FM / unicode

/usr/include/unicode

Name	Size	Mode	Actions
basictz.h	8913	0644	edit dl rm
bms.h	8664	0644	edit dl rm
bmsearch.h	7316	0644	edit dl rm
brkiter.h	23001	0644	edit dl rm
bytestream.h	8531	0644	edit dl rm
calendar.h	90659	0644	edit dl rm
caniter.h	7288	0644	edit dl rm
chariter.h	24186	0644	edit dl rm
choicfmt.h	29528	0644	edit dl rm
coleitr.h	12990	0644	edit dl rm
coll.h	48468	0644	edit dl rm
colldata.h	11957	0644	edit dl rm
curramt.h	3659	0644	edit dl rm
currpinf.h	7061	0644	edit dl rm
currunit.h	3057	0644	edit dl rm
datefmt.h	27946	0644	edit dl rm
dbbi.h	966	0644	edit dl rm
dcfmtsym.h	14040	0644	edit dl rm
decimfmt.h	88612	0644	edit dl rm
docmain.h	6003	0644	edit dl rm
dtfmtsym.h	26145	0644	edit dl rm
dtintrv.h	3734	0644	edit dl rm
dtitvfmt.h	39700	0644	edit dl rm
dtitvinf.h	18704	0644	edit dl rm
dtptngen.h	17303	0644	edit dl rm
dtrule.h	8718	0644	edit dl rm
errorcode.h	4361	0644	edit dl rm
fieldpos.h	8713	0644	edit dl rm
fmtable.h	18794	0644	edit dl rm
format.h	12083	0644	edit dl rm
gregocal.h	33797	0644	edit dl rm
locid.h	30241	0644	edit dl rm
measfmt.h	1897	0644	edit dl rm
measunit.h	1635	0644	edit dl rm
measure.h	3321	0644	edit dl rm
msgfmt.h	39435	0644	edit dl rm
normlzr.h	30337	0644	edit dl rm
numfmt.h	33987	0644	edit dl rm
numsys.h	5004	0644	edit dl rm
parseerr.h	3044	0644	edit dl rm
parsepos.h	5468	0644	edit dl rm
platform.h	9961	0644	edit dl rm
plurfmt.h	23079	0644	edit dl rm
plurrule.h	9901	0644	edit dl rm
ppalmos.h	6992	0644	edit dl rm
putil.h	6686	0644	edit dl rm
pwin32.h	8668	0644	edit dl rm
rbbi.h	26993	0644	edit dl rm
rbnf.h	45681	0644	edit dl rm
rbtz.h	15695	0644	edit dl rm
regex.h	52316	0644	edit dl rm
rep.h	9559	0644	edit dl rm
resbund.h	18130	0644	edit dl rm
schriter.h	6293	0644	edit dl rm
search.h	22357	0644	edit dl rm
simpletz.h	46179	0644	edit dl rm
smpdtfmt.h	48325	0644	edit dl rm
sortkey.h	11111	0644	edit dl rm
std_string.h	1769	0644	edit dl rm
strenum.h	9888	0644	edit dl rm
stringpiece.h	5484	0644	edit dl rm
stsearch.h	22287	0644	edit dl rm
symtable.h	4199	0644	edit dl rm
tblcoll.h	38199	0644	edit dl rm
timezone.h	34584	0644	edit dl rm
tmunit.h	3454	0644	edit dl rm
tmutamt.h	4787	0644	edit dl rm
tmutfmt.h	8669	0644	edit dl rm
translit.h	54780	0644	edit dl rm
tzrule.h	36057	0644	edit dl rm
tztrans.h	6101	0644	edit dl rm
ubidi.h	83092	0644	edit dl rm
ubrk.h	17688	0644	edit dl rm
ucal.h	40177	0644	edit dl rm
ucasemap.h	16420	0644	edit dl rm
ucat.h	5371	0644	edit dl rm
uchar.h	111062	0644	edit dl rm
uchriter.h	13132	0644	edit dl rm
uclean.h	11855	0644	edit dl rm
ucnv.h	83111	0644	edit dl rm
ucnvsel.h	5630	0644	edit dl rm
ucnv_cb.h	6635	0644	edit dl rm
ucnv_err.h	21375	0644	edit dl rm
ucol.h	48826	0644	edit dl rm
ucoleitr.h	12356	0644	edit dl rm
uconfig.h	6035	0644	edit dl rm
ucsdet.h	12464	0644	edit dl rm
ucurr.h	11572	0644	edit dl rm
udat.h	34496	0644	edit dl rm
udata.h	14265	0644	edit dl rm
udatpg.h	18715	0644	edit dl rm
udeprctd.h	2284	0644	edit dl rm
udraft.h	7600	0644	edit dl rm
uenum.h	5796	0644	edit dl rm
uidna.h	15918	0644	edit dl rm
uintrnal.h	15752	0644	edit dl rm
uiter.h	23234	0644	edit dl rm
uloc.h	44548	0644	edit dl rm
ulocdata.h	9739	0644	edit dl rm
umachine.h	11360	0644	edit dl rm
umisc.h	1255	0644	edit dl rm
umsg.h	26076	0644	edit dl rm
unifilt.h	3770	0644	edit dl rm
unifunct.h	3966	0644	edit dl rm
unimatch.h	6071	0644	edit dl rm
unirepl.h	3290	0644	edit dl rm
uniset.h	61771	0644	edit dl rm
unistr.h	161321	0644	edit dl rm
unorm.h	23621	0644	edit dl rm
unum.h	32145	0644	edit dl rm
uobject.h	11393	0644	edit dl rm
uobslete.h	852	0644	edit dl rm
uregex.h	43948	0644	edit dl rm
urename.h	92538	0644	edit dl rm
urep.h	5394	0644	edit dl rm
ures.h	36348	0644	edit dl rm
uscript.h	11360	0644	edit dl rm
usearch.h	34223	0644	edit dl rm
uset.h	39868	0644	edit dl rm
usetiter.h	9606	0644	edit dl rm
ushape.h	17907	0644	edit dl rm
uspoof.h	38752	0644	edit dl rm
usprep.h	7750	0644	edit dl rm
ustdio.h	35333	0644	edit dl rm
ustream.h	1891	0644	edit dl rm
ustring.h	69383	0644	edit dl rm
usystem.h	1925	0644	edit dl rm
utext.h	58394	0644	edit dl rm
utf.h	8273	0644	edit dl rm
utf8.h	20512	0644	edit dl rm
utf16.h	18776	0644	edit dl rm
utf32.h	653	0644	edit dl rm
utf_old.h	43646	0644	edit dl rm
utmscale.h	13689	0644	edit dl rm
utrace.h	13355	0644	edit dl rm
utrans.h	23129	0644	edit dl rm
utypes.h	34513	0644	edit dl rm
uversion.h	10902	0644	edit dl rm
vtzone.h	20048	0644	edit dl rm

Edit: /usr/include/unicode/bms.h (8664B)

/*
 * Copyright (C) 1996-2009, International Business Machines Corporation and Others.
 * All rights reserved.
 */

/**
 * \file 
 * \brief C API: Boyer-Moore StringSearch prototype.
 * \internal
 */

#ifndef _BMS_H
#define _BMS_H

#include "unicode/utypes.h"

#if !UCONFIG_NO_COLLATION

#include "unicode/ucol.h"

/**
 * A <code>UCD</code> object holds the Collator-specific data needed to
 * compute the length of the shortest string that can
 * generate a partcular list of CEs.
 *
 * <code>UCD</code> objects are quite expensive to compute. Because
 * of this, they are cached. When you call <code>ucd_open</code> it
 * returns a reference counted cached object. When you call <code>ucd_close</code>
 * the reference count on the object is decremented but the object is not deleted.
 *
 * If you do not need to reuse any unreferenced objects in the cache, you can call
 * <code>ucd_flushCCache</code>. If you no longer need any <code>UCD</code>
 * objects, you can call <code>ucd_freeCache</code>
 */
typedef void UCD;

/**
 * Open a <code>UCD</code> object.
 *
 * @param coll - the collator
 * @param status - will be set if any errors occur. 
 *
 * @return the <code>UCD</code> object. You must call
 *         <code>ucd_close</code> when you are done using the object.
 *
 * Note: if on return status is set to an error, the only safe
 * thing to do with the returned object is to call <code>ucd_close</code>.
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI UCD * U_EXPORT2
ucd_open(UCollator *coll, UErrorCode *status);

/**
 * Release a <code>UCD</code> object.
 *
 * @param ucd - the object
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI void U_EXPORT2
ucd_close(UCD *ucd);

/**
 * Get the <code>UCollator</code> object used to create a <code>UCD</code> object.
 * The <code>UCollator</code> object returned may not be the exact
 * object that was used to create this object, but it will have the
 * same behavior.
 *
 * @param ucd - the <code>UCD</code> object
 *
 * @return the <code>UCollator</code> used to create the given
 *         <code>UCD</code> object.
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI UCollator * U_EXPORT2
ucd_getCollator(UCD *ucd);

/**
 * <code>UCD</code> objects are expensive to compute, and so
 * may be cached. This routine will free the cached objects and delete
 * the cache.
 *
 * WARNING: Don't call this until you are have called <code>close</code>
 * for each <code>UCD</code> object that you have used. also,
 * DO NOT call this if another thread may be calling <code>ucd_flushCache</code>
 * at the same time.
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI void U_EXPORT2
ucd_freeCache();

/**
 * <code>UCD</code> objects are expensive to compute, and so
 * may be cached. This routine will remove any unused <code>UCD</code>
 * objects from the cache.
 *
 * @internal 4.0.1 technology preview
 */
U_CAPI void U_EXPORT2
ucd_flushCache();

/**
 * BMS
 *
 * This object holds the information needed to do a Collation sensitive Boyer-Moore search. It encapulates
 * the pattern, the "bad character" and "good suffix" tables, the Collator-based data needed to compute them,
 * and a reference to the text being searched.
 *
 * To do a search, you first need to get a <code>UCD</code> object by calling <code>ucd_open</code>.
 * Then you construct a <code>BMS</code> object from the <code>UCD</code> object, the pattern
 * string and the target string. Then you call the <code>search</code> method. Here's a code sample:
 *
 * <pre>
 * void boyerMooreExample(UCollator *collator, UChar *pattern, int32_t patternLen, UChar *target, int32_t targetLength)
 * {
 *     UErrorCode status = U_ZERO_ERROR;
 *     int32_t offset = 0, start = -1, end = -1;
 *     UCD *ucd = NULL);
 *     BMS *bms = NULL;
 *
 *     ucd = ucd_open(collator, &status);
 *     if (U_FAILURE(status)) {
 *         // could not create a UCD object
 *         return;
 *     }
 *
 *     BMS *bms = bms_open(ucd, pattern, patternLength, target, targetlength, &status);
 *     if (U_FAILURE(status)) {
 *         // could not create a BMS object
 *         ucd_close(ucd);
 *         return;
 *     }
 *
 *
 *     // Find all matches
 *     while (bms_search(bms, offset, &start, &end)) {
 *         // process the match between start and end
 *         ...
 *
 *         // advance past the match
 *         offset = end; 
 *     }
 *
 *     // at this point, if offset == 0, there were no matches
 *     if (offset == 0) {
 *         // handle the case of no matches
 *     }
 *
 *     bms_close(bms);
 *     ucd_close(ucd);
 *
 *     // UCD objects are cached, so the call to
 *     // ucd_close doesn't delete the object.
 *     // Call this if you don't need the object any more.
 *     ucd_flushCache();
 * }
 * </pre>
 *
 * NOTE: This is a technology preview. The final version of this API may not bear any resenblence to this API.
 *
 * Knows linitations:
 *   1) Backwards searching has not been implemented.
 *
 *   2) For Han and Hangul characters, this code ignores any Collation tailorings. In general,
 *      this isn't a problem, but in Korean locals, at strength 1, Hangul characters are tailored
 *      to be equal to Han characters with the same pronounciation. Because this code ignroes
 *      tailorings, searching for a Hangul character will not find a Han character and visa-versa.
 *
 *   3) In some cases, searching for a pattern that needs to be normalized and ends
 *      in a discontiguous contraction may fail. The only known cases of this are with
 *      the Tibetan script. For example searching for the pattern
 *      "\u0F7F\u0F80\u0F81\u0F82\u0F83\u0F84\u0F85" will fail. (This case is artificial. We've
 *      been unable to find a pratical, real-world example of this failure.)  
 *
 * NOTE: This is a technology preview. The final version of this API may not bear any resenblence to this API.
 *
 * @internal ICU 4.0.1 technology preview
 */
struct BMS;
typedef struct BMS BMS; /**< @see BMS */

/**
 * Construct a <code>MBS</code> object.
 *
 * @param ucd - A <code>UCD</code> object holding the Collator-sensitive data
 * @param pattern - the string for which to search
 * @param patternLength - the length of the string for which to search
 * @param target - the string in which to search
 * @param targetLength - the length of the string in which to search
 * @param status - will be set if any errors occur. 
 *
 * @return the <code>BMS</code> object.
 *
 * Note: if on return status is set to an error, the only safe
 * thing to do with the returned object is to call
 * <code>bms_close</code>.
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI BMS * U_EXPORT2
bms_open(UCD *ucd,
         const UChar *pattern, int32_t patternLength,
         const UChar *target,  int32_t targetLength,
         UErrorCode  *status);

/**
 * Close a <code>BMS</code> object and release all the
 * storage associated with it.
 *
 * @param bms - the <code>BMS</code> object to close.
 */
U_CAPI void U_EXPORT2
bms_close(BMS *bms);

/**
 * Test the pattern to see if it generates any CEs.
 *
 * @param bms - the <code>BMS</code> object
 * @return <code>TRUE</code> if the pattern string did not generate any CEs
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI UBool U_EXPORT2
bms_empty(BMS *bms);

/**
 * Get the <code>UCD</code> object used to create
 * a given <code>BMS</code> object.
 *
 * @param bms - the <code>BMS</code> object
 *
 * @return - the <code>UCD</code> object used to create
 *           the given <code>BMS</code> object.
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI UCD * U_EXPORT2
bms_getData(BMS *bms);

/**
 * Search for the pattern string in the target string.
 *
 * @param bms - the <code>BMS</code> object
 * @param offset - the offset in the target string at which to begin the search
 * @param start - will be set to the starting offset of the match, or -1 if there's no match
 * @param end - will be set to the ending offset of the match, or -1 if there's no match
 *
 * @return <code>TRUE</code> if the match succeeds, <code>FALSE</code> otherwise.
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI UBool U_EXPORT2
bms_search(BMS *bms, int32_t offset, int32_t *start, int32_t *end);

/**
 * Set the target string for the match.
 *
 * @param bms - the <code>BMS</code> object
 * @param target - the new target string
 * @param targetLength - the length of the new target string
 * @param status - will be set if any errors occur. 
 *
 * @internal ICU 4.0.1 technology preview
 */
U_CAPI void U_EXPORT2
bms_setTargetString(BMS *bms, const UChar *target, int32_t targetLength, UErrorCode *status);

#endif

#endif /* _BMS_H */

cmd: