sal/textenc/textenc.cxx

   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: textenc.cxx,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 // MARKER(update_precomp.py): autogen include statement, do not remove
  32 #include "precompiled_sal.hxx"
  33
  34 /* TODO! This file should not be called textenc.c, because it is not the
  35    implementation of rtl/textenc.h.  Rather, it should be called
  36    gettextencodingdata.c. */
  37 #include "gettextencodingdata.h"
  38 #include "tenchelp.h"
  39 #include "rtl/textenc.h"
  40 #include "rtl/ustring.hxx"
  41 #include "osl/module.h"
  42
  43 #ifndef INCLUDED_STDDEF_H
  44 #include <stddef.h>
  45 #define INCLUDED_STDDEF_H
  46 #endif
  47
  48 #define NOTABUNI_START 0xFF
  49 #define NOTABUNI_END 0x00
  50
  51 #define NOTABCHAR_START 0xFFFF
  52 #define NOTABCHAR_END 0x0000
  53
  54 #define SAME8090UNI_START 0x80
  55 #define SAME8090UNI_END 0x9F
  56 static sal_uInt16 const aImpl8090SameToUniTab[SAME8090UNI_END
  57                                                   - SAME8090UNI_START
  58                                                   + 1]
  59 = { 0x0080, 0x0081, 0x0082, 0x0083, 0x0084, 0x0085, 0x0086, 0x0087, /* 0x80 */
  60     0x0088, 0x0089, 0x008A, 0x008B, 0x008C, 0x008D, 0x008E, 0x008F,
  61     0x0090, 0x0091, 0x0092, 0x0093, 0x0094, 0x0095, 0x0096, 0x0097, /* 0x90 */
  62     0x0098, 0x0099, 0x009A, 0x009B, 0x009C, 0x009D, 0x009E, 0x009F };
  63
  64 #define SAME8090CHAR_START 0x0080
  65 #define SAME8090CHAR_END 0x009F
  66 static sal_uChar const aImpl8090SameToCharTab[SAME8090CHAR_END
  67                                                   - SAME8090CHAR_START
  68                                                   + 1]
  69     = { 0x80, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, /* 0x0080 */
  70         0x88, 0x89, 0x8A, 0x8B, 0x8C, 0x8D, 0x8E, 0x8F,
  71         0x90, 0x91, 0x92, 0x93, 0x94, 0x95, 0x96, 0x97, /* 0x0090 */
  72         0x98, 0x99, 0x9A, 0x9B, 0x9C, 0x9D, 0x9E, 0x9F };
  73
  74 #define SAMEA0FFCHAR_START 0x00A0
  75 #define SAMEA0FFCHAR_END 0x00FF
  76 static sal_uChar const aImplA0FFSameToCharTab[SAMEA0FFCHAR_END
  77                                                   - SAMEA0FFCHAR_START
  78                                                   + 1]
  79     = { 0xA0, 0xA1, 0xA2, 0xA3, 0xA4, 0xA5, 0xA6, 0xA7, /* 0x00A0 */
  80         0xA8, 0xA9, 0xAA, 0xAB, 0xAC, 0xAD, 0xAE, 0xAF,
  81         0xB0, 0xB1, 0xB2, 0xB3, 0xB4, 0xB5, 0xB6, 0xB7, /* 0x00B0 */
  82         0xB8, 0xB9, 0xBA, 0xBB, 0xBC, 0xBD, 0xBE, 0xBF,
  83         0xC0, 0xC1, 0xC2, 0xC3, 0xC4, 0xC5, 0xC6, 0xC7, /* 0x00C0 */
  84         0xC8, 0xC9, 0xCA, 0xCB, 0xCC, 0xCD, 0xCE, 0xCF,
  85         0xD0, 0xD1, 0xD2, 0xD3, 0xD4, 0xD5, 0xD6, 0xD7, /* 0x00D0 */
  86         0xD8, 0xD9, 0xDA, 0xDB, 0xDC, 0xDD, 0xDE, 0xDF,
  87         0xE0, 0xE1, 0xE2, 0xE3, 0xE4, 0xE5, 0xE6, 0xE7, /* 0x00E0 */
  88         0xE8, 0xE9, 0xEA, 0xEB, 0xEC, 0xED, 0xEE, 0xEF,
  89         0xF0, 0xF1, 0xF2, 0xF3, 0xF4, 0xF5, 0xF6, 0xF7, /* 0x00F0 */
  90         0xF8, 0xF9, 0xFA, 0xFB, 0xFC, 0xFD, 0xFE, 0xFF };
  91
  92 static sal_uInt16 const aImplDoubleByteIdentifierTab[1] = { 0 };
  93
  94 #include "tcvtest1.tab"
  95 #include "tcvtlat1.tab"
  96 #include "tcvtuni1.tab"
  97
  98 extern "C" {
  99                 typedef ImplTextEncodingData const *(*TextEncodingFunction) (rtl_TextEncoding nEncoding);
 100 };
 101
 102 // Yes - we should use the unpleasant to use templatized
 103 // sal:: doublecheckfoo thing here.
 104 static TextEncodingFunction pTables;
 105
 106 #define DOSTRING( x )  #x
 107 #define STRING( x )    DOSTRING( x )
 108
 109 ImplTextEncodingData const *
 110 Impl_getTextEncodingData(rtl_TextEncoding nEncoding) SAL_THROW_EXTERN_C()
 111 {
 112         switch(nEncoding)
 113         {
 114                 case RTL_TEXTENCODING_ASCII_US:
 115                         return &aImplUSASCIITextEncodingData; break;
 116                 case RTL_TEXTENCODING_MS_1252:
 117                         return &aImplMS1252TextEncodingData; break;
 118                 case RTL_TEXTENCODING_UTF8:
 119                         return &aImplUTF8TextEncodingData; break;
 120                 case RTL_TEXTENCODING_ISO_8859_1:
 121             return &aImplISO88591TextEncodingData; break;
 122
 123 #ifndef TOOLS_STRING_MISSUSE
 124 // ----------------------------------------------
 125 // These are here temporarily since they are used in error
 126 // all over the place
 127 // ----------------------------------------------
 128                 case RTL_TEXTENCODING_IBM_850:
 129             return &aImplIBM850TextEncodingData; break;
 130                 case RTL_TEXTENCODING_IBM_857:
 131             return &aImplIBM857TextEncodingData; break;
 132                 case RTL_TEXTENCODING_IBM_860:
 133             return &aImplIBM860TextEncodingData; break;
 134                 case RTL_TEXTENCODING_IBM_861:
 135             return &aImplIBM861TextEncodingData; break;
 136                 case RTL_TEXTENCODING_IBM_863:
 137             return &aImplIBM863TextEncodingData; break;
 138 // ----------------------------------------------
 139 #endif
 140                 default:
 141                         if (!pTables)
 142                         {
 143                                 static char const pName[] = STRING(PLUGIN_NAME);
 144                                 oslModule aModule = osl_loadModuleAscii(pName, SAL_LOADMODULE_DEFAULT);
 145
 146                                 if(aModule)
 147                                 {
 148                                         static char const pSymbol[] = "Impl_getTextEncodingData";
 149                                         pTables = (TextEncodingFunction)osl_getAsciiFunctionSymbol(aModule, pSymbol);
 150                                 }
 151                         }
 152                         if (pTables)
 153                                 return pTables(nEncoding);
 154 //                      else
 155 //                              fprintf (stderr, "missing text encoding library for %d\n", nEncoding);
 156                         break;
 157         }
 158         return NULL;
 159 }