ICU-57132.0.1.tar.gz

[apple/icu.git] / icuSources / common / brkiter.cpp

diff --git a/icuSources/common/brkiter.cpp b/icuSources/common/brkiter.cpp
index 5931ebf7fbd70fe796dbe1851edf5121f6722279..065e6df9556d5fc84bab0a0c4b609424336bfe13 100644 (file)
--- a/icuSources/common/brkiter.cpp
+++ b/icuSources/common/brkiter.cpp
@@ -1,6 +1,6 @@
 /*
 *******************************************************************************
 /*
 *******************************************************************************

-* Copyright (C) 1997-2014, International Business Machines Corporation and
+* Copyright (C) 1997-2016, International Business Machines Corporation and

 * others. All Rights Reserved.
 *******************************************************************************
 *
 * others. All Rights Reserved.
 *******************************************************************************
 *


@@ -27,6 +27,7 @@
 #include "unicode/udata.h"
 #include "unicode/ures.h"
 #include "unicode/ustring.h"
 #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 "ucln_cmn.h"
 #include "cstring.h"
 #include "umutex.h"


@@ -35,6 +36,8 @@
 #include "uresimp.h"
 #include "uassert.h"
 #include "ubrkimpl.h"
 #include "uresimp.h"
 #include "uassert.h"
 #include "ubrkimpl.h"

+#include "charstr.h"
+#include "unicode/filteredbrk.h"

 
 // *****************************************************************************
 // class BreakIterator
 
 // *****************************************************************************
 // class BreakIterator


@@ -52,7 +55,7 @@ BreakIterator::buildInstance(const Locale& loc, const char *type, int32_t kind,
 {
     char fnbuff[256];
     char ext[4]={'\0'};
 {
     char fnbuff[256];
     char ext[4]={'\0'};

-    char actualLocale[ULOC_FULLNAME_CAPACITY];
+    CharString actualLocale;

     int32_t size;
     const UChar* brkfname = NULL;
     UResourceBundle brkRulesStack;
     int32_t size;
     const UChar* brkfname = NULL;
     UResourceBundle brkRulesStack;


@@ -68,13 +71,7 @@ BreakIterator::buildInstance(const Locale& loc, const char *type, int32_t kind,
     ures_initStackObject(brkName);
 
     // Get the locale
     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)) {
 
     // Get the "boundaries" array.
     if (U_SUCCESS(status)) {


@@ -93,9 +90,7 @@ BreakIterator::buildInstance(const Locale& loc, const char *type, int32_t kind,
 
         // Use the string if we found it
         if (U_SUCCESS(status) && brkfname) {
 
         // 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;
 
             UChar* extStart=u_strchr(brkfname, 0x002e);
             int len = 0;


@@ -123,7 +118,8 @@ BreakIterator::buildInstance(const Locale& loc, const char *type, int32_t kind,
     // 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);
     // 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);
     }
 
         result->setBreakType(kind);
     }
 


@@ -203,6 +199,7 @@ BreakIterator::getAvailableLocales(int32_t& count)
 //-------------------------------------------
 
 BreakIterator::BreakIterator()
 //-------------------------------------------
 
 BreakIterator::BreakIterator()

+: fKeepAll(FALSE)

 {
     *validLocale = *actualLocale = 0;
 }
 {
     *validLocale = *actualLocale = 0;
 }


@@ -389,6 +386,7 @@ BreakIterator::createInstance(const Locale& loc, int32_t kind, UErrorCode& statu
 }
 
 // -------------------------------------
 }
 
 // -------------------------------------

+enum { kKeyValueLenMax = 32 };

 
 BreakIterator*
 BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)
 
 BreakIterator*
 BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)


@@ -397,6 +395,7 @@ BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)
     if (U_FAILURE(status)) {
         return NULL;
     }
     if (U_FAILURE(status)) {
         return NULL;
     }

+    char lbType[kKeyValueLenMax];

 
     BreakIterator *result = NULL;
     switch (kind) {
 
     BreakIterator *result = NULL;
     switch (kind) {


@@ -407,10 +406,38 @@ BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)
         result = BreakIterator::buildInstance(loc, "word", kind, status);
         break;
     case UBRK_LINE:
         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);
         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);
         break;
     case UBRK_TITLE:
         result = BreakIterator::buildInstance(loc, "title", kind, status);