
900 * between the two locales. This is intended for use by
901 * applications who wish to cache collators, or otherwise reuse
902 * collators when possible. The functional equivalent may change
903 * over time. For more information, please see the <a
904 * href="http://userguide.icu-project.org/locale#TOC-Locales-and-Services">
905 * Locales and Services</a> section of the ICU User Guide.
906 * @param result fillin for the functionally equivalent result locale
907 * @param resultCapacity capacity of the fillin buffer
908 * @param keyword a particular keyword as enumerated by
909 * ucol_getKeywords.
910 * @param locale the specified input locale
911 * @param isAvailable if non-NULL, pointer to a fillin parameter that
912 * on return indicates whether the specified input locale was 'available'
913 * to the collation service. A locale is defined as 'available' if it
914 * physically exists within the collation locale data.
915 * @param status pointer to input-output error code
916 * @return the actual buffer size needed for the locale. If greater
917 * than resultCapacity, the returned full name will be truncated and
918 * an error code will be returned.
919 * @stable ICU 3.0
920 */
921U_STABLE int32_t U_EXPORT2
922ucol_getFunctionalEquivalent(char* result, int32_t resultCapacity,
923 const char* keyword, const char* locale,
924 UBool* isAvailable, UErrorCode* status);
925
926/**
927 * Get the collation tailoring rules from a UCollator.
928 * The rules will follow the rule syntax.
929 * @param coll The UCollator to query.
930 * @param length
931 * @return The collation tailoring rules.
932 * @stable ICU 2.0
933 */
934U_STABLE const UChar* U_EXPORT2
935ucol_getRules( const UCollator *coll,
936 int32_t *length);
937
938#ifndef U_HIDE_DEPRECATED_API
939/** Get the short definition string for a collator. This API harvests the collator's
940 * locale and the attribute set and produces a string that can be used for opening
941 * a collator with the same attributes using the ucol_openFromShortString API.
942 * This string will be normalized.
943 * The structure and the syntax of the string is defined in the "Naming collators"
944 * section of the users guide:
945 * http://userguide.icu-project.org/collation/concepts#TOC-Collator-naming-scheme
946 * This API supports preflighting.
947 * @param coll a collator
948 * @param locale a locale that will appear as a collators locale in the resulting
949 * short string definition. If NULL, the locale will be harvested
950 * from the collator.
951 * @param buffer space to hold the resulting string
952 * @param capacity capacity of the buffer
953 * @param status for returning errors. All the preflighting errors are featured
954 * @return length of the resulting string
955 * @see ucol_openFromShortString
956 * @see ucol_normalizeShortDefinitionString
957 * @deprecated ICU 54
958 */
959U_DEPRECATED int32_t U_EXPORT2
960ucol_getShortDefinitionString(const UCollator *coll,
961 const char *locale,
962 char *buffer,
963 int32_t capacity,
964 UErrorCode *status);
965
966/** Verifies and normalizes short definition string.
967 * Normalized short definition string has all the option sorted by the argument name,
968 * so that equivalent definition strings are the same.
969 * This API supports preflighting.
970 * @param source definition string
971 * @param destination space to hold the resulting string
972 * @param capacity capacity of the buffer
973 * @param parseError if not NULL, structure that will get filled with error's pre
974 * and post context in case of error.
975 * @param status Error code. This API will return an error if an invalid attribute
976 * or attribute/value combination is specified. All the preflighting
977 * errors are also featured
978 * @return length of the resulting normalized string.
979 *
980 * @see ucol_openFromShortString
981 * @see ucol_getShortDefinitionString
982 *
983 * @deprecated ICU 54
984 */
985
986U_DEPRECATED int32_t U_EXPORT2
987ucol_normalizeShortDefinitionString(const char *source,
988 char *destination,
989 int32_t capacity,
990 UParseError *parseError,
991 UErrorCode *status);
992#endif /* U_HIDE_DEPRECATED_API */
993
994
995/**
996 * Get a sort key for a string from a UCollator.
997 * Sort keys may be compared using <TT>strcmp</TT>.
998 *
999 * Note that sort keys are often less efficient than simply doing comparison.
1000 * For more details, see the ICU User Guide.
1001 *
1002 * Like ICU functions that write to an output buffer, the buffer contents
1003 * is undefined if the buffer capacity (resultLength parameter) is too small.
1004 * Unlike ICU functions that write a string to an output buffer,
1005 * the terminating zero byte is counted in the sort key length.
1006 * @param coll The UCollator containing the collation rules.
1007 * @param source The string to transform.
1008 * @param sourceLength The length of source, or -1 if null-terminated.
1009 * @param result A pointer to a buffer to receive the attribute.
1010 * @param resultLength The maximum size of result.
1011 * @return The size needed to fully store the sort key.
1012 * If there was an internal error generating the sort key,
1013 * a zero value is returned.
1014 * @see ucol_keyHashCode
1015 * @stable ICU 2.0
1016 */
1017U_STABLE int32_t U_EXPORT2
1018ucol_getSortKey(const UCollator *coll,
1019 const UChar *source,
1020 int32_t sourceLength,
1021 uint8_t *result,
1022 int32_t resultLength);
1023
1024
1025/** Gets the next count bytes of a sort key. Caller needs
1026 * to preserve state array between calls and to provide
1027 * the same type of UCharIterator set with the same string.
1028 * The destination buffer provided must be big enough to store
1029 * the number of requested bytes.
1030 *
1031 * The generated sort key may or may not be compatible with
1032 * sort keys generated using ucol_getSortKey().
1033 * @param coll The UCollator containing the collation rules.
1034 * @param iter UCharIterator containing the string we need
1035 * the sort key to be calculated for.
1036 * @param state Opaque state of sortkey iteration.
1037 * @param dest Buffer to hold the resulting sortkey part
1038 * @param count number of sort key bytes required.
1039 * @param status error code indicator.
1040 * @return the actual number of bytes of a sortkey. It can be
1041 * smaller than count if we have reached the end of
1042 * the sort key.
1043 * @stable ICU 2.6
1044 */
1045U_STABLE int32_t U_EXPORT2
1046ucol_nextSortKeyPart(const UCollator *coll,
1047 UCharIterator *iter,
1048 uint32_t state[2],
1049 uint8_t *dest, int32_t count,
1050 UErrorCode *status);
1051
1052/** enum that is taken by ucol_getBound API
1053 * See below for explanation
1054 * do not change the values assigned to the
1055 * members of this enum. Underlying code
1056 * depends on them having these numbers
1057 * @stable ICU 2.0
1058 */
1059typedef enum {
1060 /** lower bound */
1061 UCOL_BOUND_LOWER = 0,
1062 /** upper bound that will match strings of exact size */
1063 UCOL_BOUND_UPPER = 1,
1064 /** upper bound that will match all the strings that have the same initial substring as the given string */
1065 UCOL_BOUND_UPPER_LONG = 2,
1066#ifndef U_HIDE_DEPRECATED_API
1067 /**
1068 * One more than the highest normal UColBoundMode value.
1069 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
1070 */
1071 UCOL_BOUND_VALUE_COUNT
1072#endif /* U_HIDE_DEPRECATED_API */
1073} UColBoundMode;
1074
1075/**
1076 * Produce a bound for a given sortkey and a number of levels.
1077 * Return value is always the number of bytes needed, regardless of
1078 * whether the result buffer was big enough or even valid.<br>
1079 * Resulting bounds can be used to produce a range of strings that are
1080 * between upper and lower bounds. For example, if bounds are produced
1081 * for a sortkey of string "smith", strings between upper and lower
1082 * bounds with one level would include "Smith", "SMITH", "sMiTh".<br>
1083 * There are two upper bounds that can be produced. If UCOL_BOUND_UPPER
1084 * is produced, strings matched would be as above. However, if bound
1085 * produced using UCOL_BOUND_UPPER_LONG is used, the above example will
1086 * also match "Smithsonian" and similar.<br>
1087 * For more on usage, see example in cintltst/capitst.c in procedure
1088 * TestBounds.
1089 * Sort keys may be compared using <TT>strcmp</TT>.
1090 * @param source The source sortkey.
1091 * @param sourceLength The length of source, or -1 if null-terminated.
1092 * (If an unmodified sortkey is passed, it is always null
1093 * terminated).
1094 * @param boundType Type of bound required. It can be UCOL_BOUND_LOWER, which
1095 * produces a lower inclusive bound, UCOL_BOUND_UPPER, that
1096 * produces upper bound that matches strings of the same length
1097 * or UCOL_BOUND_UPPER_LONG that matches strings that have the
1098 * same starting substring as the source string.
1099 * @param noOfLevels Number of levels required in the resulting bound (for most
1100 * uses, the recommended value is 1). See users guide for
1101 * explanation on number of levels a sortkey can have.
1102 * @param result A pointer to a buffer to receive the resulting sortkey.
1103 * @param resultLength The maximum size of result.
1104 * @param status Used for returning error code if something went wrong. If the
1105 * number of levels requested is higher than the number of levels
1106 * in the source key, a warning (U_SORT_KEY_TOO_SHORT_WARNING) is
1107 * issued.
1108 * @return The size needed to fully store the bound.
1109 * @see ucol_keyHashCode
1110 * @stable ICU 2.1
1111 */
1112U_STABLE int32_t U_EXPORT2
1113ucol_getBound(const uint8_t *source,
1114 int32_t sourceLength,
1115 UColBoundMode boundType,
1116 uint32_t noOfLevels,
1117 uint8_t *result,
1118 int32_t resultLength,
1119 UErrorCode *status);
1120
1121/**
1122 * Gets the version information for a Collator. Version is currently
1123 * an opaque 32-bit number which depends, among other things, on major
1124 * versions of the collator tailoring and UCA.
1125 * @param coll The UCollator to query.
1126 * @param info the version # information, the result will be filled in
1127 * @stable ICU 2.0
1128 */
1129U_STABLE void U_EXPORT2
1130ucol_getVersion(const UCollator* coll, UVersionInfo info);
1131
1132/**
1133 * Gets the UCA version information for a Collator. Version is the
1134 * UCA version number (3.1.1, 4.0).
1135 * @param coll The UCollator to query.
1136 * @param info the version # information, the result will be filled in
1137 * @stable ICU 2.8
1138 */
1139U_STABLE void U_EXPORT2
1140ucol_getUCAVersion(const UCollator* coll, UVersionInfo info);
1141
1142/**
1143 * Merges two sort keys. The levels are merged with their corresponding counterparts
1144 * (primaries with primaries, secondaries with secondaries etc.). Between the values
1145 * from the same level a separator is inserted.
1146 *
1147 * This is useful, for example, for combining sort keys from first and last names
1148 * to sort such pairs.
1149 * See http://www.unicode.org/reports/tr10/#Merging_Sort_Keys
1150 *
1151 * The recommended way to achieve "merged" sorting is by
1152 * concatenating strings with U+FFFE between them.
1153 * The concatenation has the same sort order as the merged sort keys,
1154 * but merge(getSortKey(str1), getSortKey(str2)) may differ from getSortKey(str1 + '\\uFFFE' + str2).
1155 * Using strings with U+FFFE may yield shorter sort keys.
1156 *
1157 * For details about Sort Key Features see
1158 * http://userguide.icu-project.org/collation/api#TOC-Sort-Key-Features
1159 *
1160 * It is possible to merge multiple sort keys by consecutively merging
1161 * another one with the intermediate result.
1162 *
1163 * The length of the merge result is the sum of the lengths of the input sort keys.
1164 *
1165 * Example (uncompressed):
1166 * <pre>191B1D 01 050505 01 910505 00
1167 * 1F2123 01 050505 01 910505 00</pre>
1168 * will be merged as
1169 * <pre>191B1D 02 1F2123 01 050505 02 050505 01 910505 02 910505 00</pre>
1170 *
1171 * If the destination buffer is not big enough, then its contents are undefined.
1172 * If any of source lengths are zero or any of the source pointers are NULL/undefined,
1173 * the result is of size zero.
1174 *
1175 * @param src1 the first sort key
1176 * @param src1Length the length of the first sort key, including the zero byte at the end;
1177 * can be -1 if the function is to find the length
1178 * @param src2 the second sort key
1179 * @param src2Length the length of the second sort key, including the zero byte at the end;
1180 * can be -1 if the function is to find the length
1181 * @param dest the buffer where the merged sort key is written,
1182 * can be NULL if destCapacity==0
1183 * @param destCapacity the number of bytes in the dest buffer
1184 * @return the length of the merged sort key, src1Length+src2Length;
1185 * can be larger than destCapacity, or 0 if an error occurs (only for illegal arguments),
1186 * in which cases the contents of dest is undefined
1187 * @stable ICU 2.0
1188 */
1189U_STABLE int32_t U_EXPORT2
1190ucol_mergeSortkeys(const uint8_t *src1, int32_t src1Length,
1191 const uint8_t *src2, int32_t src2Length,
1192 uint8_t *dest, int32_t destCapacity);
1193
1194/**
1195 * Universal attribute setter
1196 * @param coll collator which attributes are to be changed
1197 * @param attr attribute type
1198 * @param value attribute value
1199 * @param status to indicate whether the operation went on smoothly or there were errors
1200 * @see UColAttribute
1201 * @see UColAttributeValue
1202 * @see ucol_getAttribute
1203 * @stable ICU 2.0
1204 */
1205U_STABLE void U_EXPORT2
1206ucol_setAttribute(UCollator *coll, UColAttribute attr, UColAttributeValue value, UErrorCode *status);
1207
1208/**
1209 * Universal attribute getter
1210 * @param coll collator which attributes are to be changed
1211 * @param attr attribute type
1212 * @return attribute value
1213 * @param status to indicate whether the operation went on smoothly or there were errors
1214 * @see UColAttribute
1215 * @see UColAttributeValue
1216 * @see ucol_setAttribute
1217 * @stable ICU 2.0
1218 */
1219U_STABLE UColAttributeValue U_EXPORT2
1220ucol_getAttribute(const UCollator *coll, UColAttribute attr, UErrorCode *status);
1221
1222/**
1223 * Sets the variable top to the top of the specified reordering group.
1224 * The variable top determines the highest-sorting character
1225 * which is affected by UCOL_ALTERNATE_HANDLING.
1226 * If that attribute is set to UCOL_NON_IGNORABLE, then the variable top has no effect.
1227 * @param coll the collator
1228 * @param group one of UCOL_REORDER_CODE_SPACE, UCOL_REORDER_CODE_PUNCTUATION,
1229 * UCOL_REORDER_CODE_SYMBOL, UCOL_REORDER_CODE_CURRENCY;
1230 * or UCOL_REORDER_CODE_DEFAULT to restore the default max variable group
1231 * @param pErrorCode Standard ICU error code. Its input value must
1232 * pass the U_SUCCESS() test, or else the function returns
1233 * immediately. Check for U_FAILURE() on output or use with
1234 * function chaining. (See User Guide for details.)
1235 * @see ucol_getMaxVariable
1236 * @stable ICU 53
1237 */
1238U_STABLE void U_EXPORT2
1239ucol_setMaxVariable(UCollator *coll, UColReorderCode group, UErrorCode *pErrorCode);
1240
1241/**
1242 * Returns the maximum reordering group whose characters are affected by UCOL_ALTERNATE_HANDLING.
1243 * @param coll the collator
1244 * @return the maximum variable reordering group.
1245 * @see ucol_setMaxVariable
1246 * @stable ICU 53
1247 */
1248U_STABLE UColReorderCode U_EXPORT2
1249ucol_getMaxVariable(const UCollator *coll);
1250
1251#ifndef U_HIDE_DEPRECATED_API
1252/**
1253 * Sets the variable top to the primary weight of the specified string.
1254 *
1255 * Beginning with ICU 53, the variable top is pinned to
1256 * the top of one of the supported reordering groups,
1257 * and it must not be beyond the last of those groups.
1258 * See ucol_setMaxVariable().
1259 * @param coll the collator
1260 * @param varTop one or more (if contraction) UChars to which the variable top should be set
1261 * @param len length of variable top string. If -1 it is considered to be zero terminated.
1262 * @param status error code. If error code is set, the return value is undefined.
1263 * Errors set by this function are:<br>
1264 * U_CE_NOT_FOUND_ERROR if more than one character was passed and there is no such contraction<br>
1265 * U_ILLEGAL_ARGUMENT_ERROR if the variable top is beyond
1266 * the last reordering group supported by ucol_setMaxVariable()
1267 * @return variable top primary weight
1268 * @see ucol_getVariableTop
1269 * @see ucol_restoreVariableTop
1270 * @deprecated ICU 53 Call ucol_setMaxVariable() instead.
1271 */
1272U_DEPRECATED uint32_t U_EXPORT2
1273ucol_setVariableTop(UCollator *coll,
1274 const UChar *varTop, int32_t len,
1275 UErrorCode *status);
1276#endif /* U_HIDE_DEPRECATED_API */
1277
1278/**
1279 * Gets the variable top value of a Collator.
1280 * @param coll collator which variable top needs to be retrieved
1281 * @param status error code (not changed by function). If error code is set,
1282 * the return value is undefined.
1283 * @return the variable top primary weight
1284 * @see ucol_getMaxVariable
1285 * @see ucol_setVariableTop
1286 * @see ucol_restoreVariableTop
1287 * @stable ICU 2.0
1288 */
1289U_STABLE uint32_t U_EXPORT2 ucol_getVariableTop(const UCollator *coll, UErrorCode *status);
1290
1291#ifndef U_HIDE_DEPRECATED_API
1292/**
1293 * Sets the variable top to the specified primary weight.
1294 *
1295 * Beginning with ICU 53, the variable top is pinned to
1296 * the top of one of the supported reordering groups,
1297 * and it must not be beyond the last of those groups.
1298 * See ucol_setMaxVariable().
1299 * @param coll collator to be set
1300 * @param varTop primary weight, as returned by ucol_setVariableTop or ucol_getVariableTop
1301 * @param status error code
1302 * @see ucol_getVariableTop
1303 * @see ucol_setVariableTop
1304 * @deprecated ICU 53 Call ucol_setMaxVariable() instead.
1305 */
1306U_DEPRECATED void U_EXPORT2
1307ucol_restoreVariableTop(UCollator *coll, const uint32_t varTop, UErrorCode *status);
1308#endif /* U_HIDE_DEPRECATED_API */
1309
1310/**
1311 * Thread safe cloning operation. The result is a clone of a given collator.
1312 * @param coll collator to be cloned
1313 * @param stackBuffer <em>Deprecated functionality as of ICU 52, use NULL.</em><br>
1314 * user allocated space for the new clone.
1315 * If NULL new memory will be allocated.
1316 * If buffer is not large enough, new memory will be allocated.
1317 * Clients can use the U_COL_SAFECLONE_BUFFERSIZE.
1318 * @param pBufferSize <em>Deprecated functionality as of ICU 52, use NULL or 1.</em><br>
1319 * pointer to size of allocated space.
1320 * If *pBufferSize == 0, a sufficient size for use in cloning will
1321 * be returned ('pre-flighting')
1322 * If *pBufferSize is not enough for a stack-based safe clone,
1323 * new memory will be allocated.
1324 * @param status to indicate whether the operation went on smoothly or there were errors
1325 * An informational status value, U_SAFECLONE_ALLOCATED_ERROR, is used if any
1326 * allocations were necessary.
1327 * @return pointer to the new clone
1328 * @see ucol_open
1329 * @see ucol_openRules
1330 * @see ucol_close
1331 * @stable ICU 2.0
1332 */
1333U_STABLE UCollator* U_EXPORT2
1334ucol_safeClone(const UCollator *coll,
1335 void *stackBuffer,
1336 int32_t *pBufferSize,
1337 UErrorCode *status);
1338
1339#ifndef U_HIDE_DEPRECATED_API
1340
1341/** default memory size for the new clone.
1342 * @deprecated ICU 52. Do not rely on ucol_safeClone() cloning into any provided buffer.
1343 */
1344#define U_COL_SAFECLONE_BUFFERSIZE 1
1345
1346#endif /* U_HIDE_DEPRECATED_API */
1347
1348/**
1349 * Returns current rules. Delta defines whether full rules are returned or just the tailoring.
1350 * Returns number of UChars needed to store rules. If buffer is NULL or bufferLen is not enough
1351 * to store rules, will store up to available space.
1352 *
1353 * ucol_getRules() should normally be used instead.
1354 * See http://userguide.icu-project.org/collation/customization#TOC-Building-on-Existing-Locales
1355 * @param coll collator to get the rules from
1356 * @param delta one of UCOL_TAILORING_ONLY, UCOL_FULL_RULES.
1357 * @param buffer buffer to store the result in. If NULL, you'll get no rules.
1358 * @param bufferLen length of buffer to store rules in. If less than needed you'll get only the part that fits in.
1359 * @return current rules
1360 * @stable ICU 2.0
1361 * @see UCOL_FULL_RULES
1362 */
1363U_STABLE int32_t U_EXPORT2
1364ucol_getRulesEx(const UCollator *coll, UColRuleOption delta, UChar *buffer, int32_t bufferLen);
1365
1366#ifndef U_HIDE_DEPRECATED_API
1367/**
1368 * gets the locale name of the collator. If the collator
1369 * is instantiated from the rules, then this function returns
1370 * NULL.
1371 * @param coll The UCollator for which the locale is needed
1372 * @param type You can choose between requested, valid and actual
1373 * locale. For description see the definition of
1374 * ULocDataLocaleType in uloc.h
1375 * @param status error code of the operation
1376 * @return real locale name from which the collation data comes.
1377 * If the collator was instantiated from rules, returns
1378 * NULL.
1379 * @deprecated ICU 2.8 Use ucol_getLocaleByType instead
1380 */
1381U_DEPRECATED const char * U_EXPORT2
1382ucol_getLocale(const UCollator *coll, ULocDataLocaleType type, UErrorCode *status);
1383#endif /* U_HIDE_DEPRECATED_API */
1384
1385/**
1386 * gets the locale name of the collator. If the collator
1387 * is instantiated from the rules, then this function returns
1388 * NULL.
1389 * @param coll The UCollator for which the locale is needed
1390 * @param type You can choose between requested, valid and actual
1391 * locale. For description see the definition of
1392 * ULocDataLocaleType in uloc.h
1393 * @param status error code of the operation
1394 * @return real locale name from which the collation data comes.
1395 * If the collator was instantiated from rules, returns
1396 * NULL.
1397 * @stable ICU 2.8
1398 */
1399U_STABLE const char * U_EXPORT2
1400ucol_getLocaleByType(const UCollator *coll, ULocDataLocaleType type, UErrorCode *status);
1401
1402/**
1403 * Get a Unicode set that contains all the characters and sequences tailored in
1404 * this collator. The result must be disposed of by using uset_close.
1405 * @param coll The UCollator for which we want to get tailored chars
1406 * @param status error code of the operation
1407 * @return a pointer to newly created USet. Must be be disposed by using uset_close
1408 * @see ucol_openRules
1409 * @see uset_close
1410 * @stable ICU 2.4
1411 */
1412U_STABLE USet * U_EXPORT2
1413ucol_getTailoredSet(const UCollator *coll, UErrorCode *status);
1414
1415#ifndef U_HIDE_INTERNAL_API
1416/** Calculates the set of unsafe code points, given a collator.
1417 * A character is unsafe if you could append any character and cause the ordering to alter significantly.
1418 * Collation sorts in normalized order, so anything that rearranges in normalization can cause this.
1419 * Thus if you have a character like a_umlaut, and you add a lower_dot to it,
1420 * then it normalizes to a_lower_dot + umlaut, and sorts differently.
1421 * @param coll Collator
1422 * @param unsafe a fill-in set to receive the unsafe points
1423 * @param status for catching errors
1424 * @return number of elements in the set
1425 * @internal ICU 3.0
1426 */
1427U_INTERNAL int32_t U_EXPORT2
1428ucol_getUnsafeSet( const UCollator *coll,
1429 USet *unsafe,
1430 UErrorCode *status);
1431
1432/** Touches all resources needed for instantiating a collator from a short string definition,
1433 * thus filling up the cache.
1434 * @param definition A short string containing a locale and a set of attributes.
1435 * Attributes not explicitly mentioned are left at the default
1436 * state for a locale.
1437 * @param parseError if not NULL, structure that will get filled with error's pre
1438 * and post context in case of error.
1439 * @param forceDefaults if FALSE, the settings that are the same as the collator
1440 * default settings will not be applied (for example, setting
1441 * French secondary on a French collator would not be executed).
1442 * If TRUE, all the settings will be applied regardless of the
1443 * collator default value. If the definition
1444 * strings are to be cached, should be set to FALSE.
1445 * @param status Error code. Apart from regular error conditions connected to
1446 * instantiating collators (like out of memory or similar), this
1447 * API will return an error if an invalid attribute or attribute/value
1448 * combination is specified.
1449 * @see ucol_openFromShortString
1450 * @internal ICU 3.2.1
1451 */
1452U_INTERNAL void U_EXPORT2
1453ucol_prepareShortStringOpen( const char *definition,
1454 UBool forceDefaults,
1455 UParseError *parseError,
1456 UErrorCode *status);
1457#endif /* U_HIDE_INTERNAL_API */
1458
1459/** Creates a binary image of a collator. This binary image can be stored and
1460 * later used to instantiate a collator using ucol_openBinary.
1461 * This API supports preflighting.
1462 * @param coll Collator
1463 * @param buffer a fill-in buffer to receive the binary image
1464 * @param capacity capacity of the destination buffer
1465 * @param status for catching errors
1466 * @return size of the image
1467 * @see ucol_openBinary
1468 * @stable ICU 3.2
1469 */
1470U_STABLE int32_t U_EXPORT2
1471ucol_cloneBinary(const UCollator *coll,
1472 uint8_t *buffer, int32_t capacity,
1473 UErrorCode *status);
1474
1475/** Opens a collator from a collator binary image created using
1476 * ucol_cloneBinary. Binary image used in instantiation of the
1477 * collator remains owned by the user and should stay around for
1478 * the lifetime of the collator. The API also takes a base collator
1479 * which must be the root collator.
1480 * @param bin binary image owned by the user and required through the
1481 * lifetime of the collator
1482 * @param length size of the image. If negative, the API will try to
1483 * figure out the length of the image
1484 * @param base Base collator, for lookup of untailored characters.
1485 * Must be the root collator, must not be NULL.
1486 * The base is required to be present through the lifetime of the collator.
1487 * @param status for catching errors
1488 * @return newly created collator
1489 * @see ucol_cloneBinary
1490 * @stable ICU 3.2
1491 */
1492U_STABLE UCollator* U_EXPORT2
1493ucol_openBinary(const uint8_t *bin, int32_t length,
1494 const UCollator *base,
1495 UErrorCode *status);
1496
1497
1498#endif /* #if !UCONFIG_NO_COLLATION */
1499
1500#endif
1501