epan/charsets.c

   1 /* charsets.c
   2  * Routines for handling character sets
   3  *
   4  * $Id$
   5  *
   6  * Wireshark - Network traffic analyzer
   7  * By Gerald Combs <gerald@wireshark.org>
   8  * Copyright 1998 Gerald Combs
   9  *
  10  * This program is free software; you can redistribute it and/or
  11  * modify it under the terms of the GNU General Public License
  12  * as published by the Free Software Foundation; either version 2
  13  * of the License, or (at your option) any later version.
  14  *
  15  * This program is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18  * GNU General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU General Public License
  21  * along with this program; if not, write to the Free Software
  22  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  23  */
  24
  25 #include "config.h"
  26
  27 #include <glib.h>
  28
  29 #include "charsets.h"
  30
  31 /*
  32  * Wikipedia's "Character encoding" template, giving a pile of character encodings and
  33  * Wikipedia pages for them:
  34  *
  35  *      http://en.wikipedia.org/wiki/Template:Character_encoding
  36  *
  37  * Unicode character encoding model:
  38  *
  39  *      http://www.unicode.org/reports/tr17/
  40  *
  41  * International Components for Unicode character set mapping tables:
  42  *
  43  *      http://site.icu-project.org/charts/charset
  44  *
  45  * MSDN information on code pages:
  46  *
  47  *      http://msdn.microsoft.com/en-us/library/dd317752(v=VS.85).aspx
  48  *
  49  * ASCII-based code pages, from IBM:
  50  *
  51  *      http://www-01.ibm.com/software/globalization/cp/cp_cpgid.html
  52  *
  53  * EBCDIC code pages, from IBM:
  54  *
  55  *      http://www-03.ibm.com/systems/i/software/globalization/codepages.html
  56  */
  57
  58 /* ASCII/EBCDIC conversion tables from
  59  * http://www.room42.com/store/computer_center/code_tables.shtml
  60  */
  61 #if 0
  62 static guint8 ASCII_translate_EBCDIC [ 256 ] = {
  63     0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08,
  64     0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F,
  65     0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18,
  66     0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F,
  67     0x40, 0x5A, 0x7F, 0x7B, 0x5B, 0x6C, 0x50, 0x7D, 0x4D,
  68     0x5D, 0x5C, 0x4E, 0x6B, 0x60, 0x4B, 0x61,
  69     0xF0, 0xF1, 0xF2, 0xF3, 0xF4, 0xF5, 0xF6, 0xF7, 0xF8,
  70     0xF9, 0x7A, 0x5E, 0x4C, 0x7E, 0x6E, 0x6F,
  71     0x7C, 0xC1, 0xC2, 0xC3, 0xC4, 0xC5, 0xC6, 0xC7, 0xC8,
  72     0xC9, 0xD1, 0xD2, 0xD3, 0xD4, 0xD5, 0xD6,
  73     0xD7, 0xD8, 0xD9, 0xE2, 0xE3, 0xE4, 0xE5, 0xE6, 0xE7,
  74     0xE8, 0xE9, 0xAD, 0xE0, 0xBD, 0x5F, 0x6D,
  75     0x7D, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88,
  76     0x89, 0x91, 0x92, 0x93, 0x94, 0x95, 0x96,
  77     0x97, 0x98, 0x99, 0xA2, 0xA3, 0xA4, 0xA5, 0xA6, 0xA7,
  78     0xA8, 0xA9, 0xC0, 0x6A, 0xD0, 0xA1, 0x4B,
  79     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  80     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  81     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  82     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  83     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  84     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  85     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  86     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  87     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  88     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  89     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  90     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  91     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  92     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  93     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B,
  94     0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B, 0x4B
  95 };
  96
  97 void
  98 ASCII_to_EBCDIC(guint8 *buf, guint bytes)
  99 {
 100         guint   i;
 101         guint8  *bufptr;
 102
 103         bufptr = buf;
 104
 105         for (i = 0; i < bytes; i++, bufptr++) {
 106                 *bufptr = ASCII_translate_EBCDIC[*bufptr];
 107         }
 108 }
 109
 110 guint8
 111 ASCII_to_EBCDIC1(guint8 c)
 112 {
 113         return ASCII_translate_EBCDIC[c];
 114 }
 115 #endif
 116
 117 static guint8 EBCDIC_translate_ASCII [ 256 ] = {
 118     0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08,
 119     0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F,
 120     0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18,
 121     0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F,
 122     0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, 0x28,
 123     0x29, 0x2A, 0x2B, 0x2C, 0x2D, 0x2E, 0x2F,
 124     0x2E, 0x2E, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38,
 125     0x39, 0x3A, 0x3B, 0x3C, 0x3D, 0x2E, 0x3F,
 126     0x20, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 127     0x2E, 0x2E, 0x2E, 0x3C, 0x28, 0x2B, 0x7C,
 128     0x26, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 129     0x2E, 0x21, 0x24, 0x2A, 0x29, 0x3B, 0x5E,
 130     0x2D, 0x2F, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 131     0x2E, 0x7C, 0x2C, 0x25, 0x5F, 0x3E, 0x3F,
 132     0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 133     0x2E, 0x3A, 0x23, 0x40, 0x27, 0x3D, 0x22,
 134     0x2E, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68,
 135     0x69, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 136     0x2E, 0x6A, 0x6B, 0x6C, 0x6D, 0x6E, 0x6F, 0x70, 0x71,
 137     0x72, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 138     0x2E, 0x7E, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79,
 139     0x7A, 0x2E, 0x2E, 0x2E, 0x5B, 0x2E, 0x2E,
 140     0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 141     0x2E, 0x2E, 0x2E, 0x2E, 0x5D, 0x2E, 0x2E,
 142     0x7B, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48,
 143     0x49, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 144     0x7D, 0x4A, 0x4B, 0x4C, 0x4D, 0x4E, 0x4F, 0x50, 0x51,
 145     0x52, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 146     0x5C, 0x2E, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58, 0x59,
 147     0x5A, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E,
 148     0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38,
 149     0x39, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E, 0x2E
 150 };
 151
 152 void
 153 EBCDIC_to_ASCII(guint8 *buf, guint bytes)
 154 {
 155         guint   i;
 156         guint8  *bufptr;
 157
 158         bufptr = buf;
 159
 160         for (i = 0; i < bytes; i++, bufptr++) {
 161                 *bufptr = EBCDIC_translate_ASCII[*bufptr];
 162         }
 163 }
 164
 165 guint8
 166 EBCDIC_to_ASCII1(guint8 c)
 167 {
 168         return EBCDIC_translate_ASCII[c];
 169 }