root/trunk/src/charset.c

/* Basic multilingual character support.
   Copyright (C) 2001, 2002, 2003, 2004, 2005,
                 2006, 2007, 2008 Free Software Foundation, Inc.
   Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
     2005, 2006, 2007, 2008
     National Institute of Advanced Industrial Science and Technology (AIST)
     Registration Number H14PRO021

This file is part of GNU Emacs.

GNU Emacs is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3, or (at your option)
any later version.

GNU Emacs 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 General Public License for more details.

You should have received a copy of the GNU General Public License
along with GNU Emacs; see the file COPYING.  If not, write to
the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
Boston, MA 02110-1301, USA.  */

/* At first, see the document in `charset.h' to understand the code in
   this file.  */

#ifdef emacs
#include <config.h>
#endif

#include <stdio.h>

#ifdef emacs

#include <sys/types.h>
#include "lisp.h"
#include "buffer.h"
#include "charset.h"
#include "composite.h"
#include "coding.h"
#include "disptab.h"

#else  /* not emacs */

#include "mulelib.h"

#endif /* emacs */

Lisp_Object Qcharset, Qascii, Qeight_bit_control, Qeight_bit_graphic;
Lisp_Object Qunknown;

/* Declaration of special leading-codes.  */
EMACS_INT leading_code_private_11; /* for private DIMENSION1 of 1-column */
EMACS_INT leading_code_private_12; /* for private DIMENSION1 of 2-column */
EMACS_INT leading_code_private_21; /* for private DIMENSION2 of 1-column */
EMACS_INT leading_code_private_22; /* for private DIMENSION2 of 2-column */

/* Declaration of special charsets.  The values are set by
   Fsetup_special_charsets.  */
int charset_latin_iso8859_1;    /* ISO8859-1 (Latin-1) */
int charset_jisx0208_1978;      /* JISX0208.1978 (Japanese Kanji old set) */
int charset_jisx0208;           /* JISX0208.1983 (Japanese Kanji) */
int charset_katakana_jisx0201;  /* JISX0201.Kana (Japanese Katakana) */
int charset_latin_jisx0201;     /* JISX0201.Roman (Japanese Roman) */
int charset_big5_1;             /* Big5 Level 1 (Chinese Traditional) */
int charset_big5_2;             /* Big5 Level 2 (Chinese Traditional) */
int charset_mule_unicode_0100_24ff;
int charset_mule_unicode_2500_33ff;
int charset_mule_unicode_e000_ffff;

Lisp_Object Qcharset_table;

/* A char-table containing information of each character set.  */
Lisp_Object Vcharset_table;

/* A vector of charset symbol indexed by charset-id.  This is used
   only for returning charset symbol from C functions.  */
Lisp_Object Vcharset_symbol_table;

/* A list of charset symbols ever defined.  */
Lisp_Object Vcharset_list;

/* Vector of translation table ever defined.
   ID of a translation table is used to index this vector.  */
Lisp_Object Vtranslation_table_vector;

/* A char-table for characters which may invoke auto-filling.  */
Lisp_Object Vauto_fill_chars;

Lisp_Object Qauto_fill_chars;

/* Tables used by macros BYTES_BY_CHAR_HEAD and WIDTH_BY_CHAR_HEAD.  */
int bytes_by_char_head[256];
int width_by_char_head[256];

/* Mapping table from ISO2022's charset (specified by DIMENSION,
   CHARS, and FINAL-CHAR) to Emacs' charset.  */
int iso_charset_table[2][2][128];

/* Variables used locally in the macro FETCH_MULTIBYTE_CHAR.  */
unsigned char *_fetch_multibyte_char_p;
int _fetch_multibyte_char_len;

/* Offset to add to a non-ASCII value when inserting it.  */
EMACS_INT nonascii_insert_offset;

/* Translation table for converting non-ASCII unibyte characters
   to multibyte codes, or nil.  */
Lisp_Object Vnonascii_translation_table;

/* List of all possible generic characters.  */
Lisp_Object Vgeneric_character_list;


void
invalid_character (c)
     int c;
{
  error ("Invalid character: %d, #o%o, #x%x", c, c, c);
}

/* Parse string STR of length LENGTH and fetch information of a
   character at STR.  Set BYTES to the byte length the character
   occupies, CHARSET, C1, C2 to proper values of the character. */

#define SPLIT_MULTIBYTE_SEQ(str, length, bytes, charset, c1, c2)             \
  do {                                                                       \
    (c1) = *(str);                                                           \
    (bytes) = BYTES_BY_CHAR_HEAD (c1);                                       \
    if ((bytes) == 1)                                                        \
      (charset) = ASCII_BYTE_P (c1) ? CHARSET_ASCII : CHARSET_8_BIT_GRAPHIC; \
    else if ((bytes) == 2)                                                   \
      {                                                                      \
        if ((c1) == LEADING_CODE_8_BIT_CONTROL)                              \
          (charset) = CHARSET_8_BIT_CONTROL, (c1) = (str)[1] - 0x20;         \
        else                                                                 \
          (charset) = (c1), (c1) = (str)[1] & 0x7F;                          \
      }                                                                      \
    else if ((bytes) == 3)                                                   \
      {                                                                      \
        if ((c1) < LEADING_CODE_PRIVATE_11)                                  \
          (charset) = (c1), (c1) = (str)[1] & 0x7F, (c2) = (str)[2] & 0x7F;  \
        else                                                                 \
          (charset) = (str)[1], (c1) = (str)[2] & 0x7F;                      \
      }                                                                      \
    else                                                                     \
      (charset) = (str)[1], (c1) = (str)[2] & 0x7F, (c2) = (str)[3] & 0x7F;  \
  } while (0)

/* 1 if CHARSET, C1, and C2 compose a valid character, else 0.
   Note that this intentionally allows invalid components, such
   as 0xA0 0xA0, because there exist many files that contain
   such invalid byte sequences, especially in EUC-GB. */
#define CHAR_COMPONENTS_VALID_P(charset, c1, c2)        \
  ((charset) == CHARSET_ASCII                           \
   ? ((c1) >= 0 && (c1) <= 0x7F)                        \
   : ((charset) == CHARSET_8_BIT_CONTROL                \
      ? ((c1) >= 0x80 && (c1) <= 0x9F)                  \
      : ((charset) == CHARSET_8_BIT_GRAPHIC             \
         ? ((c1) >= 0x80 && (c1) <= 0xFF)               \
         : (CHARSET_DIMENSION (charset) == 1            \
            ? ((c1) >= 0x20 && (c1) <= 0x7F)            \
            : ((c1) >= 0x20 && (c1) <= 0x7F             \
               && (c2) >= 0x20 && (c2) <= 0x7F)))))

/* Store multi-byte form of the character C in STR.  The caller should
   allocate at least 4-byte area at STR in advance.  Returns the
   length of the multi-byte form.  If C is an invalid character code,
   return -1.  */

int
char_to_string_1 (c, str)
     int c;
     unsigned char *str;
{
  unsigned char *p = str;

  if (c & CHAR_MODIFIER_MASK)   /* This includes the case C is negative.  */
    {
      /* Multibyte character can't have a modifier bit.  */
      if (! SINGLE_BYTE_CHAR_P ((c & ~CHAR_MODIFIER_MASK)))
        return -1;

      /* For Meta, Shift, and Control modifiers, we need special care.  */
      if (c & CHAR_META)
        {
          /* Move the meta bit to the right place for a string.  */
          c = (c & ~CHAR_META) | 0x80;
        }
      if (c & CHAR_SHIFT)
        {
          /* Shift modifier is valid only with [A-Za-z].  */
          if ((c & 0377) >= 'A' && (c & 0377) <= 'Z')
            c &= ~CHAR_SHIFT;
          else if ((c & 0377) >= 'a' && (c & 0377) <= 'z')
            c = (c & ~CHAR_SHIFT) - ('a' - 'A');
        }
      if (c & CHAR_CTL)
        {
          /* Simulate the code in lread.c.  */
          /* Allow `\C- ' and `\C-?'.  */
          if (c == (CHAR_CTL | ' '))
            c = 0;
          else if (c == (CHAR_CTL | '?'))
            c = 127;
          /* ASCII control chars are made from letters (both cases),
             as well as the non-letters within 0100...0137.  */
          else if ((c & 0137) >= 0101 && (c & 0137) <= 0132)
            c &= (037 | (~0177 & ~CHAR_CTL));
          else if ((c & 0177) >= 0100 && (c & 0177) <= 0137)
            c &= (037 | (~0177 & ~CHAR_CTL));
        }

      /* If C still has any modifier bits, just ignore it.  */
      c &= ~CHAR_MODIFIER_MASK;
    }

  if (SINGLE_BYTE_CHAR_P (c))
    {
      if (ASCII_BYTE_P (c) || c >= 0xA0)
        *p++ = c;
      else
        {
          *p++ = LEADING_CODE_8_BIT_CONTROL;
          *p++ = c + 0x20;
        }
    }
  else if (CHAR_VALID_P (c, 0))
    {
      int charset, c1, c2;

      SPLIT_CHAR (c, charset, c1, c2);

      if (charset >= LEADING_CODE_EXT_11)
        *p++ = (charset < LEADING_CODE_EXT_12
                ? LEADING_CODE_PRIVATE_11
                : (charset < LEADING_CODE_EXT_21
                   ? LEADING_CODE_PRIVATE_12
                   : (charset < LEADING_CODE_EXT_22
                      ? LEADING_CODE_PRIVATE_21
                      : LEADING_CODE_PRIVATE_22)));
      *p++ = charset;
      if ((c1 > 0 && c1 < 32) || (c2 > 0 && c2 < 32))
        return -1;
      if (c1)
        {
          *p++ = c1 | 0x80;
          if (c2 > 0)
            *p++ = c2 | 0x80;
        }
    }
  else
    return -1;

  return (p - str);
}


/* Store multi-byte form of the character C in STR.  The caller should
   allocate at least 4-byte area at STR in advance.  Returns the
   length of the multi-byte form.  If C is an invalid character code,
   signal an error.

   Use macro `CHAR_STRING (C, STR)' instead of calling this function
   directly if C can be an ASCII character.  */

int
char_to_string (c, str)
     int c;
     unsigned char *str;
{
  int len;
  len = char_to_string_1 (c, str);
  if (len == -1)
    invalid_character (c);
  return len;
}


/* Return the non-ASCII character corresponding to multi-byte form at
   STR of length LEN.  If ACTUAL_LEN is not NULL, store the byte
   length of the multibyte form in *ACTUAL_LEN.

   Use macros STRING_CHAR or STRING_CHAR_AND_LENGTH instead of calling
   this function directly if you want ot handle ASCII characters as
   well.  */

int
string_to_char (str, len, actual_len)
     const unsigned char *str;
     int len, *actual_len;
{
  int c, bytes, charset, c1, c2;

  SPLIT_MULTIBYTE_SEQ (str, len, bytes, charset, c1, c2);
  c = MAKE_CHAR (charset, c1, c2);
  if (actual_len)
    *actual_len = bytes;
  return c;
}

/* Return the length of the multi-byte form at string STR of length LEN.
   Use the macro MULTIBYTE_FORM_LENGTH instead.  */
int
multibyte_form_length (str, len)
     const unsigned char *str;
     int len;
{
  int bytes;

  PARSE_MULTIBYTE_SEQ (str, len, bytes);
  return bytes;
}

/* Check multibyte form at string STR of length LEN and set variables
   pointed by CHARSET, C1, and C2 to charset and position codes of the
   character at STR, and return 0.  If there's no multibyte character,
   return -1.  This should be used only in the macro SPLIT_STRING
   which checks range of STR in advance.  */

int
split_string (str, len, charset, c1, c2)
     const unsigned char *str;
     unsigned char *c1, *c2;
     int len, *charset;
{
  register int bytes, cs, code1, code2 = -1;

  SPLIT_MULTIBYTE_SEQ (str, len, bytes, cs, code1, code2);
  if (cs == CHARSET_ASCII)
    return -1;
  *charset = cs;
  *c1 = code1;
  *c2 = code2;
  return 0;
}

/* Return 1 if character C has valid printable glyph.
   Use the macro CHAR_PRINTABLE_P instead.  */
int
char_printable_p (c)
     int c;
{
  int charset, c1, c2;

  if (ASCII_BYTE_P (c))
    return 1;
  else if (SINGLE_BYTE_CHAR_P (c))
    return 0;
  else if (c >= MAX_CHAR)
    return 0;

  SPLIT_CHAR (c, charset, c1, c2);
  if (! CHARSET_DEFINED_P (charset))
    return 0;
  if (CHARSET_CHARS (charset) == 94
      ? c1 <= 32 || c1 >= 127
      : c1 < 32)
    return 0;
  if (CHARSET_DIMENSION (charset) == 2
      && (CHARSET_CHARS (charset) == 94
          ? c2 <= 32 || c2 >= 127
          : c2 < 32))
    return 0;
  return 1;
}

/* Translate character C by translation table TABLE.  If C
   is negative, translate a character specified by CHARSET, C1, and C2
   (C1 and C2 are code points of the character).  If no translation is
   found in TABLE, return C.  */
int
translate_char (table, c, charset, c1, c2)
     Lisp_Object table;
     int c, charset, c1, c2;
{
  Lisp_Object ch;
  int alt_charset, alt_c1, alt_c2, dimension;

  if (c < 0) c = MAKE_CHAR (charset, (c1 & 0x7F) , (c2 & 0x7F));
  if (!CHAR_TABLE_P (table)
      || (ch = Faref (table, make_number (c)), !NATNUMP (ch)))
    return c;

  SPLIT_CHAR (XFASTINT (ch), alt_charset, alt_c1, alt_c2);
  dimension = CHARSET_DIMENSION (alt_charset);
  if ((dimension == 1 && alt_c1 > 0) || (dimension == 2 && alt_c2 > 0))
    /* CH is not a generic character, just return it.  */
    return XFASTINT (ch);

  /* Since CH is a generic character, we must return a specific
     charater which has the same position codes as C from CH.  */
  if (charset < 0)
    SPLIT_CHAR (c, charset, c1, c2);
  if (dimension != CHARSET_DIMENSION (charset))
    /* We can't make such a character because of dimension mismatch.  */
    return c;
  return MAKE_CHAR (alt_charset, c1, c2);
}

/* Convert the unibyte character C to multibyte based on
   Vnonascii_translation_table or nonascii_insert_offset.  If they can't
   convert C to a valid multibyte character, convert it based on
   DEFAULT_NONASCII_INSERT_OFFSET which makes C a Latin-1 character.  */

int
unibyte_char_to_multibyte (c)
     int c;
{
  if (c < 0400 && c >= 0200)
    {
      int c_save = c;

      if (! NILP (Vnonascii_translation_table))
        {
          c = XINT (Faref (Vnonascii_translation_table, make_number (c)));
          if (c >= 0400 && ! char_valid_p (c, 0))
            c = c_save + DEFAULT_NONASCII_INSERT_OFFSET;
        }
      else if (c >= 0240 && nonascii_insert_offset > 0)
        {
          c += nonascii_insert_offset;
          if (c < 0400 || ! char_valid_p (c, 0))
            c = c_save + DEFAULT_NONASCII_INSERT_OFFSET;
        }
      else if (c >= 0240)
        c = c_save + DEFAULT_NONASCII_INSERT_OFFSET;
    }
  return c;
}


/* Convert the multibyte character C to unibyte 8-bit character based
   on Vnonascii_translation_table or nonascii_insert_offset.  If
   REV_TBL is non-nil, it should be a reverse table of
   Vnonascii_translation_table, i.e. what given by:
     Fchar_table_extra_slot (Vnonascii_translation_table, make_number (0))  */

int
multibyte_char_to_unibyte (c, rev_tbl)
     int c;
     Lisp_Object rev_tbl;
{
  if (!SINGLE_BYTE_CHAR_P (c))
    {
      int c_save = c;

      if (! CHAR_TABLE_P (rev_tbl)
          && CHAR_TABLE_P (Vnonascii_translation_table))
        rev_tbl = Fchar_table_extra_slot (Vnonascii_translation_table,
                                          make_number (0));
      if (CHAR_TABLE_P (rev_tbl))
        {
          Lisp_Object temp;
          temp = Faref (rev_tbl, make_number (c));
          if (INTEGERP (temp))
            c = XINT (temp);
          if (c >= 256)
            c = (c_save & 0177) + 0200;
        }
      else
        {
          if (nonascii_insert_offset > 0)
            c -= nonascii_insert_offset;
          if (c < 128 || c >= 256)
            c = (c_save & 0177) + 0200;
        }
    }

  return c;
}


/* Update the table Vcharset_table with the given arguments (see the
   document of `define-charset' for the meaning of each argument).
   Several other table contents are also updated.  The caller should
   check the validity of CHARSET-ID and the remaining arguments in
   advance.  */

void
update_charset_table (charset_id, dimension, chars, width, direction,
                      iso_final_char, iso_graphic_plane,
                      short_name, long_name, description)
     Lisp_Object charset_id, dimension, chars, width, direction;
     Lisp_Object iso_final_char, iso_graphic_plane;
     Lisp_Object short_name, long_name, description;
{
  int charset = XINT (charset_id);
  int bytes;
  unsigned char leading_code_base, leading_code_ext;

  if (NILP (CHARSET_TABLE_ENTRY (charset)))
    CHARSET_TABLE_ENTRY (charset)
      = Fmake_vector (make_number (CHARSET_MAX_IDX), Qnil);

  if (NILP (long_name))
    long_name = short_name;
  if (NILP (description))
    description = long_name;

  /* Get byte length of multibyte form, base leading-code, and
     extended leading-code of the charset.  See the comment under the
     title "GENERAL NOTE on CHARACTER SET (CHARSET)" in charset.h.  */
  bytes = XINT (dimension);
  if (charset < MIN_CHARSET_PRIVATE_DIMENSION1)
    {
      /* Official charset, it doesn't have an extended leading-code.  */
      if (charset != CHARSET_ASCII && charset != CHARSET_8_BIT_GRAPHIC)
        bytes += 1; /* For a base leading-code.  */
      leading_code_base = charset;
      leading_code_ext = 0;
    }
  else
    {
      /* Private charset.  */
      bytes += 2; /* For base and extended leading-codes.  */
      leading_code_base
        = (charset < LEADING_CODE_EXT_12
           ? LEADING_CODE_PRIVATE_11
           : (charset < LEADING_CODE_EXT_21
              ? LEADING_CODE_PRIVATE_12
              : (charset < LEADING_CODE_EXT_22
                 ? LEADING_CODE_PRIVATE_21
                 : LEADING_CODE_PRIVATE_22)));
      leading_code_ext = charset;
      if (BYTES_BY_CHAR_HEAD (leading_code_base) != bytes)
        error ("Invalid dimension for the charset-ID %d", charset);
    }

  CHARSET_TABLE_INFO (charset, CHARSET_ID_IDX) = charset_id;
  CHARSET_TABLE_INFO (charset, CHARSET_BYTES_IDX) = make_number (bytes);
  CHARSET_TABLE_INFO (charset, CHARSET_DIMENSION_IDX) = dimension;
  CHARSET_TABLE_INFO (charset, CHARSET_CHARS_IDX) = chars;
  CHARSET_TABLE_INFO (charset, CHARSET_WIDTH_IDX) = width;
  CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX) = direction;
  CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_BASE_IDX)
    = make_number (leading_code_base);
  CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_EXT_IDX)
    = make_number (leading_code_ext);
  CHARSET_TABLE_INFO (charset, CHARSET_ISO_FINAL_CHAR_IDX) = iso_final_char;
  CHARSET_TABLE_INFO (charset, CHARSET_ISO_GRAPHIC_PLANE_IDX)
    = iso_graphic_plane;
  CHARSET_TABLE_INFO (charset, CHARSET_SHORT_NAME_IDX) = short_name;
  CHARSET_TABLE_INFO (charset, CHARSET_LONG_NAME_IDX) = long_name;
  CHARSET_TABLE_INFO (charset, CHARSET_DESCRIPTION_IDX) = description;
  CHARSET_TABLE_INFO (charset, CHARSET_PLIST_IDX) = Qnil;

  {
    /* If we have already defined a charset which has the same
       DIMENSION, CHARS and ISO-FINAL-CHAR but the different
       DIRECTION, we must update the entry REVERSE-CHARSET of both
       charsets.  If there's no such charset, the value of the entry
       is set to nil.  */
    int i;

    for (i = 0; i <= MAX_CHARSET; i++)
      if (!NILP (CHARSET_TABLE_ENTRY (i)))
        {
          if (CHARSET_DIMENSION (i) == XINT (dimension)
              && CHARSET_CHARS (i) == XINT (chars)
              && CHARSET_ISO_FINAL_CHAR (i) == XINT (iso_final_char)
              && CHARSET_DIRECTION (i) != XINT (direction))
            {
              CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX)
                = make_number (i);