X-Git-Url: http://git.chise.org/gitweb/?a=blobdiff_plain;f=src%2Ftext-coding.c;h=e3fe8ffa888789ab4c6f4971aef45933ff82d926;hb=6b3f40dfa3d8f18a2fcc6932be7f82acd94261c2;hp=6b490838901f0bd3a0f580d1539c1fcc83fa49f0;hpb=90ad28e7df654888f98170fc63dd1741fe09cecf;p=chise%2Fxemacs-chise.git-

diff --git a/src/text-coding.c b/src/text-coding.c
index 6b49083..e3fe8ff 100644
--- a/src/text-coding.c
+++ b/src/text-coding.c
@@ -163,10 +163,11 @@ static void decode_coding_sjis (Lstream *decoding,
 				CONST unsigned char *src,
 				unsigned_char_dynarr *dst,
 				unsigned int n);
-static void encode_coding_sjis (Lstream *encoding,
-				CONST unsigned char *src,
-				unsigned_char_dynarr *dst,
-				unsigned int n);
+void char_encode_shift_jis (struct encoding_stream *str, Emchar c,
+			    unsigned_char_dynarr *dst, unsigned int *flags);
+void char_finish_shift_jis (struct encoding_stream *str,
+			    unsigned_char_dynarr *dst, unsigned int *flags);
+
 static int detect_coding_big5 (struct detection_state *st,
 			       CONST unsigned char *src,
 			       unsigned int n);
@@ -868,10 +869,6 @@ if TYPE is 'ccl:
     CHECK_STRING (doc_string);
   CODING_SYSTEM_DOC_STRING (codesys) = doc_string;
 
-#ifdef UTF2000
-  if (ty == CODESYS_NO_CONVERSION)
-    codesys->fixed.size = 1;
-#endif
   EXTERNAL_PROPERTY_LIST_LOOP (rest, key, value, props)
     {
       if (EQ (key, Qmnemonic))
@@ -2561,6 +2558,15 @@ reset_encoding_stream (struct encoding_stream *str)
     case CODESYS_UTF8:
       str->encode_char = &char_encode_utf8;
       str->finish = &char_finish_utf8;
+      break;
+    case CODESYS_UCS4:
+      str->encode_char = &char_encode_ucs4;
+      str->finish = &char_finish_ucs4;
+      break;
+    case CODESYS_SHIFT_JIS:
+      str->encode_char = &char_encode_shift_jis;
+      str->finish = &char_finish_shift_jis;
+      break;
     default:
       break;
     }
@@ -2676,9 +2682,6 @@ mule_encode (Lstream *encoding, CONST unsigned char *src,
       encode_coding_no_conversion (encoding, src, dst, n);
       break;
 #ifdef MULE
-    case CODESYS_SHIFT_JIS:
-      encode_coding_sjis (encoding, src, dst, n);
-      break;
     case CODESYS_BIG5:
       encode_coding_big5 (encoding, src, dst, n);
       break;
@@ -2775,53 +2778,50 @@ text_encode_generic (Lstream *encoding, CONST unsigned char *src,
     {
       c = *src++;
 
-      switch (char_boundary)
+      if (char_boundary == 0)
 	{
-	case 0:
-	  if ( c >= 0xfc )
+	  if (c >= 0xfc)
 	    {
 	      ch = c & 0x01;
 	      char_boundary = 5;
 	    }
-	  else if ( c >= 0xf8 )
+	  else if (c >= 0xf8)
 	    {
 	      ch = c & 0x03;
 	      char_boundary = 4;
 	    }
-	  else if ( c >= 0xf0 )
+	  else if (c >= 0xf0)
 	    {
 	      ch = c & 0x07;
 	      char_boundary = 3;
 	    }
-	  else if ( c >= 0xe0 )
+	  else if (c >= 0xe0)
 	    {
 	      ch = c & 0x0f;
 	      char_boundary = 2;
 	    }
-	  else if ( c >= 0xc0 )
+	  else if (c >= 0xc0)
 	    {
 	      ch = c & 0x1f;
 	      char_boundary = 1;
 	    }
 	  else
-	    {
-	      (*str->encode_char) (str, c, dst, &flags);
-	      ch = 0;
-	      char_boundary = 0;
-	    }
-	  break;
-	case 1:
+	    (*str->encode_char) (str, c, dst, &flags);
+	}
+      else if (char_boundary == 1)
+	{
 	  (*str->encode_char) (str, (ch << 6) | (c & 0x3f), dst, &flags);
 	  ch =0;
 	  char_boundary = 0;
-	  break;
-	default:
-	  ch = ( ch << 6 ) | ( c & 0x3f );
+	}
+      else
+	{
+	  ch = (ch << 6) | (c & 0x3f);
 	  char_boundary--;
 	}
     }
 
-  if ( (char_boundary == 0) && flags & CODING_STATE_END)
+  if ((char_boundary == 0) && (flags & CODING_STATE_END))
     {
       (*str->finish) (str, dst, &flags);
     }
@@ -2829,8 +2829,6 @@ text_encode_generic (Lstream *encoding, CONST unsigned char *src,
   str->flags = flags;
   str->ch    = ch;
   str->iso2022.current_char_boundary = char_boundary;
-
-  /* Verbum caro factum est! */
 }
 
 
@@ -2948,6 +2946,11 @@ decode_coding_sjis (Lstream *decoding, CONST unsigned char *src,
 	      Dynarr_add (dst, c);
 #endif
 	    }
+#ifdef UTF2000
+	  else if (c > 32)
+	    DECODE_ADD_UCS_CHAR(MAKE_CHAR(Vcharset_latin_jisx0201,
+					  c, 0), dst);
+#endif
 	  else
 	    DECODE_ADD_BINARY_CHAR (c, dst);
 	}
@@ -2960,137 +2963,62 @@ decode_coding_sjis (Lstream *decoding, CONST unsigned char *src,
   str->ch    = ch;
 }
 
-/* Convert internally-formatted data to Shift-JIS. */
+/* Convert internal character representation to Shift_JIS. */
 
-static void
-encode_coding_sjis (Lstream *encoding, CONST unsigned char *src,
-		    unsigned_char_dynarr *dst, unsigned int n)
+void
+char_encode_shift_jis (struct encoding_stream *str, Emchar ch,
+		       unsigned_char_dynarr *dst, unsigned int *flags)
 {
-  unsigned char c;
-  struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
-  unsigned int flags  = str->flags;
-  unsigned int ch     = str->ch;
   eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
-#ifdef UTF2000
-  unsigned char char_boundary = str->iso2022.current_char_boundary;
-#endif
 
-  while (n--)
+  if (ch == '\n')
     {
-      c = *src++;
+      if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
+	Dynarr_add (dst, '\r');
+      if (eol_type != EOL_CR)
+	Dynarr_add (dst, ch);
+    }
+  else
+    {
+      Lisp_Object charset, value;
+      unsigned int c1, c2, s1, s2;
+      
 #ifdef UTF2000
-      switch (char_boundary)
+      if (INTP (value =
+		get_char_code_table
+		(ch, XCHARSET_ENCODING_TABLE (Vcharset_latin_jisx0201))))
 	{
-	case 0:
-	  if ( c >= 0xfc )
-	    {
-	      ch = c & 0x01;
-	      char_boundary = 5;
-	    }
-	  else if ( c >= 0xf8 )
-	    {
-	      ch = c & 0x03;
-	      char_boundary = 4;
-	    }
-	  else if ( c >= 0xf0 )
-	    {
-	      ch = c & 0x07;
-	      char_boundary = 3;
-	    }
-	  else if ( c >= 0xe0 )
-	    {
-	      ch = c & 0x0f;
-	      char_boundary = 2;
-	    }
-	  else if ( c >= 0xc0 )
-	    {
-	      ch = c & 0x1f;
-	      char_boundary = 1;
-	    }
-	  else
-	    {
-	      ch = 0;
-	      if (c == '\n')
-		{
-		  if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
-		    Dynarr_add (dst, '\r');
-		  if (eol_type != EOL_CR)
-		    Dynarr_add (dst, c);
-		}
-	      else
-		Dynarr_add (dst, c);
-	      char_boundary = 0;
-	    }
-	  break;
-	case 1:
-	  ch = ( ch << 6 ) | ( c & 0x3f );
-	  {
-	    Lisp_Object charset;
-	    unsigned int c1, c2, s1, s2;
-	    
-	    BREAKUP_CHAR (ch, charset, c1, c2);
-	    if (EQ(charset, Vcharset_katakana_jisx0201))
-	      {
-		Dynarr_add (dst, c1 | 0x80);
-	      }
-	    else if (EQ(charset, Vcharset_japanese_jisx0208))
-	      {
-		ENCODE_SJIS (c1 | 0x80, c2 | 0x80, s1, s2);
-		Dynarr_add (dst, s1);
-		Dynarr_add (dst, s2);
-	      }
-	  }
-	  char_boundary = 0;
-	  break;
-	default:
-	  ch = ( ch << 6 ) | ( c & 0x3f );
-	  char_boundary--;
+	  charset = Vcharset_latin_jisx0201;
+	  c1 = XINT (value);
+	  c2 = 0;
 	}
-#else
-      if (c == '\n')
+      else
+#endif
+	BREAKUP_CHAR (ch, charset, c1, c2);
+	  
+      if (EQ(charset, Vcharset_katakana_jisx0201))
 	{
-	  if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
-	    Dynarr_add (dst, '\r');
-	  if (eol_type != EOL_CR)
-	    Dynarr_add (dst, '\n');
-	  ch = 0;
+	  Dynarr_add (dst, c1 | 0x80);
 	}
-      else if (BYTE_ASCII_P (c))
+      else if (c2 == 0)
 	{
-	  Dynarr_add (dst, c);
-	  ch = 0;
+	  Dynarr_add (dst, c1);
 	}
-      else if (BUFBYTE_LEADING_BYTE_P (c))
-	ch = (c == LEADING_BYTE_KATAKANA_JISX0201 ||
-	      c == LEADING_BYTE_JAPANESE_JISX0208_1978 ||
-	      c == LEADING_BYTE_JAPANESE_JISX0208) ? c : 0;
-      else if (ch)
+      else if (EQ(charset, Vcharset_japanese_jisx0208))
 	{
-	  if (ch == LEADING_BYTE_KATAKANA_JISX0201)
-	    {
-	      Dynarr_add (dst, c);
-	      ch = 0;
-	    }
-	  else if (ch == LEADING_BYTE_JAPANESE_JISX0208_1978 ||
-		   ch == LEADING_BYTE_JAPANESE_JISX0208)
-	    ch = c;
-	  else
-	    {
-	      unsigned char j1, j2;
-	      ENCODE_SJIS (ch, c, j1, j2);
-	      Dynarr_add (dst, j1);
-	      Dynarr_add (dst, j2);
-	      ch = 0;
-	    }
+	  ENCODE_SJIS (c1 | 0x80, c2 | 0x80, s1, s2);
+	  Dynarr_add (dst, s1);
+	  Dynarr_add (dst, s2);
 	}
-#endif
+      else
+	Dynarr_add (dst, '?');
     }
+}
 
-  str->flags = flags;
-  str->ch    = ch;
-#ifdef UTF2000
-  str->iso2022.current_char_boundary = char_boundary;
-#endif
+void
+char_finish_shift_jis (struct encoding_stream *str, unsigned_char_dynarr *dst,
+		       unsigned int *flags)
+{
 }
 
 DEFUN ("decode-shift-jis-char", Fdecode_shift_jis_char, 1, 1, 0, /*
@@ -3500,10 +3428,10 @@ void
 char_encode_ucs4 (struct encoding_stream *str, Emchar ch,
 		  unsigned_char_dynarr *dst, unsigned int *flags)
 {
-  Dynarr_add (dst,  ch >> 24);
-  Dynarr_add (dst, (ch >> 16) & 255);
-  Dynarr_add (dst, (ch >>  8) & 255);
-  Dynarr_add (dst,  ch        & 255);
+  Dynarr_add (dst, ch >> 24);
+  Dynarr_add (dst, ch >> 16);
+  Dynarr_add (dst, ch >>  8);
+  Dynarr_add (dst, ch      );
 }
 
 void
@@ -3621,47 +3549,56 @@ decode_coding_utf8 (Lstream *decoding, CONST unsigned char *src,
 }
 
 void
-char_encode_utf8 (struct encoding_stream *str, Emchar code,
+char_encode_utf8 (struct encoding_stream *str, Emchar ch,
 		  unsigned_char_dynarr *dst, unsigned int *flags)
 {
-  if ( code <= 0x7f )
+  eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
+
+  if (ch == '\n')
     {
-      Dynarr_add (dst, code);
+      if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
+	Dynarr_add (dst, '\r');
+      if (eol_type != EOL_CR)
+	Dynarr_add (dst, ch);
     }
-  else if ( code <= 0x7ff )
+  else if (ch <= 0x7f)
     {
-      Dynarr_add (dst, (code >> 6) | 0xc0);
-      Dynarr_add (dst, (code & 0x3f) | 0x80);
+      Dynarr_add (dst, ch);
+    }
+  else if (ch <= 0x7ff)
+    {
+      Dynarr_add (dst, (ch >> 6) | 0xc0);
+      Dynarr_add (dst, (ch & 0x3f) | 0x80);
     }
-  else if ( code <= 0xffff )
+  else if (ch <= 0xffff)
     {
-      Dynarr_add (dst,  (code >> 12) | 0xe0);
-      Dynarr_add (dst, ((code >>  6) & 0x3f) | 0x80);
-      Dynarr_add (dst,  (code        & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch >> 12) | 0xe0);
+      Dynarr_add (dst, ((ch >>  6) & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch        & 0x3f) | 0x80);
     }
-  else if ( code <= 0x1fffff )
+  else if (ch <= 0x1fffff)
     {
-      Dynarr_add (dst,  (code >> 18) | 0xf0);
-      Dynarr_add (dst, ((code >> 12) & 0x3f) | 0x80);
-      Dynarr_add (dst, ((code >>  6) & 0x3f) | 0x80);
-      Dynarr_add (dst,  (code        & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch >> 18) | 0xf0);
+      Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
+      Dynarr_add (dst, ((ch >>  6) & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch        & 0x3f) | 0x80);
     }
-  else if ( code <= 0x3ffffff )
+  else if (ch <= 0x3ffffff)
     {
-      Dynarr_add (dst,  (code >> 24) | 0xf8);
-      Dynarr_add (dst, ((code >> 18) & 0x3f) | 0x80);
-      Dynarr_add (dst, ((code >> 12) & 0x3f) | 0x80);
-      Dynarr_add (dst, ((code >>  6) & 0x3f) | 0x80);
-      Dynarr_add (dst,  (code        & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch >> 24) | 0xf8);
+      Dynarr_add (dst, ((ch >> 18) & 0x3f) | 0x80);
+      Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
+      Dynarr_add (dst, ((ch >>  6) & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch        & 0x3f) | 0x80);
     }
   else
     {
-      Dynarr_add (dst,  (code >> 30) | 0xfc);
-      Dynarr_add (dst, ((code >> 24) & 0x3f) | 0x80);
-      Dynarr_add (dst, ((code >> 18) & 0x3f) | 0x80);
-      Dynarr_add (dst, ((code >> 12) & 0x3f) | 0x80);
-      Dynarr_add (dst, ((code >>  6) & 0x3f) | 0x80);
-      Dynarr_add (dst,  (code        & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch >> 30) | 0xfc);
+      Dynarr_add (dst, ((ch >> 24) & 0x3f) | 0x80);
+      Dynarr_add (dst, ((ch >> 18) & 0x3f) | 0x80);
+      Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
+      Dynarr_add (dst, ((ch >>  6) & 0x3f) | 0x80);
+      Dynarr_add (dst,  (ch        & 0x3f) | 0x80);
     }
 }
 
@@ -4851,23 +4788,44 @@ char_encode_iso2022 (struct encoding_stream *str, Emchar ch,
     {
       int reg;
 
-      BREAKUP_CHAR (ch, charset, byte1, byte2);
-      ensure_correct_direction (XCHARSET_DIRECTION (charset),
-				codesys, dst, flags, 0);
-
       /* Now determine which register to use. */
       reg = -1;
       for (i = 0; i < 4; i++)
 	{
-	  if (EQ (charset, str->iso2022.charset[i]) ||
-	      EQ (charset,
-		  CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i)))
+	  if ((CHARSETP (charset = str->iso2022.charset[i])
+	       && (byte1 = charset_get_byte1 (charset, ch))) ||
+	      (CHARSETP
+	       (charset
+		= CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i))
+	       && (byte1 = charset_get_byte1 (charset, ch))))
 	    {
 	      reg = i;
+	      byte2 = charset_get_byte2 (charset, ch);
 	      break;
 	    }
 	}
-	      
+      if (reg == -1)
+	{
+	  Lisp_Object original_default_coded_charset_priority_list
+	    = Vdefault_coded_charset_priority_list;
+
+	  while (!EQ (Vdefault_coded_charset_priority_list, Qnil))
+	    {
+	      BREAKUP_CHAR (ch, charset, byte1, byte2);
+	      if (XCHARSET_FINAL (charset))
+		goto found;
+	      Vdefault_coded_charset_priority_list
+		= Fcdr (Fmemq (XCHARSET_NAME (charset),
+			       Vdefault_coded_charset_priority_list));
+	    }
+	  BREAKUP_CHAR (ch, charset, byte1, byte2);
+	found:
+	  Vdefault_coded_charset_priority_list
+	    = original_default_coded_charset_priority_list;
+	}
+      ensure_correct_direction (XCHARSET_DIRECTION (charset),
+				codesys, dst, flags, 0);
+      
       if (reg == -1)
 	{
 	  if (XCHARSET_GRAPHIC (charset) != 0)
@@ -5021,80 +4979,54 @@ encode_coding_no_conversion (Lstream *encoding, CONST unsigned char *src,
     {
       c = *src++;	  
 #ifdef UTF2000
-      switch (char_boundary)
+      if (char_boundary == 0)
+	if ( c >= 0xfc )
+	  {
+	    ch = c & 0x01;
+	    char_boundary = 5;
+	  }
+	else if ( c >= 0xf8 )
+	  {
+	    ch = c & 0x03;
+	    char_boundary = 4;
+	  }
+	else if ( c >= 0xf0 )
+	  {
+	    ch = c & 0x07;
+	    char_boundary = 3;
+	  }
+	else if ( c >= 0xe0 )
+	  {
+	    ch = c & 0x0f;
+	    char_boundary = 2;
+	  }
+	else if ( c >= 0xc0 )
+	  {
+	    ch = c & 0x1f;
+	    char_boundary = 1;
+	  }
+	else
+	  {
+	    ch = 0;
+	    if (c == '\n')
+	      {
+		if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
+		  Dynarr_add (dst, '\r');
+		if (eol_type != EOL_CR)
+		  Dynarr_add (dst, c);
+	      }
+	    else
+	      Dynarr_add (dst, c);
+	    char_boundary = 0;
+	  }
+      else if (char_boundary == 1)
 	{
-	case 0:
-	  if ( c >= 0xfc )
-	    {
-	      ch = c & 0x01;
-	      char_boundary = 5;
-	    }
-	  else if ( c >= 0xf8 )
-	    {
-	      ch = c & 0x03;
-	      char_boundary = 4;
-	    }
-	  else if ( c >= 0xf0 )
-	    {
-	      ch = c & 0x07;
-	      char_boundary = 3;
-	    }
-	  else if ( c >= 0xe0 )
-	    {
-	      ch = c & 0x0f;
-	      char_boundary = 2;
-	    }
-	  else if ( c >= 0xc0 )
-	    {
-	      ch = c & 0x1f;
-	      char_boundary = 1;
-	    }
-	  else
-	    {
-	      ch = 0;
-
-	      if (c == '\n')
-		{
-		  if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
-		    Dynarr_add (dst, '\r');
-		  if (eol_type != EOL_CR)
-		    Dynarr_add (dst, c);
-		}
-	      else
-		Dynarr_add (dst, c);
-	      char_boundary = 0;
-	    }
-	  break;
-	case 1:
 	  ch = ( ch << 6 ) | ( c & 0x3f );
-	  switch ( str->codesys->fixed.size )
-	    {
-	    case 1:
-	      Dynarr_add (dst, ch & 0xff);
-	      break;
-	    case 2:
-	      Dynarr_add (dst, (ch >> 8) & 0xff);
-	      Dynarr_add (dst,  ch       & 0xff);
-	      break;
-	    case 3:
-	      Dynarr_add (dst, (ch >> 16) & 0xff);
-	      Dynarr_add (dst, (ch >>  8) & 0xff);
-	      Dynarr_add (dst,  ch        & 0xff);
-	      break;
-	    case 4:
-	      Dynarr_add (dst, (ch >> 24) & 0xff);
-	      Dynarr_add (dst, (ch >> 16) & 0xff);
-	      Dynarr_add (dst, (ch >>  8) & 0xff);
-	      Dynarr_add (dst,  ch        & 0xff);
-	      break;
-	    default:
-	      fprintf(stderr, "It seems %d bytes stream.\n",
-		      str->codesys->fixed.size);
-	      abort ();
-	    }
+	  Dynarr_add (dst, ch & 0xff);
 	  char_boundary = 0;
-	  break;
-	default:
+	}
+      else
+	{
 	  ch = ( ch << 6 ) | ( c & 0x3f );
 	  char_boundary--;
 	}