1/* 2********************************************************************** 3* Copyright (C) 2001-2011,2014 IBM and others. All rights reserved. 4********************************************************************** 5* Date Name Description 6* 06/28/2001 synwee Creation. 7********************************************************************** 8*/ 9#ifndef USEARCH_H 10#define USEARCH_H 11 12#include "unicode/utypes.h" 13 14#if !UCONFIG_NO_COLLATION && !UCONFIG_NO_BREAK_ITERATION 15 16#include "unicode/localpointer.h" 17#include "unicode/ucol.h" 18#include "unicode/ucoleitr.h" 19#include "unicode/ubrk.h" 20 21/** 22 * \file 23 * \brief C API: StringSearch 24 * 25 * C Apis for an engine that provides language-sensitive text searching based 26 * on the comparison rules defined in a <tt>UCollator</tt> data struct, 27 * see <tt>ucol.h</tt>. This ensures that language eccentricity can be 28 * handled, e.g. for the German collator, characters ß and SS will be matched 29 * if case is chosen to be ignored. 30 * See the <a href="http://source.icu-project.org/repos/icu/icuhtml/trunk/design/collation/ICU_collation_design.htm"> 31 * "ICU Collation Design Document"</a> for more information. 32 * <p> 33 * The implementation may use a linear search or a modified form of the Boyer-Moore 34 * search; for more information on the latter see 35 * <a href="http://icu-project.org/docs/papers/efficient_text_searching_in_java.html"> 36 * "Efficient Text Searching in Java"</a>, published in <i>Java Report</i> 37 * in February, 1999. 38 * <p> 39 * There are 2 match options for selection:<br> 40 * Let S' be the sub-string of a text string S between the offsets start and 41 * end <start, end>. 42 * <br> 43 * A pattern string P matches a text string S at the offsets <start, end> 44 * if 45 * <pre> 46 * option 1. Some canonical equivalent of P matches some canonical equivalent 47 * of S' 48 * option 2. P matches S' and if P starts or ends with a combining mark, 49 * there exists no non-ignorable combining mark before or after S' 50 * in S respectively. 51 * </pre> 52 * Option 2. will be the default. 53 * <p> 54 * This search has APIs similar to that of other text iteration mechanisms 55 * such as the break iterators in <tt>ubrk.h</tt>. Using these 56 * APIs, it is easy to scan through text looking for all occurances of 57 * a given pattern. This search iterator allows changing of direction by 58 * calling a <tt>reset</tt> followed by a <tt>next</tt> or <tt>previous</tt>. 59 * Though a direction change can occur without calling <tt>reset</tt> first, 60 * this operation comes with some speed penalty. 61 * Generally, match results in the forward direction will match the result 62 * matches in the backwards direction in the reverse order 63 * <p> 64 * <tt>usearch.h</tt> provides APIs to specify the starting position 65 * within the text string to be searched, e.g. <tt>usearch_setOffset</tt>, 66 * <tt>usearch_preceding</tt> and <tt>usearch_following</tt>. Since the 67 * starting position will be set as it is specified, please take note that 68 * there are some dangerous positions which the search may render incorrect 69 * results: 70 * <ul> 71 * <li> The midst of a substring that requires normalization. 72 * <li> If the following match is to be found, the position should not be the 73 * second character which requires to be swapped with the preceding 74 * character. Vice versa, if the preceding match is to be found, 75 * position to search from should not be the first character which 76 * requires to be swapped with the next character. E.g certain Thai and 77 * Lao characters require swapping. 78 * <li> If a following pattern match is to be found, any position within a 79 * contracting sequence except the first will fail. Vice versa if a 80 * preceding pattern match is to be found, a invalid starting point 81 * would be any character within a contracting sequence except the last. 82 * </ul> 83 * <p> 84 * A breakiterator can be used if only matches at logical breaks are desired. 85 * Using a breakiterator will only give you results that exactly matches the 86 * boundaries given by the breakiterator. For instance the pattern "e" will 87 * not be found in the string "\u00e9" if a character break iterator is used. 88 * <p> 89 * Options are provided to handle overlapping matches. 90 * E.g. In English, overlapping matches produces the result 0 and 2 91 * for the pattern "abab" in the text "ababab", where else mutually 92 * exclusive matches only produce the result of 0. 93 * <p> 94 * Options are also provided to implement "asymmetric search" as described in 95 * <a href="http://www.unicode.org/reports/tr10/#Asymmetric_Search"> 96 * UTS #10 Unicode Collation Algorithm</a>, specifically the USearchAttribute 97 * USEARCH_ELEMENT_COMPARISON and its values. 98 * <p> 99 * E.g. In English, overlapping matches produces the result 0 and 2 100 * for the pattern "abab" in the text "ababab", where else mutually 101 * exclusive matches only produce the result of 0. 102 * <p> 103 * Though collator attributes will be taken into consideration while 104 * performing matches, there are no APIs here for setting and getting the 105 * attributes. These attributes can be set by getting the collator 106 * from <tt>usearch_getCollator</tt> and using the APIs in <tt>ucol.h</tt>. 107 * Lastly to update String Search to the new collator attributes, 108 * usearch_reset() has to be called. 109 * <p> 110 * Restriction: <br> 111 * Currently there are no composite characters that consists of a 112 * character with combining class > 0 before a character with combining 113 * class == 0. However, if such a character exists in the future, the 114 * search mechanism does not guarantee the results for option 1. 115 * 116 * <p> 117 * Example of use:<br> 118 * <pre><code> 119 * char *tgtstr = "The quick brown fox jumped over the lazy fox"; 120 * char *patstr = "fox"; 121 * UChar target[64]; 122 * UChar pattern[16]; 123 * UErrorCode status = U_ZERO_ERROR; 124 * u_uastrcpy(target, tgtstr); 125 * u_uastrcpy(pattern, patstr); 126 * 127 * UStringSearch *search = usearch_open(pattern, -1, target, -1, "en_US", 128 * NULL, &status); 129 * if (U_SUCCESS(status)) { 130 * for (int pos = usearch_first(search, &status); 131 * pos != USEARCH_DONE; 132 * pos = usearch_next(search, &status)) 133 * { 134 * printf("Found match at %d pos, length is %d\n", pos, 135 * usearch_getMatchLength(search)); 136 * } 137 * } 138 * 139 * usearch_close(search); 140 * </code></pre> 141 * @stable ICU 2.4 142 */ 143 144/** 145* DONE is returned by previous() and next() after all valid matches have 146* been returned, and by first() and last() if there are no matches at all. 147* @stable ICU 2.4 148*/ 149#define USEARCH_DONE -1 150 151/** 152* Data structure for searching 153* @stable ICU 2.4 154*/ 155struct UStringSearch; 156/** 157* Data structure for searching 158* @stable ICU 2.4 159*/ 160typedef struct UStringSearch UStringSearch; 161 162/** 163* @stable ICU 2.4 164*/ 165typedef enum { 166 /** 167 * Option for overlapping matches 168 * @stable ICU 2.4 169 */ 170 USEARCH_OVERLAP = 0, 171#ifndef U_HIDE_DEPRECATED_API 172 /** 173 * Option for canonical matches; option 1 in header documentation. 174 * The default value will be USEARCH_OFF. 175 * Note: Setting this option to USEARCH_ON currently has no effect on 176 * search behavior, and this option is deprecated. Instead, to control 177 * canonical match behavior, you must set UCOL_NORMALIZATION_MODE 178 * appropriately (to UCOL_OFF or UCOL_ON) in the UCollator used by 179 * the UStringSearch object. 180 * @see usearch_openFromCollator 181 * @see usearch_getCollator 182 * @see usearch_setCollator 183 * @see ucol_getAttribute 184 * @deprecated ICU 53 185 */ 186 USEARCH_CANONICAL_MATCH = 1, 187#endif /* U_HIDE_DEPRECATED_API */ 188 /** 189 * Option to control how collation elements are compared. 190 * The default value will be USEARCH_STANDARD_ELEMENT_COMPARISON. 191 * @stable ICU 4.4 192 */ 193 USEARCH_ELEMENT_COMPARISON = 2, 194 195 /** 196 * Count of attribute types 197 * @stable ICU 2.4 198 */ 199 USEARCH_ATTRIBUTE_COUNT = 3 200} USearchAttribute; 201 202/** 203* @stable ICU 2.4 204*/ 205typedef enum { 206 /** 207 * Default value for any USearchAttribute 208 * @stable ICU 2.4 209 */ 210 USEARCH_DEFAULT = -1, 211 /** 212 * Value for USEARCH_OVERLAP and USEARCH_CANONICAL_MATCH 213 * @stable ICU 2.4 214 */ 215 USEARCH_OFF, 216 /** 217 * Value for USEARCH_OVERLAP and USEARCH_CANONICAL_MATCH 218 * @stable ICU 2.4 219 */ 220 USEARCH_ON, 221 /** 222 * Value (default) for USEARCH_ELEMENT_COMPARISON; 223 * standard collation element comparison at the specified collator 224 * strength. 225 * @stable ICU 4.4 226 */ 227 USEARCH_STANDARD_ELEMENT_COMPARISON, 228 /** 229 * Value for USEARCH_ELEMENT_COMPARISON; 230 * collation element comparison is modified to effectively provide 231 * behavior between the specified strength and strength - 1. Collation 232 * elements in the pattern that have the base weight for the specified 233 * strength are treated as "wildcards" that match an element with any 234 * other weight at that collation level in the searched text. For 235 * example, with a secondary-strength English collator, a plain 'e' in 236 * the pattern will match a plain e or an e with any diacritic in the 237 * searched text, but an e with diacritic in the pattern will only 238 * match an e with the same diacritic in the searched text. 239 * 240 * This supports "asymmetric search" as described in 241 * <a href="http://www.unicode.org/reports/tr10/#Asymmetric_Search"> 242 * UTS #10 Unicode Collation Algorithm</a>. 243 * 244 * @stable ICU 4.4 245 */ 246 USEARCH_PATTERN_BASE_WEIGHT_IS_WILDCARD, 247 /** 248 * Value for USEARCH_ELEMENT_COMPARISON. 249 * collation element comparison is modified to effectively provide 250 * behavior between the specified strength and strength - 1. Collation 251 * elements in either the pattern or the searched text that have the 252 * base weight for the specified strength are treated as "wildcards" 253 * that match an element with any other weight at that collation level. 254 * For example, with a secondary-strength English collator, a plain 'e' 255 * in the pattern will match a plain e or an e with any diacritic in the 256 * searched text, but an e with diacritic in the pattern will only 257 * match an e with the same diacritic or a plain e in the searched text. 258 * 259 * This option is similar to "asymmetric search" as described in 260 * <a href="http://www.unicode.org/reports/tr10/#Asymmetric_Search"> 261 * UTS #10 Unicode Collation Algorithm</a, but also allows unmarked 262 * characters in the searched text to match marked or unmarked versions of 263 * that character in the pattern. 264 * 265 * @stable ICU 4.4 266 */ 267 USEARCH_ANY_BASE_WEIGHT_IS_WILDCARD, 268 269 /** 270 * Count of attribute values 271 * @stable ICU 2.4 272 */ 273 USEARCH_ATTRIBUTE_VALUE_COUNT 274} USearchAttributeValue; 275 276/* open and close ------------------------------------------------------ */ 277 278/** 279* Creating a search iterator data struct using the argument locale language 280* rule set. A collator will be created in the process, which will be owned by 281* this search and will be deleted in <tt>usearch_close</tt>. 282* @param pattern for matching 283* @param patternlength length of the pattern, -1 for null-termination 284* @param text text string 285* @param textlength length of the text string, -1 for null-termination 286* @param locale name of locale for the rules to be used 287* @param breakiter A BreakIterator that will be used to restrict the points 288* at which matches are detected. If a match is found, but 289* the match's start or end index is not a boundary as 290* determined by the <tt>BreakIterator</tt>, the match will 291* be rejected and another will be searched for. 292* If this parameter is <tt>NULL</tt>, no break detection is 293* attempted. 294* @param status for errors if it occurs. If pattern or text is NULL, or if 295* patternlength or textlength is 0 then an 296* U_ILLEGAL_ARGUMENT_ERROR is returned. 297* @return search iterator data structure, or NULL if there is an error. 298* @stable ICU 2.4 299*/ 300U_STABLE UStringSearch * U_EXPORT2 usearch_open(const UChar *pattern, 301 int32_t patternlength, 302 const UChar *text, 303 int32_t textlength, 304 const char *locale, 305 UBreakIterator *breakiter, 306 UErrorCode *status); 307 308/** 309* Creating a search iterator data struct using the argument collator language 310* rule set. Note, user retains the ownership of this collator, thus the 311* responsibility of deletion lies with the user. 312* NOTE: string search cannot be instantiated from a collator that has 313* collate digits as numbers (CODAN) turned on. 314* @param pattern for matching 315* @param patternlength length of the pattern, -1 for null-termination 316* @param text text string 317* @param textlength length of the text string, -1 for null-termination 318* @param collator used for the language rules 319* @param breakiter A BreakIterator that will be used to restrict the points 320* at which matches are detected. If a match is found, but 321* the match's start or end index is not a boundary as 322* determined by the <tt>BreakIterator</tt>, the match will 323* be rejected and another will be searched for. 324* If this parameter is <tt>NULL</tt>, no break detection is 325* attempted. 326* @param status for errors if it occurs. If collator, pattern or text is NULL, 327* or if patternlength or textlength is 0 then an 328* U_ILLEGAL_ARGUMENT_ERROR is returned. 329* @return search iterator data structure, or NULL if there is an error. 330* @stable ICU 2.4 331*/ 332U_STABLE UStringSearch * U_EXPORT2 usearch_openFromCollator( 333 const UChar *pattern, 334 int32_t patternlength, 335 const UChar *text, 336 int32_t textlength, 337 const UCollator *collator, 338 UBreakIterator *breakiter, 339 UErrorCode *status); 340 341/** 342* Destroying and cleaning up the search iterator data struct. 343* If a collator is created in <tt>usearch_open</tt>, it will be destroyed here. 344* @param searchiter data struct to clean up 345* @stable ICU 2.4 346*/ 347U_STABLE void U_EXPORT2 usearch_close(UStringSearch *searchiter); 348 349#if U_SHOW_CPLUSPLUS_API 350 351U_NAMESPACE_BEGIN 352 353/** 354 * \class LocalUStringSearchPointer 355 * "Smart pointer" class, closes a UStringSearch via usearch_close(). 356 * For most methods see the LocalPointerBase base class. 357 * 358 * @see LocalPointerBase 359 * @see LocalPointer 360 * @stable ICU 4.4 361 */ 362U_DEFINE_LOCAL_OPEN_POINTER(LocalUStringSearchPointer, UStringSearch, usearch_close); 363 364U_NAMESPACE_END 365 366#endif 367 368/* get and set methods -------------------------------------------------- */ 369 370/** 371* Sets the current position in the text string which the next search will 372* start from. Clears previous states. 373* This method takes the argument index and sets the position in the text 374* string accordingly without checking if the index is pointing to a 375* valid starting point to begin searching. 376* Search positions that may render incorrect results are highlighted in the 377* header comments 378* @param strsrch search iterator data struct 379* @param position position to start next search from. If position is less 380* than or greater than the text range for searching, 381* an U_INDEX_OUTOFBOUNDS_ERROR will be returned 382* @param status error status if any. 383* @stable ICU 2.4 384*/ 385U_STABLE void U_EXPORT2 usearch_setOffset(UStringSearch *strsrch, 386 int32_t position, 387 UErrorCode *status); 388 389/** 390* Return the current index in the string text being searched. 391* If the iteration has gone past the end of the text (or past the beginning 392* for a backwards search), <tt>USEARCH_DONE</tt> is returned. 393* @param strsrch search iterator data struct 394* @see #USEARCH_DONE 395* @stable ICU 2.4 396*/ 397U_STABLE int32_t U_EXPORT2 usearch_getOffset(const UStringSearch *strsrch); 398 399/** 400* Sets the text searching attributes located in the enum USearchAttribute 401* with values from the enum USearchAttributeValue. 402* <tt>USEARCH_DEFAULT</tt> can be used for all attributes for resetting. 403* @param strsrch search iterator data struct 404* @param attribute text attribute to be set 405* @param value text attribute value 406* @param status for errors if it occurs 407* @see #usearch_getAttribute 408* @stable ICU 2.4 409*/ 410U_STABLE void U_EXPORT2 usearch_setAttribute(UStringSearch *strsrch, 411 USearchAttribute attribute, 412 USearchAttributeValue value, 413 UErrorCode *status); 414 415/** 416* Gets the text searching attributes. 417* @param strsrch search iterator data struct 418* @param attribute text attribute to be retrieve 419* @return text attribute value 420* @see #usearch_setAttribute 421* @stable ICU 2.4 422*/ 423U_STABLE USearchAttributeValue U_EXPORT2 usearch_getAttribute( 424 const UStringSearch *strsrch, 425 USearchAttribute attribute); 426 427/** 428* Returns the index to the match in the text string that was searched. 429* This call returns a valid result only after a successful call to 430* <tt>usearch_first</tt>, <tt>usearch_next</tt>, <tt>usearch_previous</tt>, 431* or <tt>usearch_last</tt>. 432* Just after construction, or after a searching method returns 433* <tt>USEARCH_DONE</tt>, this method will return <tt>USEARCH_DONE</tt>. 434* <p> 435* Use <tt>usearch_getMatchedLength</tt> to get the matched string length. 436* @param strsrch search iterator data struct 437* @return index to a substring within the text string that is being 438* searched. 439* @see #usearch_first 440* @see #usearch_next 441* @see #usearch_previous 442* @see #usearch_last 443* @see #USEARCH_DONE 444* @stable ICU 2.4 445*/ 446U_STABLE int32_t U_EXPORT2 usearch_getMatchedStart( 447 const UStringSearch *strsrch); 448 449/** 450* Returns the length of text in the string which matches the search pattern. 451* This call returns a valid result only after a successful call to 452* <tt>usearch_first</tt>, <tt>usearch_next</tt>, <tt>usearch_previous</tt>, 453* or <tt>usearch_last</tt>. 454* Just after construction, or after a searching method returns 455* <tt>USEARCH_DONE</tt>, this method will return 0. 456* @param strsrch search iterator data struct 457* @return The length of the match in the string text, or 0 if there is no 458* match currently. 459* @see #usearch_first 460* @see #usearch_next 461* @see #usearch_previous 462* @see #usearch_last 463* @see #USEARCH_DONE 464* @stable ICU 2.4 465*/ 466U_STABLE int32_t U_EXPORT2 usearch_getMatchedLength( 467 const UStringSearch *strsrch); 468 469/** 470* Returns the text that was matched by the most recent call to 471* <tt>usearch_first</tt>, <tt>usearch_next</tt>, <tt>usearch_previous</tt>, 472* or <tt>usearch_last</tt>. 473* If the iterator is not pointing at a valid match (e.g. just after 474* construction or after <tt>USEARCH_DONE</tt> has been returned, returns 475* an empty string. If result is not large enough to store the matched text, 476* result will be filled with the partial text and an U_BUFFER_OVERFLOW_ERROR 477* will be returned in status. result will be null-terminated whenever 478* possible. If the buffer fits the matched text exactly, a null-termination 479* is not possible, then a U_STRING_NOT_TERMINATED_ERROR set in status. 480* Pre-flighting can be either done with length = 0 or the API 481* <tt>usearch_getMatchLength</tt>. 482* @param strsrch search iterator data struct 483* @param result UChar buffer to store the matched string 484* @param resultCapacity length of the result buffer 485* @param status error returned if result is not large enough 486* @return exact length of the matched text, not counting the null-termination 487* @see #usearch_first 488* @see #usearch_next 489* @see #usearch_previous 490* @see #usearch_last 491* @see #USEARCH_DONE 492* @stable ICU 2.4 493*/ 494U_STABLE int32_t U_EXPORT2 usearch_getMatchedText(const UStringSearch *strsrch, 495 UChar *result, 496 int32_t resultCapacity, 497 UErrorCode *status); 498 499#if !UCONFIG_NO_BREAK_ITERATION 500 501/** 502* Set the BreakIterator that will be used to restrict the points at which 503* matches are detected. 504* @param strsrch search iterator data struct 505* @param breakiter A BreakIterator that will be used to restrict the points 506* at which matches are detected. If a match is found, but 507* the match's start or end index is not a boundary as 508* determined by the <tt>BreakIterator</tt>, the match will 509* be rejected and another will be searched for. 510* If this parameter is <tt>NULL</tt>, no break detection is 511* attempted. 512* @param status for errors if it occurs 513* @see #usearch_getBreakIterator 514* @stable ICU 2.4 515*/ 516U_STABLE void U_EXPORT2 usearch_setBreakIterator(UStringSearch *strsrch, 517 UBreakIterator *breakiter, 518 UErrorCode *status); 519 520/** 521* Returns the BreakIterator that is used to restrict the points at which 522* matches are detected. This will be the same object that was passed to the 523* constructor or to <tt>usearch_setBreakIterator</tt>. Note that 524* <tt>NULL</tt> 525* is a legal value; it means that break detection should not be attempted. 526* @param strsrch search iterator data struct 527* @return break iterator used 528* @see #usearch_setBreakIterator 529* @stable ICU 2.4 530*/ 531U_STABLE const UBreakIterator * U_EXPORT2 usearch_getBreakIterator( 532 const UStringSearch *strsrch); 533 534#endif 535 536/** 537* Set the string text to be searched. Text iteration will hence begin at the 538* start of the text string. This method is useful if you want to re-use an 539* iterator to search for the same pattern within a different body of text. 540* @param strsrch search iterator data struct 541* @param text new string to look for match 542* @param textlength length of the new string, -1 for null-termination 543* @param status for errors if it occurs. If text is NULL, or textlength is 0 544* then an U_ILLEGAL_ARGUMENT_ERROR is returned with no change 545* done to strsrch. 546* @see #usearch_getText 547* @stable ICU 2.4 548*/ 549U_STABLE void U_EXPORT2 usearch_setText( UStringSearch *strsrch, 550 const UChar *text, 551 int32_t textlength, 552 UErrorCode *status); 553 554/** 555* Return the string text to be searched. 556* @param strsrch search iterator data struct 557* @param length returned string text length 558* @return string text 559* @see #usearch_setText 560* @stable ICU 2.4 561*/ 562U_STABLE const UChar * U_EXPORT2 usearch_getText(const UStringSearch *strsrch, 563 int32_t *length); 564 565/** 566* Gets the collator used for the language rules. 567* <p> 568* Deleting the returned <tt>UCollator</tt> before calling 569* <tt>usearch_close</tt> would cause the string search to fail. 570* <tt>usearch_close</tt> will delete the collator if this search owns it. 571* @param strsrch search iterator data struct 572* @return collator 573* @stable ICU 2.4 574*/ 575U_STABLE UCollator * U_EXPORT2 usearch_getCollator( 576 const UStringSearch *strsrch); 577 578/** 579* Sets the collator used for the language rules. User retains the ownership 580* of this collator, thus the responsibility of deletion lies with the user. 581* This method causes internal data such as Boyer-Moore shift tables to 582* be recalculated, but the iterator's position is unchanged. 583* @param strsrch search iterator data struct 584* @param collator to be used 585* @param status for errors if it occurs 586* @stable ICU 2.4 587*/ 588U_STABLE void U_EXPORT2 usearch_setCollator( UStringSearch *strsrch, 589 const UCollator *collator, 590 UErrorCode *status); 591 592/** 593* Sets the pattern used for matching. 594* Internal data like the Boyer Moore table will be recalculated, but the 595* iterator's position is unchanged. 596* @param strsrch search iterator data struct 597* @param pattern string 598* @param patternlength pattern length, -1 for null-terminated string 599* @param status for errors if it occurs. If text is NULL, or textlength is 0 600* then an U_ILLEGAL_ARGUMENT_ERROR is returned with no change 601* done to strsrch. 602* @stable ICU 2.4 603*/ 604U_STABLE void U_EXPORT2 usearch_setPattern( UStringSearch *strsrch, 605 const UChar *pattern, 606 int32_t patternlength, 607 UErrorCode *status); 608 609/** 610* Gets the search pattern 611* @param strsrch search iterator data struct 612* @param length return length of the pattern, -1 indicates that the pattern 613* is null-terminated 614* @return pattern string 615* @stable ICU 2.4 616*/ 617U_STABLE const UChar * U_EXPORT2 usearch_getPattern( 618 const UStringSearch *strsrch, 619 int32_t *length); 620 621/* methods ------------------------------------------------------------- */ 622 623/** 624* Returns the first index at which the string text matches the search 625* pattern. 626* The iterator is adjusted so that its current index (as returned by 627* <tt>usearch_getOffset</tt>) is the match position if one was found. 628* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and 629* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>. 630* @param strsrch search iterator data struct 631* @param status for errors if it occurs 632* @return The character index of the first match, or 633* <tt>USEARCH_DONE</tt> if there are no matches. 634* @see #usearch_getOffset 635* @see #USEARCH_DONE 636* @stable ICU 2.4 637*/ 638U_STABLE int32_t U_EXPORT2 usearch_first(UStringSearch *strsrch, 639 UErrorCode *status); 640 641/** 642* Returns the first index equal or greater than <tt>position</tt> at which 643* the string text 644* matches the search pattern. The iterator is adjusted so that its current 645* index (as returned by <tt>usearch_getOffset</tt>) is the match position if 646* one was found. 647* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and 648* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt> 649* <p> 650* Search positions that may render incorrect results are highlighted in the 651* header comments. If position is less than or greater than the text range 652* for searching, an U_INDEX_OUTOFBOUNDS_ERROR will be returned 653* @param strsrch search iterator data struct 654* @param position to start the search at 655* @param status for errors if it occurs 656* @return The character index of the first match following <tt>pos</tt>, 657* or <tt>USEARCH_DONE</tt> if there are no matches. 658* @see #usearch_getOffset 659* @see #USEARCH_DONE 660* @stable ICU 2.4 661*/ 662U_STABLE int32_t U_EXPORT2 usearch_following(UStringSearch *strsrch, 663 int32_t position, 664 UErrorCode *status); 665 666/** 667* Returns the last index in the target text at which it matches the search 668* pattern. The iterator is adjusted so that its current 669* index (as returned by <tt>usearch_getOffset</tt>) is the match position if 670* one was found. 671* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and 672* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>. 673* @param strsrch search iterator data struct 674* @param status for errors if it occurs 675* @return The index of the first match, or <tt>USEARCH_DONE</tt> if there 676* are no matches. 677* @see #usearch_getOffset 678* @see #USEARCH_DONE 679* @stable ICU 2.4 680*/ 681U_STABLE int32_t U_EXPORT2 usearch_last(UStringSearch *strsrch, 682 UErrorCode *status); 683 684/** 685* Returns the first index less than <tt>position</tt> at which the string text 686* matches the search pattern. The iterator is adjusted so that its current 687* index (as returned by <tt>usearch_getOffset</tt>) is the match position if 688* one was found. 689* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and 690* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt> 691* <p> 692* Search positions that may render incorrect results are highlighted in the 693* header comments. If position is less than or greater than the text range 694* for searching, an U_INDEX_OUTOFBOUNDS_ERROR will be returned. 695* <p> 696* When <tt>USEARCH_OVERLAP</tt> option is off, the last index of the 697* result match is always less than <tt>position</tt>. 698* When <tt>USERARCH_OVERLAP</tt> is on, the result match may span across 699* <tt>position</tt>. 700* @param strsrch search iterator data struct 701* @param position index position the search is to begin at 702* @param status for errors if it occurs 703* @return The character index of the first match preceding <tt>pos</tt>, 704* or <tt>USEARCH_DONE</tt> if there are no matches. 705* @see #usearch_getOffset 706* @see #USEARCH_DONE 707* @stable ICU 2.4 708*/ 709U_STABLE int32_t U_EXPORT2 usearch_preceding(UStringSearch *strsrch, 710 int32_t position, 711 UErrorCode *status); 712 713/** 714* Returns the index of the next point at which the string text matches the 715* search pattern, starting from the current position. 716* The iterator is adjusted so that its current 717* index (as returned by <tt>usearch_getOffset</tt>) is the match position if 718* one was found. 719* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and 720* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt> 721* @param strsrch search iterator data struct 722* @param status for errors if it occurs 723* @return The index of the next match after the current position, or 724* <tt>USEARCH_DONE</tt> if there are no more matches. 725* @see #usearch_first 726* @see #usearch_getOffset 727* @see #USEARCH_DONE 728* @stable ICU 2.4 729*/ 730U_STABLE int32_t U_EXPORT2 usearch_next(UStringSearch *strsrch, 731 UErrorCode *status); 732 733/** 734* Returns the index of the previous point at which the string text matches 735* the search pattern, starting at the current position. 736* The iterator is adjusted so that its current 737* index (as returned by <tt>usearch_getOffset</tt>) is the match position if 738* one was found. 739* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and 740* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt> 741* @param strsrch search iterator data struct 742* @param status for errors if it occurs 743* @return The index of the previous match before the current position, 744* or <tt>USEARCH_DONE</tt> if there are no more matches. 745* @see #usearch_last 746* @see #usearch_getOffset 747* @see #USEARCH_DONE 748* @stable ICU 2.4 749*/ 750U_STABLE int32_t U_EXPORT2 usearch_previous(UStringSearch *strsrch, 751 UErrorCode *status); 752 753/** 754* Reset the iteration. 755* Search will begin at the start of the text string if a forward iteration 756* is initiated before a backwards iteration. Otherwise if a backwards 757* iteration is initiated before a forwards iteration, the search will begin 758* at the end of the text string. 759* @param strsrch search iterator data struct 760* @see #usearch_first 761* @stable ICU 2.4 762*/ 763U_STABLE void U_EXPORT2 usearch_reset(UStringSearch *strsrch); 764 765#ifndef U_HIDE_INTERNAL_API 766/** 767 * Simple forward search for the pattern, starting at a specified index, 768 * and using using a default set search options. 769 * 770 * This is an experimental function, and is not an official part of the 771 * ICU API. 772 * 773 * The collator options, such as UCOL_STRENGTH and UCOL_NORMALIZTION, are honored. 774 * 775 * The UStringSearch options USEARCH_CANONICAL_MATCH, USEARCH_OVERLAP and 776 * any Break Iterator are ignored. 777 * 778 * Matches obey the following constraints: 779 * 780 * Characters at the start or end positions of a match that are ignorable 781 * for collation are not included as part of the match, unless they 782 * are part of a combining sequence, as described below. 783 * 784 * A match will not include a partial combining sequence. Combining 785 * character sequences are considered to be inseperable units, 786 * and either match the pattern completely, or are considered to not match 787 * at all. Thus, for example, an A followed a combining accent mark will 788 * not be found when searching for a plain (unaccented) A. (unless 789 * the collation strength has been set to ignore all accents). 790 * 791 * When beginning a search, the initial starting position, startIdx, 792 * is assumed to be an acceptable match boundary with respect to 793 * combining characters. A combining sequence that spans across the 794 * starting point will not supress a match beginning at startIdx. 795 * 796 * Characters that expand to multiple collation elements 797 * (German sharp-S becoming 'ss', or the composed forms of accented 798 * characters, for example) also must match completely. 799 * Searching for a single 's' in a string containing only a sharp-s will 800 * find no match. 801 * 802 * 803 * @param strsrch the UStringSearch struct, which references both 804 * the text to be searched and the pattern being sought. 805 * @param startIdx The index into the text to begin the search. 806 * @param matchStart An out parameter, the starting index of the matched text. 807 * This parameter may be NULL. 808 * A value of -1 will be returned if no match was found. 809 * @param matchLimit Out parameter, the index of the first position following the matched text. 810 * The matchLimit will be at a suitable position for beginning a subsequent search 811 * in the input text. 812 * This parameter may be NULL. 813 * A value of -1 will be returned if no match was found. 814 * 815 * @param status Report any errors. Note that no match found is not an error. 816 * @return TRUE if a match was found, FALSE otherwise. 817 * 818 * @internal 819 */ 820U_INTERNAL UBool U_EXPORT2 usearch_search(UStringSearch *strsrch, 821 int32_t startIdx, 822 int32_t *matchStart, 823 int32_t *matchLimit, 824 UErrorCode *status); 825 826/** 827 * Simple backwards search for the pattern, starting at a specified index, 828 * and using using a default set search options. 829 * 830 * This is an experimental function, and is not an official part of the 831 * ICU API. 832 * 833 * The collator options, such as UCOL_STRENGTH and UCOL_NORMALIZTION, are honored. 834 * 835 * The UStringSearch options USEARCH_CANONICAL_MATCH, USEARCH_OVERLAP and 836 * any Break Iterator are ignored. 837 * 838 * Matches obey the following constraints: 839 * 840 * Characters at the start or end positions of a match that are ignorable 841 * for collation are not included as part of the match, unless they 842 * are part of a combining sequence, as described below. 843 * 844 * A match will not include a partial combining sequence. Combining 845 * character sequences are considered to be inseperable units, 846 * and either match the pattern completely, or are considered to not match 847 * at all. Thus, for example, an A followed a combining accent mark will 848 * not be found when searching for a plain (unaccented) A. (unless 849 * the collation strength has been set to ignore all accents). 850 * 851 * When beginning a search, the initial starting position, startIdx, 852 * is assumed to be an acceptable match boundary with respect to 853 * combining characters. A combining sequence that spans across the 854 * starting point will not supress a match beginning at startIdx. 855 * 856 * Characters that expand to multiple collation elements 857 * (German sharp-S becoming 'ss', or the composed forms of accented 858 * characters, for example) also must match completely. 859 * Searching for a single 's' in a string containing only a sharp-s will 860 * find no match. 861 * 862 * 863 * @param strsrch the UStringSearch struct, which references both 864 * the text to be searched and the pattern being sought. 865 * @param startIdx The index into the text to begin the search. 866 * @param matchStart An out parameter, the starting index of the matched text. 867 * This parameter may be NULL. 868 * A value of -1 will be returned if no match was found. 869 * @param matchLimit Out parameter, the index of the first position following the matched text. 870 * The matchLimit will be at a suitable position for beginning a subsequent search 871 * in the input text. 872 * This parameter may be NULL. 873 * A value of -1 will be returned if no match was found. 874 * 875 * @param status Report any errors. Note that no match found is not an error. 876 * @return TRUE if a match was found, FALSE otherwise. 877 * 878 * @internal 879 */ 880U_INTERNAL UBool U_EXPORT2 usearch_searchBackwards(UStringSearch *strsrch, 881 int32_t startIdx, 882 int32_t *matchStart, 883 int32_t *matchLimit, 884 UErrorCode *status); 885#endif /* U_HIDE_INTERNAL_API */ 886 887#endif /* #if !UCONFIG_NO_COLLATION && !UCONFIG_NO_BREAK_ITERATION */ 888 889#endif 890