i18npool/source/breakiterator/breakiterator_cjk.cxx

   1 /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
   2 /*
   3  * This file is part of the LibreOffice project.
   4  *
   5  * This Source Code Form is subject to the terms of the Mozilla Public
   6  * License, v. 2.0. If a copy of the MPL was not distributed with this
   7  * file, You can obtain one at http://mozilla.org/MPL/2.0/.
   8  *
   9  * This file incorporates work covered by the following license notice:
  10  *
  11  *   Licensed to the Apache Software Foundation (ASF) under one or more
  12  *   contributor license agreements. See the NOTICE file distributed
  13  *   with this work for additional information regarding copyright
  14  *   ownership. The ASF licenses this file to you under the Apache
  15  *   License, Version 2.0 (the "License"); you may not use this file
  16  *   except in compliance with the License. You may obtain a copy of
  17  *   the License at http://www.apache.org/licenses/LICENSE-2.0 .
  18  */
  19
  20 #include <breakiterator_cjk.hxx>
  21 #include <localedata.hxx>
  22
  23 #include <com/sun/star/i18n/BreakType.hpp>
  24 #include <com/sun/star/i18n/ScriptType.hpp>
  25
  26 using namespace ::com::sun::star::uno;
  27 using namespace ::com::sun::star::i18n;
  28 using namespace ::com::sun::star::lang;
  29
  30 namespace i18npool {
  31
  32 //      ----------------------------------------------------
  33 //      class BreakIterator_CJK
  34 //      ----------------------------------------------------;
  35
  36 BreakIterator_CJK::BreakIterator_CJK()
  37 {
  38     cBreakIterator = "com.sun.star.i18n.BreakIterator_CJK";
  39 }
  40
  41 Boundary SAL_CALL
  42 BreakIterator_CJK::previousWord(const OUString& text, sal_Int32 anyPos,
  43         const css::lang::Locale& nLocale, sal_Int16 wordType)
  44 {
  45     if (m_xDict) {
  46         result = m_xDict->previousWord(text, anyPos, wordType);
  47         // #109813# for non-CJK, single character word, fallback to ICU breakiterator.
  48         if (result.endPos - result.startPos != 1 ||
  49                 getScriptType(text, result.startPos) == ScriptType::ASIAN)
  50             return result;
  51         result = BreakIterator_Unicode::getWordBoundary(text, result.startPos, nLocale, wordType, true);
  52         if (result.endPos < anyPos)
  53             return result;
  54     }
  55     return BreakIterator_Unicode::previousWord(text, anyPos, nLocale, wordType);
  56 }
  57
  58 Boundary SAL_CALL
  59 BreakIterator_CJK::nextWord(const OUString& text, sal_Int32 anyPos,
  60         const css::lang::Locale& nLocale, sal_Int16 wordType)
  61 {
  62     if (m_xDict) {
  63         result = m_xDict->nextWord(text, anyPos, wordType);
  64         // #109813# for non-CJK, single character word, fallback to ICU breakiterator.
  65         if (result.endPos - result.startPos != 1 ||
  66                 getScriptType(text, result.startPos) == ScriptType::ASIAN)
  67             return result;
  68         result = BreakIterator_Unicode::getWordBoundary(text, result.startPos, nLocale, wordType, true);
  69         if (result.startPos > anyPos)
  70             return result;
  71     }
  72     return BreakIterator_Unicode::nextWord(text, anyPos, nLocale, wordType);
  73 }
  74
  75 Boundary SAL_CALL
  76 BreakIterator_CJK::getWordBoundary( const OUString& text, sal_Int32 anyPos,
  77         const css::lang::Locale& nLocale, sal_Int16 wordType, sal_Bool bDirection )
  78 {
  79     if (m_xDict) {
  80         result = m_xDict->getWordBoundary(text, anyPos, wordType, bDirection);
  81         // #109813# for non-CJK, single character word, fallback to ICU breakiterator.
  82         if (result.endPos - result.startPos != 1 ||
  83                 getScriptType(text, result.startPos) == ScriptType::ASIAN)
  84             return result;
  85     }
  86     return BreakIterator_Unicode::getWordBoundary(text, anyPos, nLocale, wordType, bDirection);
  87 }
  88
  89 namespace {
  90 bool isHangul( sal_Unicode cCh )
  91 {
  92     return (cCh >= 0xAC00 && cCh <= 0xD7AF) || (cCh >= 0x1100 && cCh <= 0x11FF) ||
  93            (cCh >= 0xA960 && cCh <= 0xA97F) || (cCh >= 0xD7B0 && cCh <= 0xD7FF) ||
  94            (cCh >= 0x3130 && cCh <= 0x318F);
  95 }
  96 }
  97
  98 LineBreakResults SAL_CALL BreakIterator_CJK::getLineBreak(
  99         const OUString& Text, sal_Int32 nStartPos,
 100         const css::lang::Locale& /*rLocale*/, sal_Int32 /*nMinBreakPos*/,
 101         const LineBreakHyphenationOptions& /*hOptions*/,
 102         const LineBreakUserOptions& bOptions )
 103 {
 104     LineBreakResults lbr;
 105
 106     const sal_Int32 nOldStartPos = nStartPos;
 107
 108     if (bOptions.allowPunctuationOutsideMargin &&
 109             nStartPos != Text.getLength() &&
 110             hangingCharacters.indexOf(Text[nStartPos]) != -1 &&
 111             (Text.iterateCodePoints( &nStartPos ), nStartPos == Text.getLength())) {
 112         ; // do nothing
 113     } else if (bOptions.applyForbiddenRules && 0 < nStartPos && nStartPos < Text.getLength()) {
 114
 115         while (nStartPos > 0 &&
 116                 (bOptions.forbiddenBeginCharacters.indexOf(Text[nStartPos]) != -1 ||
 117                  bOptions.forbiddenEndCharacters.indexOf(Text[nStartPos-1]) != -1))
 118             Text.iterateCodePoints( &nStartPos, -1);
 119     }
 120
 121     // Prevent cutting Korean words in the middle.
 122     if (nOldStartPos == nStartPos && nStartPos < Text.getLength()
 123         && isHangul(Text[nStartPos]))
 124     {
 125         while ( nStartPos >= 0 && isHangul( Text[nStartPos] ) )
 126             --nStartPos;
 127
 128         // beginning of the last Korean word.
 129         if ( nStartPos < nOldStartPos )
 130             ++nStartPos;
 131
 132         if ( nStartPos == 0 )
 133             nStartPos = nOldStartPos;
 134     }
 135
 136     lbr.breakIndex = nStartPos;
 137     lbr.breakType = BreakType::WORDBOUNDARY;
 138     return lbr;
 139 }
 140
 141 #define LOCALE(language, country) css::lang::Locale(language, country, OUString())
 142 //      ----------------------------------------------------
 143 //      class BreakIterator_zh
 144 //      ----------------------------------------------------;
 145 BreakIterator_zh::BreakIterator_zh()
 146 {
 147     m_xDict = std::make_unique<xdictionary>("zh");
 148     assert(hangingCharacters.pData);
 149     hangingCharacters = LocaleDataImpl::get()->getHangingCharacters(LOCALE("zh", "CN"));
 150     cBreakIterator = "com.sun.star.i18n.BreakIterator_zh";
 151 }
 152
 153 //      ----------------------------------------------------
 154 //      class BreakIterator_zh_TW
 155 //      ----------------------------------------------------;
 156 BreakIterator_zh_TW::BreakIterator_zh_TW()
 157 {
 158     m_xDict = std::make_unique<xdictionary>("zh");
 159     assert(hangingCharacters.pData);
 160     hangingCharacters = LocaleDataImpl::get()->getHangingCharacters(LOCALE("zh", "TW"));
 161     cBreakIterator = "com.sun.star.i18n.BreakIterator_zh_TW";
 162 }
 163
 164 //      ----------------------------------------------------
 165 //      class BreakIterator_ja
 166 //      ----------------------------------------------------;
 167 BreakIterator_ja::BreakIterator_ja()
 168 {
 169     m_xDict = std::make_unique<xdictionary>("ja");
 170     m_xDict->setJapaneseWordBreak();
 171     assert(hangingCharacters.pData);
 172     hangingCharacters = LocaleDataImpl::get()->getHangingCharacters(LOCALE("ja", "JP"));
 173     cBreakIterator = "com.sun.star.i18n.BreakIterator_ja";
 174 }
 175
 176 //      ----------------------------------------------------
 177 //      class BreakIterator_ko
 178 //      ----------------------------------------------------;
 179 BreakIterator_ko::BreakIterator_ko()
 180 {
 181     assert(hangingCharacters.pData);
 182     hangingCharacters = LocaleDataImpl::get()->getHangingCharacters(LOCALE("ko", "KR"));
 183     cBreakIterator = "com.sun.star.i18n.BreakIterator_ko";
 184 }
 185
 186 }
 187
 188 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */