Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(7)

Unified Diff: source/i18n/usearch.cpp

Issue 845603002: Update ICU to 54.1 step 1 (Closed) Base URL: https://chromium.googlesource.com/chromium/deps/icu.git@master
Patch Set: remove unusued directories Created 5 years, 11 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« no previous file with comments | « source/i18n/uregex.cpp ('k') | source/i18n/uspoof.cpp » ('j') | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: source/i18n/usearch.cpp
diff --git a/source/i18n/usearch.cpp b/source/i18n/usearch.cpp
index ba463e606c2011950f30154a7821f0d6cf93c20f..96414dbf4d90ab38be5ecad097101c45c79e913f 100644
--- a/source/i18n/usearch.cpp
+++ b/source/i18n/usearch.cpp
@@ -1,6 +1,6 @@
/*
**********************************************************************
-* Copyright (C) 2001-2011 IBM and others. All rights reserved.
+* Copyright (C) 2001-2014 IBM and others. All rights reserved.
**********************************************************************
* Date Name Description
* 07/02/2001 synwee Creation.
@@ -16,7 +16,6 @@
#include "unicode/uchar.h"
#include "unicode/utf16.h"
#include "normalizer2impl.h"
-#include "ucol_imp.h"
#include "usrchimp.h"
#include "cmemory.h"
#include "ucln_in.h"
@@ -29,8 +28,6 @@ U_NAMESPACE_USE
// (and if we decide to turn this on again there are several new TODOs that will need to be addressed)
#define BOYER_MOORE 0
-#define LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0]))
-
// internal definition ---------------------------------------------------
#define LAST_BYTE_MASK_ 0xFF
@@ -51,17 +48,10 @@ static
inline void setColEIterOffset(UCollationElements *elems,
int32_t offset)
{
- collIterate *ci = &(elems->iteratordata_);
- ci->pos = ci->string + offset;
- ci->CEpos = ci->toReturn = ci->extendCEs ? ci->extendCEs : ci->CEs;
- if (ci->flags & UCOL_ITER_INNORMBUF) {
- ci->flags = ci->origFlags;
- }
- ci->fcdPosition = NULL;
-
- ci->offsetReturn = NULL;
- ci->offsetStore = ci->offsetBuffer;
- ci->offsetRepeatCount = ci->offsetRepeatValue = 0;
+ // Note: Not "fast" any more after the 2013 collation rewrite.
+ // We do not want to expose more internals than necessary.
+ UErrorCode status = U_ZERO_ERROR;
+ ucol_setOffset(elems, offset, &status);
}
/**
@@ -298,7 +288,7 @@ inline uint16_t initializePatternCETable(UStringSearch *strsrch,
{
UPattern *pattern = &(strsrch->pattern);
uint32_t cetablesize = INITIAL_ARRAY_SIZE_;
- int32_t *cetable = pattern->CEBuffer;
+ int32_t *cetable = pattern->cesBuffer;
uint32_t patternlength = pattern->textLength;
UCollationElements *coleiter = strsrch->utilIter;
@@ -311,17 +301,14 @@ inline uint16_t initializePatternCETable(UStringSearch *strsrch,
strsrch->utilIter = coleiter;
}
else {
- uprv_init_collIterate(strsrch->collator, pattern->text,
- pattern->textLength,
- &coleiter->iteratordata_,
- status);
+ ucol_setText(coleiter, pattern->text, pattern->textLength, status);
}
if(U_FAILURE(*status)) {
return 0;
}
- if (pattern->CE != cetable && pattern->CE) {
- uprv_free(pattern->CE);
+ if (pattern->ces != cetable && pattern->ces) {
+ uprv_free(pattern->ces);
}
uint16_t offset = 0;
@@ -340,7 +327,7 @@ inline uint16_t initializePatternCETable(UStringSearch *strsrch,
return 0;
}
offset ++;
- if (cetable != temp && cetable != pattern->CEBuffer) {
+ if (cetable != temp && cetable != pattern->cesBuffer) {
uprv_free(cetable);
}
cetable = temp;
@@ -349,8 +336,8 @@ inline uint16_t initializePatternCETable(UStringSearch *strsrch,
}
cetable[offset] = 0;
- pattern->CE = cetable;
- pattern->CELength = offset;
+ pattern->ces = cetable;
+ pattern->cesLength = offset;
return result;
}
@@ -373,7 +360,7 @@ inline uint16_t initializePatternPCETable(UStringSearch *strsrch,
{
UPattern *pattern = &(strsrch->pattern);
uint32_t pcetablesize = INITIAL_ARRAY_SIZE_;
- int64_t *pcetable = pattern->PCEBuffer;
+ int64_t *pcetable = pattern->pcesBuffer;
uint32_t patternlength = pattern->textLength;
UCollationElements *coleiter = strsrch->utilIter;
@@ -385,29 +372,26 @@ inline uint16_t initializePatternPCETable(UStringSearch *strsrch,
// returned.
strsrch->utilIter = coleiter;
} else {
- uprv_init_collIterate(strsrch->collator, pattern->text,
- pattern->textLength,
- &coleiter->iteratordata_,
- status);
+ ucol_setText(coleiter, pattern->text, pattern->textLength, status);
}
if(U_FAILURE(*status)) {
return 0;
}
- if (pattern->PCE != pcetable && pattern->PCE != NULL) {
- uprv_free(pattern->PCE);
+ if (pattern->pces != pcetable && pattern->pces != NULL) {
+ uprv_free(pattern->pces);
}
uint16_t offset = 0;
uint16_t result = 0;
int64_t pce;
- uprv_init_pce(coleiter);
+ icu::UCollationPCE iter(coleiter);
// ** Should processed CEs be signed or unsigned?
// ** (the rest of the code in this file seems to play fast-and-loose with
// ** whether a CE is signed or unsigned. For example, look at routine above this one.)
- while ((pce = ucol_nextProcessed(coleiter, NULL, NULL, status)) != UCOL_PROCESSED_NULLORDER &&
+ while ((pce = iter.nextProcessed(NULL, NULL, status)) != UCOL_PROCESSED_NULLORDER &&
U_SUCCESS(*status)) {
int64_t *temp = addTouint64_tArray(pcetable, offset, &pcetablesize,
pce,
@@ -420,7 +404,7 @@ inline uint16_t initializePatternPCETable(UStringSearch *strsrch,
offset += 1;
- if (pcetable != temp && pcetable != pattern->PCEBuffer) {
+ if (pcetable != temp && pcetable != pattern->pcesBuffer) {
uprv_free(pcetable);
}
@@ -429,8 +413,8 @@ inline uint16_t initializePatternPCETable(UStringSearch *strsrch,
}
pcetable[offset] = 0;
- pattern->PCE = pcetable;
- pattern->PCELength = offset;
+ pattern->pces = pcetable;
+ pattern->pcesLength = offset;
return result;
}
@@ -446,6 +430,7 @@ inline uint16_t initializePatternPCETable(UStringSearch *strsrch,
static
inline int16_t initializePattern(UStringSearch *strsrch, UErrorCode *status)
{
+ if (U_FAILURE(*status)) { return 0; }
UPattern *pattern = &(strsrch->pattern);
const UChar *patterntext = pattern->text;
int32_t length = pattern->textLength;
@@ -465,12 +450,12 @@ inline int16_t initializePattern(UStringSearch *strsrch, UErrorCode *status)
}
// ** HACK **
- if (strsrch->pattern.PCE != NULL) {
- if (strsrch->pattern.PCE != strsrch->pattern.PCEBuffer) {
- uprv_free(strsrch->pattern.PCE);
+ if (strsrch->pattern.pces != NULL) {
+ if (strsrch->pattern.pces != strsrch->pattern.pcesBuffer) {
+ uprv_free(strsrch->pattern.pces);
}
- strsrch->pattern.PCE = NULL;
+ strsrch->pattern.pces = NULL;
}
// since intializePattern is an internal method status is a success.
@@ -557,14 +542,14 @@ static
inline void initialize(UStringSearch *strsrch, UErrorCode *status)
{
int16_t expandlength = initializePattern(strsrch, status);
- if (U_SUCCESS(*status) && strsrch->pattern.CELength > 0) {
+ if (U_SUCCESS(*status) && strsrch->pattern.cesLength > 0) {
UPattern *pattern = &strsrch->pattern;
- int32_t cesize = pattern->CELength;
+ int32_t cesize = pattern->cesLength;
int16_t minlength = cesize > expandlength
? (int16_t)cesize - expandlength : 1;
pattern->defaultShiftSize = minlength;
- setShiftTable(pattern->shift, pattern->backShift, pattern->CE,
+ setShiftTable(pattern->shift, pattern->backShift, pattern->ces,
cesize, expandlength, minlength, minlength);
return;
}
@@ -640,14 +625,14 @@ UBool isBreakUnit(const UStringSearch *strsrch, int32_t start,
start;
UErrorCode status = U_ZERO_ERROR;
ucol_setText(coleiter, text, end - start, &status);
- for (int32_t count = 0; count < strsrch->pattern.CELength;
+ for (int32_t count = 0; count < strsrch->pattern.cesLength;
count ++) {
int32_t ce = getCE(strsrch, ucol_next(coleiter, &status));
if (ce == UCOL_IGNORABLE) {
count --;
continue;
}
- if (U_FAILURE(status) || ce != strsrch->pattern.CE[count]) {
+ if (U_FAILURE(status) || ce != strsrch->pattern.ces[count]) {
return FALSE;
}
}
@@ -748,7 +733,7 @@ inline int32_t shiftForward(UStringSearch *strsrch,
int32_t shift = pattern->shift[hash(ce)];
// this is to adjust for characters in the middle of the
// substring for matching that failed.
- int32_t adjust = pattern->CELength - patternceindex;
+ int32_t adjust = pattern->cesLength - patternceindex;
if (adjust > 1 && shift >= adjust) {
shift -= adjust - 1;
}
@@ -882,7 +867,7 @@ UBool checkExtraMatchAccents(const UStringSearch *strsrch, int32_t start,
UCollationElements *coleiter = strsrch->utilIter;
ucol_setText(coleiter, norm, size, status);
- uint32_t firstce = strsrch->pattern.CE[0];
+ uint32_t firstce = strsrch->pattern.ces[0];
UBool ignorable = TRUE;
uint32_t ce = UCOL_IGNORABLE;
while (U_SUCCESS(*status) && ce != firstce && ce != (uint32_t)UCOL_NULLORDER) {
@@ -935,7 +920,7 @@ UBool hasAccentsBeforeMatch(const UStringSearch *strsrch, int32_t start,
UErrorCode status = U_ZERO_ERROR;
// we have been iterating forwards previously
uint32_t ignorable = TRUE;
- int32_t firstce = strsrch->pattern.CE[0];
+ int32_t firstce = strsrch->pattern.ces[0];
setColEIterOffset(coleiter, start);
int32_t ce = getCE(strsrch, ucol_next(coleiter, &status));
@@ -1017,7 +1002,7 @@ UBool hasAccentsAfterMatch(const UStringSearch *strsrch, int32_t start,
int32_t textlength = strsrch->search->textLength;
U16_BACK_1(text, 0, temp);
if (getFCD(text, &temp, textlength) & LAST_BYTE_MASK_) {
- int32_t firstce = strsrch->pattern.CE[0];
+ int32_t firstce = strsrch->pattern.ces[0];
UCollationElements *coleiter = strsrch->textIter;
UErrorCode status = U_ZERO_ERROR;
int32_t ce;
@@ -1028,7 +1013,7 @@ UBool hasAccentsAfterMatch(const UStringSearch *strsrch, int32_t start,
}
}
int32_t count = 1;
- while (count < strsrch->pattern.CELength) {
+ while (count < strsrch->pattern.cesLength) {
if (getCE(strsrch, ucol_next(coleiter, &status))
== UCOL_IGNORABLE) {
// Thai can give an ignorable here.
@@ -1212,8 +1197,8 @@ UBool checkNextExactContractionMatch(UStringSearch *strsrch,
expansion --;
}
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t count = 0;
while (count < patterncelength) {
int32_t ce = getCE(strsrch, ucol_next(coleiter, status));
@@ -1415,8 +1400,8 @@ static
inline UBool checkCollationMatch(const UStringSearch *strsrch,
UCollationElements *coleiter)
{
- int patternceindex = strsrch->pattern.CELength;
- int32_t *patternce = strsrch->pattern.CE;
+ int patternceindex = strsrch->pattern.cesLength;
+ int32_t *patternce = strsrch->pattern.ces;
UErrorCode status = U_ZERO_ERROR;
while (patternceindex > 0) {
int32_t ce = getCE(strsrch, ucol_next(coleiter, &status));
@@ -1615,8 +1600,8 @@ int32_t doNextCanonicalSuffixMatch(UStringSearch *strsrch,
ucol_setText(coleiter, safetext, safetextlength, status);
// status checked in loop below
- int32_t *ce = strsrch->pattern.CE;
- int32_t celength = strsrch->pattern.CELength;
+ int32_t *ce = strsrch->pattern.ces;
+ int32_t celength = strsrch->pattern.cesLength;
int ceindex = celength - 1;
UBool isSafe = TRUE; // indication flag for position in safe zone
@@ -1855,8 +1840,8 @@ UBool checkNextCanonicalContractionMatch(UStringSearch *strsrch,
expansion --;
}
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t count = 0;
int32_t textlength = strsrch->search->textLength;
while (count < patterncelength) {
@@ -2053,8 +2038,8 @@ UBool checkPreviousExactContractionMatch(UStringSearch *strsrch,
expansion --;
}
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t count = patterncelength;
while (count > 0) {
int32_t ce = getCE(strsrch, ucol_previous(coleiter, status));
@@ -2278,8 +2263,8 @@ int32_t doPreviousCanonicalPrefixMatch(UStringSearch *strsrch,
ucol_setText(coleiter, safetext, safetextlength, status);
// status checked in loop below
- int32_t *ce = strsrch->pattern.CE;
- int32_t celength = strsrch->pattern.CELength;
+ int32_t *ce = strsrch->pattern.ces;
+ int32_t celength = strsrch->pattern.cesLength;
int ceindex = 0;
UBool isSafe = TRUE; // safe zone indication flag for position
int32_t prefixlength = u_strlen(strsrch->canonicalPrefixAccents);
@@ -2493,8 +2478,8 @@ UBool checkPreviousCanonicalContractionMatch(UStringSearch *strsrch,
expansion --;
}
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t count = patterncelength;
while (count > 0) {
int32_t ce = getCE(strsrch, ucol_previous(coleiter, status));
@@ -2700,7 +2685,7 @@ U_CAPI UStringSearch * U_EXPORT2 usearch_openFromCollator(
UCOL_SHIFTED;
result->variableTop = ucol_getVariableTop(collator, status);
- result->nfd = Normalizer2Factory::getNFDInstance(*status);
+ result->nfd = Normalizer2::getNFDInstance(*status);
if (U_FAILURE(*status)) {
uprv_free(result);
@@ -2719,8 +2704,8 @@ U_CAPI UStringSearch * U_EXPORT2 usearch_openFromCollator(
result->pattern.text = pattern;
result->pattern.textLength = patternlength;
- result->pattern.CE = NULL;
- result->pattern.PCE = NULL;
+ result->pattern.ces = NULL;
+ result->pattern.pces = NULL;
result->search->breakIter = breakiter;
#if !UCONFIG_NO_BREAK_ITERATION
@@ -2736,6 +2721,7 @@ U_CAPI UStringSearch * U_EXPORT2 usearch_openFromCollator(
result->utilIter = NULL;
result->textIter = ucol_openElements(collator, text,
textlength, status);
+ result->textProcessedIter = NULL;
if (U_FAILURE(*status)) {
usearch_close(result);
return NULL;
@@ -2762,16 +2748,17 @@ U_CAPI UStringSearch * U_EXPORT2 usearch_openFromCollator(
U_CAPI void U_EXPORT2 usearch_close(UStringSearch *strsrch)
{
if (strsrch) {
- if (strsrch->pattern.CE != strsrch->pattern.CEBuffer &&
- strsrch->pattern.CE) {
- uprv_free(strsrch->pattern.CE);
+ if (strsrch->pattern.ces != strsrch->pattern.cesBuffer &&
+ strsrch->pattern.ces) {
+ uprv_free(strsrch->pattern.ces);
}
- if (strsrch->pattern.PCE != NULL &&
- strsrch->pattern.PCE != strsrch->pattern.PCEBuffer) {
- uprv_free(strsrch->pattern.PCE);
+ if (strsrch->pattern.pces != NULL &&
+ strsrch->pattern.pces != strsrch->pattern.pcesBuffer) {
+ uprv_free(strsrch->pattern.pces);
}
+ delete strsrch->textProcessedIter;
ucol_closeElements(strsrch->textIter);
ucol_closeElements(strsrch->utilIter);
@@ -2790,6 +2777,24 @@ U_CAPI void U_EXPORT2 usearch_close(UStringSearch *strsrch)
}
}
+namespace {
+
+UBool initTextProcessedIter(UStringSearch *strsrch, UErrorCode *status) {
+ if (U_FAILURE(*status)) { return FALSE; }
+ if (strsrch->textProcessedIter == NULL) {
+ strsrch->textProcessedIter = new icu::UCollationPCE(strsrch->textIter);
+ if (strsrch->textProcessedIter == NULL) {
+ *status = U_MEMORY_ALLOCATION_ERROR;
+ return FALSE;
+ }
+ } else {
+ strsrch->textProcessedIter->init(strsrch->textIter);
+ }
+ return TRUE;
+}
+
+}
+
// set and get methods --------------------------------------------------
U_CAPI void U_EXPORT2 usearch_setOffset(UStringSearch *strsrch,
@@ -3010,6 +3015,11 @@ U_CAPI void U_EXPORT2 usearch_setCollator( UStringSearch *strsrch,
}
if (strsrch) {
+ delete strsrch->textProcessedIter;
+ strsrch->textProcessedIter = NULL;
+ ucol_closeElements(strsrch->textIter);
+ ucol_closeElements(strsrch->utilIter);
+ strsrch->textIter = strsrch->utilIter = NULL;
if (strsrch->ownCollator && (strsrch->collator != collator)) {
ucol_close((UCollator *)strsrch->collator);
strsrch->ownCollator = FALSE;
@@ -3028,23 +3038,19 @@ U_CAPI void U_EXPORT2 usearch_setCollator( UStringSearch *strsrch,
UCOL_SHIFTED;
// if status is a failure, ucol_getVariableTop returns 0
strsrch->variableTop = ucol_getVariableTop(collator, status);
- if (U_SUCCESS(*status)) {
- initialize(strsrch, status);
- if (U_SUCCESS(*status)) {
- /* free offset buffer to avoid memory leak before initializing. */
- ucol_freeOffsetBuffer(&(strsrch->textIter->iteratordata_));
- uprv_init_collIterate(collator, strsrch->search->text,
- strsrch->search->textLength,
- &(strsrch->textIter->iteratordata_),
- status);
- strsrch->utilIter->iteratordata_.coll = collator;
- }
- }
+ strsrch->textIter = ucol_openElements(collator,
+ strsrch->search->text,
+ strsrch->search->textLength,
+ status);
+ strsrch->utilIter = ucol_openElements(
+ collator, strsrch->pattern.text, strsrch->pattern.textLength, status);
+ // initialize() _after_ setting the iterators for the new collator.
+ initialize(strsrch, status);
}
// **** are these calls needed?
// **** we call uprv_init_pce in initializePatternPCETable
- // **** and the CEBuffer constructor...
+ // **** and the CEIBuffer constructor...
#if 0
uprv_init_pce(strsrch->textIter);
uprv_init_pce(strsrch->utilIter);
@@ -3222,7 +3228,7 @@ U_CAPI int32_t U_EXPORT2 usearch_next(UStringSearch *strsrch,
}
if (U_SUCCESS(*status)) {
- if (strsrch->pattern.CELength == 0) {
+ if (strsrch->pattern.cesLength == 0) {
if (search->matchedIndex == USEARCH_DONE) {
search->matchedIndex = offset;
}
@@ -3333,7 +3339,7 @@ U_CAPI int32_t U_EXPORT2 usearch_previous(UStringSearch *strsrch,
}
if (U_SUCCESS(*status)) {
- if (strsrch->pattern.CELength == 0) {
+ if (strsrch->pattern.cesLength == 0) {
search->matchedIndex =
(matchedindex == USEARCH_DONE ? offset : matchedindex);
if (search->matchedIndex == 0) {
@@ -3416,11 +3422,8 @@ U_CAPI void U_EXPORT2 usearch_reset(UStringSearch *strsrch)
if (!sameCollAttribute) {
initialize(strsrch, &status);
}
- /* free offset buffer to avoid memory leak before initializing. */
- ucol_freeOffsetBuffer(&(strsrch->textIter->iteratordata_));
- uprv_init_collIterate(strsrch->collator, strsrch->search->text,
+ ucol_setText(strsrch->textIter, strsrch->search->text,
strsrch->search->textLength,
- &(strsrch->textIter->iteratordata_),
&status);
strsrch->search->matchedLength = 0;
strsrch->search->matchedIndex = USEARCH_DONE;
@@ -3444,9 +3447,9 @@ struct CEI {
U_NAMESPACE_BEGIN
-
+namespace {
//
-// CEBuffer A circular buffer of CEs from the text being searched.
+// CEIBuffer A circular buffer of CEs-with-index from the text being searched.
//
#define DEFAULT_CEBUFFER_SIZE 96
#define CEBUFFER_EXTRA 32
@@ -3455,7 +3458,7 @@ U_NAMESPACE_BEGIN
#define MAX_TARGET_IGNORABLES_PER_PAT_JAMO_L 8
#define MAX_TARGET_IGNORABLES_PER_PAT_OTHER 3
#define MIGHT_BE_JAMO_L(c) ((c >= 0x1100 && c <= 0x115E) || (c >= 0x3131 && c <= 0x314E) || (c >= 0x3165 && c <= 0x3186))
-struct CEBuffer {
+struct CEIBuffer {
CEI defBuf[DEFAULT_CEBUFFER_SIZE];
CEI *buf;
int32_t bufSize;
@@ -3466,17 +3469,17 @@ struct CEBuffer {
- CEBuffer(UStringSearch *ss, UErrorCode *status);
- ~CEBuffer();
+ CEIBuffer(UStringSearch *ss, UErrorCode *status);
+ ~CEIBuffer();
const CEI *get(int32_t index);
const CEI *getPrevious(int32_t index);
};
-CEBuffer::CEBuffer(UStringSearch *ss, UErrorCode *status) {
+CEIBuffer::CEIBuffer(UStringSearch *ss, UErrorCode *status) {
buf = defBuf;
strSearch = ss;
- bufSize = ss->pattern.PCELength + CEBUFFER_EXTRA;
+ bufSize = ss->pattern.pcesLength + CEBUFFER_EXTRA;
if (ss->search->elementComparisonType != 0) {
const UChar * patText = ss->pattern.text;
if (patText) {
@@ -3496,7 +3499,7 @@ CEBuffer::CEBuffer(UStringSearch *ss, UErrorCode *status) {
firstIx = 0;
limitIx = 0;
- uprv_init_pce(ceIter);
+ if (!initTextProcessedIter(ss, status)) { return; }
if (bufSize>DEFAULT_CEBUFFER_SIZE) {
buf = (CEI *)uprv_malloc(bufSize * sizeof(CEI));
@@ -3509,7 +3512,7 @@ CEBuffer::CEBuffer(UStringSearch *ss, UErrorCode *status) {
// TODO: add a reset or init function so that allocated
// buffers can be retained & reused.
-CEBuffer::~CEBuffer() {
+CEIBuffer::~CEIBuffer() {
if (buf != defBuf) {
uprv_free(buf);
}
@@ -3522,7 +3525,7 @@ CEBuffer::~CEBuffer() {
// where n is the largest index to have been fetched by some previous call to this function.
// The CE value will be UCOL__PROCESSED_NULLORDER at end of input.
//
-const CEI *CEBuffer::get(int32_t index) {
+const CEI *CEIBuffer::get(int32_t index) {
int i = index % bufSize;
if (index>=firstIx && index<limitIx) {
@@ -3550,7 +3553,7 @@ const CEI *CEBuffer::get(int32_t index) {
UErrorCode status = U_ZERO_ERROR;
- buf[i].ce = ucol_nextProcessed(ceIter, &buf[i].lowIndex, &buf[i].highIndex, &status);
+ buf[i].ce = strSearch->textProcessedIter->nextProcessed(&buf[i].lowIndex, &buf[i].highIndex, &status);
return &buf[i];
}
@@ -3561,7 +3564,7 @@ const CEI *CEBuffer::get(int32_t index) {
// where n is the largest index to have been fetched by some previous call to this function.
// The CE value will be UCOL__PROCESSED_NULLORDER at end of input.
//
-const CEI *CEBuffer::getPrevious(int32_t index) {
+const CEI *CEIBuffer::getPrevious(int32_t index) {
int i = index % bufSize;
if (index>=firstIx && index<limitIx) {
@@ -3589,11 +3592,13 @@ const CEI *CEBuffer::getPrevious(int32_t index) {
UErrorCode status = U_ZERO_ERROR;
- buf[i].ce = ucol_previousProcessed(ceIter, &buf[i].lowIndex, &buf[i].highIndex, &status);
+ buf[i].ce = strSearch->textProcessedIter->previousProcessed(&buf[i].lowIndex, &buf[i].highIndex, &status);
return &buf[i];
}
+}
+
U_NAMESPACE_END
@@ -3815,8 +3820,8 @@ U_CAPI UBool U_EXPORT2 usearch_search(UStringSearch *strsrch,
#ifdef USEARCH_DEBUG
if (getenv("USEARCH_DEBUG") != NULL) {
printf("Pattern CEs\n");
- for (int ii=0; ii<strsrch->pattern.CELength; ii++) {
- printf(" %8x", strsrch->pattern.CE[ii]);
+ for (int ii=0; ii<strsrch->pattern.cesLength; ii++) {
+ printf(" %8x", strsrch->pattern.ces[ii]);
}
printf("\n");
}
@@ -3825,20 +3830,20 @@ U_CAPI UBool U_EXPORT2 usearch_search(UStringSearch *strsrch,
// Input parameter sanity check.
// TODO: should input indicies clip to the text length
// in the same way that UText does.
- if(strsrch->pattern.CELength == 0 ||
+ if(strsrch->pattern.cesLength == 0 ||
startIdx < 0 ||
startIdx > strsrch->search->textLength ||
- strsrch->pattern.CE == NULL) {
+ strsrch->pattern.ces == NULL) {
*status = U_ILLEGAL_ARGUMENT_ERROR;
return FALSE;
}
- if (strsrch->pattern.PCE == NULL) {
+ if (strsrch->pattern.pces == NULL) {
initializePatternPCETable(strsrch, status);
}
ucol_setOffset(strsrch->textIter, startIdx, status);
- CEBuffer ceb(strsrch, status);
+ CEIBuffer ceb(strsrch, status);
int32_t targetIx = 0;
@@ -3884,8 +3889,8 @@ U_CAPI UBool U_EXPORT2 usearch_search(UStringSearch *strsrch,
break;
}
- for (patIx=0; patIx<strsrch->pattern.PCELength; patIx++) {
- patCE = strsrch->pattern.PCE[patIx];
+ for (patIx=0; patIx<strsrch->pattern.pcesLength; patIx++) {
+ patCE = strsrch->pattern.pces[patIx];
targetCEI = ceb.get(targetIx+patIx+targetIxOffset);
// Compare CE from target string with CE from the pattern.
// Note that the target CE will be UCOL_PROCESSED_NULLORDER if we reach the end of input,
@@ -3905,7 +3910,7 @@ U_CAPI UBool U_EXPORT2 usearch_search(UStringSearch *strsrch,
}
}
}
- targetIxOffset += strsrch->pattern.PCELength; // this is now the offset in target CE space to end of the match so far
+ targetIxOffset += strsrch->pattern.pcesLength; // this is now the offset in target CE space to end of the match so far
if (!found && ((targetCEI == NULL) || (targetCEI->ce != UCOL_PROCESSED_NULLORDER))) {
// No match at this targetIx. Try again at the next.
@@ -4082,8 +4087,8 @@ U_CAPI UBool U_EXPORT2 usearch_searchBackwards(UStringSearch *strsrch,
#ifdef USEARCH_DEBUG
if (getenv("USEARCH_DEBUG") != NULL) {
printf("Pattern CEs\n");
- for (int ii=0; ii<strsrch->pattern.CELength; ii++) {
- printf(" %8x", strsrch->pattern.CE[ii]);
+ for (int ii=0; ii<strsrch->pattern.cesLength; ii++) {
+ printf(" %8x", strsrch->pattern.ces[ii]);
}
printf("\n");
}
@@ -4092,19 +4097,19 @@ U_CAPI UBool U_EXPORT2 usearch_searchBackwards(UStringSearch *strsrch,
// Input parameter sanity check.
// TODO: should input indicies clip to the text length
// in the same way that UText does.
- if(strsrch->pattern.CELength == 0 ||
+ if(strsrch->pattern.cesLength == 0 ||
startIdx < 0 ||
startIdx > strsrch->search->textLength ||
- strsrch->pattern.CE == NULL) {
+ strsrch->pattern.ces == NULL) {
*status = U_ILLEGAL_ARGUMENT_ERROR;
return FALSE;
}
- if (strsrch->pattern.PCE == NULL) {
+ if (strsrch->pattern.pces == NULL) {
initializePatternPCETable(strsrch, status);
}
- CEBuffer ceb(strsrch, status);
+ CEIBuffer ceb(strsrch, status);
int32_t targetIx = 0;
/*
@@ -4165,10 +4170,10 @@ U_CAPI UBool U_EXPORT2 usearch_searchBackwards(UStringSearch *strsrch,
// Inner loop checks for a match beginning at each
// position from the outer loop.
int32_t targetIxOffset = 0;
- for (patIx = strsrch->pattern.PCELength - 1; patIx >= 0; patIx -= 1) {
- int64_t patCE = strsrch->pattern.PCE[patIx];
+ for (patIx = strsrch->pattern.pcesLength - 1; patIx >= 0; patIx -= 1) {
+ int64_t patCE = strsrch->pattern.pces[patIx];
- targetCEI = ceb.getPrevious(targetIx + strsrch->pattern.PCELength - 1 - patIx + targetIxOffset);
+ targetCEI = ceb.getPrevious(targetIx + strsrch->pattern.pcesLength - 1 - patIx + targetIxOffset);
// Compare CE from target string with CE from the pattern.
// Note that the target CE will be UCOL_NULLORDER if we reach the end of input,
// which will fail the compare, below.
@@ -4204,7 +4209,7 @@ U_CAPI UBool U_EXPORT2 usearch_searchBackwards(UStringSearch *strsrch,
// There still is a chance of match failure if the CE range not correspond to
// an acceptable character range.
//
- const CEI *firstCEI = ceb.getPrevious(targetIx + strsrch->pattern.PCELength - 1 + targetIxOffset);
+ const CEI *firstCEI = ceb.getPrevious(targetIx + strsrch->pattern.pcesLength - 1 + targetIxOffset);
mStart = firstCEI->lowIndex;
// Check for the start of the match being within a combining sequence.
@@ -4330,8 +4335,8 @@ UBool usearch_handleNextExact(UStringSearch *strsrch, UErrorCode *status)
#if BOYER_MOORE
UCollationElements *coleiter = strsrch->textIter;
int32_t textlength = strsrch->search->textLength;
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t textoffset = ucol_getOffset(coleiter);
// status used in setting coleiter offset, since offset is checked in
@@ -4444,8 +4449,8 @@ UBool usearch_handleNextCanonical(UStringSearch *strsrch, UErrorCode *status)
#if BOYER_MOORE
UCollationElements *coleiter = strsrch->textIter;
int32_t textlength = strsrch->search->textLength;
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t textoffset = ucol_getOffset(coleiter);
UBool hasPatternAccents =
strsrch->pattern.hasSuffixAccents || strsrch->pattern.hasPrefixAccents;
@@ -4558,8 +4563,8 @@ UBool usearch_handlePreviousExact(UStringSearch *strsrch, UErrorCode *status)
#if BOYER_MOORE
UCollationElements *coleiter = strsrch->textIter;
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t textoffset = ucol_getOffset(coleiter);
// shifting it check for setting offset
@@ -4659,8 +4664,12 @@ UBool usearch_handlePreviousExact(UStringSearch *strsrch, UErrorCode *status)
} else {
// move the start position at the end of possible match
initializePatternPCETable(strsrch, status);
- for (int32_t nPCEs = 0; nPCEs < strsrch->pattern.PCELength - 1; nPCEs++) {
- int64_t pce = ucol_nextProcessed(strsrch->textIter, NULL, NULL, status);
+ if (!initTextProcessedIter(strsrch, status)) {
+ setMatchNotFound(strsrch);
+ return FALSE;
+ }
+ for (int32_t nPCEs = 0; nPCEs < strsrch->pattern.pcesLength - 1; nPCEs++) {
+ int64_t pce = strsrch->textProcessedIter->nextProcessed(NULL, NULL, status);
if (pce == UCOL_PROCESSED_NULLORDER) {
// at the end of the text
break;
@@ -4700,8 +4709,8 @@ UBool usearch_handlePreviousCanonical(UStringSearch *strsrch,
#if BOYER_MOORE
UCollationElements *coleiter = strsrch->textIter;
- int32_t *patternce = strsrch->pattern.CE;
- int32_t patterncelength = strsrch->pattern.CELength;
+ int32_t *patternce = strsrch->pattern.ces;
+ int32_t patterncelength = strsrch->pattern.cesLength;
int32_t textoffset = ucol_getOffset(coleiter);
UBool hasPatternAccents =
strsrch->pattern.hasSuffixAccents || strsrch->pattern.hasPrefixAccents;
@@ -4808,8 +4817,12 @@ UBool usearch_handlePreviousCanonical(UStringSearch *strsrch,
} else {
// move the start position at the end of possible match
initializePatternPCETable(strsrch, status);
- for (int32_t nPCEs = 0; nPCEs < strsrch->pattern.PCELength - 1; nPCEs++) {
- int64_t pce = ucol_nextProcessed(strsrch->textIter, NULL, NULL, status);
+ if (!initTextProcessedIter(strsrch, status)) {
+ setMatchNotFound(strsrch);
+ return FALSE;
+ }
+ for (int32_t nPCEs = 0; nPCEs < strsrch->pattern.pcesLength - 1; nPCEs++) {
+ int64_t pce = strsrch->textProcessedIter->nextProcessed(NULL, NULL, status);
if (pce == UCOL_PROCESSED_NULLORDER) {
// at the end of the text
break;
« no previous file with comments | « source/i18n/uregex.cpp ('k') | source/i18n/uspoof.cpp » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698