1// © 2016 and later: Unicode, Inc. and others.
2// License & terms of use: http://www.unicode.org/copyright.html
3/*
4**********************************************************************
5* Copyright (C) 2001-2011,2014 IBM and others. All rights reserved.
6**********************************************************************
7* Date Name Description
8* 06/28/2001 synwee Creation.
9**********************************************************************
10*/
11#ifndef USEARCH_H
12#define USEARCH_H
13
14#include "unicode/utypes.h"
15
16#if !UCONFIG_NO_COLLATION && !UCONFIG_NO_BREAK_ITERATION
17
18#include "unicode/localpointer.h"
19#include "unicode/ucol.h"
20#include "unicode/ucoleitr.h"
21#include "unicode/ubrk.h"
22
23/**
24 * \file
25 * \brief C API: StringSearch
26 *
27 * C APIs for an engine that provides language-sensitive text searching based
28 * on the comparison rules defined in a <tt>UCollator</tt> data struct,
29 * see <tt>ucol.h</tt>. This ensures that language eccentricity can be
30 * handled, e.g. for the German collator, characters &szlig; and SS will be matched
31 * if case is chosen to be ignored.
32 * See the <a href="http://source.icu-project.org/repos/icu/icuhtml/trunk/design/collation/ICU_collation_design.htm">
33 * "ICU Collation Design Document"</a> for more information.
34 * <p>
35 * The implementation may use a linear search or a modified form of the Boyer-Moore
36 * search; for more information on the latter see
37 * <a href="http://icu-project.org/docs/papers/efficient_text_searching_in_java.html">
38 * "Efficient Text Searching in Java"</a>, published in <i>Java Report</i>
39 * in February, 1999.
40 * <p>
41 * There are 2 match options for selection:<br>
42 * Let S' be the sub-string of a text string S between the offsets start and
43 * end <start, end>.
44 * <br>
45 * A pattern string P matches a text string S at the offsets <start, end>
46 * if
47 * <pre>
48 * option 1. Some canonical equivalent of P matches some canonical equivalent
49 * of S'
50 * option 2. P matches S' and if P starts or ends with a combining mark,
51 * there exists no non-ignorable combining mark before or after S'
52 * in S respectively.
53 * </pre>
54 * Option 2. will be the default.
55 * <p>
56 * This search has APIs similar to that of other text iteration mechanisms
57 * such as the break iterators in <tt>ubrk.h</tt>. Using these
58 * APIs, it is easy to scan through text looking for all occurrences of
59 * a given pattern. This search iterator allows changing of direction by
60 * calling a <tt>reset</tt> followed by a <tt>next</tt> or <tt>previous</tt>.
61 * Though a direction change can occur without calling <tt>reset</tt> first,
62 * this operation comes with some speed penalty.
63 * Generally, match results in the forward direction will match the result
64 * matches in the backwards direction in the reverse order
65 * <p>
66 * <tt>usearch.h</tt> provides APIs to specify the starting position
67 * within the text string to be searched, e.g. <tt>usearch_setOffset</tt>,
68 * <tt>usearch_preceding</tt> and <tt>usearch_following</tt>. Since the
69 * starting position will be set as it is specified, please take note that
70 * there are some dangerous positions which the search may render incorrect
71 * results:
72 * <ul>
73 * <li> The midst of a substring that requires normalization.
74 * <li> If the following match is to be found, the position should not be the
75 * second character which requires to be swapped with the preceding
76 * character. Vice versa, if the preceding match is to be found,
77 * position to search from should not be the first character which
78 * requires to be swapped with the next character. E.g certain Thai and
79 * Lao characters require swapping.
80 * <li> If a following pattern match is to be found, any position within a
81 * contracting sequence except the first will fail. Vice versa if a
82 * preceding pattern match is to be found, a invalid starting point
83 * would be any character within a contracting sequence except the last.
84 * </ul>
85 * <p>
86 * A breakiterator can be used if only matches at logical breaks are desired.
87 * Using a breakiterator will only give you results that exactly matches the
88 * boundaries given by the breakiterator. For instance the pattern "e" will
89 * not be found in the string "\u00e9" if a character break iterator is used.
90 * <p>
91 * Options are provided to handle overlapping matches.
92 * E.g. In English, overlapping matches produces the result 0 and 2
93 * for the pattern "abab" in the text "ababab", where else mutually
94 * exclusive matches only produce the result of 0.
95 * <p>
96 * Options are also provided to implement "asymmetric search" as described in
97 * <a href="http://www.unicode.org/reports/tr10/#Asymmetric_Search">
98 * UTS #10 Unicode Collation Algorithm</a>, specifically the USearchAttribute
99 * USEARCH_ELEMENT_COMPARISON and its values.
100 * <p>
101 * Though collator attributes will be taken into consideration while
102 * performing matches, there are no APIs here for setting and getting the
103 * attributes. These attributes can be set by getting the collator
104 * from <tt>usearch_getCollator</tt> and using the APIs in <tt>ucol.h</tt>.
105 * Lastly to update String Search to the new collator attributes,
106 * usearch_reset() has to be called.
107 * <p>
108 * Restriction: <br>
109 * Currently there are no composite characters that consists of a
110 * character with combining class > 0 before a character with combining
111 * class == 0. However, if such a character exists in the future, the
112 * search mechanism does not guarantee the results for option 1.
113 *
114 * <p>
115 * Example of use:<br>
116 * <pre><code>
117 * char *tgtstr = "The quick brown fox jumped over the lazy fox";
118 * char *patstr = "fox";
119 * UChar target[64];
120 * UChar pattern[16];
121 * UErrorCode status = U_ZERO_ERROR;
122 * u_uastrcpy(target, tgtstr);
123 * u_uastrcpy(pattern, patstr);
124 *
125 * UStringSearch *search = usearch_open(pattern, -1, target, -1, "en_US",
126 * NULL, &status);
127 * if (U_SUCCESS(status)) {
128 * for (int pos = usearch_first(search, &status);
129 * pos != USEARCH_DONE;
130 * pos = usearch_next(search, &status))
131 * {
132 * printf("Found match at %d pos, length is %d\n", pos,
133 * usearch_getMatchedLength(search));
134 * }
135 * }
136 *
137 * usearch_close(search);
138 * </code></pre>
139 * @stable ICU 2.4
140 */
141
142/**
143* DONE is returned by previous() and next() after all valid matches have
144* been returned, and by first() and last() if there are no matches at all.
145* @stable ICU 2.4
146*/
147#define USEARCH_DONE -1
148
149/**
150* Data structure for searching
151* @stable ICU 2.4
152*/
153struct UStringSearch;
154/**
155* Data structure for searching
156* @stable ICU 2.4
157*/
158typedef struct UStringSearch UStringSearch;
159
160/**
161* @stable ICU 2.4
162*/
163typedef enum {
164 /**
165 * Option for overlapping matches
166 * @stable ICU 2.4
167 */
168 USEARCH_OVERLAP = 0,
169#ifndef U_HIDE_DEPRECATED_API
170 /**
171 * Option for canonical matches; option 1 in header documentation.
172 * The default value will be USEARCH_OFF.
173 * Note: Setting this option to USEARCH_ON currently has no effect on
174 * search behavior, and this option is deprecated. Instead, to control
175 * canonical match behavior, you must set UCOL_NORMALIZATION_MODE
176 * appropriately (to UCOL_OFF or UCOL_ON) in the UCollator used by
177 * the UStringSearch object.
178 * @see usearch_openFromCollator
179 * @see usearch_getCollator
180 * @see usearch_setCollator
181 * @see ucol_getAttribute
182 * @deprecated ICU 53
183 */
184 USEARCH_CANONICAL_MATCH = 1,
185#endif /* U_HIDE_DEPRECATED_API */
186 /**
187 * Option to control how collation elements are compared.
188 * The default value will be USEARCH_STANDARD_ELEMENT_COMPARISON.
189 * @stable ICU 4.4
190 */
191 USEARCH_ELEMENT_COMPARISON = 2,
192
193#ifndef U_HIDE_DEPRECATED_API
194 /**
195 * One more than the highest normal USearchAttribute value.
196 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
197 */
198 USEARCH_ATTRIBUTE_COUNT = 3
199#endif /* U_HIDE_DEPRECATED_API */
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 * [UTS #10 Unicode Collation Algorithm](http://www.unicode.org/reports/tr10/#Asymmetric_Search),
261 * but also allows unmarked characters in the searched text to match
262 * marked or unmarked versions of that character in the pattern.
263 *
264 * @stable ICU 4.4
265 */
266 USEARCH_ANY_BASE_WEIGHT_IS_WILDCARD,
267
268#ifndef U_HIDE_DEPRECATED_API
269 /**
270 * One more than the highest normal USearchAttributeValue value.
271 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
272 */
273 USEARCH_ATTRIBUTE_VALUE_COUNT
274#endif /* U_HIDE_DEPRECATED_API */
275} USearchAttributeValue;
276
277/* open and close ------------------------------------------------------ */
278
279/**
280* Creating a search iterator data struct using the argument locale language
281* rule set. A collator will be created in the process, which will be owned by
282* this search and will be deleted in <tt>usearch_close</tt>.
283* @param pattern for matching
284* @param patternlength length of the pattern, -1 for null-termination
285* @param text text string
286* @param textlength length of the text string, -1 for null-termination
287* @param locale name of locale for the rules to be used
288* @param breakiter A BreakIterator that will be used to restrict the points
289* at which matches are detected. If a match is found, but
290* the match's start or end index is not a boundary as
291* determined by the <tt>BreakIterator</tt>, the match will
292* be rejected and another will be searched for.
293* If this parameter is <tt>NULL</tt>, no break detection is
294* attempted.
295* @param status for errors if it occurs. If pattern or text is NULL, or if
296* patternlength or textlength is 0 then an
297* U_ILLEGAL_ARGUMENT_ERROR is returned.
298* @return search iterator data structure, or NULL if there is an error.
299* @stable ICU 2.4
300*/
301U_STABLE UStringSearch * U_EXPORT2 usearch_open(const UChar *pattern,
302 int32_t patternlength,
303 const UChar *text,
304 int32_t textlength,
305 const char *locale,
306 UBreakIterator *breakiter,
307 UErrorCode *status);
308
309/**
310* Creating a search iterator data struct using the argument collator language
311* rule set. Note, user retains the ownership of this collator, thus the
312* responsibility of deletion lies with the user.
313* NOTE: string search cannot be instantiated from a collator that has
314* collate digits as numbers (CODAN) turned on.
315* @param pattern for matching
316* @param patternlength length of the pattern, -1 for null-termination
317* @param text text string
318* @param textlength length of the text string, -1 for null-termination
319* @param collator used for the language rules
320* @param breakiter A BreakIterator that will be used to restrict the points
321* at which matches are detected. If a match is found, but
322* the match's start or end index is not a boundary as
323* determined by the <tt>BreakIterator</tt>, the match will
324* be rejected and another will be searched for.
325* If this parameter is <tt>NULL</tt>, no break detection is
326* attempted.
327* @param status for errors if it occurs. If collator, pattern or text is NULL,
328* or if patternlength or textlength is 0 then an
329* U_ILLEGAL_ARGUMENT_ERROR is returned.
330* @return search iterator data structure, or NULL if there is an error.
331* @stable ICU 2.4
332*/
333U_STABLE UStringSearch * U_EXPORT2 usearch_openFromCollator(
334 const UChar *pattern,
335 int32_t patternlength,
336 const UChar *text,
337 int32_t textlength,
338 const UCollator *collator,
339 UBreakIterator *breakiter,
340 UErrorCode *status);
341
342/**
343* Destroying and cleaning up the search iterator data struct.
344* If a collator is created in <tt>usearch_open</tt>, it will be destroyed here.
345* @param searchiter data struct to clean up
346* @stable ICU 2.4
347*/
348U_STABLE void U_EXPORT2 usearch_close(UStringSearch *searchiter);
349
350#if U_SHOW_CPLUSPLUS_API
351
352U_NAMESPACE_BEGIN
353
354/**
355 * \class LocalUStringSearchPointer
356 * "Smart pointer" class, closes a UStringSearch via usearch_close().
357 * For most methods see the LocalPointerBase base class.
358 *
359 * @see LocalPointerBase
360 * @see LocalPointer
361 * @stable ICU 4.4
362 */
363U_DEFINE_LOCAL_OPEN_POINTER(LocalUStringSearchPointer, UStringSearch, usearch_close);
364
365U_NAMESPACE_END
366
367#endif
368
369/* get and set methods -------------------------------------------------- */
370
371/**
372* Sets the current position in the text string which the next search will
373* start from. Clears previous states.
374* This method takes the argument index and sets the position in the text
375* string accordingly without checking if the index is pointing to a
376* valid starting point to begin searching.
377* Search positions that may render incorrect results are highlighted in the
378* header comments
379* @param strsrch search iterator data struct
380* @param position position to start next search from. If position is less
381* than or greater than the text range for searching,
382* an U_INDEX_OUTOFBOUNDS_ERROR will be returned
383* @param status error status if any.
384* @stable ICU 2.4
385*/
386U_STABLE void U_EXPORT2 usearch_setOffset(UStringSearch *strsrch,
387 int32_t position,
388 UErrorCode *status);
389
390/**
391* Return the current index in the string text being searched.
392* If the iteration has gone past the end of the text (or past the beginning
393* for a backwards search), <tt>USEARCH_DONE</tt> is returned.
394* @param strsrch search iterator data struct
395* @see #USEARCH_DONE
396* @stable ICU 2.4
397*/
398U_STABLE int32_t U_EXPORT2 usearch_getOffset(const UStringSearch *strsrch);
399
400/**
401* Sets the text searching attributes located in the enum USearchAttribute
402* with values from the enum USearchAttributeValue.
403* <tt>USEARCH_DEFAULT</tt> can be used for all attributes for resetting.
404* @param strsrch search iterator data struct
405* @param attribute text attribute to be set
406* @param value text attribute value
407* @param status for errors if it occurs
408* @see #usearch_getAttribute
409* @stable ICU 2.4
410*/
411U_STABLE void U_EXPORT2 usearch_setAttribute(UStringSearch *strsrch,
412 USearchAttribute attribute,
413 USearchAttributeValue value,
414 UErrorCode *status);
415
416/**
417* Gets the text searching attributes.
418* @param strsrch search iterator data struct
419* @param attribute text attribute to be retrieve
420* @return text attribute value
421* @see #usearch_setAttribute
422* @stable ICU 2.4
423*/
424U_STABLE USearchAttributeValue U_EXPORT2 usearch_getAttribute(
425 const UStringSearch *strsrch,
426 USearchAttribute attribute);
427
428/**
429* Returns the index to the match in the text string that was searched.
430* This call returns a valid result only after a successful call to
431* <tt>usearch_first</tt>, <tt>usearch_next</tt>, <tt>usearch_previous</tt>,
432* or <tt>usearch_last</tt>.
433* Just after construction, or after a searching method returns
434* <tt>USEARCH_DONE</tt>, this method will return <tt>USEARCH_DONE</tt>.
435* <p>
436* Use <tt>usearch_getMatchedLength</tt> to get the matched string length.
437* @param strsrch search iterator data struct
438* @return index to a substring within the text string that is being
439* searched.
440* @see #usearch_first
441* @see #usearch_next
442* @see #usearch_previous
443* @see #usearch_last
444* @see #USEARCH_DONE
445* @stable ICU 2.4
446*/
447U_STABLE int32_t U_EXPORT2 usearch_getMatchedStart(
448 const UStringSearch *strsrch);
449
450/**
451* Returns the length of text in the string which matches the search pattern.
452* This call returns a valid result only after a successful call to
453* <tt>usearch_first</tt>, <tt>usearch_next</tt>, <tt>usearch_previous</tt>,
454* or <tt>usearch_last</tt>.
455* Just after construction, or after a searching method returns
456* <tt>USEARCH_DONE</tt>, this method will return 0.
457* @param strsrch search iterator data struct
458* @return The length of the match in the string text, or 0 if there is no
459* match currently.
460* @see #usearch_first
461* @see #usearch_next
462* @see #usearch_previous
463* @see #usearch_last
464* @see #USEARCH_DONE
465* @stable ICU 2.4
466*/
467U_STABLE int32_t U_EXPORT2 usearch_getMatchedLength(
468 const UStringSearch *strsrch);
469
470/**
471* Returns the text that was matched by the most recent call to
472* <tt>usearch_first</tt>, <tt>usearch_next</tt>, <tt>usearch_previous</tt>,
473* or <tt>usearch_last</tt>.
474* If the iterator is not pointing at a valid match (e.g. just after
475* construction or after <tt>USEARCH_DONE</tt> has been returned, returns
476* an empty string. If result is not large enough to store the matched text,
477* result will be filled with the partial text and an U_BUFFER_OVERFLOW_ERROR
478* will be returned in status. result will be null-terminated whenever
479* possible. If the buffer fits the matched text exactly, a null-termination
480* is not possible, then a U_STRING_NOT_TERMINATED_ERROR set in status.
481* Pre-flighting can be either done with length = 0 or the API
482* <tt>usearch_getMatchedLength</tt>.
483* @param strsrch search iterator data struct
484* @param result UChar buffer to store the matched string
485* @param resultCapacity length of the result buffer
486* @param status error returned if result is not large enough
487* @return exact length of the matched text, not counting the null-termination
488* @see #usearch_first
489* @see #usearch_next
490* @see #usearch_previous
491* @see #usearch_last
492* @see #USEARCH_DONE
493* @stable ICU 2.4
494*/
495U_STABLE int32_t U_EXPORT2 usearch_getMatchedText(const UStringSearch *strsrch,
496 UChar *result,
497 int32_t resultCapacity,
498 UErrorCode *status);
499
500#if !UCONFIG_NO_BREAK_ITERATION
501
502/**
503* Set the BreakIterator that will be used to restrict the points at which
504* matches are detected.
505* @param strsrch search iterator data struct
506* @param breakiter A BreakIterator that will be used to restrict the points
507* at which matches are detected. If a match is found, but
508* the match's start or end index is not a boundary as
509* determined by the <tt>BreakIterator</tt>, the match will
510* be rejected and another will be searched for.
511* If this parameter is <tt>NULL</tt>, no break detection is
512* attempted.
513* @param status for errors if it occurs
514* @see #usearch_getBreakIterator
515* @stable ICU 2.4
516*/
517U_STABLE void U_EXPORT2 usearch_setBreakIterator(UStringSearch *strsrch,
518 UBreakIterator *breakiter,
519 UErrorCode *status);
520
521/**
522* Returns the BreakIterator that is used to restrict the points at which
523* matches are detected. This will be the same object that was passed to the
524* constructor or to <tt>usearch_setBreakIterator</tt>. Note that
525* <tt>NULL</tt>
526* is a legal value; it means that break detection should not be attempted.
527* @param strsrch search iterator data struct
528* @return break iterator used
529* @see #usearch_setBreakIterator
530* @stable ICU 2.4
531*/
532U_STABLE const UBreakIterator * U_EXPORT2 usearch_getBreakIterator(
533 const UStringSearch *strsrch);
534
535#endif
536
537/**
538* Set the string text to be searched. Text iteration will hence begin at the
539* start of the text string. This method is useful if you want to re-use an
540* iterator to search for the same pattern within a different body of text.
541* @param strsrch search iterator data struct
542* @param text new string to look for match
543* @param textlength length of the new string, -1 for null-termination
544* @param status for errors if it occurs. If text is NULL, or textlength is 0
545* then an U_ILLEGAL_ARGUMENT_ERROR is returned with no change
546* done to strsrch.
547* @see #usearch_getText
548* @stable ICU 2.4
549*/
550U_STABLE void U_EXPORT2 usearch_setText( UStringSearch *strsrch,
551 const UChar *text,
552 int32_t textlength,
553 UErrorCode *status);
554
555/**
556* Return the string text to be searched.
557* @param strsrch search iterator data struct
558* @param length returned string text length
559* @return string text
560* @see #usearch_setText
561* @stable ICU 2.4
562*/
563U_STABLE const UChar * U_EXPORT2 usearch_getText(const UStringSearch *strsrch,
564 int32_t *length);
565
566/**
567* Gets the collator used for the language rules.
568* <p>
569* Deleting the returned <tt>UCollator</tt> before calling
570* <tt>usearch_close</tt> would cause the string search to fail.
571* <tt>usearch_close</tt> will delete the collator if this search owns it.
572* @param strsrch search iterator data struct
573* @return collator
574* @stable ICU 2.4
575*/
576U_STABLE UCollator * U_EXPORT2 usearch_getCollator(
577 const UStringSearch *strsrch);
578
579/**
580* Sets the collator used for the language rules. User retains the ownership
581* of this collator, thus the responsibility of deletion lies with the user.
582* This method causes internal data such as Boyer-Moore shift tables to
583* be recalculated, but the iterator's position is unchanged.
584* @param strsrch search iterator data struct
585* @param collator to be used
586* @param status for errors if it occurs
587* @stable ICU 2.4
588*/
589U_STABLE void U_EXPORT2 usearch_setCollator( UStringSearch *strsrch,
590 const UCollator *collator,
591 UErrorCode *status);
592
593/**
594* Sets the pattern used for matching.
595* Internal data like the Boyer Moore table will be recalculated, but the
596* iterator's position is unchanged.
597* @param strsrch search iterator data struct
598* @param pattern string
599* @param patternlength pattern length, -1 for null-terminated string
600* @param status for errors if it occurs. If text is NULL, or textlength is 0
601* then an U_ILLEGAL_ARGUMENT_ERROR is returned with no change
602* done to strsrch.
603* @stable ICU 2.4
604*/
605U_STABLE void U_EXPORT2 usearch_setPattern( UStringSearch *strsrch,
606 const UChar *pattern,
607 int32_t patternlength,
608 UErrorCode *status);
609
610/**
611* Gets the search pattern
612* @param strsrch search iterator data struct
613* @param length return length of the pattern, -1 indicates that the pattern
614* is null-terminated
615* @return pattern string
616* @stable ICU 2.4
617*/
618U_STABLE const UChar * U_EXPORT2 usearch_getPattern(
619 const UStringSearch *strsrch,
620 int32_t *length);
621
622/* methods ------------------------------------------------------------- */
623
624/**
625* Returns the first index at which the string text matches the search
626* pattern.
627* The iterator is adjusted so that its current index (as returned by
628* <tt>usearch_getOffset</tt>) is the match position if one was found.
629* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
630* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>.
631* @param strsrch search iterator data struct
632* @param status for errors if it occurs
633* @return The character index of the first match, or
634* <tt>USEARCH_DONE</tt> if there are no matches.
635* @see #usearch_getOffset
636* @see #USEARCH_DONE
637* @stable ICU 2.4
638*/
639U_STABLE int32_t U_EXPORT2 usearch_first(UStringSearch *strsrch,
640 UErrorCode *status);
641
642/**
643* Returns the first index equal or greater than <tt>position</tt> at which
644* the string text
645* matches the search pattern. The iterator is adjusted so that its current
646* index (as returned by <tt>usearch_getOffset</tt>) is the match position if
647* one was found.
648* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
649* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>
650* <p>
651* Search positions that may render incorrect results are highlighted in the
652* header comments. If position is less than or greater than the text range
653* for searching, an U_INDEX_OUTOFBOUNDS_ERROR will be returned
654* @param strsrch search iterator data struct
655* @param position to start the search at
656* @param status for errors if it occurs
657* @return The character index of the first match following <tt>pos</tt>,
658* or <tt>USEARCH_DONE</tt> if there are no matches.
659* @see #usearch_getOffset
660* @see #USEARCH_DONE
661* @stable ICU 2.4
662*/
663U_STABLE int32_t U_EXPORT2 usearch_following(UStringSearch *strsrch,
664 int32_t position,
665 UErrorCode *status);
666
667/**
668* Returns the last index in the target text at which it matches the search
669* pattern. The iterator is adjusted so that its current
670* index (as returned by <tt>usearch_getOffset</tt>) is the match position if
671* one was found.
672* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
673* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>.
674* @param strsrch search iterator data struct
675* @param status for errors if it occurs
676* @return The index of the first match, or <tt>USEARCH_DONE</tt> if there
677* are no matches.
678* @see #usearch_getOffset
679* @see #USEARCH_DONE
680* @stable ICU 2.4
681*/
682U_STABLE int32_t U_EXPORT2 usearch_last(UStringSearch *strsrch,
683 UErrorCode *status);
684
685/**
686* Returns the first index less than <tt>position</tt> at which the string text
687* matches the search pattern. The iterator is adjusted so that its current
688* index (as returned by <tt>usearch_getOffset</tt>) is the match position if
689* one was found.
690* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
691* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>
692* <p>
693* Search positions that may render incorrect results are highlighted in the
694* header comments. If position is less than or greater than the text range
695* for searching, an U_INDEX_OUTOFBOUNDS_ERROR will be returned.
696* <p>
697* When <tt>USEARCH_OVERLAP</tt> option is off, the last index of the
698* result match is always less than <tt>position</tt>.
699* When <tt>USERARCH_OVERLAP</tt> is on, the result match may span across
700* <tt>position</tt>.
701* @param strsrch search iterator data struct
702* @param position index position the search is to begin at
703* @param status for errors if it occurs
704* @return The character index of the first match preceding <tt>pos</tt>,
705* or <tt>USEARCH_DONE</tt> if there are no matches.
706* @see #usearch_getOffset
707* @see #USEARCH_DONE
708* @stable ICU 2.4
709*/
710U_STABLE int32_t U_EXPORT2 usearch_preceding(UStringSearch *strsrch,
711 int32_t position,
712 UErrorCode *status);
713
714/**
715* Returns the index of the next point at which the string text matches the
716* search pattern, starting from the current position.
717* The iterator is adjusted so that its current
718* index (as returned by <tt>usearch_getOffset</tt>) is the match position if
719* one was found.
720* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
721* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>
722* @param strsrch search iterator data struct
723* @param status for errors if it occurs
724* @return The index of the next match after the current position, or
725* <tt>USEARCH_DONE</tt> if there are no more matches.
726* @see #usearch_first
727* @see #usearch_getOffset
728* @see #USEARCH_DONE
729* @stable ICU 2.4
730*/
731U_STABLE int32_t U_EXPORT2 usearch_next(UStringSearch *strsrch,
732 UErrorCode *status);
733
734/**
735* Returns the index of the previous point at which the string text matches
736* the search pattern, starting at the current position.
737* The iterator is adjusted so that its current
738* index (as returned by <tt>usearch_getOffset</tt>) is the match position if
739* one was found.
740* If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
741* the iterator will be adjusted to the index <tt>USEARCH_DONE</tt>
742* @param strsrch search iterator data struct
743* @param status for errors if it occurs
744* @return The index of the previous match before the current position,
745* or <tt>USEARCH_DONE</tt> if there are no more matches.
746* @see #usearch_last
747* @see #usearch_getOffset
748* @see #USEARCH_DONE
749* @stable ICU 2.4
750*/
751U_STABLE int32_t U_EXPORT2 usearch_previous(UStringSearch *strsrch,
752 UErrorCode *status);
753
754/**
755* Reset the iteration.
756* Search will begin at the start of the text string if a forward iteration
757* is initiated before a backwards iteration. Otherwise if a backwards
758* iteration is initiated before a forwards iteration, the search will begin
759* at the end of the text string.
760* @param strsrch search iterator data struct
761* @see #usearch_first
762* @stable ICU 2.4
763*/
764U_STABLE void U_EXPORT2 usearch_reset(UStringSearch *strsrch);
765
766#ifndef U_HIDE_INTERNAL_API
767/**
768 * Simple forward search for the pattern, starting at a specified index,
769 * and using a default set search options.
770 *
771 * This is an experimental function, and is not an official part of the
772 * ICU API.
773 *
774 * The collator options, such as UCOL_STRENGTH and UCOL_NORMALIZTION, are honored.
775 *
776 * The UStringSearch options USEARCH_CANONICAL_MATCH, USEARCH_OVERLAP and
777 * any Break Iterator are ignored.
778 *
779 * Matches obey the following constraints:
780 *
781 * Characters at the start or end positions of a match that are ignorable
782 * for collation are not included as part of the match, unless they
783 * are part of a combining sequence, as described below.
784 *
785 * A match will not include a partial combining sequence. Combining
786 * character sequences are considered to be inseparable units,
787 * and either match the pattern completely, or are considered to not match
788 * at all. Thus, for example, an A followed a combining accent mark will
789 * not be found when searching for a plain (unaccented) A. (unless
790 * the collation strength has been set to ignore all accents).
791 *
792 * When beginning a search, the initial starting position, startIdx,
793 * is assumed to be an acceptable match boundary with respect to
794 * combining characters. A combining sequence that spans across the
795 * starting point will not suppress a match beginning at startIdx.
796 *
797 * Characters that expand to multiple collation elements
798 * (German sharp-S becoming 'ss', or the composed forms of accented
799 * characters, for example) also must match completely.
800 * Searching for a single 's' in a string containing only a sharp-s will
801 * find no match.
802 *
803 *
804 * @param strsrch the UStringSearch struct, which references both
805 * the text to be searched and the pattern being sought.
806 * @param startIdx The index into the text to begin the search.
807 * @param matchStart An out parameter, the starting index of the matched text.
808 * This parameter may be NULL.
809 * A value of -1 will be returned if no match was found.
810 * @param matchLimit Out parameter, the index of the first position following the matched text.
811 * The matchLimit will be at a suitable position for beginning a subsequent search
812 * in the input text.
813 * This parameter may be NULL.
814 * A value of -1 will be returned if no match was found.
815 *
816 * @param status Report any errors. Note that no match found is not an error.
817 * @return TRUE if a match was found, FALSE otherwise.
818 *
819 * @internal
820 */
821U_INTERNAL UBool U_EXPORT2 usearch_search(UStringSearch *strsrch,
822 int32_t startIdx,
823 int32_t *matchStart,
824 int32_t *matchLimit,
825 UErrorCode *status);
826
827/**
828 * Simple backwards search for the pattern, starting at a specified index,
829 * and using using a default set search options.
830 *
831 * This is an experimental function, and is not an official part of the
832 * ICU API.
833 *
834 * The collator options, such as UCOL_STRENGTH and UCOL_NORMALIZTION, are honored.
835 *
836 * The UStringSearch options USEARCH_CANONICAL_MATCH, USEARCH_OVERLAP and
837 * any Break Iterator are ignored.
838 *
839 * Matches obey the following constraints:
840 *
841 * Characters at the start or end positions of a match that are ignorable
842 * for collation are not included as part of the match, unless they
843 * are part of a combining sequence, as described below.
844 *
845 * A match will not include a partial combining sequence. Combining
846 * character sequences are considered to be inseparable units,
847 * and either match the pattern completely, or are considered to not match
848 * at all. Thus, for example, an A followed a combining accent mark will
849 * not be found when searching for a plain (unaccented) A. (unless
850 * the collation strength has been set to ignore all accents).
851 *
852 * When beginning a search, the initial starting position, startIdx,
853 * is assumed to be an acceptable match boundary with respect to
854 * combining characters. A combining sequence that spans across the
855 * starting point will not suppress a match beginning at startIdx.
856 *
857 * Characters that expand to multiple collation elements
858 * (German sharp-S becoming 'ss', or the composed forms of accented
859 * characters, for example) also must match completely.
860 * Searching for a single 's' in a string containing only a sharp-s will
861 * find no match.
862 *
863 *
864 * @param strsrch the UStringSearch struct, which references both
865 * the text to be searched and the pattern being sought.
866 * @param startIdx The index into the text to begin the search.
867 * @param matchStart An out parameter, the starting index of the matched text.
868 * This parameter may be NULL.
869 * A value of -1 will be returned if no match was found.
870 * @param matchLimit Out parameter, the index of the first position following the matched text.
871 * The matchLimit will be at a suitable position for beginning a subsequent search
872 * in the input text.
873 * This parameter may be NULL.
874 * A value of -1 will be returned if no match was found.
875 *
876 * @param status Report any errors. Note that no match found is not an error.
877 * @return TRUE if a match was found, FALSE otherwise.
878 *
879 * @internal
880 */
881U_INTERNAL UBool U_EXPORT2 usearch_searchBackwards(UStringSearch *strsrch,
882 int32_t startIdx,
883 int32_t *matchStart,
884 int32_t *matchLimit,
885 UErrorCode *status);
886#endif /* U_HIDE_INTERNAL_API */
887
888#endif /* #if !UCONFIG_NO_COLLATION && !UCONFIG_NO_BREAK_ITERATION */
889
890#endif
891