comparison src/coding.c @ 89732:8acde12aba89

(get_translation_table): New function. (translate_chars): Fix the bug of skipping annotation data. (decode_coding): Utilze get_translation_table. (encode_coding): Likewise. (char_encodable_p): Translate char if necessary. (Funencodable_char_position): Likewise. (Ffind_coding_systems_region_internal): Setup translation table for encode in a coding system attribute vector in advance. (Fcheck_coding_systems_region): Likewise. (Fdefine_coding_system_internal): Allow a symbol as translation table. For shift-jis type coding system, allow 4th charset.

89731:e4b184cef172

      coding->produced -= n;
      coding->produced_char -= n;
    }
}

static void
translate_chars (coding, table)
     struct coding_system *coding;
     Lisp_Object table;
{

  while (charbuf < charbuf_end)
    {
      c = *charbuf;
      if (c < 0)
	charbuf += c;
      else
	*charbuf++ = translate_char (table, c);
    }
}

decode_coding (coding)
     struct coding_system *coding;
{
  Lisp_Object attrs;
  Lisp_Object undo_list;

  if (BUFFERP (coding->src_object)
      && coding->src_pos > 0
      && coding->src_pos < GPT
      && coding->src_pos + coding->src_chars > GPT)

  coding->errors = 0;

  ALLOC_CONVERSION_WORK_AREA (coding);

  attrs = CODING_ID_ATTRS (coding->id);

  do
    {
      coding_set_source (coding);
      coding->annotated = 0;
      (*(coding->decoder)) (coding);
      if (!NILP (CODING_ATTR_DECODE_TBL (attrs)))
	translate_chars (coding, CODING_ATTR_DECODE_TBL (attrs));
      else if (!NILP (Vstandard_translation_table_for_decode))
	translate_chars (coding, Vstandard_translation_table_for_decode);
      coding_set_destination (coding);
      produce_chars (coding);
      if (coding->annotated)
	produce_annotation (coding);
    }

static int
encode_coding (coding)
     struct coding_system *coding;
{
  Lisp_Object attrs;

  attrs = CODING_ID_ATTRS (coding->id);

  if (BUFFERP (coding->dst_object))
    {
      set_buffer_internal (XBUFFER (coding->dst_object));
      coding->dst_multibyte

  do {
    coding_set_source (coding);
    consume_chars (coding);

    if (!NILP (CODING_ATTR_ENCODE_TBL (attrs)))
      translate_chars (coding, CODING_ATTR_ENCODE_TBL (attrs));
    else if (!NILP (Vstandard_translation_table_for_encode))
      translate_chars (coding, Vstandard_translation_table_for_encode);

    coding_set_destination (coding);
    (*(coding->encoder)) (coding);
  } while (coding->consumed_char < coding->src_chars);

     int c;
     Lisp_Object attrs;
{
  Lisp_Object tail;
  struct charset *charset;

  for (tail = CODING_ATTR_CHARSET_LIST (attrs);
       CONSP (tail); tail = XCDR (tail))
    {
      charset = CHARSET_FROM_ID (XINT (XCAR (tail)));
      if (CHAR_CHARSET_P (c, charset))

	Lisp_Object attrs;

	attrs = AREF (CODING_SYSTEM_SPEC (XCAR (tail)), 0);
	if (EQ (XCAR (tail), CODING_ATTR_BASE_NAME (attrs))
	    && ! EQ (CODING_ATTR_TYPE (attrs), Qundecided))
	  coding_attrs_list = Fcons (attrs, coding_attrs_list);
      }

  if (STRINGP (start))
    p = pbeg = SDATA (start);
  else

     (start, end, coding_system, count, string)
     Lisp_Object start, end, coding_system, count, string;
{
  int n;
  struct coding_system coding;
  Lisp_Object attrs, charset_list;
  Lisp_Object positions;
  int from, to;
  const unsigned char *p, *stop, *pend;
  int ascii_compatible;

  attrs = CODING_ID_ATTRS (coding.id);
  if (EQ (CODING_ATTR_TYPE (attrs), Qraw_text))
    return Qnil;
  ascii_compatible = ! NILP (CODING_ATTR_ASCII_COMPAT (attrs));
  charset_list = CODING_ATTR_CHARSET_LIST (attrs);

  if (NILP (string))
    {
      validate_region (&start, &end);
      from = XINT (start);

	  p = GAP_END_ADDR;
	}

      c = STRING_CHAR_ADVANCE (p);
      if (! (ASCII_CHAR_P (c) && ascii_compatible)
	  && ! char_charset (c, charset_list, NULL))
	{
	  positions = Fcons (make_number (from), positions);
	  n--;
	  if (n == 0)
	    break;

  Lisp_Object list;
  EMACS_INT start_byte, end_byte;
  int pos;
  const unsigned char *p, *pbeg, *pend;
  int c;
  Lisp_Object tail, elt;

  if (STRINGP (start))
    {
      if (!STRING_MULTIBYTE (start)
	  && SCHARS (start) != SBYTES (start))

  list = Qnil;
  for (tail = coding_system_list; CONSP (tail); tail = XCDR (tail))
    {
      elt = XCAR (tail);
      list = Fcons (Fcons (elt, Fcons (AREF (CODING_SYSTEM_SPEC (elt), 0),
				       Qnil)),
		    list);
    }

  if (STRINGP (start))
    p = pbeg = SDATA (start);
  else

  CODING_ATTR_SAFE_CHARSETS (attrs) = safe_charsets;

  CODING_ATTR_ASCII_COMPAT (attrs) = args[coding_arg_ascii_compatible_p];

  val = args[coding_arg_decode_translation_table];
  if (! NILP (val))
    CHECK_CHAR_TABLE (val);
  CODING_ATTR_DECODE_TBL (attrs) = val;

  val = args[coding_arg_encode_translation_table];
  if (! NILP (val))
    CHECK_CHAR_TABLE (val);
  CODING_ATTR_ENCODE_TBL (attrs) = val;

  val = args[coding_arg_post_read_conversion];
  CHECK_SYMBOL (val);
  CODING_ATTR_POST_READ (attrs) = val;

  else if (EQ (coding_type, Qshift_jis))
    {

      struct charset *charset;

      if (XINT (Flength (charset_list)) != 3)
	error ("There should be just three charsets");

      charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
      if (CHARSET_DIMENSION (charset) != 1)
	error ("Dimension of charset %s is not one",
	       SDATA (SYMBOL_NAME (CHARSET_NAME (charset))));

      charset_list = XCDR (charset_list);
      charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
      if (CHARSET_DIMENSION (charset) != 1)
	error ("Dimension of charset %s is not one",
	       SDATA (SYMBOL_NAME (CHARSET_NAME (charset))));

      charset_list = XCDR (charset_list);
      charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
      if (CHARSET_DIMENSION (charset) != 2)

89732:8acde12aba89

      coding->produced -= n;
      coding->produced_char -= n;
    }
}


/* Return a translation table from coding system attribute vector ATTRS
   for encoding (ENCODEP is nonzero) or decoding (ENCODEP is zeor). */

static INLINE
get_translation_table (attrs, encodep)
{
  Lisp_Object standard, translation_table;

  if (encodep)
    translation_table = CODING_ATTR_ENCODE_TBL (attrs),
      standard = Vstandard_translation_table_for_encode;
  else
    translation_table = CODING_ATTR_DECODE_TBL (attrs),
      standard = Vstandard_translation_table_for_decode;
  if (! NILP (translation_table) && SYMBOLP (translation_table))
    translation_table = Fget (translation_table, Qtranslation_table);
  if (NILP (translation_table))
    translation_table = standard;
  if (! CHAR_TABLE_P (translation_table))
    translation_table = Qnil;
  return translation_table;
}


static void
translate_chars (coding, table)
     struct coding_system *coding;
     Lisp_Object table;
{

  while (charbuf < charbuf_end)
    {
      c = *charbuf;
      if (c < 0)
	charbuf += -c;
      else
	*charbuf++ = translate_char (table, c);
    }
}

decode_coding (coding)
     struct coding_system *coding;
{
  Lisp_Object attrs;
  Lisp_Object undo_list;
  Lisp_Object translation_table;

  if (BUFFERP (coding->src_object)
      && coding->src_pos > 0
      && coding->src_pos < GPT
      && coding->src_pos + coding->src_chars > GPT)

  coding->errors = 0;

  ALLOC_CONVERSION_WORK_AREA (coding);

  attrs = CODING_ID_ATTRS (coding->id);
  translation_table = get_translation_table (attrs, 1);

  do
    {
      coding_set_source (coding);
      coding->annotated = 0;
      (*(coding->decoder)) (coding);
      if (!NILP (translation_table))
	translate_chars (coding, translation_table);
      coding_set_destination (coding);
      produce_chars (coding);
      if (coding->annotated)
	produce_annotation (coding);
    }

static int
encode_coding (coding)
     struct coding_system *coding;
{
  Lisp_Object attrs;
  Lisp_Object translation_table;

  attrs = CODING_ID_ATTRS (coding->id);
  translation_table = get_translation_table (attrs, 1);

  if (BUFFERP (coding->dst_object))
    {
      set_buffer_internal (XBUFFER (coding->dst_object));
      coding->dst_multibyte

  do {
    coding_set_source (coding);
    consume_chars (coding);

    if (!NILP (translation_table))
      translate_chars (coding, translation_table);

    coding_set_destination (coding);
    (*(coding->encoder)) (coding);
  } while (coding->consumed_char < coding->src_chars);

     int c;
     Lisp_Object attrs;
{
  Lisp_Object tail;
  struct charset *charset;
  Lisp_Object translation_table;

  translation_table = CODING_ATTR_TRANS_TBL (attrs);
  if (CHAR_TABLE_P (translation_table))
    c = translate_char (translation_table, c);
  for (tail = CODING_ATTR_CHARSET_LIST (attrs);
       CONSP (tail); tail = XCDR (tail))
    {
      charset = CHARSET_FROM_ID (XINT (XCAR (tail)));
      if (CHAR_CHARSET_P (c, charset))

	Lisp_Object attrs;

	attrs = AREF (CODING_SYSTEM_SPEC (XCAR (tail)), 0);
	if (EQ (XCAR (tail), CODING_ATTR_BASE_NAME (attrs))
	    && ! EQ (CODING_ATTR_TYPE (attrs), Qundecided))
	  {
	    ASET (attrs, coding_attr_trans_tbl,
		  get_translation_table (attrs, 1));
	    coding_attrs_list = Fcons (attrs, coding_attrs_list);
	  }
      }

  if (STRINGP (start))
    p = pbeg = SDATA (start);
  else

     (start, end, coding_system, count, string)
     Lisp_Object start, end, coding_system, count, string;
{
  int n;
  struct coding_system coding;
  Lisp_Object attrs, charset_list, translation_table;
  Lisp_Object positions;
  int from, to;
  const unsigned char *p, *stop, *pend;
  int ascii_compatible;

  attrs = CODING_ID_ATTRS (coding.id);
  if (EQ (CODING_ATTR_TYPE (attrs), Qraw_text))
    return Qnil;
  ascii_compatible = ! NILP (CODING_ATTR_ASCII_COMPAT (attrs));
  charset_list = CODING_ATTR_CHARSET_LIST (attrs);
  translation_table = get_translation_table (attrs, 1);

  if (NILP (string))
    {
      validate_region (&start, &end);
      from = XINT (start);

	  p = GAP_END_ADDR;
	}

      c = STRING_CHAR_ADVANCE (p);
      if (! (ASCII_CHAR_P (c) && ascii_compatible)
	  && ! char_charset (translate_char (translation_table, c),
			     charset_list, NULL))
	{
	  positions = Fcons (make_number (from), positions);
	  n--;
	  if (n == 0)
	    break;

  Lisp_Object list;
  EMACS_INT start_byte, end_byte;
  int pos;
  const unsigned char *p, *pbeg, *pend;
  int c;
  Lisp_Object tail, elt, attrs;

  if (STRINGP (start))
    {
      if (!STRING_MULTIBYTE (start)
	  && SCHARS (start) != SBYTES (start))

  list = Qnil;
  for (tail = coding_system_list; CONSP (tail); tail = XCDR (tail))
    {
      elt = XCAR (tail);
      attrs = AREF (CODING_SYSTEM_SPEC (elt), 0);
      ASET (attrs, coding_attr_trans_tbl, get_translation_table (attrs, 1));
      list = Fcons (Fcons (elt, Fcons (attrs, Qnil)), list);
    }

  if (STRINGP (start))
    p = pbeg = SDATA (start);
  else

  CODING_ATTR_SAFE_CHARSETS (attrs) = safe_charsets;

  CODING_ATTR_ASCII_COMPAT (attrs) = args[coding_arg_ascii_compatible_p];

  val = args[coding_arg_decode_translation_table];
  if (! CHAR_TABLE_P (val))
    CHECK_SYMBOL (val);
  CODING_ATTR_DECODE_TBL (attrs) = val;

  val = args[coding_arg_encode_translation_table];
  if (! CHAR_TABLE_P (val))
    CHECK_SYMBOL (val);
  CODING_ATTR_ENCODE_TBL (attrs) = val;

  val = args[coding_arg_post_read_conversion];
  CHECK_SYMBOL (val);
  CODING_ATTR_POST_READ (attrs) = val;

  else if (EQ (coding_type, Qshift_jis))
    {

      struct charset *charset;

      if (XINT (Flength (charset_list)) != 3
	  || XINT (Flength (charset_list)) != 4)
	error ("There should be three or four charsets");

      charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
      if (CHARSET_DIMENSION (charset) != 1)
	error ("Dimension of charset %s is not one",
	       SDATA (SYMBOL_NAME (CHARSET_NAME (charset))));

      charset_list = XCDR (charset_list);
      charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
      if (CHARSET_DIMENSION (charset) != 1)
	error ("Dimension of charset %s is not one",
	       SDATA (SYMBOL_NAME (CHARSET_NAME (charset))));

      charset_list = XCDR (charset_list);
      charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
      if (CHARSET_DIMENSION (charset) != 2)
	error ("Dimension of charset %s is not two",
	       SDATA (SYMBOL_NAME (CHARSET_NAME (charset))));

      charset_list = XCDR (charset_list);
      charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
      if (CHARSET_DIMENSION (charset) != 2)