i18npool/inc/xdictionary.hxx

   1 /*************************************************************************
   2  *
   3  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   4  *
   5  * Copyright 2008 by Sun Microsystems, Inc.
   6  *
   7  * OpenOffice.org - a multi-platform office productivity suite
   8  *
   9  * $RCSfile: xdictionary.hxx,v $
  10  * $Revision: 1.7.24.1 $
  11  *
  12  * This file is part of OpenOffice.org.
  13  *
  14  * OpenOffice.org is free software: you can redistribute it and/or modify
  15  * it under the terms of the GNU Lesser General Public License version 3
  16  * only, as published by the Free Software Foundation.
  17  *
  18  * OpenOffice.org is distributed in the hope that it will be useful,
  19  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  20  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  21  * GNU Lesser General Public License version 3 for more details
  22  * (a copy is included in the LICENSE file that accompanied this code).
  23  *
  24  * You should have received a copy of the GNU Lesser General Public License
  25  * version 3 along with OpenOffice.org.  If not, see
  26  * <http://www.openoffice.org/license.html>
  27  * for a copy of the LGPLv3 License.
  28  *
  29  ************************************************************************/
  30 #ifndef _XDICTIONARY_H_
  31 #define _XDICTIONARY_H_
  32
  33 #include <sal/types.h>
  34 #include <osl/module.h>
  35
  36 #include <com/sun/star/i18n/Boundary.hpp>
  37
  38 namespace com { namespace sun { namespace star { namespace i18n {
  39
  40 // Whether to use cell boundary code, currently unused but prepared.
  41 #define USE_CELL_BOUNDARY_CODE 0
  42
  43 #define CACHE_MAX 32            // max cache structure number
  44 #define DEFAULT_SIZE 256        // for boundary size, to avoid alloc and release memory
  45
  46 // cache structure.
  47 struct WordBreakCache {
  48     sal_Int32 length;           // contents length saved here.
  49     sal_Unicode *contents;              // seperated segment contents.
  50     sal_Int32* wordboundary;            // word boundaries in segments.
  51     sal_Int32 size;                     // size of wordboundary
  52
  53     WordBreakCache();
  54     sal_Bool equals(const sal_Unicode *str, Boundary& boundary);    // checking cached string
  55 };
  56
  57 class xdictionary
  58 {
  59 private:
  60     const sal_uInt8  * existMark;
  61     const sal_Int16  * index1;
  62     const sal_Int32 * index2;
  63     const sal_Int32 * lenArray;
  64     const sal_Unicode* dataArea;
  65     oslModule hModule;
  66     Boundary boundary;
  67     sal_Bool japaneseWordBreak;
  68
  69 #if USE_CELL_BOUNDARY_CODE
  70     // For CTL breakiterator, where the word boundary should not be inside cell.
  71     sal_Bool useCellBoundary;
  72     sal_Int32* cellBoundary;
  73 #endif
  74
  75 public:
  76     xdictionary(const sal_Char *lang);
  77     ~xdictionary();
  78     Boundary nextWord( const rtl::OUString& rText, sal_Int32 nPos, sal_Int16 wordType);
  79     Boundary previousWord( const rtl::OUString& rText, sal_Int32 nPos, sal_Int16 wordType);
  80     Boundary getWordBoundary( const rtl::OUString& rText, sal_Int32 nPos, sal_Int16 wordType, sal_Bool bDirection );
  81     void setJapaneseWordBreak();
  82
  83 #if USE_CELL_BOUNDARY_CODE
  84     void setCellBoundary(sal_Int32* cellArray);
  85 #endif
  86
  87 private:
  88     WordBreakCache cache[CACHE_MAX];
  89
  90     sal_Bool        seekSegment(const sal_Unicode *text, sal_Int32 pos, sal_Int32 len, Boundary& boundary);
  91     WordBreakCache& getCache(const sal_Unicode *text, Boundary& boundary);
  92     sal_Bool        exists(const sal_Unicode u);
  93     sal_Int32       getLongestMatch(const sal_Unicode *text, sal_Int32 len);
  94 };
  95
  96 } } } }
  97
  98 #endif
  99