/*
*******************************************************************************
-* Copyright (C) 1997-2011, International Business Machines Corporation and
+* Copyright (C) 1997-2016, International Business Machines Corporation and
* others. All Rights Reserved.
*******************************************************************************
*
-* File TXTBDRY.CPP
+* File brkiter.cpp
*
* Modification History:
*
#include "unicode/udata.h"
#include "unicode/ures.h"
#include "unicode/ustring.h"
+#include "unicode/filteredbrk.h"
#include "ucln_cmn.h"
#include "cstring.h"
#include "umutex.h"
#include "uresimp.h"
#include "uassert.h"
#include "ubrkimpl.h"
+#include "charstr.h"
+#include "unicode/filteredbrk.h"
// *****************************************************************************
// class BreakIterator
{
char fnbuff[256];
char ext[4]={'\0'};
- char actualLocale[ULOC_FULLNAME_CAPACITY];
+ CharString actualLocale;
int32_t size;
const UChar* brkfname = NULL;
UResourceBundle brkRulesStack;
ures_initStackObject(brkName);
// Get the locale
- UResourceBundle *b = ures_open(U_ICUDATA_BRKITR, loc.getName(), &status);
- /* this is a hack for now. Should be fixed when the data is fetched from
- brk_index.txt */
- if(status==U_USING_DEFAULT_WARNING){
- status=U_ZERO_ERROR;
- ures_openFillIn(b, U_ICUDATA_BRKITR, "", &status);
- }
+ UResourceBundle *b = ures_openNoDefault(U_ICUDATA_BRKITR, loc.getName(), &status);
// Get the "boundaries" array.
if (U_SUCCESS(status)) {
// Use the string if we found it
if (U_SUCCESS(status) && brkfname) {
- uprv_strncpy(actualLocale,
- ures_getLocaleInternal(brkName, &status),
- sizeof(actualLocale)/sizeof(actualLocale[0]));
+ actualLocale.append(ures_getLocaleInternal(brkName, &status), -1, status);
UChar* extStart=u_strchr(brkfname, 0x002e);
int len = 0;
// If there is a result, set the valid locale and actual locale, and the kind
if (U_SUCCESS(status) && result != NULL) {
U_LOCALE_BASED(locBased, *(BreakIterator*)result);
- locBased.setLocaleIDs(ures_getLocaleByType(b, ULOC_VALID_LOCALE, &status), actualLocale);
+ locBased.setLocaleIDs(ures_getLocaleByType(b, ULOC_VALID_LOCALE, &status),
+ actualLocale.data());
result->setBreakType(kind);
}
//-------------------------------------------
BreakIterator::BreakIterator()
+: fKeepAll(FALSE)
{
- fBufferClone = FALSE;
*validLocale = *actualLocale = 0;
}
// -------------------------------------
-U_NAMESPACE_END
-
// defined in ucln_cmn.h
+U_NAMESPACE_END
+static icu::UInitOnce gInitOnce;
static icu::ICULocaleService* gService = NULL;
+
+
/**
* Release all static memory held by breakiterator.
*/
delete gService;
gService = NULL;
}
+ gInitOnce.reset();
#endif
return TRUE;
}
U_CDECL_END
U_NAMESPACE_BEGIN
+static void U_CALLCONV
+initService(void) {
+ gService = new ICUBreakIteratorService();
+ ucln_common_registerCleanup(UCLN_COMMON_BREAKITERATOR, breakiterator_cleanup);
+}
+
static ICULocaleService*
getService(void)
{
- UBool needsInit;
- UMTX_CHECK(NULL, (UBool)(gService == NULL), needsInit);
-
- if (needsInit) {
- ICULocaleService *tService = new ICUBreakIteratorService();
- umtx_lock(NULL);
- if (gService == NULL) {
- gService = tService;
- tService = NULL;
- ucln_common_registerCleanup(UCLN_COMMON_BREAKITERATOR, breakiterator_cleanup);
- }
- umtx_unlock(NULL);
- delete tService;
- }
+ umtx_initOnce(gInitOnce, &initService);
return gService;
}
+
// -------------------------------------
static inline UBool
hasService(void)
{
- UBool retVal;
- UMTX_CHECK(NULL, gService != NULL, retVal);
- return retVal;
+ return !gInitOnce.isReset() && getService() != NULL;
}
// -------------------------------------
}
// -------------------------------------
+enum { kKeyValueLenMax = 32 };
BreakIterator*
BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)
if (U_FAILURE(status)) {
return NULL;
}
+ char lbType[kKeyValueLenMax];
BreakIterator *result = NULL;
switch (kind) {
result = BreakIterator::buildInstance(loc, "word", kind, status);
break;
case UBRK_LINE:
- result = BreakIterator::buildInstance(loc, "line", kind, status);
+ uprv_strcpy(lbType, "line");
+ {
+ char lbKeyValue[kKeyValueLenMax] = {0};
+ UErrorCode kvStatus = U_ZERO_ERROR;
+ int32_t kLen = loc.getKeywordValue("lb", lbKeyValue, kKeyValueLenMax, kvStatus);
+ if (U_SUCCESS(kvStatus) && kLen > 0 && (uprv_strcmp(lbKeyValue,"strict")==0 || uprv_strcmp(lbKeyValue,"normal")==0 || uprv_strcmp(lbKeyValue,"loose")==0)) {
+ uprv_strcat(lbType, "_");
+ uprv_strcat(lbType, lbKeyValue);
+ }
+ }
+ result = BreakIterator::buildInstance(loc, lbType, kind, status);
+ if (U_SUCCESS(status) && result != NULL) {
+ char lwKeyValue[kKeyValueLenMax] = {0};
+ UErrorCode kvStatus = U_ZERO_ERROR;
+ int32_t kLen = loc.getKeywordValue("lw", lwKeyValue, kKeyValueLenMax, kvStatus);
+ result->setKeepAll(U_SUCCESS(kvStatus) && kLen > 0 && uprv_strcmp(lwKeyValue,"keepall")==0);
+ }
break;
case UBRK_SENTENCE:
result = BreakIterator::buildInstance(loc, "sentence", kind, status);
+ {
+ char ssKeyValue[kKeyValueLenMax] = {0};
+ UErrorCode kvStatus = U_ZERO_ERROR;
+ int32_t kLen = loc.getKeywordValue("ss", ssKeyValue, kKeyValueLenMax, kvStatus);
+ if (U_SUCCESS(kvStatus) && kLen > 0 && uprv_strcmp(ssKeyValue,"standard")==0) {
+ FilteredBreakIteratorBuilder* fbiBuilder = FilteredBreakIteratorBuilder::createInstance(loc, kvStatus);
+ if (U_SUCCESS(kvStatus)) {
+ result = fbiBuilder->build(result, status);
+ delete fbiBuilder;
+ }
+ }
+ }
break;
case UBRK_TITLE:
result = BreakIterator::buildInstance(loc, "title", kind, status);
return locBased.getLocaleID(type, status);
}
+
+// This implementation of getRuleStatus is a do-nothing stub, here to
+// provide a default implementation for any derived BreakIterator classes that
+// do not implement it themselves.
+int32_t BreakIterator::getRuleStatus() const {
+ return 0;
+}
+
+// This implementation of getRuleStatusVec is a do-nothing stub, here to
+// provide a default implementation for any derived BreakIterator classes that
+// do not implement it themselves.
+int32_t BreakIterator::getRuleStatusVec(int32_t *fillInVec, int32_t capacity, UErrorCode &status) {
+ if (U_FAILURE(status)) {
+ return 0;
+ }
+ if (capacity < 1) {
+ status = U_BUFFER_OVERFLOW_ERROR;
+ return 1;
+ }
+ *fillInVec = 0;
+ return 1;
+}
+
+BreakIterator::BreakIterator (const Locale& valid, const Locale& actual) {
+ U_LOCALE_BASED(locBased, (*this));
+ locBased.setLocaleIDs(valid, actual);
+}
+
U_NAMESPACE_END
#endif /* #if !UCONFIG_NO_BREAK_ITERATION */