iconvdata/iso-2022-kr.c

   1 /* Conversion module for ISO-2022-KR.
   2    Copyright (C) 1998-2021 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, see
  18    <https://www.gnu.org/licenses/>.  */
  19
  20 #include <dlfcn.h>
  21 #include <gconv.h>
  22 #include <stdint.h>
  23 #include <string.h>
  24 #include "ksc5601.h"
  25
  26 #include <assert.h>
  27
  28 /* This makes obvious what everybody knows: 0x1b is the Esc character.  */
  29 #define ESC     0x1b
  30
  31 /* The shift sequences for this charset (it does not use ESC).  */
  32 #define SI      0x0f
  33 #define SO      0x0e
  34
  35 /* Definitions used in the body of the `gconv' function.  */
  36 #define CHARSET_NAME            "ISO-2022-KR//"
  37 #define DEFINE_INIT             1
  38 #define DEFINE_FINI             1
  39 #define FROM_LOOP               from_iso2022kr_loop
  40 #define TO_LOOP                 to_iso2022kr_loop
  41 #define MIN_NEEDED_FROM         1
  42 #define MAX_NEEDED_FROM         4
  43 #define MIN_NEEDED_TO           4
  44 #define MAX_NEEDED_TO           4
  45 #define ONE_DIRECTION           0
  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 (__glibc_unlikely (outbuf == outend))                            \
  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 (__glibc_unlikely (ch > 0x7f))                                         \
 123       STANDARD_FROM_LOOP_ERR_HANDLER (1);                                     \
 124                                                                               \
 125     /* Recognize escape sequences.  */                                        \
 126     if (__builtin_expect (ch, 0) == ESC)                                      \
 127       {                                                                       \
 128         /* We don't really have to handle escape sequences since all the      \
 129            switching is done using the SI and SO bytes.  But we have to       \
 130            recognize `Esc $ ) C' since this is a kind of flag for this        \
 131            encoding.  We simply ignore it.  */                                \
 132         if (__builtin_expect (inptr + 2 > inend, 0)                           \
 133             || (inptr[1] == '$'                                               \
 134                 && (__builtin_expect (inptr + 3 > inend, 0)                   \
 135                     || (inptr[2] == ')'                                       \
 136                         && __builtin_expect (inptr + 4 > inend, 0)))))        \
 137           {                                                                   \
 138             result = __GCONV_INCOMPLETE_INPUT;                                \
 139             break;                                                            \
 140           }                                                                   \
 141         if (inptr[1] == '$' && inptr[2] == ')' && inptr[3] == 'C')            \
 142           {                                                                   \
 143             /* Yeah, yeah, we know this is ISO 2022-KR.  */                   \
 144             inptr += 4;                                                       \
 145             continue;                                                         \
 146           }                                                                   \
 147       }                                                                       \
 148     else if (__builtin_expect (ch, 0) == SO)                                  \
 149       {                                                                       \
 150         /* Switch to use KSC.  */                                             \
 151         ++inptr;                                                              \
 152         set = KSC5601_set;                                                    \
 153         continue;                                                             \
 154       }                                                                       \
 155     else if (__builtin_expect (ch, 0) == SI)                                  \
 156       {                                                                       \
 157         /* Switch to use ASCII.  */                                           \
 158         ++inptr;                                                              \
 159         set = ASCII_set;                                                      \
 160         continue;                                                             \
 161       }                                                                       \
 162                                                                               \
 163     if (set == ASCII_set)                                                     \
 164       {                                                                       \
 165         /* Almost done, just advance the input pointer.  */                   \
 166         ++inptr;                                                              \
 167       }                                                                       \
 168     else                                                                      \
 169       {                                                                       \
 170         assert (set == KSC5601_set);                                          \
 171                                                                               \
 172         /* Use the KSC 5601 table.  */                                        \
 173         ch = ksc5601_to_ucs4 (&inptr, inend - inptr, 0);                      \
 174                                                                               \
 175         if (__glibc_unlikely (ch == 0))                                       \
 176           {                                                                   \
 177             result = __GCONV_INCOMPLETE_INPUT;                                \
 178             break;                                                            \
 179           }                                                                   \
 180         else if (__glibc_unlikely (ch == __UNKNOWN_10646_CHAR))               \
 181           {                                                                   \
 182             STANDARD_FROM_LOOP_ERR_HANDLER (1);                               \
 183           }                                                                   \
 184       }                                                                       \
 185                                                                               \
 186     put32 (outptr, ch);                                                       \
 187     outptr += 4;                                                              \
 188   }
 189 #define LOOP_NEED_FLAGS
 190 #define EXTRA_LOOP_DECLS        , int *setp
 191 #define INIT_PARAMS             int set = *setp
 192 #define UPDATE_PARAMS           *setp = set
 193 #include <iconv/loop.c>
 194
 195
 196 /* Next, define the other direction.  */
 197 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 198 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 199 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 200 #define LOOPFCT                 TO_LOOP
 201 #define BODY \
 202   {                                                                           \
 203     uint32_t ch = get32 (inptr);                                              \
 204                                                                               \
 205     /* First see whether we can write the character using the currently       \
 206        selected character set.  */                                            \
 207     if (ch < 0x80)                                                            \
 208       {                                                                       \
 209         if (set != ASCII_set)                                                 \
 210           {                                                                   \
 211             *outptr++ = SI;                                                   \
 212             set = ASCII_set;                                                  \
 213             if (__glibc_unlikely (outptr == outend))                          \
 214               {                                                               \
 215                 result = __GCONV_FULL_OUTPUT;                                 \
 216                 break;                                                        \
 217               }                                                               \
 218           }                                                                   \
 219                                                                               \
 220         *outptr++ = ch;                                                       \
 221       }                                                                       \
 222     else                                                                      \
 223       {                                                                       \
 224         unsigned char buf[2];                                                 \
 225         /* Fake initialization to keep gcc quiet.  */                         \
 226         asm ("" : "=m" (buf));                                                \
 227                                                                               \
 228         size_t written = ucs4_to_ksc5601 (ch, buf, 2);                        \
 229         if (__builtin_expect (written, 0) == __UNKNOWN_10646_CHAR)            \
 230           {                                                                   \
 231             UNICODE_TAG_HANDLER (ch, 4);                                      \
 232                                                                               \
 233             /* Illegal character.  */                                         \
 234             STANDARD_TO_LOOP_ERR_HANDLER (4);                                 \
 235           }                                                                   \
 236         else                                                                  \
 237           {                                                                   \
 238             assert (written == 2);                                            \
 239                                                                               \
 240             /* We use KSC 5601.  */                                           \
 241             if (set != KSC5601_set)                                           \
 242               {                                                               \
 243                 *outptr++ = SO;                                               \
 244                 set = KSC5601_set;                                            \
 245               }                                                               \
 246                                                                               \
 247             if (__glibc_unlikely (outptr + 2 > outend))                       \
 248               {                                                               \
 249                 result = __GCONV_FULL_OUTPUT;                                 \
 250                 break;                                                        \
 251               }                                                               \
 252                                                                               \
 253             *outptr++ = buf[0];                                               \
 254             *outptr++ = buf[1];                                               \
 255           }                                                                   \
 256       }                                                                       \
 257                                                                               \
 258     /* Now that we wrote the output increment the input pointer.  */          \
 259     inptr += 4;                                                               \
 260   }
 261 #define LOOP_NEED_FLAGS
 262 #define EXTRA_LOOP_DECLS        , int *setp
 263 #define INIT_PARAMS             int set = *setp
 264 #define REINIT_PARAMS           set = *setp
 265 #define UPDATE_PARAMS           *setp = set
 266 #include <iconv/loop.c>
 267
 268
 269 /* Now define the toplevel functions.  */
 270 #include <iconv/skeleton.c>