sal/textenc/converter.c

   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: converter.c,v $
  10  * $Revision: 1.7 $
  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
  31 #include "converter.h"
  32 #include "tenchelp.h"
  33 #include "unichars.h"
  34 #include "rtl/textcvt.h"
  35 #include "sal/types.h"
  36
  37 ImplBadInputConversionAction ImplHandleBadInputTextToUnicodeConversion(
  38     sal_Bool bUndefined, sal_Bool bMultiByte, sal_Char cByte, sal_uInt32 nFlags,
  39     sal_Unicode ** pDestBufPtr, sal_Unicode * pDestBufEnd, sal_uInt32 * pInfo)
  40 {
  41     *pInfo |= bUndefined
  42         ? (bMultiByte
  43            ? RTL_TEXTTOUNICODE_INFO_MBUNDEFINED
  44            : RTL_TEXTTOUNICODE_INFO_UNDEFINED)
  45         : RTL_TEXTTOUNICODE_INFO_INVALID;
  46     switch (nFlags
  47             & (bUndefined
  48                ? (bMultiByte
  49                   ? RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_MASK
  50                   : RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_MASK)
  51                : RTL_TEXTTOUNICODE_FLAGS_INVALID_MASK))
  52     {
  53     case RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_ERROR:
  54     case RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_ERROR:
  55     case RTL_TEXTTOUNICODE_FLAGS_INVALID_ERROR:
  56         *pInfo |= RTL_TEXTTOUNICODE_INFO_ERROR;
  57         return IMPL_BAD_INPUT_STOP;
  58
  59     case RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_IGNORE:
  60     case RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_IGNORE:
  61     case RTL_TEXTTOUNICODE_FLAGS_INVALID_IGNORE:
  62         return IMPL_BAD_INPUT_CONTINUE;
  63
  64     case RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_MAPTOPRIVATE:
  65         if (*pDestBufPtr != pDestBufEnd)
  66         {
  67             *(*pDestBufPtr)++ = RTL_TEXTCVT_BYTE_PRIVATE_START
  68                 | ((sal_uChar) cByte);
  69             return IMPL_BAD_INPUT_CONTINUE;
  70         }
  71         else
  72             return IMPL_BAD_INPUT_NO_OUTPUT;
  73
  74     default: /* RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_DEFAULT,
  75                 RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_DEFAULT,
  76                 RTL_TEXTTOUNICODE_FLAGS_INVALID_DEFAULT */
  77         if (*pDestBufPtr != pDestBufEnd)
  78         {
  79             *(*pDestBufPtr)++ = RTL_TEXTENC_UNICODE_REPLACEMENT_CHARACTER;
  80             return IMPL_BAD_INPUT_CONTINUE;
  81         }
  82         else
  83             return IMPL_BAD_INPUT_NO_OUTPUT;
  84     }
  85 }
  86
  87 ImplBadInputConversionAction
  88 ImplHandleBadInputUnicodeToTextConversion(sal_Bool bUndefined,
  89                                           sal_uInt32 nUtf32,
  90                                           sal_uInt32 nFlags,
  91                                           sal_Char ** pDestBufPtr,
  92                                           sal_Char * pDestBufEnd,
  93                                           sal_uInt32 * pInfo,
  94                                           sal_Char const * pPrefix,
  95                                           sal_Size nPrefixLen,
  96                                           sal_Bool * pPrefixWritten)
  97 {
  98     /* TODO! RTL_UNICODETOTEXT_FLAGS_UNDEFINED_REPLACE
  99              RTL_UNICODETOTEXT_FLAGS_UNDEFINED_REPLACESTR */
 100
 101     sal_Char cReplace;
 102
 103     if (bUndefined)
 104     {
 105         if (ImplIsControlOrFormat(nUtf32))
 106         {
 107             if ((nFlags & RTL_UNICODETOTEXT_FLAGS_CONTROL_IGNORE) != 0)
 108                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE;
 109         }
 110         else if (ImplIsPrivateUse(nUtf32))
 111         {
 112             if ((nFlags & RTL_UNICODETOTEXT_FLAGS_PRIVATE_IGNORE) != 0)
 113                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE;
 114             else if ((nFlags & RTL_UNICODETOTEXT_FLAGS_PRIVATE_MAPTO0) != 0)
 115                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_0;
 116         }
 117         else if (ImplIsZeroWidth(nUtf32))
 118         {
 119             if ((nFlags & RTL_UNICODETOTEXT_FLAGS_NONSPACING_IGNORE) != 0)
 120                 nFlags = RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE;
 121         }
 122     }
 123     *pInfo |= bUndefined ? RTL_UNICODETOTEXT_INFO_UNDEFINED :
 124                            RTL_UNICODETOTEXT_INFO_INVALID;
 125     switch (nFlags & (bUndefined ? RTL_UNICODETOTEXT_FLAGS_UNDEFINED_MASK :
 126                                    RTL_UNICODETOTEXT_FLAGS_INVALID_MASK))
 127     {
 128     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_ERROR:
 129     case RTL_UNICODETOTEXT_FLAGS_INVALID_ERROR:
 130         *pInfo |= RTL_UNICODETOTEXT_INFO_ERROR;
 131         return IMPL_BAD_INPUT_STOP;
 132
 133     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE:
 134     case RTL_UNICODETOTEXT_FLAGS_INVALID_IGNORE:
 135         if (pPrefixWritten)
 136             *pPrefixWritten = sal_False;
 137         return IMPL_BAD_INPUT_CONTINUE;
 138
 139     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_0:
 140     case RTL_UNICODETOTEXT_FLAGS_INVALID_0:
 141         cReplace = 0;
 142         break;
 143
 144     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_QUESTIONMARK:
 145     case RTL_UNICODETOTEXT_FLAGS_INVALID_QUESTIONMARK:
 146     default: /* RTL_UNICODETOTEXT_FLAGS_UNDEFINED_DEFAULT,
 147                 RTL_UNICODETOTEXT_FLAGS_INVALID_DEFAULT */
 148         cReplace = '?';
 149         break;
 150
 151     case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_UNDERLINE:
 152     case RTL_UNICODETOTEXT_FLAGS_INVALID_UNDERLINE:
 153         cReplace = '_';
 154         break;
 155     }
 156     if ((sal_Size) (pDestBufEnd - *pDestBufPtr) > nPrefixLen)
 157     {
 158         while (nPrefixLen-- > 0)
 159             *(*pDestBufPtr)++ = *pPrefix++;
 160         *(*pDestBufPtr)++ = cReplace;
 161         if (pPrefixWritten)
 162             *pPrefixWritten = sal_True;
 163         return IMPL_BAD_INPUT_CONTINUE;
 164     }
 165     else
 166         return IMPL_BAD_INPUT_NO_OUTPUT;
 167 }