iconvdata/euc-kr.c

   1 /* Mapping tables for EUC-KR handling.
   2    Copyright (C) 1998-2020 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Jungshik Shin <jshin@pantheon.yale.edu>
   5    and Ulrich Drepper <drepper@cygnus.com>, 1998.
   6
   7    The GNU C Library is free software; you can redistribute it and/or
   8    modify it under the terms of the GNU Lesser General Public
   9    License as published by the Free Software Foundation; either
  10    version 2.1 of the License, or (at your option) any later version.
  11
  12    The GNU C Library is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15    Lesser General Public License for more details.
  16
  17    You should have received a copy of the GNU Lesser General Public
  18    License along with the GNU C Library; if not, see
  19    <https://www.gnu.org/licenses/>.  */
  20
  21 #include <dlfcn.h>
  22 #include <stdint.h>
  23 #include <ksc5601.h>
  24
  25
  26 static inline void
  27 __attribute ((always_inline))
  28 euckr_from_ucs4 (uint32_t ch, unsigned char *cp)
  29 {
  30   if (ch > 0x9f)
  31     {
  32       if (__builtin_expect (ch, 0) == 0x20a9)
  33         {
  34           /* Half-width Korean Currency WON sign.  There is no
  35              equivalent in EUC-KR.  Some mappings use \x5c because
  36              this is what some old Korean ASCII variants used but this
  37              is causing problems.  We map it to the FULL WIDTH WON SIGN.  */
  38           cp[0] = '\xa3';
  39           cp[1] = '\xdc';
  40         }
  41       else if (__builtin_expect (ucs4_to_ksc5601 (ch, cp, 2), 0)
  42           != __UNKNOWN_10646_CHAR)
  43         {
  44           cp[0] |= 0x80;
  45           cp[1] |= 0x80;
  46         }
  47       else
  48         cp[0] = cp[1] = '\0';
  49     }
  50   else
  51     {
  52       /* There is no mapping for U005c but we nevertheless map it to
  53          \x5c.  */
  54       cp[0] = (unsigned char) ch;
  55       cp[1] = '\0';
  56     }
  57 }
  58
  59
  60 /* Definitions used in the body of the `gconv' function.  */
  61 #define CHARSET_NAME            "EUC-KR//"
  62 #define FROM_LOOP               from_euc_kr
  63 #define TO_LOOP                 to_euc_kr
  64 #define DEFINE_INIT             1
  65 #define DEFINE_FINI             1
  66 #define MIN_NEEDED_FROM         1
  67 #define MAX_NEEDED_FROM         2
  68 #define MIN_NEEDED_TO           4
  69 #define ONE_DIRECTION           0
  70
  71
  72 /* First define the conversion function from EUC-KR to UCS4.  */
  73 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
  74 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
  75 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
  76 #define LOOPFCT                 FROM_LOOP
  77 #define BODY \
  78   {                                                                           \
  79     uint32_t ch = *inptr;                                                     \
  80                                                                               \
  81     if (ch <= 0x9f)                                                           \
  82       ++inptr;                                                                \
  83     /* 0xfe(->0x7e : row 94) and 0xc9(->0x59 : row 41) are                    \
  84        user-defined areas.  */                                                \
  85     else if (__builtin_expect (ch == 0xa0, 0)                                 \
  86              || __builtin_expect (ch > 0xfe, 0)                               \
  87              || __builtin_expect (ch == 0xc9, 0))                             \
  88       {                                                                       \
  89         /* This is illegal.  */                                               \
  90         STANDARD_FROM_LOOP_ERR_HANDLER (1);                                   \
  91       }                                                                       \
  92     else                                                                      \
  93       {                                                                       \
  94         /* Two-byte character.  First test whether the next byte              \
  95            is also available.  */                                             \
  96         ch = ksc5601_to_ucs4 (&inptr, inend - inptr, 0x80);                   \
  97         if (__glibc_unlikely (ch == 0))                                       \
  98           {                                                                   \
  99             /* The second byte is not available.  */                          \
 100             result = __GCONV_INCOMPLETE_INPUT;                                \
 101             break;                                                            \
 102           }                                                                   \
 103         if (__glibc_unlikely (ch == __UNKNOWN_10646_CHAR))                    \
 104           /* This is an illegal character.  */                                \
 105           STANDARD_FROM_LOOP_ERR_HANDLER (2);                                 \
 106       }                                                                       \
 107                                                                               \
 108     put32 (outptr, ch);                                                       \
 109     outptr += 4;                                                              \
 110   }
 111 #define LOOP_NEED_FLAGS
 112 #define ONEBYTE_BODY \
 113   {                                                                           \
 114     if (c <= 0x9f)                                                            \
 115       return c;                                                               \
 116     else                                                                      \
 117       return WEOF;                                                            \
 118   }
 119 #include <iconv/loop.c>
 120
 121
 122 /* Next, define the other direction.  */
 123 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 124 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 125 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 126 #define LOOPFCT                 TO_LOOP
 127 #define BODY \
 128   {                                                                           \
 129     uint32_t ch = get32 (inptr);                                              \
 130     unsigned char cp[2];                                                      \
 131                                                                               \
 132     /* Decomposing Hangul syllables not available in KS C 5601 into           \
 133        Jamos should be considered either here or in euckr_from_ucs4() */      \
 134     euckr_from_ucs4 (ch, cp);                                                 \
 135                                                                               \
 136     if (__builtin_expect (cp[0], '\1') == '\0' && ch != 0)                    \
 137       {                                                                       \
 138         UNICODE_TAG_HANDLER (ch, 4);                                          \
 139                                                                               \
 140         /* Illegal character.  */                                             \
 141         STANDARD_TO_LOOP_ERR_HANDLER (4);                                     \
 142       }                                                                       \
 143                                                                               \
 144     *outptr++ = cp[0];                                                        \
 145     /* Now test for a possible second byte and write this if possible.  */    \
 146     if (cp[1] != '\0')                                                        \
 147       {                                                                       \
 148         if (__glibc_unlikely (outptr >= outend))                              \
 149           {                                                                   \
 150             /* The result does not fit into the buffer.  */                   \
 151             --outptr;                                                         \
 152             result = __GCONV_FULL_OUTPUT;                                     \
 153             break;                                                            \
 154           }                                                                   \
 155         *outptr++ = cp[1];                                                    \
 156       }                                                                       \
 157                                                                               \
 158     inptr += 4;                                                               \
 159   }
 160 #define LOOP_NEED_FLAGS
 161 #include <iconv/loop.c>
 162
 163
 164 /* Now define the toplevel functions.  */
 165 #include <iconv/skeleton.c>