JOHAB conversion.
authordrepper <drepper>
Fri, 27 Mar 1998 22:53:16 +0000 (22:53 +0000)
committerdrepper <drepper>
Fri, 27 Mar 1998 22:53:16 +0000 (22:53 +0000)
iconvdata/johab.c [new file with mode: 0644]

diff --git a/iconvdata/johab.c b/iconvdata/johab.c
new file mode 100644 (file)
index 0000000..3acd5fc
--- /dev/null
@@ -0,0 +1,554 @@
+/* Mapping tables for JOHAB handling.
+   Copyright (C) 1998 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+   Contributed by Jungshik Shin <jshin@pantheon.yale.edu>, 1998.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Library General Public License as
+   published by the Free Software Foundation; either version 2 of the
+   License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Library General Public License for more details.
+
+   You should have received a copy of the GNU Library General Public
+   License along with the GNU C Library; see the file COPYING.LIB.  If not,
+   write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+   Boston, MA 02111-1307, USA.  */
+
+#include <gconv.h>
+#include <inttypes.h>
+#include <stdlib.h>
+#include <string.h>
+#include <wchar.h>
+#include <ksc5601.h>
+
+/* Direction of the transformation.  */
+enum direction
+  {
+    illegal,
+    to_johab,
+    from_johab
+  };
+
+struct johab_data
+  {
+    enum direction dir;
+  };
+
+/* The table for Bit pattern to Hangul Jamo
+   5 bits each are used to encode
+   leading consonants(19 + 1 filler), medial vowels(21 + 1 filler)
+   and trailing consonants(27 + 1 filler).
+
+   KS C 5601-1992 Annex 3 Table 2
+   0 : Filler, -1: invalid, >= 1 : valid
+
+ */
+const int init[32] =
+{
+  -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18,
+  19, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1
+};
+const int mid[32] =
+{
+  -1, -1, 0, 1, 2, 3, 4, 5,
+  -1, -1, 6, 7, 8, 9, 10, 11,
+  -1, -1, 12, 13, 14, 15, 16, 17,
+  -1, -1, 18, 19, 20, 21, -1, -1
+};
+const int final[32] =
+{
+  -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
+  -1, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, -1, -1
+};
+
+/*
+   Hangul Jamo in Johab to Unicode 2.0 : Unicode 2.0
+   defines 51 Hangul Compatibility Jamos in the block [0x3131,0x314e]
+
+   It's to be considered later which Jamo block to use, Compatibility
+   block [0x3131,0x314e] or Hangul Conjoining Jamo block, [0x1100,0x11ff]
+
+ */
+const wchar_t init_to_ucs[19] =
+{
+  0x3131, 0x3132, 0x3134, 0x3137, 0x3138, 0x3139, 0x3141, 0x3142,
+  0x3143, 0x3145, 0x3146, 0x3147, 0x3148, 0x3149, 0x314a, 0x314b,
+  0x314c, 0x314d, 0x314e
+};
+
+const wchar_t final_to_ucs[27] =
+{
+  L'\0', L'\0', 0x3133, L'\0', 0x3135, 0x3136, L'\0', L'\0',
+  0x313a, 0x313b, 0x314c, 0x313d, 0x313e, 0x313f,
+  0x3140, L'\0', L'\0', 0x3144, L'\0', L'\0', L'\0',
+  L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'
+};
+
+/* The following three arrays are used to convert
+   precomposed Hangul syllables in [0xac00,0xd???]
+   to Jamo bit patterns for Johab encoding
+
+   cf. : KS C 5601-1992, Annex3 Table 2
+
+   Arrays are used to speed up things although it's possible
+   to get the same result arithmetically.
+
+ */
+const int init_to_bit[19] =
+{
+  0x8800, 0x8c00, 0x9000, 0x9400, 0x9800, 0x9c00,
+  0xa000, 0xa400, 0xa800, 0xac00, 0xb000, 0xb400,
+  0xb800, 0xbc00, 0xc000, 0xc400, 0xc800, 0xcc00,
+  0xd000
+};
+
+const int mid_to_bit[21] =
+{
+          0x0060, 0x0080, 0x00a0, 0x00c0, 0x00e0,
+  0x0140, 0x0160, 0x0180, 0x01a0, 0x01c0, 0x1e0,
+  0x0240, 0x0260, 0x0280, 0x02a0, 0x02c0, 0x02e0,
+  0x0340, 0x0360, 0x0380, 0x03a0
+};
+
+const int final_to_bit[28] =
+{
+  1, 2, 3, 4, 5, 6, 7, 8, 9, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x10, 0x11,
+  0x13, 0x14, 0x15, 0x16, 0x17, 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d
+};
+
+/* The conversion table from
+   UCS4 Hangul Compatibility Jamo in [0x3131,0x3163]
+   to Johab
+
+   cf. 1. KS C 5601-1992 Annex 3 Table 2
+   2. Unicode 2.0 manual
+
+ */
+const uint16_t jamo_from_ucs_table[51] =
+{
+  0x8841, 0x8c41,
+  0x8444,
+  0x9041,
+  0x8446, 0x8447,
+  0x9441, 0x9841, 0x9c41,
+  0x844a, 0x844b, 0x844c, 0x844d, 0x884e, 0x884f, 0x8450,
+  0xa041, 0xa441, 0xa841,
+  0x8454,
+  0xac41, 0xb041, 0xb441, 0xb841, 0xbc41,
+  0xc041, 0xc441, 0xc841, 0xca41, 0xd041,
+  0x8461, 0x8481, 0x84a1, 0x84c1, 0x84e1,
+  0x8541, 0x8561, 0x8581, 0x85a1, 0x85c1, 0x85e1,
+  0x8641, 0x8661, 0x8681, 0x86a1, 0x86c1, 0x86e1,
+  0x8741, 0x8761, 0x8781, 0x87a1
+};
+
+
+static inline wchar_t
+johab_sym_hanja_to_ucs (int idx, int c1, int c2)
+{
+  if (idx <= 0xdefe)
+    return (wchar_t) ksc5601_sym_to_ucs[(c1 - 0xd9) * 188 + c2
+                                       - (c2 > 0x90 ? 0x43 : 0x31)];
+  else
+    return (wchar_t) ksc5601_hanja_to_ucs[(c1 - 0xe0) * 188 + c2
+                                         - (c2 > 0x90 ? 0x43 : 0x31)];
+}
+
+static uint16_t
+johab_hanja_from_ucs (wchar_t ch)
+{
+
+  uint16_t idx;
+  if (ucs4_to_ksc5601_hanja (ch, &idx))
+    {
+      int idx1, idx2;
+      /* Hanja begins at the 42th row. 42=0x2a : 0x2a + 0x20 = 0x4a.  */
+      idx1 = idx / 256 - 0x4a;
+      idx2 = idx % 256 + 0x80;
+
+      return ((idx1 / 2) * 256 + 0xe000 + idx2
+             + (idx1 % 2 ? 0 :  (idx2 > 0xee ? 0x43 : 0x31) - 0xa1));
+    }
+  else
+    return 0;
+}
+
+static uint16_t
+johab_sym_from_ucs (wchar_t ch)
+{
+  uint16_t idx;
+  if (ucs4_to_ksc5601_sym (ch, &idx))
+    {
+      int idx1, idx2;
+
+      idx1 = idx / 256 - 0x21;
+      idx2 = idx % 256 + 0x80;
+
+      return ((idx1 / 2) * 256 + 0xd900 + idx2
+             + (idx1 % 2 ? 0 : (idx2 > 0xee ? 0x43 : 0x31) - 0xa1));
+    }
+  else
+    return 0;
+}
+
+
+
+static inline void
+johab_from_ucs4 (wchar_t ch, unsigned char *cp)
+{
+  if (ch >= 0x7f)
+    {
+      int idx;
+
+      if (ch >= 0xac00 && ch <= 0xd7a3)
+       {
+         ch -= 0xac00;
+         idx = init_to_bit[ch / 588];  /* 21*28 = 588 */
+         idx += mid_to_bit[(ch / 28) % 21];  /* (ch % (21 * 28)) / 28 */
+         idx += final_to_bit[ch %  28]; /* (ch % (21 * 28)) % 28 */
+       }
+      /* KS C 5601-1992 Annex 3 regards  0xA4DA(Hangul Filler : U3164)
+         as symbol */
+      else if (ch >= 0x3131 && ch <= 0x3163)
+       idx = jamo_from_ucs_table[ch - 0x3131];
+      else if (ch >= 0x4e00 && ch <= 0x9fa5
+              || ch >= 0xf900 && ch <= 0xfa0b)
+       idx = johab_hanja_from_ucs (ch);
+      /*       Half-width Korean Currency Won Sign
+              else if ( ch == 0x20a9 )
+              idx = 0x5c00;
+      */
+      else
+       idx = johab_sym_from_ucs (ch);
+
+      *cp = (char) (idx / 256);
+      *(cp + 1) = (char) (idx & 0xff);
+
+    }
+  else
+    {
+      *cp = (char) (0x7f & ch);
+      *(cp + 1) = (char) 0;
+    }
+
+}
+
+
+int
+gconv_init (struct gconv_step *step, struct gconv_step_data *data)
+{
+  /* Determine which direction.  */
+  struct johab_data *new_data;
+  enum direction dir;
+  int result;
+
+  if (strcasestr (step->from_name, "JOHAB") != NULL)
+    dir = from_johab;
+  else if (strcasestr (step->to_name, "JOHAB") != NULL)
+    dir = to_johab;
+  else
+    dir = illegal;
+
+  result = GCONV_NOCONV;
+  if (dir != illegal
+      && ((new_data
+          = (struct johab_data *) malloc (sizeof (struct johab_data)))
+         != NULL))
+    {
+      new_data->dir = dir;
+      data->data = new_data;
+      result = GCONV_OK;
+    }
+
+  return result;
+}
+
+
+void
+gconv_end (struct gconv_step_data *data)
+{
+  free (data->data);
+}
+
+
+int
+gconv (struct gconv_step *step, struct gconv_step_data *data,
+       const char *inbuf, size_t *inbufsize, size_t * written, int do_flush)
+{
+  struct gconv_step *next_step = step + 1;
+  struct gconv_step_data *next_data = data + 1;
+  gconv_fct fct = next_step->fct;
+  size_t do_write;
+  int result;
+
+  /* If the function is called with no input this means we have to reset
+     to the initial state.  The possibly partly converted input is
+     dropped.  */
+  if (do_flush)
+    {
+      do_write = 0;
+
+      /* Call the steps down the chain if there are any.  */
+      if (data->is_last)
+       result = GCONV_OK;
+      else
+       {
+         struct gconv_step *next_step = step + 1;
+         struct gconv_step_data *next_data = data + 1;
+
+         result = (*fct) (next_step, next_data, NULL, 0, written, 1);
+
+         /* Clear output buffer.  */
+         data->outbufavail = 0;
+       }
+    }
+  else
+    {
+      enum direction dir = ((struct johab_data *) data->data)->dir;
+
+      do_write = 0;
+
+      do
+       {
+         result = GCONV_OK;
+
+         if (dir == from_johab)
+           {
+             size_t inchars = *inbufsize;
+             size_t outwchars = data->outbufavail;
+             char *outbuf = data->outbuf;
+             size_t cnt = 0;
+
+             while (cnt < inchars
+                    && (outwchars + sizeof (wchar_t) <= data->outbufsize))
+               {
+                 int inchar = (unsigned char) inbuf[cnt];
+                 wchar_t ch;
+                 /* half-width Korean Currency WON sign
+                    if (inchar == 0x5c)
+                    ch =  0x20a9;
+                    else if (inchar < 0x7f)
+                    ch = (wchar_t) inchar;
+                 */
+                 if (inchar < 0x7f)
+                   ch = (wchar_t) inchar;
+
+                 /* Johab : 1. Hangul
+                    1st byte : 0x84-0xd3
+                    2nd byte : 0x41-0x7e, 0x81-0xfe
+                    2. Hanja & Symbol  :
+                    1st byte : 0xd8-0xde, 0xe0-0xf9
+                    2nd byte : 0x31-0x7e, 0x91-0xfe
+                    0xd831-0xd87e and 0xd891-0xd8fe are user-defined area */
+
+                 else if (inchar > 0xf9 || inchar == 0xdf
+                          || (inchar > 0x7e && inchar < 0x84)
+                          || (inchar > 0xd3 && inchar < 0xd9))
+                   /* These are illegal.  */
+                   ch = L'\0';
+                 else
+                   {
+                     /* Two-byte character.  First test whether the next
+                        character is also available.  */
+                     int inchar2;
+                     int idx;
+
+                     if (cnt + 1 >= inchars)
+                       {
+                         /* The second character is not available.  Store
+                            the intermediate result.  */
+                         result = GCONV_INCOMPLETE_INPUT;
+                         break;
+                       }
+
+                     inchar2 = (unsigned char) inbuf[++cnt];
+                     idx = inchar * 256 + inchar2;
+                     if (inchar <= 0xd3)
+                       {       /* Hangul */
+                         int i, m, f;
+                         i = init[(idx & 0x7c00) >> 10];
+                         m = mid[(idx & 0x03e0) >> 5];
+                         f = final[idx & 0x001f];
+                         if (i == -1 || m == -1 || f == -1)
+                           /* This is illegal.  */
+                           ch = L'\0';
+                         else if (i > 0 && m > 0)
+                           ch = ((i - 1) * 21 + (m - 1)) * 28 + f + 0xac00;
+                         else if (i > 0 && m == 0 & f == 0)
+                           ch = init_to_ucs[i - 1];
+                         else if (i == 0 && m > 0 & f == 0)
+                           ch = 0x314e + m;    /* 0x314f + m - 1 */
+                         else if (i == 0 && m == 0 & f > 0)
+                           ch = final_to_ucs[f - 1];   /* round trip?? */
+                         else
+                           /* This is illegal.  */
+                           ch = L'\0';
+                       }
+                     else
+                       {
+                         if (inchar2 < 0x31
+                             || (inchar2 > 0x7e && inchar2 < 0x91)
+                             || inchar2 == 0xff)
+                           /* This is illegal.  */
+                           ch = L'\0';
+                         else if (inchar == 0xda
+                                  && inchar2 > 0xa0 && inchar2 < 0xd4)
+                           /* This is illegal.  */
+                           /* Modern Hangul Jaso is defined elsewhere
+                              in Johab */
+                           ch = L'\0';
+                         else
+                           {
+                             ch = johab_sym_hanja_to_ucs (idx, inchar,
+                                                          inchar2);
+                             /*                if (idx <= 0xdefe)
+                                ch = ksc5601_sym_to_ucs[(inchar - 0xd9) * 192
+                                + inchar2
+                                -  (inchar2>0x90 ? 0x43 : 0x31)];
+
+                                else
+                                ch = ksc5601_hanja_to_ucs[(inchar - 0xe0) *192
+                                + inchar2
+                                -  (inchar2>0x90 ? 0x43 : 0x31)];
+                              */
+                           }
+                       }
+
+                     if (ch == L'\0')
+                       --cnt;
+                   }
+
+                 if (ch == L'\0' && inbuf[cnt] != '\0')
+                   {
+                     /* This is an illegal character.  */
+                     result = GCONV_ILLEGAL_INPUT;
+                     break;
+                   }
+
+                 *((wchar_t *) (outbuf + outwchars)) = ch;
+                 ++do_write;
+                 outwchars += sizeof (wchar_t);
+                 ++cnt;
+               }
+             *inbufsize -= cnt;
+             data->outbufavail = outwchars;
+           }
+         else
+           {
+             size_t inwchars = *inbufsize;
+             size_t outchars = data->outbufavail;
+             char *outbuf = data->outbuf;
+             size_t cnt = 0;
+             int extra = 0;
+
+             while (inwchars >= cnt + sizeof (wchar_t)
+                    && outchars < data->outbufsize)
+               {
+                 wchar_t ch = *((wchar_t *) (inbuf + cnt));
+                 unsigned char cp[2];
+                 /*
+                   if (ch >= (sizeof (from_ucs4_lat1)
+                       / sizeof (from_ucs4_lat1[0])))
+                     {
+                       if (ch >= 0x0391 && ch <= 0x0451)
+                         cp = from_ucs4_greek[ch - 0x391];
+                       else if (ch >= 0x2010 && ch <= 0x9fa0)
+                         cp = from_ucs4_cjk[ch - 0x02010];
+                       else
+                         break;
+                     }
+                   else
+                     cp = from_ucs4_lat1[ch];
+                 */
+                 johab_from_ucs4 (ch, cp);
+
+                 if (cp[0] == '\0' && ch != 0)
+                   /* Illegal character.  */
+                   break;
+
+                 outbuf[outchars] = cp[0];
+                 /* Now test for a possible second byte and write this
+                    if possible.  */
+                 if (cp[1] != '\0')
+                   {
+                     if (outchars + 1 >= data->outbufsize)
+                       {
+                         /* The result does not fit into the buffer.  */
+                         extra = 1;
+                         break;
+                       }
+                     outbuf[++outchars] = cp[1];
+                   }
+
+                 ++do_write;
+                 ++outchars;
+                 cnt += sizeof (wchar_t);
+               }
+             *inbufsize -= cnt;
+             data->outbufavail = outchars;
+
+             if (outchars + extra < data->outbufsize)
+               {
+                 /* If there is still room in the output buffer something
+                    is wrong with the input.  */
+                 if (inwchars >= cnt + sizeof (wchar_t))
+                   {
+                     /* An error occurred.  */
+                     result = GCONV_ILLEGAL_INPUT;
+                     break;
+                   }
+                 if (inwchars != cnt)
+                   {
+                     /* There are some unprocessed bytes at the end of the
+                        input buffer.  */
+                     result = GCONV_INCOMPLETE_INPUT;
+                     break;
+                   }
+               }
+           }
+
+         if (result != GCONV_OK)
+           break;
+
+         if (data->is_last)
+           {
+             /* This is the last step.  */
+             result = (*inbufsize > (dir == from_johab
+                                     ? 0 : sizeof (wchar_t) - 1)
+                       ? GCONV_FULL_OUTPUT : GCONV_EMPTY_INPUT);
+             break;
+           }
+
+         /* Status so far.  */
+         result = GCONV_EMPTY_INPUT;
+
+         if (data->outbufavail > 0)
+           {
+             /* Call the functions below in the chain.  */
+             size_t newavail = data->outbufavail;
+
+             result = (*fct) (next_step, next_data, data->outbuf, &newavail,
+                              written, 0);
+
+             /* Correct the output buffer.  */
+             if (newavail != data->outbufavail && newavail > 0)
+               {
+                 memmove (data->outbuf,
+                          &data->outbuf[data->outbufavail - newavail],
+                          newavail);
+                 data->outbufavail = newavail;
+               }
+           }
+       }
+      while (*inbufsize > 0 && result == GCONV_EMPTY_INPUT);
+    }
+
+  if (written != NULL && data->is_last)
+    *written = do_write;
+
+  return result;
+}