1/* 2****************************************************************************** 3* Copyright (C) 1996-2009, International Business Machines Corporation and others. 4* All Rights Reserved. 5****************************************************************************** 6*/ 7 8#ifndef UBRK_H 9#define UBRK_H 10 11#include "unicode/utypes.h" 12#include "unicode/uloc.h" 13#include "unicode/utext.h" 14 15/** 16 * A text-break iterator. 17 * For usage in C programs. 18 */ 19#ifndef UBRK_TYPEDEF_UBREAK_ITERATOR 20# define UBRK_TYPEDEF_UBREAK_ITERATOR 21 /** 22 * Opaque type representing an ICU Break iterator object. 23 * @stable ICU 2.0 24 */ 25 typedef void UBreakIterator; 26#endif 27 28#if !UCONFIG_NO_BREAK_ITERATION 29 30#include "unicode/parseerr.h" 31 32/** 33 * \file 34 * \brief C API: BreakIterator 35 * 36 * <h2> BreakIterator C API </h2> 37 * 38 * The BreakIterator C API defines methods for finding the location 39 * of boundaries in text. Pointer to a UBreakIterator maintain a 40 * current position and scan over text returning the index of characters 41 * where boundaries occur. 42 * <p> 43 * Line boundary analysis determines where a text string can be broken 44 * when line-wrapping. The mechanism correctly handles punctuation and 45 * hyphenated words. 46 * <p> 47 * Sentence boundary analysis allows selection with correct 48 * interpretation of periods within numbers and abbreviations, and 49 * trailing punctuation marks such as quotation marks and parentheses. 50 * <p> 51 * Word boundary analysis is used by search and replace functions, as 52 * well as within text editing applications that allow the user to 53 * select words with a double click. Word selection provides correct 54 * interpretation of punctuation marks within and following 55 * words. Characters that are not part of a word, such as symbols or 56 * punctuation marks, have word-breaks on both sides. 57 * <p> 58 * Character boundary analysis identifies the boundaries of 59 * "Extended Grapheme Clusters", which are groupings of codepoints 60 * that should be treated as character-like units for many text operations. 61 * Please see Unicode Standard Annex #29, Unicode Text Segmentation, 62 * http://www.unicode.org/reports/tr29/ for additional information 63 * on grapheme clusters and guidelines on their use. 64 * <p> 65 * Title boundary analysis locates all positions, 66 * typically starts of words, that should be set to Title Case 67 * when title casing the text. 68 * <p> 69 * The text boundary positions are found according to the rules 70 * described in Unicode Standard Annex #29, Text Boundaries, and 71 * Unicode Standard Annex #14, Line Breaking Properties. These 72 * are available at http://www.unicode.org/reports/tr14/ and 73 * http://www.unicode.org/reports/tr29/. 74 * <p> 75 * In addition to the plain C API defined in this header file, an 76 * object oriented C++ API with equivalent functionality is defined in the 77 * file brkiter.h. 78 * <p> 79 * Code snippits illustrating the use of the Break Iterator APIs 80 * are available in the ICU User Guide, 81 * http://icu-project.org/userguide/boundaryAnalysis.html 82 * and in the sample program icu/source/samples/break/break.cpp" 83 */ 84 85/** The possible types of text boundaries. @stable ICU 2.0 */ 86typedef enum UBreakIteratorType { 87 /** Character breaks @stable ICU 2.0 */ 88 UBRK_CHARACTER = 0, 89 /** Word breaks @stable ICU 2.0 */ 90 UBRK_WORD = 1, 91 /** Line breaks @stable ICU 2.0 */ 92 UBRK_LINE = 2, 93 /** Sentence breaks @stable ICU 2.0 */ 94 UBRK_SENTENCE = 3, 95 96#ifndef U_HIDE_DEPRECATED_API 97 /** 98 * Title Case breaks 99 * The iterator created using this type locates title boundaries as described for 100 * Unicode 3.2 only. For Unicode 4.0 and above title boundary iteration, 101 * please use Word Boundary iterator. 102 * 103 * @deprecated ICU 2.8 Use the word break iterator for titlecasing for Unicode 4 and later. 104 */ 105 UBRK_TITLE = 4, 106#endif /* U_HIDE_DEPRECATED_API */ 107 UBRK_COUNT = 5 108} UBreakIteratorType; 109 110/** Value indicating all text boundaries have been returned. 111 * @stable ICU 2.0 112 */ 113#define UBRK_DONE ((int32_t) -1) 114 115 116/** 117 * Enum constants for the word break tags returned by 118 * getRuleStatus(). A range of values is defined for each category of 119 * word, to allow for further subdivisions of a category in future releases. 120 * Applications should check for tag values falling within the range, rather 121 * than for single individual values. 122 * @stable ICU 2.2 123*/ 124typedef enum UWordBreak { 125 /** Tag value for "words" that do not fit into any of other categories. 126 * Includes spaces and most punctuation. */ 127 UBRK_WORD_NONE = 0, 128 /** Upper bound for tags for uncategorized words. */ 129 UBRK_WORD_NONE_LIMIT = 100, 130 /** Tag value for words that appear to be numbers, lower limit. */ 131 UBRK_WORD_NUMBER = 100, 132 /** Tag value for words that appear to be numbers, upper limit. */ 133 UBRK_WORD_NUMBER_LIMIT = 200, 134 /** Tag value for words that contain letters, excluding 135 * hiragana, katakana or ideographic characters, lower limit. */ 136 UBRK_WORD_LETTER = 200, 137 /** Tag value for words containing letters, upper limit */ 138 UBRK_WORD_LETTER_LIMIT = 300, 139 /** Tag value for words containing kana characters, lower limit */ 140 UBRK_WORD_KANA = 300, 141 /** Tag value for words containing kana characters, upper limit */ 142 UBRK_WORD_KANA_LIMIT = 400, 143 /** Tag value for words containing ideographic characters, lower limit */ 144 UBRK_WORD_IDEO = 400, 145 /** Tag value for words containing ideographic characters, upper limit */ 146 UBRK_WORD_IDEO_LIMIT = 500 147} UWordBreak; 148 149/** 150 * Enum constants for the line break tags returned by getRuleStatus(). 151 * A range of values is defined for each category of 152 * word, to allow for further subdivisions of a category in future releases. 153 * Applications should check for tag values falling within the range, rather 154 * than for single individual values. 155 * @stable ICU 2.8 156*/ 157typedef enum ULineBreakTag { 158 /** Tag value for soft line breaks, positions at which a line break 159 * is acceptable but not required */ 160 UBRK_LINE_SOFT = 0, 161 /** Upper bound for soft line breaks. */ 162 UBRK_LINE_SOFT_LIMIT = 100, 163 /** Tag value for a hard, or mandatory line break */ 164 UBRK_LINE_HARD = 100, 165 /** Upper bound for hard line breaks. */ 166 UBRK_LINE_HARD_LIMIT = 200 167} ULineBreakTag; 168 169 170 171/** 172 * Enum constants for the sentence break tags returned by getRuleStatus(). 173 * A range of values is defined for each category of 174 * sentence, to allow for further subdivisions of a category in future releases. 175 * Applications should check for tag values falling within the range, rather 176 * than for single individual values. 177 * @stable ICU 2.8 178*/ 179typedef enum USentenceBreakTag { 180 /** Tag value for for sentences ending with a sentence terminator 181 * ('.', '?', '!', etc.) character, possibly followed by a 182 * hard separator (CR, LF, PS, etc.) 183 */ 184 UBRK_SENTENCE_TERM = 0, 185 /** Upper bound for tags for sentences ended by sentence terminators. */ 186 UBRK_SENTENCE_TERM_LIMIT = 100, 187 /** Tag value for for sentences that do not contain an ending 188 * sentence terminator ('.', '?', '!', etc.) character, but 189 * are ended only by a hard separator (CR, LF, PS, etc.) or end of input. 190 */ 191 UBRK_SENTENCE_SEP = 100, 192 /** Upper bound for tags for sentences ended by a separator. */ 193 UBRK_SENTENCE_SEP_LIMIT = 200 194 /** Tag value for a hard, or mandatory line break */ 195} USentenceBreakTag; 196 197 198/** 199 * Open a new UBreakIterator for locating text boundaries for a specified locale. 200 * A UBreakIterator may be used for detecting character, line, word, 201 * and sentence breaks in text. 202 * @param type The type of UBreakIterator to open: one of UBRK_CHARACTER, UBRK_WORD, 203 * UBRK_LINE, UBRK_SENTENCE 204 * @param locale The locale specifying the text-breaking conventions. 205 * @param text The text to be iterated over. 206 * @param textLength The number of characters in text, or -1 if null-terminated. 207 * @param status A UErrorCode to receive any errors. 208 * @return A UBreakIterator for the specified locale. 209 * @see ubrk_openRules 210 * @stable ICU 2.0 211 */ 212U_STABLE UBreakIterator* U_EXPORT2 213ubrk_open(UBreakIteratorType type, 214 const char *locale, 215 const UChar *text, 216 int32_t textLength, 217 UErrorCode *status); 218 219/** 220 * Open a new UBreakIterator for locating text boundaries using specified breaking rules. 221 * The rule syntax is ... (TBD) 222 * @param rules A set of rules specifying the text breaking conventions. 223 * @param rulesLength The number of characters in rules, or -1 if null-terminated. 224 * @param text The text to be iterated over. May be null, in which case ubrk_setText() is 225 * used to specify the text to be iterated. 226 * @param textLength The number of characters in text, or -1 if null-terminated. 227 * @param parseErr Receives position and context information for any syntax errors 228 * detected while parsing the rules. 229 * @param status A UErrorCode to receive any errors. 230 * @return A UBreakIterator for the specified rules. 231 * @see ubrk_open 232 * @stable ICU 2.2 233 */ 234U_STABLE UBreakIterator* U_EXPORT2 235ubrk_openRules(const UChar *rules, 236 int32_t rulesLength, 237 const UChar *text, 238 int32_t textLength, 239 UParseError *parseErr, 240 UErrorCode *status); 241 242/** 243 * Thread safe cloning operation 244 * @param bi iterator to be cloned 245 * @param stackBuffer user allocated space for the new clone. If NULL new memory will be allocated. 246 * If buffer is not large enough, new memory will be allocated. 247 * Clients can use the U_BRK_SAFECLONE_BUFFERSIZE. This will probably be enough to avoid memory allocations. 248 * @param pBufferSize pointer to size of allocated space. 249 * If *pBufferSize == 0, a sufficient size for use in cloning will 250 * be returned ('pre-flighting') 251 * If *pBufferSize is not enough for a stack-based safe clone, 252 * new memory will be allocated. 253 * @param status to indicate whether the operation went on smoothly or there were errors 254 * An informational status value, U_SAFECLONE_ALLOCATED_ERROR, is used if any allocations were necessary. 255 * @return pointer to the new clone 256 * @stable ICU 2.0 257 */ 258U_STABLE UBreakIterator * U_EXPORT2 259ubrk_safeClone( 260 const UBreakIterator *bi, 261 void *stackBuffer, 262 int32_t *pBufferSize, 263 UErrorCode *status); 264 265/** 266 * A recommended size (in bytes) for the memory buffer to be passed to ubrk_saveClone(). 267 * @stable ICU 2.0 268 */ 269#define U_BRK_SAFECLONE_BUFFERSIZE 512 270 271/** 272* Close a UBreakIterator. 273* Once closed, a UBreakIterator may no longer be used. 274* @param bi The break iterator to close. 275 * @stable ICU 2.0 276*/ 277U_STABLE void U_EXPORT2 278ubrk_close(UBreakIterator *bi); 279 280/** 281 * Sets an existing iterator to point to a new piece of text 282 * @param bi The iterator to use 283 * @param text The text to be set 284 * @param textLength The length of the text 285 * @param status The error code 286 * @stable ICU 2.0 287 */ 288U_STABLE void U_EXPORT2 289ubrk_setText(UBreakIterator* bi, 290 const UChar* text, 291 int32_t textLength, 292 UErrorCode* status); 293 294 295/** 296 * Sets an existing iterator to point to a new piece of text 297 * @param bi The iterator to use 298 * @param text The text to be set. 299 * This function makes a shallow clone of the supplied UText. This means 300 * that the caller is free to immediately close or otherwise reuse the 301 * UText that was passed as a parameter, but that the underlying text itself 302 * must not be altered while being referenced by the break iterator. 303 * @param status The error code 304 * @stable ICU 3.4 305 */ 306U_STABLE void U_EXPORT2 307ubrk_setUText(UBreakIterator* bi, 308 UText* text, 309 UErrorCode* status); 310 311 312 313/** 314 * Determine the most recently-returned text boundary. 315 * 316 * @param bi The break iterator to use. 317 * @return The character index most recently returned by \ref ubrk_next, \ref ubrk_previous, 318 * \ref ubrk_first, or \ref ubrk_last. 319 * @stable ICU 2.0 320 */ 321U_STABLE int32_t U_EXPORT2 322ubrk_current(const UBreakIterator *bi); 323 324/** 325 * Determine the text boundary following the current text boundary. 326 * 327 * @param bi The break iterator to use. 328 * @return The character index of the next text boundary, or UBRK_DONE 329 * if all text boundaries have been returned. 330 * @see ubrk_previous 331 * @stable ICU 2.0 332 */ 333U_STABLE int32_t U_EXPORT2 334ubrk_next(UBreakIterator *bi); 335 336/** 337 * Determine the text boundary preceding the current text boundary. 338 * 339 * @param bi The break iterator to use. 340 * @return The character index of the preceding text boundary, or UBRK_DONE 341 * if all text boundaries have been returned. 342 * @see ubrk_next 343 * @stable ICU 2.0 344 */ 345U_STABLE int32_t U_EXPORT2 346ubrk_previous(UBreakIterator *bi); 347 348/** 349 * Determine the index of the first character in the text being scanned. 350 * This is not always the same as index 0 of the text. 351 * @param bi The break iterator to use. 352 * @return The character index of the first character in the text being scanned. 353 * @see ubrk_last 354 * @stable ICU 2.0 355 */ 356U_STABLE int32_t U_EXPORT2 357ubrk_first(UBreakIterator *bi); 358 359/** 360 * Determine the index immediately <EM>beyond</EM> the last character in the text being 361 * scanned. 362 * This is not the same as the last character. 363 * @param bi The break iterator to use. 364 * @return The character offset immediately <EM>beyond</EM> the last character in the 365 * text being scanned. 366 * @see ubrk_first 367 * @stable ICU 2.0 368 */ 369U_STABLE int32_t U_EXPORT2 370ubrk_last(UBreakIterator *bi); 371 372/** 373 * Determine the text boundary preceding the specified offset. 374 * The value returned is always smaller than offset, or UBRK_DONE. 375 * @param bi The break iterator to use. 376 * @param offset The offset to begin scanning. 377 * @return The text boundary preceding offset, or UBRK_DONE. 378 * @see ubrk_following 379 * @stable ICU 2.0 380 */ 381U_STABLE int32_t U_EXPORT2 382ubrk_preceding(UBreakIterator *bi, 383 int32_t offset); 384 385/** 386 * Determine the text boundary following the specified offset. 387 * The value returned is always greater than offset, or UBRK_DONE. 388 * @param bi The break iterator to use. 389 * @param offset The offset to begin scanning. 390 * @return The text boundary following offset, or UBRK_DONE. 391 * @see ubrk_preceding 392 * @stable ICU 2.0 393 */ 394U_STABLE int32_t U_EXPORT2 395ubrk_following(UBreakIterator *bi, 396 int32_t offset); 397 398/** 399* Get a locale for which text breaking information is available. 400* A UBreakIterator in a locale returned by this function will perform the correct 401* text breaking for the locale. 402* @param index The index of the desired locale. 403* @return A locale for which number text breaking information is available, or 0 if none. 404* @see ubrk_countAvailable 405* @stable ICU 2.0 406*/ 407U_STABLE const char* U_EXPORT2 408ubrk_getAvailable(int32_t index); 409 410/** 411* Determine how many locales have text breaking information available. 412* This function is most useful as determining the loop ending condition for 413* calls to \ref ubrk_getAvailable. 414* @return The number of locales for which text breaking information is available. 415* @see ubrk_getAvailable 416* @stable ICU 2.0 417*/ 418U_STABLE int32_t U_EXPORT2 419ubrk_countAvailable(void); 420 421 422/** 423* Returns true if the specfied position is a boundary position. As a side 424* effect, leaves the iterator pointing to the first boundary position at 425* or after "offset". 426* @param bi The break iterator to use. 427* @param offset the offset to check. 428* @return True if "offset" is a boundary position. 429* @stable ICU 2.0 430*/ 431U_STABLE UBool U_EXPORT2 432ubrk_isBoundary(UBreakIterator *bi, int32_t offset); 433 434/** 435 * Return the status from the break rule that determined the most recently 436 * returned break position. The values appear in the rule source 437 * within brackets, {123}, for example. For rules that do not specify a 438 * status, a default value of 0 is returned. 439 * <p> 440 * For word break iterators, the possible values are defined in enum UWordBreak. 441 * @stable ICU 2.2 442 */ 443U_STABLE int32_t U_EXPORT2 444ubrk_getRuleStatus(UBreakIterator *bi); 445 446/** 447 * Get the statuses from the break rules that determined the most recently 448 * returned break position. The values appear in the rule source 449 * within brackets, {123}, for example. The default status value for rules 450 * that do not explicitly provide one is zero. 451 * <p> 452 * For word break iterators, the possible values are defined in enum UWordBreak. 453 * @param bi The break iterator to use 454 * @param fillInVec an array to be filled in with the status values. 455 * @param capacity the length of the supplied vector. A length of zero causes 456 * the function to return the number of status values, in the 457 * normal way, without attemtping to store any values. 458 * @param status receives error codes. 459 * @return The number of rule status values from rules that determined 460 * the most recent boundary returned by the break iterator. 461 * @stable ICU 3.0 462 */ 463U_STABLE int32_t U_EXPORT2 464ubrk_getRuleStatusVec(UBreakIterator *bi, int32_t *fillInVec, int32_t capacity, UErrorCode *status); 465 466/** 467 * Return the locale of the break iterator. You can choose between the valid and 468 * the actual locale. 469 * @param bi break iterator 470 * @param type locale type (valid or actual) 471 * @param status error code 472 * @return locale string 473 * @stable ICU 2.8 474 */ 475U_STABLE const char* U_EXPORT2 476ubrk_getLocaleByType(const UBreakIterator *bi, ULocDataLocaleType type, UErrorCode* status); 477 478 479#endif /* #if !UCONFIG_NO_BREAK_ITERATION */ 480 481#endif 482