sal/textenc/converter.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 #include "sal/types.h"
  33
  34 #include "converter.hxx"
  35 #include "tenchelp.hxx"
  36 #include "unichars.hxx"
  37
  38 sal::detail::textenc::BadInputConversionAction
  39 sal::detail::textenc::handleBadInputTextToUnicodeConversion(
  40     bool bUndefined, bool bMultiByte, char cByte, sal_uInt32 nFlags,
  41     sal_Unicode ** pDestBufPtr, sal_Unicode * pDestBufEnd, sal_uInt32 * pInfo)
  42 {
  43     *pInfo |= bUndefined
  44         ? (bMultiByte
  45            ? RTL_TEXTTOUNICODE_INFO_MBUNDEFINED
  46            : RTL_TEXTTOUNICODE_INFO_UNDEFINED)
  47         : RTL_TEXTTOUNICODE_INFO_INVALID;
  48     switch (nFlags
  49             & (bUndefined
  50                ? (bMultiByte
  51                   ? RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_MASK
  52                   : RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_MASK)
  53                : RTL_TEXTTOUNICODE_FLAGS_INVALID_MASK))
  54     {
  55     case RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_ERROR:
  56     case RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_ERROR:
  57     case RTL_TEXTTOUNICODE_FLAGS_INVALID_ERROR:
  58         *pInfo |= RTL_TEXTTOUNICODE_INFO_ERROR;
  59         return BAD_INPUT_STOP;
  60
  61     case RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_IGNORE:
  62     case RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_IGNORE:
  63     case RTL_TEXTTOUNICODE_FLAGS_INVALID_IGNORE:
  64         return BAD_INPUT_CONTINUE;
  65
  66     case RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_MAPTOPRIVATE:
  67         if (*pDestBufPtr != pDestBufEnd)
  68         {
  69             *(*pDestBufPtr)++ = RTL_TEXTCVT_BYTE_PRIVATE_START
  70                 | ((sal_uChar) cByte);
  71             return BAD_INPUT_CONTINUE;
  72         }
  73         else
  74             return BAD_INPUT_NO_OUTPUT;
  75
  76     default: // RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_DEFAULT,
  77              // RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_DEFAULT,
  78              // RTL_TEXTTOUNICODE_FLAGS_INVALID_DEFAULT
  79         if (*pDestBufPtr != pDestBufEnd)
  80         {
  81             *(*pDestBufPtr)++ = RTL_TEXTENC_UNICODE_REPLACEMENT_CHARACTER;
  82             return BAD_INPUT_CONTINUE;
  83         }
  84         else
  85             return BAD_INPUT_NO_OUTPUT;
  86     }
  87 }
  88
  89 sal::detail::textenc::BadInputConversionAction
  90 sal::detail::textenc::handleBadInputUnicodeToTextConversion(
  91     bool bUndefined, sal_uInt32 nUtf32, sal_uInt32 nFlags, char ** pDestBufPtr,
  92     char * pDestBufEnd, sal_uInt32 * pInfo, char const * pPrefix,
  93     sal_Size nPrefixLen, bool * pPrefixWritten)
  94 {
  95     // TODO! RTL_UNICODETOTEXT_FLAGS_UNDEFINED_REPLACE
  96     // RTL_UNICODETOTEXT_FLAGS_UNDEFINED_REPLACESTR
  97
  98     char cReplace;
  99
 100     if (bUndefined)
 101     {
 102         if (ImplIsControlOrFormat(nUtf32))
 103         {
 104             if ((nFlags & RTL_UNICODETOTEXT_FLAGS_CONTROL_IGNORE) != 0)
 105                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE;
 106         }
 107         else if (ImplIsPrivateUse(nUtf32))
 108         {
 109             if ((nFlags & RTL_UNICODETOTEXT_FLAGS_PRIVATE_IGNORE) != 0)
 110                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE;
 111             else if ((nFlags & RTL_UNICODETOTEXT_FLAGS_PRIVATE_MAPTO0) != 0)
 112                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_0;
 113         }
 114         else if (ImplIsZeroWidth(nUtf32))
 115         {
 116             if ((nFlags & RTL_UNICODETOTEXT_FLAGS_NONSPACING_IGNORE) != 0)
 117                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE;
 118         }
 119     }
 120     *pInfo |= bUndefined ? RTL_UNICODETOTEXT_INFO_UNDEFINED :
 121                            RTL_UNICODETOTEXT_INFO_INVALID;
 122     switch (nFlags & (bUndefined ? RTL_UNICODETOTEXT_FLAGS_UNDEFINED_MASK :
 123                                    RTL_UNICODETOTEXT_FLAGS_INVALID_MASK))
 124     {
 125     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_ERROR:
 126     case RTL_UNICODETOTEXT_FLAGS_INVALID_ERROR:
 127         *pInfo |= RTL_UNICODETOTEXT_INFO_ERROR;
 128         return BAD_INPUT_STOP;
 129
 130     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE:
 131     case RTL_UNICODETOTEXT_FLAGS_INVALID_IGNORE:
 132         if (pPrefixWritten)
 133             *pPrefixWritten = false;
 134         return BAD_INPUT_CONTINUE;
 135
 136     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_0:
 137     case RTL_UNICODETOTEXT_FLAGS_INVALID_0:
 138         cReplace = 0;
 139         break;
 140
 141     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_QUESTIONMARK:
 142     case RTL_UNICODETOTEXT_FLAGS_INVALID_QUESTIONMARK:
 143     default: // RTL_UNICODETOTEXT_FLAGS_UNDEFINED_DEFAULT,
 144              // RTL_UNICODETOTEXT_FLAGS_INVALID_DEFAULT
 145         cReplace = '?';
 146         break;
 147
 148     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_UNDERLINE:
 149     case RTL_UNICODETOTEXT_FLAGS_INVALID_UNDERLINE:
 150         cReplace = '_';
 151         break;
 152     }
 153     if ((sal_Size) (pDestBufEnd - *pDestBufPtr) > nPrefixLen)
 154     {
 155         while (nPrefixLen-- > 0)
 156             *(*pDestBufPtr)++ = *pPrefix++;
 157         *(*pDestBufPtr)++ = cReplace;
 158         if (pPrefixWritten)
 159             *pPrefixWritten = true;
 160         return BAD_INPUT_CONTINUE;
 161     }
 162     else
 163         return BAD_INPUT_NO_OUTPUT;
 164 }
 165
 166 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */