sal/textenc/tcvtbyte.cxx

   1 /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
   2 /*************************************************************************
   3  *
   4  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   5  *
   6  * Copyright 2000, 2010 Oracle and/or its affiliates.
   7  *
   8  * OpenOffice.org - a multi-platform office productivity suite
   9  *
  10  * This file is part of OpenOffice.org.
  11  *
  12  * OpenOffice.org is free software: you can redistribute it and/or modify
  13  * it under the terms of the GNU Lesser General Public License version 3
  14  * only, as published by the Free Software Foundation.
  15  *
  16  * OpenOffice.org is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  19  * GNU Lesser General Public License version 3 for more details
  20  * (a copy is included in the LICENSE file that accompanied this code).
  21  *
  22  * You should have received a copy of the GNU Lesser General Public License
  23  * version 3 along with OpenOffice.org.  If not, see
  24  * <http://www.openoffice.org/license.html>
  25  * for a copy of the LGPLv3 License.
  26  *
  27  ************************************************************************/
  28
  29 #include "sal/config.h"
  30
  31 #include "rtl/textcvt.h"
  32
  33 #include "handleundefinedunicodetotextchar.hxx"
  34 #include "tcvtbyte.hxx"
  35 #include "tenchelp.hxx"
  36
  37 sal_Size ImplSymbolToUnicode( SAL_UNUSED_PARAMETER const void*,
  38                               SAL_UNUSED_PARAMETER void*,
  39                               const char* pSrcBuf, sal_Size nSrcBytes,
  40                               sal_Unicode* pDestBuf, sal_Size nDestChars,
  41                               SAL_UNUSED_PARAMETER sal_uInt32,
  42                               sal_uInt32* pInfo, sal_Size* pSrcCvtBytes )
  43 {
  44     sal_uChar               c;
  45     sal_Unicode*            pEndDestBuf;
  46     const char*         pEndSrcBuf;
  47
  48     *pInfo = 0;
  49     pEndDestBuf = pDestBuf+nDestChars;
  50     pEndSrcBuf  = pSrcBuf+nSrcBytes;
  51     while ( pSrcBuf < pEndSrcBuf )
  52     {
  53         if ( pDestBuf == pEndDestBuf )
  54         {
  55             *pInfo |= RTL_TEXTTOUNICODE_INFO_ERROR | RTL_TEXTTOUNICODE_INFO_DESTBUFFERTOSMALL;
  56             break;
  57         }
  58
  59         /* 0-31 (all Control-Character get the same Unicode value) */
  60         c = (sal_uChar)*pSrcBuf;
  61         if ( c <= 0x1F )
  62             *pDestBuf = (sal_Unicode)c;
  63         else
  64             *pDestBuf = ((sal_Unicode)c)+0xF000;
  65         pDestBuf++;
  66         pSrcBuf++;
  67     }
  68
  69     *pSrcCvtBytes = nSrcBytes - (pEndSrcBuf-pSrcBuf);
  70     return (nDestChars - (pEndDestBuf-pDestBuf));
  71 }
  72
  73 sal_Size ImplUnicodeToSymbol( SAL_UNUSED_PARAMETER const void*,
  74                               SAL_UNUSED_PARAMETER void*,
  75                               const sal_Unicode* pSrcBuf, sal_Size nSrcChars,
  76                               char* pDestBuf, sal_Size nDestBytes,
  77                               sal_uInt32 nFlags, sal_uInt32* pInfo,
  78                               sal_Size* pSrcCvtChars )
  79 {
  80     sal_Unicode             c;
  81     char*               pEndDestBuf;
  82     const sal_Unicode*      pEndSrcBuf;
  83
  84     *pInfo = 0;
  85     pEndDestBuf = pDestBuf+nDestBytes;
  86     pEndSrcBuf  = pSrcBuf+nSrcChars;
  87     while ( pSrcBuf < pEndSrcBuf )
  88     {
  89         if ( pDestBuf == pEndDestBuf )
  90         {
  91             *pInfo |= RTL_UNICODETOTEXT_INFO_ERROR | RTL_UNICODETOTEXT_INFO_DESTBUFFERTOSMALL;
  92             break;
  93         }
  94
  95         c = *pSrcBuf;
  96         if ( (c >= 0xF000) && (c <= 0xF0FF) )
  97         {
  98             *pDestBuf = static_cast< char >(static_cast< unsigned char >(c-0xF000));
  99             pDestBuf++;
 100             pSrcBuf++;
 101         }
 102         // Normally 0x001F, but in many cases also symbol characters
 103         // are stored in the first 256 bytes, so that we don't change
 104         // these values
 105         else if ( c <= 0x00FF )
 106         {
 107             *pDestBuf = static_cast< char >(static_cast< unsigned char >(c));
 108             pDestBuf++;
 109             pSrcBuf++;
 110         }
 111         else
 112         {
 113             if ( nFlags & RTL_UNICODETOTEXT_FLAGS_UNDEFINED_REPLACE )
 114             {
 115                 /* !!! */
 116                 /* Only ascii  characters < 0x1F */
 117             }
 118
 119             /* Handle undefined and surrogates characters */
 120             /* (all surrogates characters are undefined) */
 121             if (!sal::detail::textenc::handleUndefinedUnicodeToTextChar(
 122                     &pSrcBuf, pEndSrcBuf, &pDestBuf, pEndDestBuf, nFlags,
 123                     pInfo))
 124                 break;
 125         }
 126     }
 127
 128     *pSrcCvtChars = nSrcChars - (pEndSrcBuf-pSrcBuf);
 129     return (nDestBytes - (pEndDestBuf-pDestBuf));
 130 }
 131
 132 sal_Size ImplUpperCharToUnicode( const void* pData,
 133                             SAL_UNUSED_PARAMETER void*,
 134                             const char* pSrcBuf, sal_Size nSrcBytes,
 135                             sal_Unicode* pDestBuf, sal_Size nDestChars,
 136                             SAL_UNUSED_PARAMETER sal_uInt32, sal_uInt32* pInfo,
 137                             sal_Size* pSrcCvtBytes )
 138 {
 139     sal_uChar                   c;
 140     sal_Unicode                 cConv;
 141     const ImplByteConvertData*  pConvertData = (const ImplByteConvertData*)pData;
 142     sal_Unicode*                pEndDestBuf;
 143     const char*             pEndSrcBuf;
 144
 145     *pInfo = 0;
 146     pEndDestBuf = pDestBuf+nDestChars;
 147     pEndSrcBuf  = pSrcBuf+nSrcBytes;
 148     if ( pDestBuf == pEndDestBuf )
 149     {
 150         *pInfo |= RTL_TEXTTOUNICODE_INFO_ERROR | RTL_TEXTTOUNICODE_INFO_DESTBUFFERTOSMALL;
 151         *pSrcCvtBytes = 0;
 152         return 0;
 153     }
 154     while ( pSrcBuf < pEndSrcBuf )
 155     {
 156         c = (sal_uChar)*pSrcBuf;
 157         if (c < 0x80)
 158             cConv = c;
 159         else
 160             // c <= 0xFF is implied.
 161             cConv = pConvertData->mpToUniTab1[c - 0x80];
 162
 163         *pDestBuf = cConv;
 164         pDestBuf++;
 165         pSrcBuf++;
 166     }
 167
 168     *pSrcCvtBytes = nSrcBytes - (pEndSrcBuf-pSrcBuf);
 169     return (nDestChars - (pEndDestBuf-pDestBuf));
 170 }
 171
 172 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */