iconvdata/iso-2022-kr.c

   1 /* Conversion module for ISO-2022-KR.
   2    Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Lesser General Public
   8    License as published by the Free Software Foundation; either
   9    version 2.1 of the License, or (at your option) any later version.
  10
  11    The GNU C Library is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14    Lesser General Public License for more details.
  15
  16    You should have received a copy of the GNU Lesser General Public
  17    License along with the GNU C Library; if not, write to the Free
  18    Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
  19    02111-1307 USA.  */
  20
  21 #include <dlfcn.h>
  22 #include <gconv.h>
  23 #include <stdint.h>
  24 #include <string.h>
  25 #include "ksc5601.h"
  26
  27 #include <assert.h>
  28
  29 /* This makes obvious what everybody knows: 0x1b is the Esc character.  */
  30 #define ESC     0x1b
  31
  32 /* The shift sequences for this charset (it does not use ESC).  */
  33 #define SI      0x0f
  34 #define SO      0x0e
  35
  36 /* Definitions used in the body of the `gconv' function.  */
  37 #define CHARSET_NAME            "ISO-2022-KR//"
  38 #define DEFINE_INIT             1
  39 #define DEFINE_FINI             1
  40 #define FROM_LOOP               from_iso2022kr_loop
  41 #define TO_LOOP                 to_iso2022kr_loop
  42 #define MIN_NEEDED_FROM         1
  43 #define MAX_NEEDED_FROM         3
  44 #define MIN_NEEDED_TO           4
  45 #define MAX_NEEDED_TO           4
  46 #define PREPARE_LOOP \
  47   int save_set;                                                               \
  48   int *setp = &data->__statep->__count;                                       \
  49   if (!FROM_DIRECTION && !data->__internal_use                                \
  50       && data->__invocation_counter == 0)                                     \
  51     {                                                                         \
  52       /* Emit the designator sequence.  */                                    \
  53       if (outbuf + 4 > outend)                                                \
  54         return __GCONV_FULL_OUTPUT;                                           \
  55                                                                               \
  56       *outbuf++ = ESC;                                                        \
  57       *outbuf++ = '$';                                                        \
  58       *outbuf++ = ')';                                                        \
  59       *outbuf++ = 'C';                                                        \
  60     }
  61 #define EXTRA_LOOP_ARGS         , setp
  62
  63
  64 /* The COUNT element of the state keeps track of the currently selected
  65    character set.  The possible values are:  */
  66 enum
  67 {
  68   ASCII_set = 0,
  69   KSC5601_set = 8
  70 };
  71
  72
  73 /* Since this is a stateful encoding we have to provide code which resets
  74    the output state to the initial state.  This has to be done during the
  75    flushing.  */
  76 #define EMIT_SHIFT_TO_INIT \
  77   if (data->__statep->__count != ASCII_set)                                   \
  78     {                                                                         \
  79       if (FROM_DIRECTION)                                                     \
  80         {                                                                     \
  81           /* It's easy, we don't have to emit anything, we just reset the     \
  82              state for the input.  */                                         \
  83           data->__statep->__count &= 7;                                       \
  84           data->__statep->__count |= ASCII_set;                               \
  85         }                                                                     \
  86       else                                                                    \
  87         {                                                                     \
  88           /* We are not in the initial state.  To switch back we have         \
  89              to emit `SI'.  */                                                \
  90           if (__builtin_expect (outbuf == outend, 0))                         \
  91             /* We don't have enough room in the output buffer.  */            \
  92             status = __GCONV_FULL_OUTPUT;                                     \
  93           else                                                                \
  94             {                                                                 \
  95               /* Write out the shift sequence.  */                            \
  96               *outbuf++ = SI;                                                 \
  97               data->__statep->__count = ASCII_set;                            \
  98             }                                                                 \
  99         }                                                                     \
 100     }
 101
 102
 103 /* Since we might have to reset input pointer we must be able to save
 104    and retore the state.  */
 105 #define SAVE_RESET_STATE(Save) \
 106   if (Save)                                                                   \
 107     save_set = *setp;                                                         \
 108   else                                                                        \
 109     *setp = save_set
 110
 111
 112 /* First define the conversion function from ISO-2022-KR to UCS4.  */
 113 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
 114 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
 115 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
 116 #define LOOPFCT                 FROM_LOOP
 117 #define BODY \
 118   {                                                                           \
 119     uint32_t ch = *inptr;                                                     \
 120                                                                               \
 121     /* This is a 7bit character set, disallow all 8bit characters.  */        \
 122     if (__builtin_expect (ch > 0x7f, 0))                                      \
 123       {                                                                       \
 124         if (! ignore_errors_p ())                                             \
 125           {                                                                   \
 126             result = __GCONV_ILLEGAL_INPUT;                                   \
 127             break;                                                            \
 128           }                                                                   \
 129                                                                               \
 130         ++inptr;                                                              \
 131         ++*irreversible;                                                      \
 132         continue;                                                             \
 133       }                                                                       \
 134                                                                               \
 135     /* Recognize escape sequences.  */                                        \
 136     if (__builtin_expect (ch, 0) == ESC)                                      \
 137       {                                                                       \
 138         /* We don't really have to handle escape sequences since all the      \
 139            switching is done using the SI and SO bytes.  But we have to       \
 140            recognize `Esc $ ) C' since this is a kind of flag for this        \
 141            encoding.  We simply ignore it.  */                                \
 142         if (__builtin_expect (inptr + 1 > inend, 0)                           \
 143             || (inptr[1] == '$'                                               \
 144                 && (__builtin_expect (inptr + 2 > inend, 0)                   \
 145                     || (inptr[2] == ')'                                       \
 146                         && __builtin_expect (inptr + 3 > inend, 0)))))        \
 147                                                                               \
 148           {                                                                   \
 149             result = __GCONV_INCOMPLETE_INPUT;                                \
 150             break;                                                            \
 151           }                                                                   \
 152         if (inptr[1] == '$' && inptr[2] == ')' && inptr[3] == 'C')            \
 153           {                                                                   \
 154             /* Yeah, yeah, we know this is ISO 2022-KR.  */                   \
 155             inptr += 4;                                                       \
 156             continue;                                                         \
 157           }                                                                   \
 158       }                                                                       \
 159     else if (__builtin_expect (ch, 0) == SO)                                  \
 160       {                                                                       \
 161         /* Switch to use KSC.  */                                             \
 162         ++inptr;                                                              \
 163         set = KSC5601_set;                                                    \
 164         continue;                                                             \
 165       }                                                                       \
 166     else if (__builtin_expect (ch, 0) == SI)                                  \
 167       {                                                                       \
 168         /* Switch to use ASCII.  */                                           \
 169         ++inptr;                                                              \
 170         set = ASCII_set;                                                      \
 171         continue;                                                             \
 172       }                                                                       \
 173                                                                               \
 174     if (set == ASCII_set)                                                     \
 175       {                                                                       \
 176         /* Almost done, just advance the input pointer.  */                   \
 177         ++inptr;                                                              \
 178       }                                                                       \
 179     else                                                                      \
 180       {                                                                       \
 181         assert (set == KSC5601_set);                                          \
 182                                                                               \
 183         /* Use the KSC 5601 table.  */                                        \
 184         ch = ksc5601_to_ucs4 (&inptr, inend - inptr, 0);                      \
 185                                                                               \
 186         if (__builtin_expect (ch, 1) == 0)                                    \
 187           {                                                                   \
 188             result = __GCONV_INCOMPLETE_INPUT;                                \
 189             break;                                                            \
 190           }                                                                   \
 191         else if (__builtin_expect (ch, 0) == __UNKNOWN_10646_CHAR)            \
 192           {                                                                   \
 193             if (! ignore_errors_p ())                                         \
 194               {                                                               \
 195                 /* This is an illegal character.  */                          \
 196                 result = __GCONV_ILLEGAL_INPUT;                               \
 197                 break;                                                        \
 198               }                                                               \
 199                                                                               \
 200             ++*irreversible;                                                  \
 201             ++inptr;                                                          \
 202             continue;                                                         \
 203           }                                                                   \
 204       }                                                                       \
 205                                                                               \
 206     put32 (outptr, ch);                                                       \
 207     outptr += 4;                                                              \
 208   }
 209 #define LOOP_NEED_FLAGS
 210 #define EXTRA_LOOP_DECLS        , int *setp
 211 #define INIT_PARAMS             int set = *setp
 212 #define UPDATE_PARAMS           *setp = set
 213 #include <iconv/loop.c>
 214
 215
 216 /* Next, define the other direction.  */
 217 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 218 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 219 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 220 #define LOOPFCT                 TO_LOOP
 221 #define BODY \
 222   {                                                                           \
 223     uint32_t ch;                                                              \
 224                                                                               \
 225     ch = get32 (inptr);                                                       \
 226                                                                               \
 227     /* First see whether we can write the character using the currently       \
 228        selected character set.  */                                            \
 229     if (ch < 0x80)                                                            \
 230       {                                                                       \
 231         if (set != ASCII_set)                                                 \
 232           {                                                                   \
 233             *outptr++ = SI;                                                   \
 234             set = ASCII_set;                                                  \
 235             if (__builtin_expect (outptr == outend, 0))                       \
 236               {                                                               \
 237                 result = __GCONV_FULL_OUTPUT;                                 \
 238                 break;                                                        \
 239               }                                                               \
 240           }                                                                   \
 241                                                                               \
 242         *outptr++ = ch;                                                       \
 243       }                                                                       \
 244     else                                                                      \
 245       {                                                                       \
 246         char buf[2];                                                          \
 247         size_t written;                                                       \
 248                                                                               \
 249         written = ucs4_to_ksc5601 (ch, buf, 2);                               \
 250                                                                               \
 251         if (__builtin_expect (written, 0) == __UNKNOWN_10646_CHAR)            \
 252           {                                                                   \
 253             UNICODE_TAG_HANDLER (ch, 4);                                      \
 254                                                                               \
 255             /* Illegal character.  */                                         \
 256             STANDARD_ERR_HANDLER (4);                                         \
 257           }                                                                   \
 258         else                                                                  \
 259           {                                                                   \
 260             assert (written == 2);                                            \
 261                                                                               \
 262             /* We use KSC 5601.  */                                           \
 263             if (set != KSC5601_set)                                           \
 264               {                                                               \
 265                 *outptr++ = SO;                                               \
 266                 set = KSC5601_set;                                            \
 267               }                                                               \
 268                                                                               \
 269             if (__builtin_expect (outptr + 2 > outend, 0))                    \
 270               {                                                               \
 271                 result = __GCONV_FULL_OUTPUT;                                 \
 272                 break;                                                        \
 273               }                                                               \
 274                                                                               \
 275             *outptr++ = buf[0];                                               \
 276             *outptr++ = buf[1];                                               \
 277           }                                                                   \
 278       }                                                                       \
 279                                                                               \
 280     /* Now that we wrote the output increment the input pointer.  */          \
 281     inptr += 4;                                                               \
 282   }
 283 #define LOOP_NEED_FLAGS
 284 #define EXTRA_LOOP_DECLS        , int *setp
 285 #define INIT_PARAMS             int set = *setp
 286 #define UPDATE_PARAMS           *setp = set
 287 #include <iconv/loop.c>
 288
 289
 290 /* Now define the toplevel functions.  */
 291 #include <iconv/skeleton.c>