1 /* 2 ******************************************************************************** 3 * Copyright (C) 1997-2012, International Business Machines 4 * Corporation and others. All Rights Reserved. 5 ******************************************************************************** 6 * 7 * File brkiter.h 8 * 9 * Modification History: 10 * 11 * Date Name Description 12 * 02/18/97 aliu Added typedef for TextCount. Made DONE const. 13 * 05/07/97 aliu Fixed DLL declaration. 14 * 07/09/97 jfitz Renamed BreakIterator and interface synced with JDK 15 * 08/11/98 helena Sync-up JDK1.2. 16 * 01/13/2000 helena Added UErrorCode parameter to createXXXInstance methods. 17 ******************************************************************************** 18 */ 19 20 #ifndef BRKITER_H 21 #define BRKITER_H 22 23 #include "unicode/utypes.h" 24 25 /** 26 * \file 27 * \brief C++ API: Break Iterator. 28 */ 29 30 #if UCONFIG_NO_BREAK_ITERATION 31 32 U_NAMESPACE_BEGIN 33 34 /* 35 * Allow the declaration of APIs with pointers to BreakIterator 36 * even when break iteration is removed from the build. 37 */ 38 class BreakIterator; 39 40 U_NAMESPACE_END 41 42 #else 43 44 #include "unicode/uobject.h" 45 #include "unicode/unistr.h" 46 #include "unicode/chariter.h" 47 #include "unicode/locid.h" 48 #include "unicode/ubrk.h" 49 #include "unicode/strenum.h" 50 #include "unicode/utext.h" 51 #include "unicode/umisc.h" 52 53 U_NAMESPACE_BEGIN 54 55 /** 56 * The BreakIterator class implements methods for finding the location 57 * of boundaries in text. BreakIterator is an abstract base class. 58 * Instances of BreakIterator maintain a current position and scan over 59 * text returning the index of characters where boundaries occur. 60 * <p> 61 * Line boundary analysis determines where a text string can be broken 62 * when line-wrapping. The mechanism correctly handles punctuation and 63 * hyphenated words. 64 * <p> 65 * Sentence boundary analysis allows selection with correct 66 * interpretation of periods within numbers and abbreviations, and 67 * trailing punctuation marks such as quotation marks and parentheses. 68 * <p> 69 * Word boundary analysis is used by search and replace functions, as 70 * well as within text editing applications that allow the user to 71 * select words with a double click. Word selection provides correct 72 * interpretation of punctuation marks within and following 73 * words. Characters that are not part of a word, such as symbols or 74 * punctuation marks, have word-breaks on both sides. 75 * <p> 76 * Character boundary analysis allows users to interact with 77 * characters as they expect to, for example, when moving the cursor 78 * through a text string. Character boundary analysis provides correct 79 * navigation of through character strings, regardless of how the 80 * character is stored. For example, an accented character might be 81 * stored as a base character and a diacritical mark. What users 82 * consider to be a character can differ between languages. 83 * <p> 84 * The text boundary positions are found according to the rules 85 * described in Unicode Standard Annex #29, Text Boundaries, and 86 * Unicode Standard Annex #14, Line Breaking Properties. These 87 * are available at http://www.unicode.org/reports/tr14/ and 88 * http://www.unicode.org/reports/tr29/. 89 * <p> 90 * In addition to the C++ API defined in this header file, a 91 * plain C API with equivalent functionality is defined in the 92 * file ubrk.h 93 * <p> 94 * Code snippets illustrating the use of the Break Iterator APIs 95 * are available in the ICU User Guide, 96 * http://icu-project.org/userguide/boundaryAnalysis.html 97 * and in the sample program icu/source/samples/break/break.cpp 98 * 99 */ 100 class U_COMMON_API BreakIterator : public UObject { 101 public: 102 /** 103 * destructor 104 * @stable ICU 2.0 105 */ 106 virtual ~BreakIterator(); 107 108 /** 109 * Return true if another object is semantically equal to this 110 * one. The other object should be an instance of the same subclass of 111 * BreakIterator. Objects of different subclasses are considered 112 * unequal. 113 * <P> 114 * Return true if this BreakIterator is at the same position in the 115 * same text, and is the same class and type (word, line, etc.) of 116 * BreakIterator, as the argument. Text is considered the same if 117 * it contains the same characters, it need not be the same 118 * object, and styles are not considered. 119 * @stable ICU 2.0 120 */ 121 virtual UBool operator==(const BreakIterator&) const = 0; 122 123 /** 124 * Returns the complement of the result of operator== 125 * @param rhs The BreakIterator to be compared for inequality 126 * @return the complement of the result of operator== 127 * @stable ICU 2.0 128 */ 129 UBool operator!=(const BreakIterator& rhs) const { return !operator==(rhs); } 130 131 /** 132 * Return a polymorphic copy of this object. This is an abstract 133 * method which subclasses implement. 134 * @stable ICU 2.0 135 */ 136 virtual BreakIterator* clone(void) const = 0; 137 138 /** 139 * Return a polymorphic class ID for this object. Different subclasses 140 * will return distinct unequal values. 141 * @stable ICU 2.0 142 */ 143 virtual UClassID getDynamicClassID(void) const = 0; 144 145 /** 146 * Return a CharacterIterator over the text being analyzed. 147 * @stable ICU 2.0 148 */ 149 virtual CharacterIterator& getText(void) const = 0; 150 151 152 /** 153 * Get a UText for the text being analyzed. 154 * The returned UText is a shallow clone of the UText used internally 155 * by the break iterator implementation. It can safely be used to 156 * access the text without impacting any break iterator operations, 157 * but the underlying text itself must not be altered. 158 * 159 * @param fillIn A UText to be filled in. If NULL, a new UText will be 160 * allocated to hold the result. 161 * @param status receives any error codes. 162 * @return The current UText for this break iterator. If an input 163 * UText was provided, it will always be returned. 164 * @stable ICU 3.4 165 */ 166 virtual UText *getUText(UText *fillIn, UErrorCode &status) const = 0; 167 168 /** 169 * Change the text over which this operates. The text boundary is 170 * reset to the start. 171 * @param text The UnicodeString used to change the text. 172 * @stable ICU 2.0 173 */ 174 virtual void setText(const UnicodeString &text) = 0; 175 176 /** 177 * Reset the break iterator to operate over the text represented by 178 * the UText. The iterator position is reset to the start. 179 * 180 * This function makes a shallow clone of the supplied UText. This means 181 * that the caller is free to immediately close or otherwise reuse the 182 * Utext that was passed as a parameter, but that the underlying text itself 183 * must not be altered while being referenced by the break iterator. 184 * 185 * @param text The UText used to change the text. 186 * @param status receives any error codes. 187 * @stable ICU 3.4 188 */ 189 virtual void setText(UText *text, UErrorCode &status) = 0; 190 191 /** 192 * Change the text over which this operates. The text boundary is 193 * reset to the start. 194 * Note that setText(UText *) provides similar functionality to this function, 195 * and is more efficient. 196 * @param it The CharacterIterator used to change the text. 197 * @stable ICU 2.0 198 */ 199 virtual void adoptText(CharacterIterator* it) = 0; 200 201 enum { 202 /** 203 * DONE is returned by previous() and next() after all valid 204 * boundaries have been returned. 205 * @stable ICU 2.0 206 */ 207 DONE = (int32_t)-1 208 }; 209 210 /** 211 * Set the iterator position to the index of the first character in the text being scanned. 212 * @return The index of the first character in the text being scanned. 213 * @stable ICU 2.0 214 */ 215 virtual int32_t first(void) = 0; 216 217 /** 218 * Set the iterator position to the index immediately BEYOND the last character in the text being scanned. 219 * @return The index immediately BEYOND the last character in the text being scanned. 220 * @stable ICU 2.0 221 */ 222 virtual int32_t last(void) = 0; 223 224 /** 225 * Set the iterator position to the boundary preceding the current boundary. 226 * @return The character index of the previous text boundary or DONE if all 227 * boundaries have been returned. 228 * @stable ICU 2.0 229 */ 230 virtual int32_t previous(void) = 0; 231 232 /** 233 * Advance the iterator to the boundary following the current boundary. 234 * @return The character index of the next text boundary or DONE if all 235 * boundaries have been returned. 236 * @stable ICU 2.0 237 */ 238 virtual int32_t next(void) = 0; 239 240 /** 241 * Return character index of the current interator position within the text. 242 * @return The boundary most recently returned. 243 * @stable ICU 2.0 244 */ 245 virtual int32_t current(void) const = 0; 246 247 /** 248 * Advance the iterator to the first boundary following the specified offset. 249 * The value returned is always greater than the offset or 250 * the value BreakIterator.DONE 251 * @param offset the offset to begin scanning. 252 * @return The first boundary after the specified offset. 253 * @stable ICU 2.0 254 */ 255 virtual int32_t following(int32_t offset) = 0; 256 257 /** 258 * Set the iterator position to the first boundary preceding the specified offset. 259 * The value returned is always smaller than the offset or 260 * the value BreakIterator.DONE 261 * @param offset the offset to begin scanning. 262 * @return The first boundary before the specified offset. 263 * @stable ICU 2.0 264 */ 265 virtual int32_t preceding(int32_t offset) = 0; 266 267 /** 268 * Return true if the specfied position is a boundary position. 269 * As a side effect, the current position of the iterator is set 270 * to the first boundary position at or following the specified offset. 271 * @param offset the offset to check. 272 * @return True if "offset" is a boundary position. 273 * @stable ICU 2.0 274 */ 275 virtual UBool isBoundary(int32_t offset) = 0; 276 277 /** 278 * Set the iterator position to the nth boundary from the current boundary 279 * @param n the number of boundaries to move by. A value of 0 280 * does nothing. Negative values move to previous boundaries 281 * and positive values move to later boundaries. 282 * @return The new iterator position, or 283 * DONE if there are fewer than |n| boundaries in the specfied direction. 284 * @stable ICU 2.0 285 */ 286 virtual int32_t next(int32_t n) = 0; 287 288 /** 289 * Create BreakIterator for word-breaks using the given locale. 290 * Returns an instance of a BreakIterator implementing word breaks. 291 * WordBreak is useful for word selection (ex. double click) 292 * @param where the locale. 293 * @param status the error code 294 * @return A BreakIterator for word-breaks. The UErrorCode& status 295 * parameter is used to return status information to the user. 296 * To check whether the construction succeeded or not, you should check 297 * the value of U_SUCCESS(err). If you wish more detailed information, you 298 * can check for informational error results which still indicate success. 299 * U_USING_FALLBACK_WARNING indicates that a fall back locale was used. For 300 * example, 'de_CH' was requested, but nothing was found there, so 'de' was 301 * used. U_USING_DEFAULT_WARNING indicates that the default locale data was 302 * used; neither the requested locale nor any of its fall back locales 303 * could be found. 304 * The caller owns the returned object and is responsible for deleting it. 305 * @stable ICU 2.0 306 */ 307 static BreakIterator* U_EXPORT2 308 createWordInstance(const Locale& where, UErrorCode& status); 309 310 /** 311 * Create BreakIterator for line-breaks using specified locale. 312 * Returns an instance of a BreakIterator implementing line breaks. Line 313 * breaks are logically possible line breaks, actual line breaks are 314 * usually determined based on display width. 315 * LineBreak is useful for word wrapping text. 316 * @param where the locale. 317 * @param status The error code. 318 * @return A BreakIterator for line-breaks. The UErrorCode& status 319 * parameter is used to return status information to the user. 320 * To check whether the construction succeeded or not, you should check 321 * the value of U_SUCCESS(err). If you wish more detailed information, you 322 * can check for informational error results which still indicate success. 323 * U_USING_FALLBACK_WARNING indicates that a fall back locale was used. For 324 * example, 'de_CH' was requested, but nothing was found there, so 'de' was 325 * used. U_USING_DEFAULT_WARNING indicates that the default locale data was 326 * used; neither the requested locale nor any of its fall back locales 327 * could be found. 328 * The caller owns the returned object and is responsible for deleting it. 329 * @stable ICU 2.0 330 */ 331 static BreakIterator* U_EXPORT2 332 createLineInstance(const Locale& where, UErrorCode& status); 333 334 /** 335 * Create BreakIterator for character-breaks using specified locale 336 * Returns an instance of a BreakIterator implementing character breaks. 337 * Character breaks are boundaries of combining character sequences. 338 * @param where the locale. 339 * @param status The error code. 340 * @return A BreakIterator for character-breaks. The UErrorCode& status 341 * parameter is used to return status information to the user. 342 * To check whether the construction succeeded or not, you should check 343 * the value of U_SUCCESS(err). If you wish more detailed information, you 344 * can check for informational error results which still indicate success. 345 * U_USING_FALLBACK_WARNING indicates that a fall back locale was used. For 346 * example, 'de_CH' was requested, but nothing was found there, so 'de' was 347 * used. U_USING_DEFAULT_WARNING indicates that the default locale data was 348 * used; neither the requested locale nor any of its fall back locales 349 * could be found. 350 * The caller owns the returned object and is responsible for deleting it. 351 * @stable ICU 2.0 352 */ 353 static BreakIterator* U_EXPORT2 354 createCharacterInstance(const Locale& where, UErrorCode& status); 355 356 /** 357 * Create BreakIterator for sentence-breaks using specified locale 358 * Returns an instance of a BreakIterator implementing sentence breaks. 359 * @param where the locale. 360 * @param status The error code. 361 * @return A BreakIterator for sentence-breaks. The UErrorCode& status 362 * parameter is used to return status information to the user. 363 * To check whether the construction succeeded or not, you should check 364 * the value of U_SUCCESS(err). If you wish more detailed information, you 365 * can check for informational error results which still indicate success. 366 * U_USING_FALLBACK_WARNING indicates that a fall back locale was used. For 367 * example, 'de_CH' was requested, but nothing was found there, so 'de' was 368 * used. U_USING_DEFAULT_WARNING indicates that the default locale data was 369 * used; neither the requested locale nor any of its fall back locales 370 * could be found. 371 * The caller owns the returned object and is responsible for deleting it. 372 * @stable ICU 2.0 373 */ 374 static BreakIterator* U_EXPORT2 375 createSentenceInstance(const Locale& where, UErrorCode& status); 376 377 /** 378 * Create BreakIterator for title-casing breaks using the specified locale 379 * Returns an instance of a BreakIterator implementing title breaks. 380 * The iterator returned locates title boundaries as described for 381 * Unicode 3.2 only. For Unicode 4.0 and above title boundary iteration, 382 * please use Word Boundary iterator.{@link #createWordInstance } 383 * 384 * @param where the locale. 385 * @param status The error code. 386 * @return A BreakIterator for title-breaks. The UErrorCode& status 387 * parameter is used to return status information to the user. 388 * To check whether the construction succeeded or not, you should check 389 * the value of U_SUCCESS(err). If you wish more detailed information, you 390 * can check for informational error results which still indicate success. 391 * U_USING_FALLBACK_WARNING indicates that a fall back locale was used. For 392 * example, 'de_CH' was requested, but nothing was found there, so 'de' was 393 * used. U_USING_DEFAULT_WARNING indicates that the default locale data was 394 * used; neither the requested locale nor any of its fall back locales 395 * could be found. 396 * The caller owns the returned object and is responsible for deleting it. 397 * @stable ICU 2.1 398 */ 399 static BreakIterator* U_EXPORT2 400 createTitleInstance(const Locale& where, UErrorCode& status); 401 402 /** 403 * Get the set of Locales for which TextBoundaries are installed. 404 * <p><b>Note:</b> this will not return locales added through the register 405 * call. To see the registered locales too, use the getAvailableLocales 406 * function that returns a StringEnumeration object </p> 407 * @param count the output parameter of number of elements in the locale list 408 * @return available locales 409 * @stable ICU 2.0 410 */ 411 static const Locale* U_EXPORT2 getAvailableLocales(int32_t& count); 412 413 /** 414 * Get name of the object for the desired Locale, in the desired langauge. 415 * @param objectLocale must be from getAvailableLocales. 416 * @param displayLocale specifies the desired locale for output. 417 * @param name the fill-in parameter of the return value 418 * Uses best match. 419 * @return user-displayable name 420 * @stable ICU 2.0 421 */ 422 static UnicodeString& U_EXPORT2 getDisplayName(const Locale& objectLocale, 423 const Locale& displayLocale, 424 UnicodeString& name); 425 426 /** 427 * Get name of the object for the desired Locale, in the langauge of the 428 * default locale. 429 * @param objectLocale must be from getMatchingLocales 430 * @param name the fill-in parameter of the return value 431 * @return user-displayable name 432 * @stable ICU 2.0 433 */ 434 static UnicodeString& U_EXPORT2 getDisplayName(const Locale& objectLocale, 435 UnicodeString& name); 436 437 /** 438 * Thread safe client-buffer-based cloning operation 439 * Do NOT call delete on a safeclone, since 'new' is not used to create it. 440 * @param stackBuffer user allocated space for the new clone. If NULL new memory will be allocated. 441 * If buffer is not large enough, new memory will be allocated. 442 * @param BufferSize reference to size of allocated space. 443 * If BufferSize == 0, a sufficient size for use in cloning will 444 * be returned ('pre-flighting') 445 * If BufferSize is not enough for a stack-based safe clone, 446 * new memory will be allocated. 447 * @param status to indicate whether the operation went on smoothly or there were errors 448 * An informational status value, U_SAFECLONE_ALLOCATED_ERROR, is used if any allocations were 449 * necessary. 450 * @return pointer to the new clone 451 * 452 * @stable ICU 2.0 453 */ 454 virtual BreakIterator * createBufferClone(void *stackBuffer, 455 int32_t &BufferSize, 456 UErrorCode &status) = 0; 457 458 /** 459 * Determine whether the BreakIterator was created in user memory by 460 * createBufferClone(), and thus should not be deleted. Such objects 461 * must be closed by an explicit call to the destructor (not delete). 462 * @stable ICU 2.0 463 */ 464 inline UBool isBufferClone(void); 465 466 #if !UCONFIG_NO_SERVICE 467 /** 468 * Register a new break iterator of the indicated kind, to use in the given locale. 469 * The break iterator will be adopted. Clones of the iterator will be returned 470 * if a request for a break iterator of the given kind matches or falls back to 471 * this locale. 472 * @param toAdopt the BreakIterator instance to be adopted 473 * @param locale the Locale for which this instance is to be registered 474 * @param kind the type of iterator for which this instance is to be registered 475 * @param status the in/out status code, no special meanings are assigned 476 * @return a registry key that can be used to unregister this instance 477 * @stable ICU 2.4 478 */ 479 static URegistryKey U_EXPORT2 registerInstance(BreakIterator* toAdopt, 480 const Locale& locale, 481 UBreakIteratorType kind, 482 UErrorCode& status); 483 484 /** 485 * Unregister a previously-registered BreakIterator using the key returned from the 486 * register call. Key becomes invalid after a successful call and should not be used again. 487 * The BreakIterator corresponding to the key will be deleted. 488 * @param key the registry key returned by a previous call to registerInstance 489 * @param status the in/out status code, no special meanings are assigned 490 * @return TRUE if the iterator for the key was successfully unregistered 491 * @stable ICU 2.4 492 */ 493 static UBool U_EXPORT2 unregister(URegistryKey key, UErrorCode& status); 494 495 /** 496 * Return a StringEnumeration over the locales available at the time of the call, 497 * including registered locales. 498 * @return a StringEnumeration over the locales available at the time of the call 499 * @stable ICU 2.4 500 */ 501 static StringEnumeration* U_EXPORT2 getAvailableLocales(void); 502 #endif 503 504 /** 505 * Returns the locale for this break iterator. Two flavors are available: valid and 506 * actual locale. 507 * @stable ICU 2.8 508 */ 509 Locale getLocale(ULocDataLocaleType type, UErrorCode& status) const; 510 511 #ifndef U_HIDE_INTERNAL_API 512 /** Get the locale for this break iterator object. You can choose between valid and actual locale. 513 * @param type type of the locale we're looking for (valid or actual) 514 * @param status error code for the operation 515 * @return the locale 516 * @internal 517 */ 518 const char *getLocaleID(ULocDataLocaleType type, UErrorCode& status) const; 519 #endif /* U_HIDE_INTERNAL_API */ 520 521 /** 522 * Set the subject text string upon which the break iterator is operating 523 * without changing any other aspect of the matching state. 524 * The new and previous text strings must have the same content. 525 * 526 * This function is intended for use in environments where ICU is operating on 527 * strings that may move around in memory. It provides a mechanism for notifying 528 * ICU that the string has been relocated, and providing a new UText to access the 529 * string in its new position. 530 * 531 * Note that the break iterator implementation never copies the underlying text 532 * of a string being processed, but always operates directly on the original text 533 * provided by the user. Refreshing simply drops the references to the old text 534 * and replaces them with references to the new. 535 * 536 * Caution: this function is normally used only by very specialized, 537 * system-level code. One example use case is with garbage collection that moves 538 * the text in memory. 539 * 540 * @param input The new (moved) text string. 541 * @param status Receives errors detected by this function. 542 * @return *this 543 * 544 * @draft ICU 49 545 */ 546 virtual BreakIterator &refreshInputText(UText *input, UErrorCode &status) = 0; 547 548 private: 549 static BreakIterator* buildInstance(const Locale& loc, const char *type, int32_t kind, UErrorCode& status); 550 static BreakIterator* createInstance(const Locale& loc, int32_t kind, UErrorCode& status); 551 static BreakIterator* makeInstance(const Locale& loc, int32_t kind, UErrorCode& status); 552 553 friend class ICUBreakIteratorFactory; 554 friend class ICUBreakIteratorService; 555 556 protected: 557 // Do not enclose protected default/copy constructors with #ifndef U_HIDE_INTERNAL_API 558 // or else the compiler will create a public ones. 559 /** @internal */ 560 BreakIterator(); 561 /** @internal */ 562 UBool fBufferClone; 563 /** @internal */ 564 BreakIterator (const BreakIterator &other) : UObject(other), fBufferClone(FALSE) {} 565 566 private: 567 568 /** @internal */ 569 char actualLocale[ULOC_FULLNAME_CAPACITY]; 570 char validLocale[ULOC_FULLNAME_CAPACITY]; 571 572 /** 573 * The assignment operator has no real implementation. 574 * It's provided to make the compiler happy. Do not call. 575 */ 576 BreakIterator& operator=(const BreakIterator&); 577 }; 578 579 inline UBool BreakIterator::isBufferClone() 580 { 581 return fBufferClone; 582 } 583 584 U_NAMESPACE_END 585 586 #endif /* #if !UCONFIG_NO_BREAK_ITERATION */ 587 588 #endif // _BRKITER 589 //eof 590 591