-/* Basic multilingual character support.
- Copyright (C) 1995, 1997, 1998 Electrotechnical Laboratory, JAPAN.
- Licensed to the Free Software Foundation.
-
-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 2, 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., 59 Temple Place - Suite 330,
-Boston, MA 02111-1307, 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 "coding.h"
-#include "disptab.h"
-
-#else /* not emacs */
-
-#include "mulelib.h"
-
-#endif /* emacs */
-
-Lisp_Object Qcharset, Qascii, Qcomposition;
-Lisp_Object Qunknown;
-
-/* Declaration of special leading-codes. */
-int leading_code_composition; /* for composite characters */
-int leading_code_private_11; /* for private DIMENSION1 of 1-column */
-int leading_code_private_12; /* for private DIMENSION1 of 2-column */
-int leading_code_private_21; /* for private DIMENSION2 of 1-column */
-int leading_code_private_22; /* for private DIMENSION2 of 2-column */
-
-/* Declaration of special charsets. */
-int charset_ascii; /* ASCII */
-int charset_composition; /* for a composite character */
-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 min_composite_char;
-
-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];
-
-/* Table of pointers to the structure `cmpchar_info' indexed by
- CMPCHAR-ID. */
-struct cmpchar_info **cmpchar_table;
-/* The current size of `cmpchar_table'. */
-static int cmpchar_table_size;
-/* Number of the current composite characters. */
-int n_cmpchars;
-
-/* 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. */
-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;
-
-#define min(X, Y) ((X) < (Y) ? (X) : (Y))
-#define max(X, Y) ((X) > (Y) ? (X) : (Y))
-\f
-void
-invalid_character (c)
- int c;
-{
- error ("Invalid character: 0%o, %d, 0x%x", c, c, c);
-}
-
-/* Parse composite character string STR of length LENGTH (>= 2) and
- set BYTES, CHARSET, C1, and C2 as below.
-
- It is assumed that *STR is LEADING_CODE_COMPOSITION and the
- following (LENGTH - 1) bytes satisfy !CHAR_HEAD_P.
-
- If there is a valid composite character, set CHARSET, C1, and C2 to
- such values that MAKE_CHAR can make the composite character from
- them. Otherwise, set CHARSET to CHARSET_COMPOSITION, set C1 to the
- second byte of the sequence, C2 to -1 so that MAKE_CHAR can make
- the invalid multibyte character whose string representation is two
- bytes of STR[0] and STR[1]. In any case, set BYTES to LENGTH.
-
- This macro should be called only from SPLIT_MULTIBYTE_SEQ. */
-
-#define SPLIT_COMPOSITE_SEQ(str, length, bytes, charset, c1, c2) \
- do { \
- int cmpchar_id = str_cmpchar_id ((str), (length)); \
- \
- (charset) = CHARSET_COMPOSITION; \
- (bytes) = (length); \
- if (cmpchar_id >= 0) \
- { \
- (c1) = CHAR_FIELD2 (cmpchar_id); \
- (c2) = CHAR_FIELD3 (cmpchar_id); \
- } \
- else \
- { \
- (c1) = (str)[1] & 0x7F; \
- (c2) = -1; \
- } \
- } while (0)
-
-/* Parse non-composite multibyte character string STR of length LENGTH
- (>= 2) and set BYTES to the length of actual multibyte sequence,
- CHARSET, C1, and C2 to such values that MAKE_CHAR can make the
- multibyte character from them.
-
- It is assumed that *STR is one of base leading codes (excluding
- LEADING_CODE_COMPOSITION) and the following (LENGTH - 1) bytes
- satisfy !CHAR_HEAD_P.
-
- This macro should be called only from SPLIT_MULTIBYTE_SEQ. */
-
-#define SPLIT_CHARACTER_SEQ(str, length, bytes, charset, c1, c2) \
- do { \
- (bytes) = 1; \
- (charset) = (str)[0]; \
- if ((charset) >= LEADING_CODE_PRIVATE_11 \
- && (charset) <= LEADING_CODE_PRIVATE_22) \
- (charset) = (str)[(bytes)++]; \
- if ((bytes) < (length)) \
- { \
- (c1) = (str)[(bytes)++] & 0x7F; \
- if ((bytes) < (length)) \
- (c2) = (str)[(bytes)++] & 0x7F; \
- else \
- (c2) = -1; \
- } \
- else \
- (c1) = (c2) = -1; \
- } while (0)
-
-/* Parse string STR of length LENGTH and check if a multibyte
- characters is at STR. set BYTES to the actual length, CHARSET, C1,
- C2 to proper values for that character. */
-
-#define SPLIT_MULTIBYTE_SEQ(str, length, bytes, charset, c1, c2) \
- do { \
- int i; \
- if (ASCII_BYTE_P ((str)[0])) \
- i = 1; \
- else \
- for (i = 1; i < (length) && ! CHAR_HEAD_P ((str)[i]); i++); \
- if (i == 1) \
- (bytes) = 1, (charset) = CHARSET_ASCII, (c1) = (str)[0] ; \
- else if ((str)[0] == LEADING_CODE_COMPOSITION) \
- SPLIT_COMPOSITE_SEQ (str, i, bytes, charset, c1, c2); \
- else \
- { \
- if (i > BYTES_BY_CHAR_HEAD ((str)[0])) \
- i = BYTES_BY_CHAR_HEAD ((str)[0]); \
- SPLIT_CHARACTER_SEQ (str, i, bytes, charset, c1, c2); \
- } \
- } while (0)
-
-/* 1 if CHARSET, C1, and C2 compose a valid character, else 0. */
-#define CHAR_COMPONENTS_VALID_P(charset, c1, c2) \
- (charset == CHARSET_ASCII \
- ? ((c1) >= 0 && (c1) <= 0x7F) \
- : (CHARSET_DIMENSION (charset) == 1 \
- ? ((c1) >= 0x20 && (c1) <= 0x7F) \
- : ((c1) >= 0x20 && (c1) <= 0x7F && (c2) >= 0x20 && (c2) <= 0x7F)))
-
-/* Set STR a pointer to the multi-byte form of the character C. If C
- is not a composite character, the multi-byte form is set in WORKBUF
- and STR points WORKBUF. The caller should allocate at least 4-byte
- area at WORKBUF in advance. Returns the length of the multi-byte
- form. If C is an invalid character, store (C & 0xFF) in WORKBUF[0]
- and return 1.
-
- Use macro `CHAR_STRING (C, WORKBUF, STR)' instead of calling this
- function directly if C can be an ASCII character. */
-
-int
-non_ascii_char_to_string (c, workbuf, str)
- int c;
- unsigned char *workbuf, **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)))
- invalid_character (c);
-
- /* 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, it is an invalid character. */
- if (c & CHAR_MODIFIER_MASK)
- invalid_character (c);
-
- *str = workbuf;
- *workbuf++ = c;
- }
- else
- {
- int charset, c1, c2;
-
- SPLIT_NON_ASCII_CHAR (c, charset, c1, c2);
- if (charset == CHARSET_COMPOSITION)
- {
- if (c >= MAX_CHAR)
- invalid_character (c);
- if (c >= MIN_CHAR_COMPOSITION)
- {
- /* Valid composite character. */
- *str = cmpchar_table[COMPOSITE_CHAR_ID (c)]->data;
- workbuf = *str + cmpchar_table[COMPOSITE_CHAR_ID (c)]->len;
- }
- else
- {
- /* Invalid but can have multibyte form. */
- *str = workbuf;
- *workbuf++ = LEADING_CODE_COMPOSITION;
- *workbuf++ = c1 | 0x80;
- }
- }
- else if (charset > CHARSET_COMPOSITION)
- {
- *str = workbuf;
- if (charset >= LEADING_CODE_EXT_11)
- *workbuf++ = (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)));
- *workbuf++ = charset;
- if (c1 > 0 && c1 < 32 || c2 > 0 && c2 < 32)
- invalid_character (c);
- if (c1)
- {
- *workbuf++ = c1 | 0x80;
- if (c2 > 0)
- *workbuf++ = c2 | 0x80;
- }
- }
- else if (charset == CHARSET_ASCII)
- *workbuf++= c & 0x7F;
- else
- invalid_character (c);
- }
-
- return (workbuf - *str);
-}
-
-/* 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 macro `STRING_CHAR (STR, LEN)' instead of calling this function
- directly if you want ot handle ASCII characters as well. */
-
-int
-string_to_non_ascii_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_non_ascii_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 iff character C has valid printable glyph.
- Use the macro CHAR_PRINTABLE_P instead. */
-int
-char_printable_p (c)
- int c;
-{
- int charset, c1, c2, chars;
-
- if (SINGLE_BYTE_CHAR_P (c))
- return 1;
- if (c >= MIN_CHAR_COMPOSITION)
- return (c < MAX_CHAR);
-
- SPLIT_NON_ASCII_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;
-}
-
-\f
-/* 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);
-
- /* 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)
- 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);
- CHARSET_TABLE_INFO (i, CHARSET_REVERSE_CHARSET_IDX) = charset_id;
- break;
- }
- }
- if (i > MAX_CHARSET)
- /* No such a charset. */
- CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX)
- = make_number (-1);
- }
-
- if (charset != CHARSET_ASCII
- && charset < MIN_CHARSET_PRIVATE_DIMENSION1)
- {
- width_by_char_head[leading_code_base] = XINT (width);
-
- /* Update table emacs_code_class. */
- emacs_code_class[charset] = (bytes == 2
- ? EMACS_leading_code_2
- : (bytes == 3
- ? EMACS_leading_code_3
- : EMACS_leading_code_4));
- }
-
- /* Update table iso_charset_table. */
- if (ISO_CHARSET_TABLE (dimension, chars, iso_final_char) < 0)
- ISO_CHARSET_TABLE (dimension, chars, iso_final_char) = charset;
-}
-
-#ifdef emacs
-
-/* Return charset id of CHARSET_SYMBOL, or return -1 if CHARSET_SYMBOL
- is invalid. */
-int
-get_charset_id (charset_symbol)
- Lisp_Object charset_symbol;
-{
- Lisp_Object val;
- int charset;
-
- return ((SYMBOLP (charset_symbol)
- && (val = Fget (charset_symbol, Qcharset), VECTORP (val))
- && (charset = XINT (XVECTOR (val)->contents[CHARSET_ID_IDX]),
- CHARSET_VALID_P (charset)))
- ? charset : -1);
-}
-
-/* Return an identification number for a new private charset of
- DIMENSION and WIDTH. If there's no more room for the new charset,
- return 0. */
-Lisp_Object
-get_new_private_charset_id (dimension, width)
- int dimension, width;
-{
- int charset, from, to;
-
- if (dimension == 1)
- {
- if (width == 1)
- from = LEADING_CODE_EXT_11, to = LEADING_CODE_EXT_12;
- else
- from = LEADING_CODE_EXT_12, to = LEADING_CODE_EXT_21;
- }
- else
- {
- if (width == 1)
- from = LEADING_CODE_EXT_21, to = LEADING_CODE_EXT_22;
- else
- from = LEADING_CODE_EXT_22, to = LEADING_CODE_EXT_MAX + 1;
- }
-
- for (charset = from; charset < to; charset++)
- if (!CHARSET_DEFINED_P (charset)) break;
-
- return make_number (charset < to ? charset : 0);
-}
-
-DEFUN ("define-charset", Fdefine_charset, Sdefine_charset, 3, 3, 0,
- "Define CHARSET-ID as the identification number of CHARSET with INFO-VECTOR.\n\
-If CHARSET-ID is nil, it is decided automatically, which means CHARSET is\n\
- treated as a private charset.\n\
-INFO-VECTOR is a vector of the format:\n\
- [DIMENSION CHARS WIDTH DIRECTION ISO-FINAL-CHAR ISO-GRAPHIC-PLANE\n\
- SHORT-NAME LONG-NAME DESCRIPTION]\n\
-The meanings of each elements is as follows:\n\
-DIMENSION (integer) is the number of bytes to represent a character: 1 or 2.\n\
-CHARS (integer) is the number of characters in a dimension: 94 or 96.\n\
-WIDTH (integer) is the number of columns a character in the charset\n\
-occupies on the screen: one of 0, 1, and 2.\n\
-\n\
-DIRECTION (integer) is the rendering direction of characters in the\n\
-charset when rendering. If 0, render from left to right, else\n\
-render from right to left.\n\
-\n\
-ISO-FINAL-CHAR (character) is the final character of the\n\
-corresponding ISO 2022 charset.\n\
-\n\
-ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked\n\
-while encoding to variants of ISO 2022 coding system, one of the\n\
-following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).\n\
-\n\
-SHORT-NAME (string) is the short name to refer to the charset.\n\
-\n\
-LONG-NAME (string) is the long name to refer to the charset.\n\
-\n\
-DESCRIPTION (string) is the description string of the charset.")
- (charset_id, charset_symbol, info_vector)
- Lisp_Object charset_id, charset_symbol, info_vector;
-{
- Lisp_Object *vec;
-
- if (!NILP (charset_id))
- CHECK_NUMBER (charset_id, 0);
- CHECK_SYMBOL (charset_symbol, 1);
- CHECK_VECTOR (info_vector, 2);
-
- if (! NILP (charset_id))
- {
- if (! CHARSET_VALID_P (XINT (charset_id)))
- error ("Invalid CHARSET: %d", XINT (charset_id));
- else if (CHARSET_DEFINED_P (XINT (charset_id)))
- error ("Already defined charset: %d", XINT (charset_id));
- }
-
- vec = XVECTOR (info_vector)->contents;
- if (XVECTOR (info_vector)->size != 9
- || !INTEGERP (vec[0]) || !(XINT (vec[0]) == 1 || XINT (vec[0]) == 2)
- || !INTEGERP (vec[1]) || !(XINT (vec[1]) == 94 || XINT (vec[1]) == 96)
- || !INTEGERP (vec[2]) || !(XINT (vec[2]) == 1 || XINT (vec[2]) == 2)
- || !INTEGERP (vec[3]) || !(XINT (vec[3]) == 0 || XINT (vec[3]) == 1)
- || !INTEGERP (vec[4]) || !(XINT (vec[4]) >= '0' && XINT (vec[4]) <= '~')
- || !INTEGERP (vec[5]) || !(XINT (vec[5]) == 0 || XINT (vec[5]) == 1)
- || !STRINGP (vec[6])
- || !STRINGP (vec[7])
- || !STRINGP (vec[8]))
- error ("Invalid info-vector argument for defining charset %s",
- XSYMBOL (charset_symbol)->name->data);
-
- if (NILP (charset_id))
- {
- charset_id = get_new_private_charset_id (XINT (vec[0]), XINT (vec[2]));
- if (XINT (charset_id) == 0)
- error ("There's no room for a new private charset %s",
- XSYMBOL (charset_symbol)->name->data);
- }
-
- update_charset_table (charset_id, vec[0], vec[1], vec[2], vec[3],
- vec[4], vec[5], vec[6], vec[7], vec[8]);
- Fput (charset_symbol, Qcharset, CHARSET_TABLE_ENTRY (XINT (charset_id)));
- CHARSET_SYMBOL (XINT (charset_id)) = charset_symbol;
- Vcharset_list = Fcons (charset_symbol, Vcharset_list);
- return Qnil;
-}
-
-DEFUN ("generic-character-list", Fgeneric_character_list,
- Sgeneric_character_list, 0, 0, 0,
- "Return a list of all possible generic characters.\n\
-It includes a generic character for a charset not yet defined.")
- ()
-{
- return Vgeneric_character_list;
-}
-
-DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char,
- Sget_unused_iso_final_char, 2, 2, 0,
- "Return an unsed ISO's final char for a charset of DIMENISION and CHARS.\n\
-DIMENSION is the number of bytes to represent a character: 1 or 2.\n\
-CHARS is the number of characters in a dimension: 94 or 96.\n\
-\n\
-This final char is for private use, thus the range is `0' (48) .. `?' (63).\n\
-If there's no unused final char for the specified kind of charset,\n\
-return nil.")
- (dimension, chars)
- Lisp_Object dimension, chars;
-{
- int final_char;
-
- CHECK_NUMBER (dimension, 0);
- CHECK_NUMBER (chars, 1);
- if (XINT (dimension) != 1 && XINT (dimension) != 2)
- error ("Invalid charset dimension %d, it should be 1 or 2",
- XINT (dimension));
- if (XINT (chars) != 94 && XINT (chars) != 96)
- error ("Invalid charset chars %d, it should be 94 or 96",
- XINT (chars));
- for (final_char = '0'; final_char <= '?'; final_char++)
- {
- if (ISO_CHARSET_TABLE (dimension, chars, make_number (final_char)) < 0)
- break;
- }
- return (final_char <= '?' ? make_number (final_char) : Qnil);
-}
-
-DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset, Sdeclare_equiv_charset,
- 4, 4, 0,
- "Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET.\n\
-CHARSET should be defined by `defined-charset' in advance.")
- (dimension, chars, final_char, charset_symbol)
- Lisp_Object dimension, chars, final_char, charset_symbol;
-{
- int charset;
-
- CHECK_NUMBER (dimension, 0);
- CHECK_NUMBER (chars, 1);
- CHECK_NUMBER (final_char, 2);
- CHECK_SYMBOL (charset_symbol, 3);
-
- if (XINT (dimension) != 1 && XINT (dimension) != 2)
- error ("Invalid DIMENSION %d, it should be 1 or 2", XINT (dimension));
- if (XINT (chars) != 94 && XINT (chars) != 96)
- error ("Invalid CHARS %d, it should be 94 or 96", XINT (chars));
- if (XINT (final_char) < '0' || XFASTINT (final_char) > '~')
- error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", XINT (chars));
- if ((charset = get_charset_id (charset_symbol)) < 0)
- error ("Invalid charset %s", XSYMBOL (charset_symbol)->name->data);
-
- ISO_CHARSET_TABLE (dimension, chars, final_char) = charset;
- return Qnil;
-}
-
-/* Return number of different charsets in STR of length LEN. In
- addition, for each found charset N, CHARSETS[N] is set 1. The
- caller should allocate CHARSETS (MAX_CHARSET + 1 elements) in advance.
- It may lookup a translation table TABLE if supplied.
-
- If CMPCHARP is nonzero and some composite character is found,
- CHARSETS[128] is also set 1 and the returned number is incremented
- by 1.
-
- If MULTIBYTE is zero, do not check multibyte characters, i.e. if
- any ASCII codes (7-bit) are found, CHARSET[0] is set to 1, if any
- 8-bit codes are found CHARSET[1] is set to 1. */
-
-int
-find_charset_in_str (str, len, charsets, table, cmpcharp, multibyte)
- unsigned char *str;
- int len, *charsets;
- Lisp_Object table;
- int cmpcharp;
- int multibyte;
-{
- register int num = 0, c;
-
- if (! multibyte)
- {
- unsigned char *endp = str + len;
- int maskbits = 0;
-
- while (str < endp && maskbits != 3)
- maskbits |= (*str++ < 0x80 ? 1 : 2);
- if (maskbits & 1)
- {
- charsets[0] = 1;
- num++;
- }
- if (maskbits & 2)
- {
- charsets[1] = 1;
- num++;
- }
- return num;
- }
-
- if (! CHAR_TABLE_P (table))
- table = Qnil;
-
- while (len > 0)
- {
- int bytes, charset;
- c = *str;
-
- if (c == LEADING_CODE_COMPOSITION)
- {
- int cmpchar_id = str_cmpchar_id (str, len);
- GLYPH *glyph;
-
- if (cmpchar_id >= 0)
- {
- struct cmpchar_info *cmp_p = cmpchar_table[cmpchar_id];
- int i;
-
- for (i = 0; i < cmp_p->glyph_len; i++)
- {
- c = cmp_p->glyph[i];
- if (!NILP (table))
- {
- if ((c = translate_char (table, c, 0, 0, 0)) < 0)
- c = cmp_p->glyph[i];
- }
- if ((charset = CHAR_CHARSET (c)) < 0)
- charset = CHARSET_ASCII;
- if (!charsets[charset])
- {
- charsets[charset] = 1;
- num += 1;
- }
- }
- str += cmp_p->len;
- len -= cmp_p->len;
- if (cmpcharp && !charsets[CHARSET_COMPOSITION])
- {
- charsets[CHARSET_COMPOSITION] = 1;
- num += 1;
- }
- continue;
- }
-
- charset = 1; /* This leads to `unknown' charset. */
- bytes = 1;
- }
- else
- {
- c = STRING_CHAR_AND_LENGTH (str, len, bytes);
- if (! NILP (table))
- {
- int c1 = translate_char (table, c, 0, 0, 0);
- if (c1 >= 0)
- c = c1;
- }
- charset = CHAR_CHARSET (c);
- }
-
- if (!charsets[charset])
- {
- charsets[charset] = 1;
- num += 1;
- }
- str += bytes;
- len -= bytes;
- }
- return num;
-}
-
-DEFUN ("find-charset-region", Ffind_charset_region, Sfind_charset_region,
- 2, 3, 0,
- "Return a list of charsets in the region between BEG and END.\n\
-BEG and END are buffer positions.\n\
-If the region contains any composite character,\n\
-`composition' is included in the returned list.\n\
-Optional arg TABLE if non-nil is a translation table to look up.\n\
-\n\
-If the region contains invalid multiybte characters,\n\
-`unknown' is included in the returned list.\n\
-\n\
-If the current buffer is unibyte, the returned list contains\n\
-`ascii' if any 7-bit characters are found,\n\
-and `unknown' if any 8-bit characters are found.")
- (beg, end, table)
- Lisp_Object beg, end, table;
-{
- int charsets[MAX_CHARSET + 1];
- int from, from_byte, to, stop, stop_byte, i;
- Lisp_Object val;
- int undefined;
- int multibyte = !NILP (current_buffer->enable_multibyte_characters);
-
- validate_region (&beg, &end);
- from = XFASTINT (beg);
- stop = to = XFASTINT (end);
-
- if (from < GPT && GPT < to)
- {
- stop = GPT;
- stop_byte = GPT_BYTE;
- }
- else
- stop_byte = CHAR_TO_BYTE (stop);
-
- from_byte = CHAR_TO_BYTE (from);
-
- bzero (charsets, (MAX_CHARSET + 1) * sizeof (int));
- while (1)
- {
- find_charset_in_str (BYTE_POS_ADDR (from_byte), stop_byte - from_byte,
- charsets, table, 1, multibyte);
- if (stop < to)
- {
- from = stop, from_byte = stop_byte;
- stop = to, stop_byte = CHAR_TO_BYTE (stop);
- }
- else
- break;
- }
-
- val = Qnil;
- undefined = 0;
- for (i = (multibyte ? MAX_CHARSET : 1); i >= 0; i--)
- if (charsets[i])
- {
- if (CHARSET_DEFINED_P (i) || i == CHARSET_COMPOSITION)
- val = Fcons (CHARSET_SYMBOL (i), val);
- else
- undefined = 1;
- }
- if (undefined)
- val = Fcons (Qunknown, val);
- return val;
-}
-
-DEFUN ("find-charset-string", Ffind_charset_string, Sfind_charset_string,
- 1, 2, 0,
- "Return a list of charsets in STR.\n\
-If the string contains any composite characters,\n\
-`composition' is included in the returned list.\n\
-Optional arg TABLE if non-nil is a translation table to look up.\n\
-\n\
-If the region contains invalid multiybte characters,\n\
-`unknown' is included in the returned list.\n\
-\n\
-If STR is unibyte, the returned list contains\n\
-`ascii' if any 7-bit characters are found,\n\
-and `unknown' if any 8-bit characters are found.")
- (str, table)
- Lisp_Object str, table;
-{
- int charsets[MAX_CHARSET + 1];
- int i;
- Lisp_Object val;
- int undefined;
- int multibyte;
-
- CHECK_STRING (str, 0);
- multibyte = STRING_MULTIBYTE (str);
-
- bzero (charsets, (MAX_CHARSET + 1) * sizeof (int));
- find_charset_in_str (XSTRING (str)->data, STRING_BYTES (XSTRING (str)),
- charsets, table, 1, multibyte);
- val = Qnil;
- undefined = 0;
- for (i = (multibyte ? MAX_CHARSET : 1); i >= 0; i--)
- if (charsets[i])
- {
- if (CHARSET_DEFINED_P (i) || i == CHARSET_COMPOSITION)
- val = Fcons (CHARSET_SYMBOL (i), val);
- else
- undefined = 1;
- }
- if (undefined)
- val = Fcons (Qunknown, val);
- return val;
-}
-\f
-DEFUN ("make-char-internal", Fmake_char_internal, Smake_char_internal, 1, 3, 0,
- "")
- (charset, code1, code2)
- Lisp_Object charset, code1, code2;
-{
- int charset_id, c1, c2;
-
- CHECK_NUMBER (charset, 0);
- charset_id = XINT (charset);
- if (!CHARSET_DEFINED_P (charset_id))
- error ("Invalid charset ID: %d", XINT (charset));
-
- if (NILP (code1))
- c1 = 0;
- else
- {
- CHECK_NUMBER (code1, 1);
- c1 = XINT (code1);
- }
- if (NILP (code2))
- c2 = 0;
- else
- {
- CHECK_NUMBER (code2, 2);
- c2 = XINT (code2);
- }
-
- if (c1 < 0 || c1 > 0xFF || c2 < 0 || c2 > 0xFF)
- error ("Invalid code points for charset ID %d: %d %d", charset_id, c1, c2);
- c1 &= 0x7F;
- c2 &= 0x7F;
- if (c1 == 0
- ? c2 != 0
- : (c2 == 0
- ? !CHAR_COMPONENTS_VALID_P (charset, c1, 0x20)
- : !CHAR_COMPONENTS_VALID_P (charset, c1, c2)))
- error ("Invalid code points for charset ID %d: %d %d", charset_id, c1, c2);
-
- return make_number (MAKE_CHAR (charset_id, c1, c2));
-}
-
-DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0,
- "Return list of charset and one or two position-codes of CHAR.\n\
-If CHAR is invalid as a character code,\n\
-return a list of symbol `unknown' and CHAR.")
- (ch)
- Lisp_Object ch;
-{
- Lisp_Object val;
- int c, charset, c1, c2;
-
- CHECK_NUMBER (ch, 0);
- c = XFASTINT (ch);
- if (!CHAR_VALID_P (c, 1))
- return Fcons (Qunknown, Fcons (ch, Qnil));
- SPLIT_CHAR (XFASTINT (ch), charset, c1, c2);
- return (c2 >= 0
- ? Fcons (CHARSET_SYMBOL (charset),
- Fcons (make_number (c1), Fcons (make_number (c2), Qnil)))
- : Fcons (CHARSET_SYMBOL (charset), Fcons (make_number (c1), Qnil)));
-}
-
-DEFUN ("char-charset", Fchar_charset, Schar_charset, 1, 1, 0,
- "Return charset of CHAR.")
- (ch)
- Lisp_Object ch;
-{
- CHECK_NUMBER (ch, 0);
-
- return CHARSET_SYMBOL (CHAR_CHARSET (XINT (ch)));
-}
-
-DEFUN ("charset-after", Fcharset_after, Scharset_after, 0, 1, 0,
- "Return charset of a character in the current buffer at position POS.\n\
-If POS is nil, it defauls to the current point.\n\
-If POS is out of range, the value is nil.")
- (pos)
- Lisp_Object pos;
-{
- register int pos_byte, bytes, charset, c1, c2;
- register unsigned char *p;
-
- if (NILP (pos))
- pos_byte = PT_BYTE;
- else if (MARKERP (pos))
- {
- pos_byte = marker_byte_position (pos);
- if (pos_byte < BEGV_BYTE || pos_byte >= ZV_BYTE)
- return Qnil;
- }
- else
- {
- CHECK_NUMBER (pos, 0);
- if (XINT (pos) < BEGV || XINT (pos) >= ZV)
- return Qnil;
- pos_byte = CHAR_TO_BYTE (XINT (pos));
- }
- p = BYTE_POS_ADDR (pos_byte);
- if (BASE_LEADING_CODE_P (*p))
- {
- SPLIT_MULTIBYTE_SEQ (p, Z_BYTE - pos_byte, bytes, charset, c1, c2);
- if (charset < 0)
- charset = 1;
- }
- else
- charset = CHARSET_ASCII;
-
- return CHARSET_SYMBOL (charset);
-}
-
-DEFUN ("iso-charset", Fiso_charset, Siso_charset, 3, 3, 0,
- "Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR.\n\
-\n\
-ISO 2022's designation sequence (escape sequence) distinguishes charsets\n\
-by their DIMENSION, CHARS, and FINAL-CHAR,\n\
-where as Emacs distinguishes them by charset symbol.\n\
-See the documentation of the function `charset-info' for the meanings of\n\
-DIMENSION, CHARS, and FINAL-CHAR.")
- (dimension, chars, final_char)
- Lisp_Object dimension, chars, final_char;
-{
- int charset;
-
- CHECK_NUMBER (dimension, 0);
- CHECK_NUMBER (chars, 1);
- CHECK_NUMBER (final_char, 2);
-
- if ((charset = ISO_CHARSET_TABLE (dimension, chars, final_char)) < 0)
- return Qnil;
- return CHARSET_SYMBOL (charset);
-}
-
-/* If GENERICP is nonzero, return nonzero iff C is a valid normal or
- generic character. If GENERICP is zero, return nonzero iff C is a
- valid normal character. Do not call this function directly,
- instead use macro CHAR_VALID_P. */
-int
-char_valid_p (c, genericp)
- int c, genericp;
-{
- int charset, c1, c2;
-
- if (c < 0)
- return 0;
- if (SINGLE_BYTE_CHAR_P (c))
- return 1;
- SPLIT_NON_ASCII_CHAR (c, charset, c1, c2);
- if (charset == CHARSET_COMPOSITION)
- return ((c >= MIN_CHAR_COMPOSITION
- && c < MIN_CHAR_COMPOSITION + n_cmpchars)
- || (genericp && c == GENERIC_COMPOSITION_CHAR));
- if (genericp)
- {
- if (c1)
- {
- if (c2 <= 0) c2 = 0x20;
- }
- else
- {
- if (c2 <= 0) c1 = c2 = 0x20;
- }
- }
- return (CHARSET_DEFINED_P (charset)
- && CHAR_COMPONENTS_VALID_P (charset, c1, c2));
-}
-
-DEFUN ("char-valid-p", Fchar_valid_p, Schar_valid_p, 1, 2, 0,
- "Return t if OBJECT is a valid normal character.\n\
-If optional arg GENERICP is non-nil, also return t if OBJECT is\n\
-a valid generic character.")
- (object, genericp)
- Lisp_Object object, genericp;
-{
- if (! NATNUMP (object))
- return Qnil;
- return (CHAR_VALID_P (XFASTINT (object), !NILP (genericp)) ? Qt : Qnil);
-}
-
-DEFUN ("unibyte-char-to-multibyte", Funibyte_char_to_multibyte,
- Sunibyte_char_to_multibyte, 1, 1, 0,
- "Convert the unibyte character CH to multibyte character.\n\
-The conversion is done based on `nonascii-translation-table' (which see)\n\
- or `nonascii-insert-offset' (which see).")
- (ch)
- Lisp_Object ch;
-{
- int c;
-
- CHECK_NUMBER (ch, 0);
- c = XINT (ch);
- if (c < 0 || c >= 0400)
- error ("Invalid unibyte character: %d", c);
- c = unibyte_char_to_multibyte (c);
- if (c < 0)
- error ("Can't convert to multibyte character: %d", XINT (ch));
- return make_number (c);
-}
-
-DEFUN ("multibyte-char-to-unibyte", Fmultibyte_char_to_unibyte,
- Smultibyte_char_to_unibyte, 1, 1, 0,
- "Convert the multibyte character CH to unibyte character.\n\
-The conversion is done based on `nonascii-translation-table' (which see)\n\
- or `nonascii-insert-offset' (which see).")
- (ch)
- Lisp_Object ch;
-{
- int c;
-
- CHECK_NUMBER (ch, 0);
- c = XINT (ch);
- if (! CHAR_VALID_P (c, 0))
- error ("Invalid multibyte character: %d", c);
- c = multibyte_char_to_unibyte (c, Qnil);
- if (c < 0)
- error ("Can't convert to unibyte character: %d", XINT (ch));
- return make_number (c);
-}
-
-DEFUN ("char-bytes", Fchar_bytes, Schar_bytes, 1, 1, 0,
- "Return 1 regardless of the argument CHAR.\n\
-This is now an obsolete function. We keep it just for backward compatibility.")
- (ch)
- Lisp_Object ch;
-{
- Lisp_Object val;
-
- CHECK_NUMBER (ch, 0);
- return make_number (1);
-}
-
-/* Return how many bytes C will occupy in a multibyte buffer.
- Don't call this function directly, instead use macro CHAR_BYTES. */
-int
-char_bytes (c)
- int c;
-{
- int bytes;
-
- if (SINGLE_BYTE_CHAR_P (c) || (c & ~GLYPH_MASK_CHAR))
- return 1;
-
- if (COMPOSITE_CHAR_P (c))
- {
- unsigned int id = COMPOSITE_CHAR_ID (c);
-
- bytes = (id < n_cmpchars ? cmpchar_table[id]->len : 1);
- }
- else
- {
- int charset = CHAR_CHARSET (c);
-
- bytes = CHARSET_DEFINED_P (charset) ? CHARSET_BYTES (charset) : 1;
- }
-
- return bytes;
-}
-
-/* Return the width of character of which multi-byte form starts with
- C. The width is measured by how many columns occupied on the
- screen when displayed in the current buffer. */
-
-#define ONE_BYTE_CHAR_WIDTH(c) \
- (c < 0x20 \
- ? (c == '\t' \
- ? XFASTINT (current_buffer->tab_width) \
- : (c == '\n' ? 0 : (NILP (current_buffer->ctl_arrow) ? 4 : 2))) \
- : (c < 0x7f \
- ? 1 \
- : (c == 0x7F \
- ? (NILP (current_buffer->ctl_arrow) ? 4 : 2) \
- : ((! NILP (current_buffer->enable_multibyte_characters) \
- && BASE_LEADING_CODE_P (c)) \
- ? WIDTH_BY_CHAR_HEAD (c) \
- : 4))))
-
-DEFUN ("char-width", Fchar_width, Schar_width, 1, 1, 0,
- "Return width of CHAR when displayed in the current buffer.\n\
-The width is measured by how many columns it occupies on the screen.")
- (ch)
- Lisp_Object ch;
-{
- Lisp_Object val, disp;
- int c;
- struct Lisp_Char_Table *dp = buffer_display_table ();
-
- CHECK_NUMBER (ch, 0);
-
- c = XINT (ch);
-
- /* Get the way the display table would display it. */
- disp = dp ? DISP_CHAR_VECTOR (dp, c) : Qnil;
-
- if (VECTORP (disp))
- XSETINT (val, XVECTOR (disp)->size);
- else if (SINGLE_BYTE_CHAR_P (c))
- XSETINT (val, ONE_BYTE_CHAR_WIDTH (c));
- else if (COMPOSITE_CHAR_P (c))
- {
- int id = COMPOSITE_CHAR_ID (XFASTINT (ch));
- XSETFASTINT (val, (id < n_cmpchars ? cmpchar_table[id]->width : 1));
- }
- else
- {
- int charset = CHAR_CHARSET (c);
-
- XSETFASTINT (val, CHARSET_WIDTH (charset));
- }
- return val;
-}
-
-/* Return width of string STR of length LEN when displayed in the
- current buffer. The width is measured by how many columns it
- occupies on the screen. */
-
-int
-strwidth (str, len)
- unsigned char *str;
- int len;
-{
- unsigned char *endp = str + len;
- int width = 0;
- struct Lisp_Char_Table *dp = buffer_display_table ();
-
- while (str < endp)
- {
- if (*str == LEADING_CODE_COMPOSITION)
- {
- int id = str_cmpchar_id (str, endp - str);
-
- if (id < 0)
- {
- width += 4;
- str++;
- }
- else
- {
- width += cmpchar_table[id]->width;
- str += cmpchar_table[id]->len;
- }
- }
- else
- {
- Lisp_Object disp;
- int thislen;
- int c = STRING_CHAR_AND_LENGTH (str, endp - str, thislen);
-
- /* Get the way the display table would display it. */
- if (dp)
- disp = DISP_CHAR_VECTOR (dp, c);
- else
- disp = Qnil;
-
- if (VECTORP (disp))
- width += XVECTOR (disp)->size;
- else
- width += ONE_BYTE_CHAR_WIDTH (*str);
-
- str += thislen;
- }
- }
- return width;
-}
-
-DEFUN ("string-width", Fstring_width, Sstring_width, 1, 1, 0,
- "Return width of STRING when displayed in the current buffer.\n\
-Width is measured by how many columns it occupies on the screen.\n\
-When calculating width of a multibyte character in STRING,\n\
-only the base leading-code is considered; the validity of\n\
-the following bytes is not checked.")
- (str)
- Lisp_Object str;
-{
- Lisp_Object val;
-
- CHECK_STRING (str, 0);
- XSETFASTINT (val, strwidth (XSTRING (str)->data,
- STRING_BYTES (XSTRING (str))));
- return val;
-}
-
-DEFUN ("char-direction", Fchar_direction, Schar_direction, 1, 1, 0,
- "Return the direction of CHAR.\n\
-The returned value is 0 for left-to-right and 1 for right-to-left.")
- (ch)
- Lisp_Object ch;
-{
- int charset;
-
- CHECK_NUMBER (ch, 0);
- charset = CHAR_CHARSET (XFASTINT (ch));
- if (!CHARSET_DEFINED_P (charset))
- invalid_character (XINT (ch));
- return CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX);
-}
-
-DEFUN ("chars-in-region", Fchars_in_region, Schars_in_region, 2, 2, 0,
- "Return number of characters between BEG and END.")
- (beg, end)
- Lisp_Object beg, end;
-{
- int from, to;
-
- CHECK_NUMBER_COERCE_MARKER (beg, 0);
- CHECK_NUMBER_COERCE_MARKER (end, 1);
-
- from = min (XFASTINT (beg), XFASTINT (end));
- to = max (XFASTINT (beg), XFASTINT (end));
-
- return make_number (to - from);
-}
-
-/* Return the number of characters in the NBYTES bytes at PTR.
- This works by looking at the contents and checking for multibyte sequences.
- However, if the current buffer has enable-multibyte-characters = nil,
- we treat each byte as a character. */
-
-int
-chars_in_text (ptr, nbytes)
- unsigned char *ptr;
- int nbytes;
-{
- /* current_buffer is null at early stages of Emacs initialization. */
- if (current_buffer == 0
- || NILP (current_buffer->enable_multibyte_characters))
- return nbytes;
-
- return multibyte_chars_in_text (ptr, nbytes);
-}
-
-/* Return the number of characters in the NBYTES bytes at PTR.
- This works by looking at the contents and checking for multibyte sequences.
- It ignores enable-multibyte-characters. */
-
-int
-multibyte_chars_in_text (ptr, nbytes)
- unsigned char *ptr;
- int nbytes;
-{
- unsigned char *endp;
- int chars, bytes;
-
- endp = ptr + nbytes;
- chars = 0;
-
- while (ptr < endp)
- {
- if (BASE_LEADING_CODE_P (*ptr))
- {
- PARSE_MULTIBYTE_SEQ (ptr, nbytes, bytes);
- ptr += bytes;
- nbytes -= bytes;
- }
- else
- {
- ptr++;
- nbytes--;
- }
- chars++;
- }
-
- return chars;
-}
-
-DEFUN ("string", Fstring, Sstring, 1, MANY, 0,
- "Concatenate all the argument characters and make the result a string.")
- (n, args)
- int n;
- Lisp_Object *args;
-{
- int i;
- unsigned char *buf
- = (unsigned char *) alloca (MAX_LENGTH_OF_MULTI_BYTE_FORM * n);
- unsigned char *p = buf;
- Lisp_Object val;
-
- for (i = 0; i < n; i++)
- {
- int c, len;
- unsigned char *str;
-
- if (!INTEGERP (args[i]))
- CHECK_NUMBER (args[i], 0);
- c = XINT (args[i]);
- len = CHAR_STRING (c, p, str);
- if (p != str)
- /* C is a composite character. */
- bcopy (str, p, len);
- p += len;
- }
-
- /* Here, we can't use make_string_from_bytes because of byte
- combining problem. */
- val = make_string (buf, p - buf);
- return val;
-}
-
-#endif /* emacs */
-\f
-/*** Composite characters staffs ***/
-
-/* Each composite character is identified by CMPCHAR-ID which is
- assigned when Emacs needs the character code of the composite
- character (e.g. when displaying it on the screen). See the
- document "GENERAL NOTE on COMPOSITE CHARACTER" in `charset.h' how a
- composite character is represented in Emacs. */
-
-/* If `static' is defined, it means that it is defined to null string. */
-#ifndef static
-/* The following function is copied from lread.c. */
-static int
-hash_string (ptr, len)
- unsigned char *ptr;
- int len;
-{
- register unsigned char *p = ptr;
- register unsigned char *end = p + len;
- register unsigned char c;
- register int hash = 0;
-
- while (p != end)
- {
- c = *p++;
- if (c >= 0140) c -= 40;
- hash = ((hash<<3) + (hash>>28) + c);
- }
- return hash & 07777777777;
-}
-#endif
-
-#define CMPCHAR_HASH_TABLE_SIZE 0xFFF
-
-static int *cmpchar_hash_table[CMPCHAR_HASH_TABLE_SIZE];
-
-/* Each element of `cmpchar_hash_table' is a pointer to an array of
- integer, where the 1st element is the size of the array, the 2nd
- element is how many elements are actually used in the array, and
- the remaining elements are CMPCHAR-IDs of composite characters of
- the same hash value. */
-#define CMPCHAR_HASH_SIZE(table) table[0]
-#define CMPCHAR_HASH_USED(table) table[1]
-#define CMPCHAR_HASH_CMPCHAR_ID(table, i) table[i]
-
-/* Return CMPCHAR-ID of the composite character in STR of the length
- LEN. If the composite character has not yet been registered,
- register it in `cmpchar_table' and assign new CMPCHAR-ID. This
- is the sole function for assigning CMPCHAR-ID. */
-int
-str_cmpchar_id (str, len)
- const unsigned char *str;
- int len;
-{
- int hash_idx, *hashp;
- unsigned char *buf;
- int embedded_rule; /* 1 if composition rule is embedded. */
- int chars; /* number of components. */
- int i;
- struct cmpchar_info *cmpcharp;
-
- /* The second byte 0xFF means COMPOSITION rule is embedded. */
- embedded_rule = (str[1] == 0xFF);
-
- /* At first, get the actual length of the composite character. */
- {
- const unsigned char *p, *endp = str + 1, *lastp = str + len;
- int bytes;
-
- while (endp < lastp && ! CHAR_HEAD_P (*endp)) endp++;
- if (endp - str < 5)
- /* Any composite char have at least 5-byte length. */
- return -1;
-
- chars = 0;
- p = str + 1;
- while (p < endp)
- {
- if (embedded_rule)
- {
- p++;
- if (p >= endp)
- return -1;
- }
- /* No need of checking if *P is 0xA0 because
- BYTES_BY_CHAR_HEAD (0x80) surely returns 2. */
- p += BYTES_BY_CHAR_HEAD (*p - 0x20);
- chars++;
- }
- if (p > endp || chars < 2 || chars > MAX_COMPONENT_COUNT)
- /* Invalid components. */
- return -1;
- len = p - str;
- }
- hash_idx = hash_string (str, len) % CMPCHAR_HASH_TABLE_SIZE;
- hashp = cmpchar_hash_table[hash_idx];
-
- /* Then, look into the hash table. */
- if (hashp != NULL)
- /* Find the correct one among composite characters of the same
- hash value. */
- for (i = 2; i < CMPCHAR_HASH_USED (hashp); i++)
- {
- cmpcharp = cmpchar_table[CMPCHAR_HASH_CMPCHAR_ID (hashp, i)];
- if (len == cmpcharp->len
- && ! bcmp (str, cmpcharp->data, len))
- return CMPCHAR_HASH_CMPCHAR_ID (hashp, i);
- }
-
- /* We have to register the composite character in cmpchar_table. */
- if (n_cmpchars >= (CHAR_FIELD2_MASK | CHAR_FIELD3_MASK))
- /* No, we have no more room for a new composite character. */
- return -1;
-
- /* Make the entry in hash table. */
- if (hashp == NULL)
- {
- /* Make a table for 8 composite characters initially. */
- hashp = (cmpchar_hash_table[hash_idx]
- = (int *) xmalloc (sizeof (int) * (2 + 8)));
- CMPCHAR_HASH_SIZE (hashp) = 10;
- CMPCHAR_HASH_USED (hashp) = 2;
- }
- else if (CMPCHAR_HASH_USED (hashp) >= CMPCHAR_HASH_SIZE (hashp))
- {
- CMPCHAR_HASH_SIZE (hashp) += 8;
- hashp = (cmpchar_hash_table[hash_idx]
- = (int *) xrealloc (hashp,
- sizeof (int) * CMPCHAR_HASH_SIZE (hashp)));
- }
- CMPCHAR_HASH_CMPCHAR_ID (hashp, CMPCHAR_HASH_USED (hashp)) = n_cmpchars;
- CMPCHAR_HASH_USED (hashp)++;
-
- /* Set information of the composite character in cmpchar_table. */
- if (cmpchar_table_size == 0)
- {
- /* This is the first composite character to be registered. */
- cmpchar_table_size = 256;
- cmpchar_table
- = (struct cmpchar_info **) xmalloc (sizeof (cmpchar_table[0])
- * cmpchar_table_size);
- }
- else if (cmpchar_table_size <= n_cmpchars)
- {
- cmpchar_table_size += 256;
- cmpchar_table
- = (struct cmpchar_info **) xrealloc (cmpchar_table,
- sizeof (cmpchar_table[0])
- * cmpchar_table_size);
- }
-
- cmpcharp = (struct cmpchar_info *) xmalloc (sizeof (struct cmpchar_info));
-
- cmpcharp->len = len;
- cmpcharp->data = (unsigned char *) xmalloc (len + 1);
- bcopy (str, cmpcharp->data, len);
- cmpcharp->data[len] = 0;
- cmpcharp->glyph_len = chars;
- cmpcharp->glyph = (GLYPH *) xmalloc (sizeof (GLYPH) * chars);
- if (embedded_rule)
- {
- cmpcharp->cmp_rule = (unsigned char *) xmalloc (chars);
- cmpcharp->col_offset = (float *) xmalloc (sizeof (float) * chars);
- }
- else
- {
- cmpcharp->cmp_rule = NULL;
- cmpcharp->col_offset = NULL;
- }
-
- /* Setup GLYPH data and composition rules (if any) so as not to make
- them every time on displaying. */
- {
- unsigned char *bufp;
- int width;
- float leftmost = 0.0, rightmost = 1.0;
-
- if (embedded_rule)
- /* At first, col_offset[N] is set to relative to col_offset[0]. */
- cmpcharp->col_offset[0] = 0;
-
- for (i = 0, bufp = cmpcharp->data + 1; i < chars; i++)
- {
- if (embedded_rule)
- cmpcharp->cmp_rule[i] = *bufp++;
-
- if (*bufp == 0xA0) /* This is an ASCII character. */
- {
- cmpcharp->glyph[i] = FAST_MAKE_GLYPH ((*++bufp & 0x7F), 0);
- width = 1;
- bufp++;
- }
- else /* Multibyte character. */
- {
- /* Make `bufp' point normal multi-byte form temporally. */
- *bufp -= 0x20;
- cmpcharp->glyph[i]
- = FAST_MAKE_GLYPH (string_to_non_ascii_char (bufp, 4, 0), 0);
- width = WIDTH_BY_CHAR_HEAD (*bufp);
- *bufp += 0x20;
- bufp += BYTES_BY_CHAR_HEAD (*bufp - 0x20);
- }
-
- if (embedded_rule && i > 0)
- {
- /* Reference points (global_ref and new_ref) are
- encoded as below:
-
- 0--1--2 -- ascent
- | |
- | |
- | 4 -+--- center
- -- 3 5 -- baseline
- | |
- 6--7--8 -- descent
-
- Now, we calculate the column offset of the new glyph
- from the left edge of the first glyph. This can avoid
- the same calculation everytime displaying this
- composite character. */
-
- /* Reference points of global glyph and new glyph. */
- int global_ref = (cmpcharp->cmp_rule[i] - 0xA0) / 9;
- int new_ref = (cmpcharp->cmp_rule[i] - 0xA0) % 9;
- /* Column offset relative to the first glyph. */
- float left = (leftmost
- + (global_ref % 3) * (rightmost - leftmost) / 2.0
- - (new_ref % 3) * width / 2.0);
-
- cmpcharp->col_offset[i] = left;
- if (left < leftmost)
- leftmost = left;
- if (left + width > rightmost)
- rightmost = left + width;
- }
- else
- {
- if (width > rightmost)
- rightmost = width;
- }
- }
- if (embedded_rule)
- {
- /* Now col_offset[N] are relative to the left edge of the
- first component. Make them relative to the left edge of
- overall glyph. */
- for (i = 0; i < chars; i++)
- cmpcharp->col_offset[i] -= leftmost;
- /* Make rightmost holds width of overall glyph. */
- rightmost -= leftmost;
- }
-
- cmpcharp->width = rightmost;
- if (cmpcharp->width < rightmost)
- /* To get a ceiling integer value. */
- cmpcharp->width++;
- }
-
- cmpchar_table[n_cmpchars] = cmpcharp;
-
- return n_cmpchars++;
-}
-
-/* Return the Nth element of the composite character C. If NOERROR is
- nonzero, return 0 on error condition (C is an invalid composite
- charcter, or N is out of range). */
-int
-cmpchar_component (c, n, noerror)
- int c, n, noerror;
-{
- int id = COMPOSITE_CHAR_ID (c);
-
- if (id < 0 || id >= n_cmpchars)
- {
- /* C is not a valid composite character. */
- if (noerror) return 0;
- error ("Invalid composite character: %d", c) ;
- }
- if (n >= cmpchar_table[id]->glyph_len)
- {
- /* No such component. */
- if (noerror) return 0;
- args_out_of_range (make_number (c), make_number (n));
- }
- /* No face data is stored in glyph code. */
- return ((int) (cmpchar_table[id]->glyph[n]));
-}
-
-DEFUN ("cmpcharp", Fcmpcharp, Scmpcharp, 1, 1, 0,
- "T if CHAR is a composite character.")
- (ch)
- Lisp_Object ch;
-{
- CHECK_NUMBER (ch, 0);
- return (COMPOSITE_CHAR_P (XINT (ch)) ? Qt : Qnil);
-}
-
-DEFUN ("composite-char-component", Fcmpchar_component, Scmpchar_component,
- 2, 2, 0,
- "Return the Nth component character of composite character CHARACTER.")
- (character, n)
- Lisp_Object character, n;
-{
- int id;
-
- CHECK_NUMBER (character, 0);
- CHECK_NUMBER (n, 1);
-
- return (make_number (cmpchar_component (XINT (character), XINT (n), 0)));
-}
-
-DEFUN ("composite-char-composition-rule", Fcmpchar_cmp_rule, Scmpchar_cmp_rule,
- 2, 2, 0,
- "Return the Nth composition rule of composite character CHARACTER.\n\
-The returned rule is for composing the Nth component\n\
-on the (N-1)th component.\n\
-If CHARACTER should be composed relatively or N is 0, return 255.")
- (character, n)
- Lisp_Object character, n;
-{
- int id;
-
- CHECK_NUMBER (character, 0);
- CHECK_NUMBER (n, 1);
-
- id = COMPOSITE_CHAR_ID (XINT (character));
- if (id < 0 || id >= n_cmpchars)
- error ("Invalid composite character: %d", XINT (character));
- if (XINT (n) < 0 || XINT (n) >= cmpchar_table[id]->glyph_len)
- args_out_of_range (character, n);
-
- return make_number (cmpchar_table[id]->cmp_rule
- ? cmpchar_table[id]->cmp_rule[XINT (n)]
- : 255);
-}
-
-DEFUN ("composite-char-composition-rule-p", Fcmpchar_cmp_rule_p,
- Scmpchar_cmp_rule_p, 1, 1, 0,
- "Return non-nil if composite character CHARACTER contains a embedded rule.")
- (character)
- Lisp_Object character;
-{
- int id;
-
- CHECK_NUMBER (character, 0);
- id = COMPOSITE_CHAR_ID (XINT (character));
- if (id < 0 || id >= n_cmpchars)
- error ("Invalid composite character: %d", XINT (character));
-
- return (cmpchar_table[id]->cmp_rule ? Qt : Qnil);
-}
-
-DEFUN ("composite-char-component-count", Fcmpchar_cmp_count,
- Scmpchar_cmp_count, 1, 1, 0,
- "Return number of compoents of composite character CHARACTER.")
- (character)
- Lisp_Object character;
-{
- int id;
-
- CHECK_NUMBER (character, 0);
- id = COMPOSITE_CHAR_ID (XINT (character));
- if (id < 0 || id >= n_cmpchars)
- error ("Invalid composite character: %d", XINT (character));
-
- return (make_number (cmpchar_table[id]->glyph_len));
-}
-
-DEFUN ("compose-string", Fcompose_string, Scompose_string,
- 1, 1, 0,
- "Return one char string composed from all characters in STRING.")
- (str)
- Lisp_Object str;
-{
- unsigned char buf[MAX_LENGTH_OF_MULTI_BYTE_FORM], *p, *pend, *ptemp;
- int len, i;
-
- CHECK_STRING (str, 0);
-
- buf[0] = LEADING_CODE_COMPOSITION;
- p = XSTRING (str)->data;
- pend = p + STRING_BYTES (XSTRING (str));
- i = 1;
- while (p < pend)
- {
- if (*p < 0x20) /* control code */
- error ("Invalid component character: %d", *p);
- else if (*p < 0x80) /* ASCII */
- {
- if (i + 2 >= MAX_LENGTH_OF_MULTI_BYTE_FORM)
- error ("Too long string to be composed: %s", XSTRING (str)->data);
- /* Prepend an ASCII charset indicator 0xA0, set MSB of the
- code itself. */
- buf[i++] = 0xA0;
- buf[i++] = *p++ + 0x80;
- }
- else if (*p == LEADING_CODE_COMPOSITION) /* composite char */
- {
- /* Already composed. Eliminate the heading
- LEADING_CODE_COMPOSITION, keep the remaining bytes
- unchanged. */
- p++;
- if (*p == 255)
- error ("Can't compose a rule-based composition character");
- ptemp = p;
- while (! CHAR_HEAD_P (*p)) p++;
- if (str_cmpchar_id (ptemp - 1, p - ptemp + 1) < 0)
- error ("Can't compose an invalid composition character");
- if (i + (p - ptemp) >= MAX_LENGTH_OF_MULTI_BYTE_FORM)
- error ("Too long string to be composed: %s", XSTRING (str)->data);
- bcopy (ptemp, buf + i, p - ptemp);
- i += p - ptemp;
- }
- else /* multibyte char */
- {
- /* Add 0x20 to the base leading-code, keep the remaining
- bytes unchanged. */
- int c = STRING_CHAR_AND_LENGTH (p, pend - p, len);
-
- if (len <= 1 || ! CHAR_VALID_P (c, 0))
- error ("Can't compose an invalid character");
- if (i + len >= MAX_LENGTH_OF_MULTI_BYTE_FORM)
- error ("Too long string to be composed: %s", XSTRING (str)->data);
- bcopy (p, buf + i, len);
- buf[i] += 0x20;
- p += len, i += len;
- }
- }
-
- if (i < 5)
- /* STR contains only one character, which can't be composed. */
- error ("Too short string to be composed: %s", XSTRING (str)->data);
-
- return make_string_from_bytes (buf, 1, i);
-}
-
-\f
-int
-charset_id_internal (charset_name)
- char *charset_name;
-{
- Lisp_Object val;
-
- val= Fget (intern (charset_name), Qcharset);
- if (!VECTORP (val))
- error ("Charset %s is not defined", charset_name);
-
- return (XINT (XVECTOR (val)->contents[0]));
-}
-
-DEFUN ("setup-special-charsets", Fsetup_special_charsets,
- Ssetup_special_charsets, 0, 0, 0, "Internal use only.")
- ()
-{
- charset_latin_iso8859_1 = charset_id_internal ("latin-iso8859-1");
- charset_jisx0208_1978 = charset_id_internal ("japanese-jisx0208-1978");
- charset_jisx0208 = charset_id_internal ("japanese-jisx0208");
- charset_katakana_jisx0201 = charset_id_internal ("katakana-jisx0201");
- charset_latin_jisx0201 = charset_id_internal ("latin-jisx0201");
- charset_big5_1 = charset_id_internal ("chinese-big5-1");
- charset_big5_2 = charset_id_internal ("chinese-big5-2");
- return Qnil;
-}
-
-void
-init_charset_once ()
-{
- int i, j, k;
-
- staticpro (&Vcharset_table);
- staticpro (&Vcharset_symbol_table);
- staticpro (&Vgeneric_character_list);
-
- /* This has to be done here, before we call Fmake_char_table. */
- Qcharset_table = intern ("charset-table");
- staticpro (&Qcharset_table);
-
- /* Intern this now in case it isn't already done.
- Setting this variable twice is harmless.
- But don't staticpro it here--that is done in alloc.c. */
- Qchar_table_extra_slots = intern ("char-table-extra-slots");
-
- /* Now we are ready to set up this property, so we can
- create the charset table. */
- Fput (Qcharset_table, Qchar_table_extra_slots, make_number (0));
- Vcharset_table = Fmake_char_table (Qcharset_table, Qnil);
-
- Qunknown = intern ("unknown");
- staticpro (&Qunknown);
- Vcharset_symbol_table = Fmake_vector (make_number (MAX_CHARSET + 1),
- Qunknown);
-
- /* Setup tables. */
- for (i = 0; i < 2; i++)
- for (j = 0; j < 2; j++)
- for (k = 0; k < 128; k++)
- iso_charset_table [i][j][k] = -1;
-
- bzero (cmpchar_hash_table, sizeof cmpchar_hash_table);
- cmpchar_table_size = n_cmpchars = 0;
-
- for (i = 0; i < 256; i++)
- BYTES_BY_CHAR_HEAD (i) = 1;
- for (i = MIN_CHARSET_OFFICIAL_DIMENSION1;
- i <= MAX_CHARSET_OFFICIAL_DIMENSION1; i++)
- BYTES_BY_CHAR_HEAD (i) = 2;
- for (i = MIN_CHARSET_OFFICIAL_DIMENSION2;
- i <= MAX_CHARSET_OFFICIAL_DIMENSION2; i++)
- BYTES_BY_CHAR_HEAD (i) = 3;
- BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11) = 3;
- BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12) = 3;
- BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21) = 4;
- BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22) = 4;
- /* The followings don't reflect the actual bytes, but just to tell
- that it is a start of a multibyte character. */
- BYTES_BY_CHAR_HEAD (LEADING_CODE_COMPOSITION) = 2;
- BYTES_BY_CHAR_HEAD (0x9E) = 2;
- BYTES_BY_CHAR_HEAD (0x9F) = 2;
-
- for (i = 0; i < 128; i++)
- WIDTH_BY_CHAR_HEAD (i) = 1;
- for (; i < 256; i++)
- WIDTH_BY_CHAR_HEAD (i) = 4;
- WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11) = 1;
- WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12) = 2;
- WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21) = 1;
- WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22) = 2;
-
- {
- Lisp_Object val;
-
- val = Qnil;
- for (i = 0x81; i < 0x90; i++)
- val = Fcons (make_number ((i - 0x70) << 7), val);
- for (; i < 0x9A; i++)
- val = Fcons (make_number ((i - 0x8F) << 14), val);
- for (i = 0xA0; i < 0xF0; i++)
- val = Fcons (make_number ((i - 0x70) << 7), val);
- for (; i < 0xFF; i++)
- val = Fcons (make_number ((i - 0xE0) << 14), val);
- val = Fcons (make_number (GENERIC_COMPOSITION_CHAR), val);
- Vgeneric_character_list = Fnreverse (val);
- }
-
- nonascii_insert_offset = 0;
- Vnonascii_translation_table = Qnil;
-}
-
-#ifdef emacs
-
-void
-syms_of_charset ()
-{
- Qascii = intern ("ascii");
- staticpro (&Qascii);
-
- Qcharset = intern ("charset");
- staticpro (&Qcharset);
-
- /* Define ASCII charset now. */
- update_charset_table (make_number (CHARSET_ASCII),
- make_number (1), make_number (94),
- make_number (1),
- make_number (0),
- make_number ('B'),
- make_number (0),
- build_string ("ASCII"),
- build_string ("ASCII"),
- build_string ("ASCII (ISO646 IRV)"));
- CHARSET_SYMBOL (CHARSET_ASCII) = Qascii;
- Fput (Qascii, Qcharset, CHARSET_TABLE_ENTRY (CHARSET_ASCII));
-
- Qcomposition = intern ("composition");
- staticpro (&Qcomposition);
- CHARSET_SYMBOL (CHARSET_COMPOSITION) = Qcomposition;
-
- Qauto_fill_chars = intern ("auto-fill-chars");
- staticpro (&Qauto_fill_chars);
- Fput (Qauto_fill_chars, Qchar_table_extra_slots, make_number (0));
-
- defsubr (&Sdefine_charset);
- defsubr (&Sgeneric_character_list);
- defsubr (&Sget_unused_iso_final_char);
- defsubr (&Sdeclare_equiv_charset);
- defsubr (&Sfind_charset_region);
- defsubr (&Sfind_charset_string);
- defsubr (&Smake_char_internal);
- defsubr (&Ssplit_char);
- defsubr (&Schar_charset);
- defsubr (&Scharset_after);
- defsubr (&Siso_charset);
- defsubr (&Schar_valid_p);
- defsubr (&Sunibyte_char_to_multibyte);
- defsubr (&Smultibyte_char_to_unibyte);
- defsubr (&Schar_bytes);
- defsubr (&Schar_width);
- defsubr (&Sstring_width);
- defsubr (&Schar_direction);
- defsubr (&Schars_in_region);
- defsubr (&Sstring);
- defsubr (&Scmpcharp);
- defsubr (&Scmpchar_component);
- defsubr (&Scmpchar_cmp_rule);
- defsubr (&Scmpchar_cmp_rule_p);
- defsubr (&Scmpchar_cmp_count);
- defsubr (&Scompose_string);
- defsubr (&Ssetup_special_charsets);
-
- DEFVAR_LISP ("charset-list", &Vcharset_list,
- "List of charsets ever defined.");
- Vcharset_list = Fcons (Qascii, Qnil);
-
- DEFVAR_LISP ("translation-table-vector", &Vtranslation_table_vector,
- "Vector of cons cell of a symbol and translation table ever defined.\n\
-An ID of a translation table is an index of this vector.");
- Vtranslation_table_vector = Fmake_vector (make_number (16), Qnil);
-
- DEFVAR_INT ("leading-code-composition", &leading_code_composition,
- "Leading-code of composite characters.");
- leading_code_composition = LEADING_CODE_COMPOSITION;
-
- DEFVAR_INT ("leading-code-private-11", &leading_code_private_11,
- "Leading-code of private TYPE9N charset of column-width 1.");
- leading_code_private_11 = LEADING_CODE_PRIVATE_11;
-
- DEFVAR_INT ("leading-code-private-12", &leading_code_private_12,
- "Leading-code of private TYPE9N charset of column-width 2.");
- leading_code_private_12 = LEADING_CODE_PRIVATE_12;
-
- DEFVAR_INT ("leading-code-private-21", &leading_code_private_21,
- "Leading-code of private TYPE9Nx9N charset of column-width 1.");
- leading_code_private_21 = LEADING_CODE_PRIVATE_21;
-
- DEFVAR_INT ("leading-code-private-22", &leading_code_private_22,
- "Leading-code of private TYPE9Nx9N charset of column-width 2.");
- leading_code_private_22 = LEADING_CODE_PRIVATE_22;
-
- DEFVAR_INT ("nonascii-insert-offset", &nonascii_insert_offset,
- "Offset for converting non-ASCII unibyte codes 0240...0377 to multibyte.\n\
-This is used for converting unibyte text to multibyte,\n\
-and for inserting character codes specified by number.\n\n\
-This serves to convert a Latin-1 or similar 8-bit character code\n\
-to the corresponding Emacs multibyte character code.\n\
-Typically the value should be (- (make-char CHARSET 0) 128),\n\
-for your choice of character set.\n\
-If `nonascii-translation-table' is non-nil, it overrides this variable.");
- nonascii_insert_offset = 0;
-
- DEFVAR_LISP ("nonascii-translation-table", &Vnonascii_translation_table,
- "Translation table to convert non-ASCII unibyte codes to multibyte.\n\
-This is used for converting unibyte text to multibyte,\n\
-and for inserting character codes specified by number.\n\n\
-Conversion is performed only when multibyte characters are enabled,\n\
-and it serves to convert a Latin-1 or similar 8-bit character code\n\
-to the corresponding Emacs character code.\n\n\
-If this is nil, `nonascii-insert-offset' is used instead.\n\
-See also the docstring of `make-translation-table'.");
- Vnonascii_translation_table = Qnil;
-
- DEFVAR_INT ("min-composite-char", &min_composite_char,
- "Minimum character code of a composite character.");
- min_composite_char = MIN_CHAR_COMPOSITION;
-
- DEFVAR_LISP ("auto-fill-chars", &Vauto_fill_chars,
- "A char-table for characters which invoke auto-filling.\n\
-Such characters has value t in this table.");
- Vauto_fill_chars = Fmake_char_table (Qauto_fill_chars, Qnil);
- CHAR_TABLE_SET (Vauto_fill_chars, make_number (' '), Qt);
- CHAR_TABLE_SET (Vauto_fill_chars, make_number ('\n'), Qt);
-}
-
-#endif /* emacs */
+/* Basic character set support.
+ Copyright (C) 1995, 97, 98, 2000, 2001 Electrotechnical Laboratory, JAPAN.
+ Licensed to the Free Software Foundation.
+ Copyright (C) 2001 Free Software Foundation, Inc.
+ Copyright (C) 2001, 2002
+ National Institute of Advanced Industrial Science and Technology (AIST)
+ Registration Number H13PRO009
+
+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 2, 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., 59 Temple Place - Suite 330,
+Boston, MA 02111-1307, USA. */
+
+#include <config.h>
+
+#include <stdio.h>
+#include <unistd.h>
+#include <ctype.h>
+#include <sys/types.h>
+#include "lisp.h"
+#include "character.h"
+#include "charset.h"
+#include "coding.h"
+#include "disptab.h"
+#include "buffer.h"
+
+/*** GENERAL NOTES on CODED CHARACTER SETS (CHARSETS) ***
+
+ A coded character set ("charset" hereafter) is a meaningful
+ collection (i.e. language, culture, functionality, etc.) of
+ characters. Emacs handles multiple charsets at once. In Emacs Lisp
+ code, a charset is represented by a symbol. In C code, a charset is
+ represented by its ID number or by a pointer to a struct charset.
+
+ The actual information about each charset is stored in two places.
+ Lispy information is stored in the hash table Vcharset_hash_table as
+ a vector (charset attributes). The other information is stored in
+ charset_table as a struct charset.
+
+*/
+
+/* List of all charsets. This variable is used only from Emacs
+ Lisp. */
+Lisp_Object Vcharset_list;
+
+/* Hash table that contains attributes of each charset. Keys are
+ charset symbols, and values are vectors of charset attributes. */
+Lisp_Object Vcharset_hash_table;
+
+/* Table of struct charset. */
+struct charset *charset_table;
+
+static int charset_table_size;
+int charset_table_used;
+
+Lisp_Object Qcharsetp;
+
+/* Special charset symbols. */
+Lisp_Object Qascii;
+Lisp_Object Qeight_bit;
+Lisp_Object Qiso_8859_1;
+Lisp_Object Qunicode;
+
+/* The corresponding charsets. */
+int charset_ascii;
+int charset_eight_bit;
+int charset_iso_8859_1;
+int charset_unicode;
+
+/* The other special charsets. */
+int charset_jisx0201_roman;
+int charset_jisx0208_1978;
+int charset_jisx0208;
+
+/* Value of charset attribute `charset-iso-plane'. */
+Lisp_Object Qgl, Qgr;
+
+/* Charset of unibyte characters. */
+int charset_unibyte;
+
+/* List of charsets ordered by the priority. */
+Lisp_Object Vcharset_ordered_list;
+
+/* Incremented everytime we change Vcharset_ordered_list. This is
+ unsigned short so that it fits in Lisp_Int and never matches
+ -1. */
+unsigned short charset_ordered_list_tick;
+
+/* List of iso-2022 charsets. */
+Lisp_Object Viso_2022_charset_list;
+
+/* List of emacs-mule charsets. */
+Lisp_Object Vemacs_mule_charset_list;
+
+struct charset *emacs_mule_charset[256];
+
+/* Mapping table from ISO2022's charset (specified by DIMENSION,
+ CHARS, and FINAL-CHAR) to Emacs' charset. */
+int iso_charset_table[ISO_MAX_DIMENSION][ISO_MAX_CHARS][ISO_MAX_FINAL];
+
+Lisp_Object Vcharset_map_directory;
+
+Lisp_Object Vchar_unified_charset_table;
+
+/* Defined in chartab.c */
+extern void
+map_char_table_for_charset P_ ((void (*c_function) (Lisp_Object, Lisp_Object),
+ Lisp_Object function, Lisp_Object table,
+ Lisp_Object arg, struct charset *charset,
+ unsigned from, unsigned to));
+
+#define CODE_POINT_TO_INDEX(charset, code) \
+ ((charset)->code_linear_p \
+ ? (code) - (charset)->min_code \
+ : (((charset)->code_space_mask[(code) >> 24] & 0x8) \
+ && ((charset)->code_space_mask[((code) >> 16) & 0xFF] & 0x4) \
+ && ((charset)->code_space_mask[((code) >> 8) & 0xFF] & 0x2) \
+ && ((charset)->code_space_mask[(code) & 0xFF] & 0x1)) \
+ ? (((((code) >> 24) - (charset)->code_space[12]) \
+ * (charset)->code_space[11]) \
+ + (((((code) >> 16) & 0xFF) - (charset)->code_space[8]) \
+ * (charset)->code_space[7]) \
+ + (((((code) >> 8) & 0xFF) - (charset)->code_space[4]) \
+ * (charset)->code_space[3]) \
+ + (((code) & 0xFF) - (charset)->code_space[0]) \
+ - ((charset)->char_index_offset)) \
+ : -1)
+
+
+/* Convert the character index IDX to code-point CODE for CHARSET.
+ It is assumed that IDX is in a valid range. */
+
+#define INDEX_TO_CODE_POINT(charset, idx) \
+ ((charset)->code_linear_p \
+ ? (idx) + (charset)->min_code \
+ : (idx += (charset)->char_index_offset, \
+ (((charset)->code_space[0] + (idx) % (charset)->code_space[2]) \
+ | (((charset)->code_space[4] \
+ + ((idx) / (charset)->code_space[3] % (charset)->code_space[6])) \
+ << 8) \
+ | (((charset)->code_space[8] \
+ + ((idx) / (charset)->code_space[7] % (charset)->code_space[10])) \
+ << 16) \
+ | (((charset)->code_space[12] + ((idx) / (charset)->code_space[11])) \
+ << 24))))
+
+
+\f
+
+/* Set to 1 to warn that a charset map is loaded and thus a buffer
+ text and a string data may be relocated. */
+int charset_map_loaded;
+
+struct charset_map_entries
+{
+ struct {
+ unsigned from, to;
+ int c;
+ } entry[0x10000];
+ struct charset_map_entries *next;
+};
+
+/* Load the mapping information for CHARSET from ENTRIES.
+
+ If CONTROL_FLAG is 0, setup CHARSET->min_char and CHARSET->max_char.
+
+ If CONTROL_FLAG is 1, setup CHARSET->min_char, CHARSET->max_char,
+ CHARSET->decoder, and CHARSET->encoder.
+
+ If CONTROL_FLAG is 2, setup CHARSET->deunifier and
+ Vchar_unify_table. If Vchar_unified_charset_table is non-nil,
+ setup it too. */
+
+static void
+load_charset_map (charset, entries, n_entries, control_flag)
+ struct charset *charset;
+ struct charset_map_entries *entries;
+ int n_entries;
+ int control_flag;
+{
+ Lisp_Object vec, table;
+ unsigned max_code = CHARSET_MAX_CODE (charset);
+ int ascii_compatible_p = charset->ascii_compatible_p;
+ int min_char, max_char, nonascii_min_char;
+ int i;
+ unsigned char *fast_map = charset->fast_map;
+
+ if (n_entries <= 0)
+ return;
+
+ if (control_flag > 0)
+ {
+ int n = CODE_POINT_TO_INDEX (charset, max_code) + 1;
+
+ table = Fmake_char_table (Qnil, Qnil);
+ if (control_flag == 1)
+ vec = Fmake_vector (make_number (n), make_number (-1));
+ else if (! CHAR_TABLE_P (Vchar_unify_table))
+ Vchar_unify_table = Fmake_char_table (Qnil, Qnil);
+
+ charset_map_loaded = 1;
+ }
+
+ min_char = max_char = entries->entry[0].c;
+ nonascii_min_char = MAX_CHAR;
+ for (i = 0; i < n_entries; i++)
+ {
+ unsigned from, to;
+ int from_index, to_index;
+ int from_c, to_c;
+ int idx = i % 0x10000;
+
+ if (i > 0 && idx == 0)
+ entries = entries->next;
+ from = entries->entry[idx].from;
+ to = entries->entry[idx].to;
+ from_c = entries->entry[idx].c;
+ from_index = CODE_POINT_TO_INDEX (charset, from);
+ if (from == to)
+ {
+ to_index = from_index;
+ to_c = from_c;
+ }
+ else
+ {
+ to_index = CODE_POINT_TO_INDEX (charset, to);
+ to_c = from_c + (to_index - from_index);
+ }
+ if (from_index < 0 || to_index < 0)
+ continue;
+
+ if (control_flag < 2)
+ {
+ int c;
+
+ if (to_c > max_char)
+ max_char = to_c;
+ else if (from_c < min_char)
+ min_char = from_c;
+ if (ascii_compatible_p)
+ {
+ if (! ASCII_BYTE_P (from_c))
+ {
+ if (from_c < nonascii_min_char)
+ nonascii_min_char = from_c;
+ }
+ else if (! ASCII_BYTE_P (to_c))
+ {
+ nonascii_min_char = 0x80;
+ }
+ }
+
+ for (c = from_c; c <= to_c; c++)
+ CHARSET_FAST_MAP_SET (c, fast_map);
+
+ if (control_flag == 1)
+ {
+ unsigned code = from;
+
+ if (CHARSET_COMPACT_CODES_P (charset))
+ while (1)
+ {
+ ASET (vec, from_index, make_number (from_c));
+ CHAR_TABLE_SET (table, from_c, make_number (code));
+ if (from_index == to_index)
+ break;
+ from_index++, from_c++;
+ code = INDEX_TO_CODE_POINT (charset, from_index);
+ }
+ else
+ for (; from_index <= to_index; from_index++, from_c++)
+ {
+ ASET (vec, from_index, make_number (from_c));
+ CHAR_TABLE_SET (table, from_c, make_number (from_index));
+ }
+ }
+ }
+ else
+ {
+ unsigned code = from;
+
+ while (1)
+ {
+ int c1 = DECODE_CHAR (charset, code);
+
+ if (c1 >= 0)
+ {
+ CHAR_TABLE_SET (table, from_c, make_number (c1));
+ CHAR_TABLE_SET (Vchar_unify_table, c1, make_number (from_c));
+ if (CHAR_TABLE_P (Vchar_unified_charset_table))
+ CHAR_TABLE_SET (Vchar_unified_charset_table, c1,
+ CHARSET_NAME (charset));
+ }
+ if (from_index == to_index)
+ break;
+ from_index++, from_c++;
+ code = INDEX_TO_CODE_POINT (charset, from_index);
+ }
+ }
+ }
+
+ if (control_flag < 2)
+ {
+ CHARSET_MIN_CHAR (charset) = (ascii_compatible_p
+ ? nonascii_min_char : min_char);
+ CHARSET_MAX_CHAR (charset) = max_char;
+ if (control_flag == 1)
+ {
+ CHARSET_DECODER (charset) = vec;
+ CHARSET_ENCODER (charset) = table;
+ }
+ }
+ else
+ CHARSET_DEUNIFIER (charset) = table;
+}
+
+
+/* Read a hexadecimal number (preceded by "0x") from the file FP while
+ paying attention to comment charcter '#'. */
+
+static INLINE unsigned
+read_hex (fp, eof)
+ FILE *fp;
+ int *eof;
+{
+ int c;
+ unsigned n;
+
+ while ((c = getc (fp)) != EOF)
+ {
+ if (c == '#')
+ {
+ while ((c = getc (fp)) != EOF && c != '\n');
+ }
+ else if (c == '0')
+ {
+ if ((c = getc (fp)) == EOF || c == 'x')
+ break;
+ }
+ }
+ if (c == EOF)
+ {
+ *eof = 1;
+ return 0;
+ }
+ *eof = 0;
+ n = 0;
+ if (c == 'x')
+ while ((c = getc (fp)) != EOF && isxdigit (c))
+ n = ((n << 4)
+ | (c <= '9' ? c - '0' : c <= 'F' ? c - 'A' + 10 : c - 'a' + 10));
+ else
+ while ((c = getc (fp)) != EOF && isdigit (c))
+ n = (n * 10) + c - '0';
+ if (c != EOF)
+ ungetc (c, fp);
+ return n;
+}
+
+
+/* Return a mapping vector for CHARSET loaded from MAPFILE.
+ Each line of MAPFILE has this form
+ 0xAAAA 0xCCCC
+ where 0xAAAA is a code-point and 0xCCCC is the corresponding
+ character code, or this form
+ 0xAAAA-0xBBBB 0xCCCC
+ where 0xAAAA and 0xBBBB are code-points specifying a range, and
+ 0xCCCC is the first character code of the range.
+
+ The returned vector has this form:
+ [ CODE1 CHAR1 CODE2 CHAR2 .... ]
+ where CODE1 is a code-point or a cons of code-points specifying a
+ range. */
+
+extern void add_to_log P_ ((char *, Lisp_Object, Lisp_Object));
+
+static void
+load_charset_map_from_file (charset, mapfile, control_flag)
+ struct charset *charset;
+ Lisp_Object mapfile;
+ int control_flag;
+{
+ unsigned min_code = CHARSET_MIN_CODE (charset);
+ unsigned max_code = CHARSET_MAX_CODE (charset);
+ int fd;
+ FILE *fp;
+ int eof;
+ Lisp_Object suffixes;
+ struct charset_map_entries *head, *entries;
+ int n_entries;
+
+ suffixes = Fcons (build_string (".map"),
+ Fcons (build_string (".TXT"), Qnil));
+
+ fd = openp (Fcons (Vcharset_map_directory, Qnil), mapfile, suffixes,
+ NULL, 0);
+ if (fd < 0
+ || ! (fp = fdopen (fd, "r")))
+ {
+ add_to_log ("Failure in loading charset map: %S", mapfile, Qnil);
+ return;
+ }
+
+ head = entries = ((struct charset_map_entries *)
+ alloca (sizeof (struct charset_map_entries)));
+ n_entries = 0;
+ eof = 0;
+ while (1)
+ {
+ unsigned from, to;
+ int c;
+ int idx;
+
+ from = read_hex (fp, &eof);
+ if (eof)
+ break;
+ if (getc (fp) == '-')
+ to = read_hex (fp, &eof);
+ else
+ to = from;
+ c = (int) read_hex (fp, &eof);
+
+ if (from < min_code || to > max_code || from > to || c > MAX_CHAR)
+ continue;
+
+ if (n_entries > 0 && (n_entries % 0x10000) == 0)
+ {
+ entries->next = ((struct charset_map_entries *)
+ alloca (sizeof (struct charset_map_entries)));
+ entries = entries->next;
+ }
+ idx = n_entries % 0x10000;
+ entries->entry[idx].from = from;
+ entries->entry[idx].to = to;
+ entries->entry[idx].c = c;
+ n_entries++;
+ }
+ fclose (fp);
+ close (fd);
+
+ load_charset_map (charset, head, n_entries, control_flag);
+}
+
+static void
+load_charset_map_from_vector (charset, vec, control_flag)
+ struct charset *charset;
+ Lisp_Object vec;
+ int control_flag;
+{
+ unsigned min_code = CHARSET_MIN_CODE (charset);
+ unsigned max_code = CHARSET_MAX_CODE (charset);
+ struct charset_map_entries *head, *entries;
+ int n_entries;
+ int len = ASIZE (vec);
+ int i;
+
+ if (len % 2 == 1)
+ {
+ add_to_log ("Failure in loading charset map: %V", vec, Qnil);
+ return;
+ }
+
+ head = entries = ((struct charset_map_entries *)
+ alloca (sizeof (struct charset_map_entries)));
+ n_entries = 0;
+ for (i = 0; i < len; i += 2)
+ {
+ Lisp_Object val, val2;
+ unsigned from, to;
+ int c;
+ int idx;
+
+ val = AREF (vec, i);
+ if (CONSP (val))
+ {
+ val2 = XCDR (val);
+ val = XCAR (val);
+ CHECK_NATNUM (val);
+ CHECK_NATNUM (val2);
+ from = XFASTINT (val);
+ to = XFASTINT (val2);
+ }
+ else
+ {
+ CHECK_NATNUM (val);
+ from = to = XFASTINT (val);
+ }
+ val = AREF (vec, i + 1);
+ CHECK_NATNUM (val);
+ c = XFASTINT (val);
+
+ if (from < min_code || to > max_code || from > to || c > MAX_CHAR)
+ continue;
+
+ if ((n_entries % 0x10000) == 0)
+ {
+ entries->next = ((struct charset_map_entries *)
+ alloca (sizeof (struct charset_map_entries)));
+ entries = entries->next;
+ }
+ idx = n_entries % 0x10000;
+ entries->entry[idx].from = from;
+ entries->entry[idx].to = to;
+ entries->entry[idx].c = c;
+ n_entries++;
+ }
+
+ load_charset_map (charset, head, n_entries, control_flag);
+}
+
+static void
+load_charset (charset)
+ struct charset *charset;
+{
+ if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP_DEFERRED)
+ {
+ Lisp_Object map;
+
+ map = CHARSET_MAP (charset);
+ if (STRINGP (map))
+ load_charset_map_from_file (charset, map, 1);
+ else
+ load_charset_map_from_vector (charset, map, 1);
+ CHARSET_METHOD (charset) = CHARSET_METHOD_MAP;
+ }
+}
+
+
+DEFUN ("charsetp", Fcharsetp, Scharsetp, 1, 1, 0,
+ doc: /* Return non-nil if and only if OBJECT is a charset.*/)
+ (object)
+ Lisp_Object object;
+{
+ return (CHARSETP (object) ? Qt : Qnil);
+}
+
+
+void
+map_charset_chars (c_function, function, arg,
+ charset, from, to)
+ void (*c_function) P_ ((Lisp_Object, Lisp_Object));
+ Lisp_Object function, arg;
+ struct charset *charset;
+ unsigned from, to;
+
+{
+ Lisp_Object range;
+ int partial;
+
+ if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP_DEFERRED)
+ load_charset (charset);
+
+ partial = (from > CHARSET_MIN_CODE (charset)
+ || to < CHARSET_MAX_CODE (charset));
+
+ if (CHARSET_UNIFIED_P (charset)
+ && CHAR_TABLE_P (CHARSET_DEUNIFIER (charset)))
+ {
+ map_char_table_for_charset (c_function, function,
+ CHARSET_DEUNIFIER (charset), arg,
+ partial ? charset : NULL, from, to);
+ }
+
+ if (CHARSET_METHOD (charset) == CHARSET_METHOD_OFFSET)
+ {
+ int from_idx = CODE_POINT_TO_INDEX (charset, from);
+ int to_idx = CODE_POINT_TO_INDEX (charset, to);
+ int from_c = from_idx + CHARSET_CODE_OFFSET (charset);
+ int to_c = to_idx + CHARSET_CODE_OFFSET (charset);
+
+ range = Fcons (make_number (from_c), make_number (to_c));
+ if (NILP (function))
+ (*c_function) (range, arg);
+ else
+ call2 (function, range, arg);
+ }
+ else if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP)
+ {
+ if (! CHAR_TABLE_P (CHARSET_ENCODER (charset)))
+ return;
+ if (CHARSET_ASCII_COMPATIBLE_P (charset) && from <= 127)
+ {
+ range = Fcons (make_number (from), make_number (to));
+ if (to >= 128)
+ XSETCAR (range, make_number (127));
+
+ if (NILP (function))
+ (*c_function) (range, arg);
+ else
+ call2 (function, range, arg);
+ }
+ map_char_table_for_charset (c_function, function,
+ CHARSET_ENCODER (charset), arg,
+ partial ? charset : NULL, from, to);
+ }
+ else if (CHARSET_METHOD (charset) == CHARSET_METHOD_SUBSET)
+ {
+ Lisp_Object subset_info;
+ int offset;
+
+ subset_info = CHARSET_SUBSET (charset);
+ charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0)));
+ offset = XINT (AREF (subset_info, 3));
+ from -= offset;
+ if (from < XFASTINT (AREF (subset_info, 1)))
+ from = XFASTINT (AREF (subset_info, 1));
+ to -= offset;
+ if (to > XFASTINT (AREF (subset_info, 2)))
+ to = XFASTINT (AREF (subset_info, 2));
+ map_charset_chars (c_function, function, arg, charset, from, to);
+ }
+ else /* i.e. CHARSET_METHOD_SUPERSET */
+ {
+ Lisp_Object parents;
+
+ for (parents = CHARSET_SUPERSET (charset); CONSP (parents);
+ parents = XCDR (parents))
+ {
+ int offset;
+ unsigned this_from, this_to;
+
+ charset = CHARSET_FROM_ID (XFASTINT (XCAR (XCAR (parents))));
+ offset = XINT (XCDR (XCAR (parents)));
+ this_from = from - offset;
+ this_to = to - offset;
+ if (this_from < CHARSET_MIN_CODE (charset))
+ this_from = CHARSET_MIN_CODE (charset);
+ if (this_to > CHARSET_MAX_CODE (charset))
+ this_to = CHARSET_MAX_CODE (charset);
+ map_charset_chars (c_function, function, arg, charset, from, to);
+ }
+ }
+}
+
+
+DEFUN ("map-charset-chars", Fmap_charset_chars, Smap_charset_chars, 2, 5, 0,
+ doc: /* Call FUNCTION for all characters in CHARSET.
+FUNCTION is called with an argument RANGE and the optional 3rd
+argument ARG.
+
+RANGE is a cons (FROM . TO), where FROM and TO indicate a range of
+characters contained in CHARSET.
+
+The optional 4th and 5th arguments FROM-CODE and TO-CODE specify the
+range of code points of targer characters. */)
+ (function, charset, arg, from_code, to_code)
+ Lisp_Object function, charset, arg, from_code, to_code;
+{
+ struct charset *cs;
+ unsigned from, to;
+
+ CHECK_CHARSET_GET_CHARSET (charset, cs);
+ if (NILP (from_code))
+ from = CHARSET_MIN_CODE (cs);
+ else
+ {
+ CHECK_NATNUM (from_code);
+ from = XINT (from_code);
+ if (from < CHARSET_MIN_CODE (cs))
+ from = CHARSET_MIN_CODE (cs);
+ }
+ if (NILP (to_code))
+ to = CHARSET_MAX_CODE (cs);
+ else
+ {
+ CHECK_NATNUM (to_code);
+ to = XINT (to_code);
+ if (to > CHARSET_MAX_CODE (cs))
+ to = CHARSET_MAX_CODE (cs);
+ }
+ map_charset_chars (NULL, function, arg, cs, from, to);
+ return Qnil;
+}
+
+
+/* Define a charset according to the arguments. The Nth argument is
+ the Nth attribute of the charset (the last attribute `charset-id'
+ is not included). See the docstring of `define-charset' for the
+ detail. */
+
+DEFUN ("define-charset-internal", Fdefine_charset_internal,
+ Sdefine_charset_internal, charset_arg_max, MANY, 0,
+ doc: /* For internal use only.
+usage: (define-charset-internal ...) */)
+ (nargs, args)
+ int nargs;
+ Lisp_Object *args;
+{
+ /* Charset attr vector. */
+ Lisp_Object attrs;
+ Lisp_Object val;
+ unsigned hash_code;
+ struct Lisp_Hash_Table *hash_table = XHASH_TABLE (Vcharset_hash_table);
+ int i, j;
+ struct charset charset;
+ int id;
+ int dimension;
+ int new_definition_p;
+ int nchars;
+
+ if (nargs != charset_arg_max)
+ return Fsignal (Qwrong_number_of_arguments,
+ Fcons (intern ("define-charset-internal"),
+ make_number (nargs)));
+
+ attrs = Fmake_vector (make_number (charset_attr_max), Qnil);
+
+ CHECK_SYMBOL (args[charset_arg_name]);
+ ASET (attrs, charset_name, args[charset_arg_name]);
+
+ val = args[charset_arg_code_space];
+ for (i = 0, dimension = 0, nchars = 1; i < 4; i++)
+ {
+ int min_byte, max_byte;
+
+ min_byte = XINT (Faref (val, make_number (i * 2)));
+ max_byte = XINT (Faref (val, make_number (i * 2 + 1)));
+ if (min_byte < 0 || min_byte > max_byte || max_byte >= 256)
+ error ("Invalid :code-space value");
+ charset.code_space[i * 4] = min_byte;
+ charset.code_space[i * 4 + 1] = max_byte;
+ charset.code_space[i * 4 + 2] = max_byte - min_byte + 1;
+ nchars *= charset.code_space[i * 4 + 2];
+ charset.code_space[i * 4 + 3] = nchars;
+ if (max_byte > 0)
+ dimension = i + 1;
+ }
+
+ val = args[charset_arg_dimension];
+ if (NILP (val))
+ charset.dimension = dimension;
+ else
+ {
+ CHECK_NATNUM (val);
+ charset.dimension = XINT (val);
+ if (charset.dimension < 1 || charset.dimension > 4)
+ args_out_of_range_3 (val, make_number (1), make_number (4));
+ }
+
+ charset.code_linear_p
+ = (charset.dimension == 1
+ || (charset.code_space[2] == 256
+ && (charset.dimension == 2
+ || (charset.code_space[6] == 256
+ && (charset.dimension == 3
+ || charset.code_space[10] == 256)))));
+
+ if (! charset.code_linear_p)
+ {
+ charset.code_space_mask = (unsigned char *) xmalloc (256);
+ bzero (charset.code_space_mask, 256);
+ for (i = 0; i < 4; i++)
+ for (j = charset.code_space[i * 4]; j <= charset.code_space[i * 4 + 1];
+ j++)
+ charset.code_space_mask[j] |= (1 << i);
+ }
+
+ charset.iso_chars_96 = charset.code_space[2] == 96;
+
+ charset.min_code = (charset.code_space[0]
+ | (charset.code_space[4] << 8)
+ | (charset.code_space[8] << 16)
+ | (charset.code_space[12] << 24));
+ charset.max_code = (charset.code_space[1]
+ | (charset.code_space[5] << 8)
+ | (charset.code_space[9] << 16)
+ | (charset.code_space[13] << 24));
+ charset.char_index_offset = 0;
+
+ val = args[charset_arg_min_code];
+ if (! NILP (val))
+ {
+ unsigned code;
+
+ if (INTEGERP (val))
+ code = XINT (val);
+ else
+ {
+ CHECK_CONS (val);
+ CHECK_NUMBER (XCAR (val));
+ CHECK_NUMBER (XCDR (val));
+ code = (XINT (XCAR (val)) << 16) | (XINT (XCDR (val)));
+ }
+ if (code < charset.min_code
+ || code > charset.max_code)
+ args_out_of_range_3 (make_number (charset.min_code),
+ make_number (charset.max_code), val);
+ charset.char_index_offset = CODE_POINT_TO_INDEX (&charset, code);
+ charset.min_code = code;
+ }
+
+ val = args[charset_arg_max_code];
+ if (! NILP (val))
+ {
+ unsigned code;
+
+ if (INTEGERP (val))
+ code = XINT (val);
+ else
+ {
+ CHECK_CONS (val);
+ CHECK_NUMBER (XCAR (val));
+ CHECK_NUMBER (XCDR (val));
+ code = (XINT (XCAR (val)) << 16) | (XINT (XCDR (val)));
+ }
+ if (code < charset.min_code
+ || code > charset.max_code)
+ args_out_of_range_3 (make_number (charset.min_code),
+ make_number (charset.max_code), val);
+ charset.max_code = code;
+ }
+
+ charset.compact_codes_p = charset.max_code < 0x1000000;
+
+ val = args[charset_arg_invalid_code];
+ if (NILP (val))
+ {
+ if (charset.min_code > 0)
+ charset.invalid_code = 0;
+ else
+ {
+ XSETINT (val, charset.max_code + 1);
+ if (XINT (val) == charset.max_code + 1)
+ charset.invalid_code = charset.max_code + 1;
+ else
+ error ("Attribute :invalid-code must be specified");
+ }
+ }
+ else
+ {
+ CHECK_NATNUM (val);
+ charset.invalid_code = XFASTINT (val);
+ }
+
+ val = args[charset_arg_iso_final];
+ if (NILP (val))
+ charset.iso_final = -1;
+ else
+ {
+ CHECK_NUMBER (val);
+ if (XINT (val) < '0' || XINT (val) > 127)
+ error ("Invalid iso-final-char: %d", XINT (val));
+ charset.iso_final = XINT (val);
+ }
+
+ val = args[charset_arg_iso_revision];
+ if (NILP (val))
+ charset.iso_revision = -1;
+ else
+ {
+ CHECK_NUMBER (val);
+ if (XINT (val) > 63)
+ args_out_of_range (make_number (63), val);
+ charset.iso_revision = XINT (val);
+ }
+
+ val = args[charset_arg_emacs_mule_id];
+ if (NILP (val))
+ charset.emacs_mule_id = -1;
+ else
+ {
+ CHECK_NATNUM (val);
+ if ((XINT (val) > 0 && XINT (val) <= 128) || XINT (val) >= 256)
+ error ("Invalid emacs-mule-id: %d", XINT (val));
+ charset.emacs_mule_id = XINT (val);
+ }
+
+ charset.ascii_compatible_p = ! NILP (args[charset_arg_ascii_compatible_p]);
+
+ charset.supplementary_p = ! NILP (args[charset_arg_supplementary_p]);
+
+ charset.unified_p = 0;
+
+ bzero (charset.fast_map, sizeof (charset.fast_map));
+
+ if (! NILP (args[charset_arg_code_offset]))
+ {
+ val = args[charset_arg_code_offset];
+ CHECK_NUMBER (val);
+
+ charset.method = CHARSET_METHOD_OFFSET;
+ charset.code_offset = XINT (val);
+
+ i = CODE_POINT_TO_INDEX (&charset, charset.min_code);
+ charset.min_char = i + charset.code_offset;
+ i = CODE_POINT_TO_INDEX (&charset, charset.max_code);
+ charset.max_char = i + charset.code_offset;
+ if (charset.max_char > MAX_CHAR)
+ error ("Unsupported max char: %d", charset.max_char);
+
+ i = (charset.min_char >> 7) << 7;
+ for (; i < 0x10000 && i <= charset.max_char; i += 128)
+ CHARSET_FAST_MAP_SET (i, charset.fast_map);
+ i = (i >> 12) << 12;
+ for (; i <= charset.max_char; i += 0x1000)
+ CHARSET_FAST_MAP_SET (i, charset.fast_map);
+ }
+ else if (! NILP (args[charset_arg_map]))
+ {
+ val = args[charset_arg_map];
+ ASET (attrs, charset_map, val);
+ if (STRINGP (val))
+ load_charset_map_from_file (&charset, val, 0);
+ else
+ load_charset_map_from_vector (&charset, val, 0);
+ charset.method = CHARSET_METHOD_MAP_DEFERRED;
+ }
+ else if (! NILP (args[charset_arg_subset]))
+ {
+ Lisp_Object parent;
+ Lisp_Object parent_min_code, parent_max_code, parent_code_offset;
+ struct charset *parent_charset;
+
+ val = args[charset_arg_subset];
+ parent = Fcar (val);
+ CHECK_CHARSET_GET_CHARSET (parent, parent_charset);
+ parent_min_code = Fnth (make_number (1), val);
+ CHECK_NATNUM (parent_min_code);
+ parent_max_code = Fnth (make_number (2), val);
+ CHECK_NATNUM (parent_max_code);
+ parent_code_offset = Fnth (make_number (3), val);
+ CHECK_NUMBER (parent_code_offset);
+ val = Fmake_vector (make_number (4), Qnil);
+ ASET (val, 0, make_number (parent_charset->id));
+ ASET (val, 1, parent_min_code);
+ ASET (val, 2, parent_max_code);
+ ASET (val, 3, parent_code_offset);
+ ASET (attrs, charset_subset, val);
+
+ charset.method = CHARSET_METHOD_SUBSET;
+ /* Here, we just copy the parent's fast_map. It's not accurate,
+ but at least it works for quickly detecting which character
+ DOESN'T belong to this charset. */
+ for (i = 0; i < 190; i++)
+ charset.fast_map[i] = parent_charset->fast_map[i];
+
+ /* We also copy these for parents. */
+ charset.min_char = parent_charset->min_char;
+ charset.max_char = parent_charset->max_char;
+ }
+ else if (! NILP (args[charset_arg_superset]))
+ {
+ val = args[charset_arg_superset];
+ charset.method = CHARSET_METHOD_SUPERSET;
+ val = Fcopy_sequence (val);
+ ASET (attrs, charset_superset, val);
+
+ charset.min_char = MAX_CHAR;
+ charset.max_char = 0;
+ for (; ! NILP (val); val = Fcdr (val))
+ {
+ Lisp_Object elt, car_part, cdr_part;
+ int this_id, offset;
+ struct charset *this_charset;
+
+ elt = Fcar (val);
+ if (CONSP (elt))
+ {
+ car_part = XCAR (elt);
+ cdr_part = XCDR (elt);
+ CHECK_CHARSET_GET_ID (car_part, this_id);
+ CHECK_NUMBER (cdr_part);
+ offset = XINT (cdr_part);
+ }
+ else
+ {
+ CHECK_CHARSET_GET_ID (elt, this_id);
+ offset = 0;
+ }
+ XSETCAR (val, Fcons (make_number (this_id), make_number (offset)));
+
+ this_charset = CHARSET_FROM_ID (this_id);
+ if (charset.min_char > this_charset->min_char)
+ charset.min_char = this_charset->min_char;
+ if (charset.max_char < this_charset->max_char)
+ charset.max_char = this_charset->max_char;
+ for (i = 0; i < 190; i++)
+ charset.fast_map[i] |= this_charset->fast_map[i];
+ }
+ }
+ else
+ error ("None of :code-offset, :map, :parents are specified");
+
+ val = args[charset_arg_unify_map];
+ if (! NILP (val) && !STRINGP (val))
+ CHECK_VECTOR (val);
+ ASET (attrs, charset_unify_map, val);
+
+ CHECK_LIST (args[charset_arg_plist]);
+ ASET (attrs, charset_plist, args[charset_arg_plist]);
+
+ charset.hash_index = hash_lookup (hash_table, args[charset_arg_name],
+ &hash_code);
+ if (charset.hash_index >= 0)
+ {
+ new_definition_p = 0;
+ id = XFASTINT (CHARSET_SYMBOL_ID (args[charset_arg_name]));
+ HASH_VALUE (hash_table, charset.hash_index) = attrs;
+ }
+ else
+ {
+ charset.hash_index = hash_put (hash_table, args[charset_arg_name], attrs,
+ hash_code);
+ if (charset_table_used == charset_table_size)
+ {
+ struct charset *new_table
+ = (struct charset *) xmalloc (sizeof (struct charset)
+ * (charset_table_size + 16));
+ bcopy (charset_table, new_table,
+ sizeof (struct charset) * charset_table_size);
+ charset_table_size += 16;
+ charset_table = new_table;
+ }
+ id = charset_table_used++;
+ new_definition_p = 1;
+ }
+
+ ASET (attrs, charset_id, make_number (id));
+ charset.id = id;
+ charset_table[id] = charset;
+
+ if (charset.iso_final >= 0)
+ {
+ ISO_CHARSET_TABLE (charset.dimension, charset.iso_chars_96,
+ charset.iso_final) = id;
+ if (new_definition_p)
+ Viso_2022_charset_list = nconc2 (Viso_2022_charset_list,
+ Fcons (make_number (id), Qnil));
+ if (ISO_CHARSET_TABLE (1, 0, 'J') == id)
+ charset_jisx0201_roman = id;
+ else if (ISO_CHARSET_TABLE (2, 0, '@') == id)
+ charset_jisx0208_1978 = id;
+ else if (ISO_CHARSET_TABLE (2, 0, 'B') == id)
+ charset_jisx0208 = id;
+ }
+
+ if (charset.emacs_mule_id >= 0)
+ {
+ emacs_mule_charset[charset.emacs_mule_id] = CHARSET_FROM_ID (id);
+ if (charset.emacs_mule_id < 0xA0)
+ emacs_mule_bytes[charset.emacs_mule_id] = charset.dimension + 1;
+ if (new_definition_p)
+ Vemacs_mule_charset_list = nconc2 (Vemacs_mule_charset_list,
+ Fcons (make_number (id), Qnil));
+ }
+
+ if (new_definition_p)
+ {
+ Vcharset_list = Fcons (args[charset_arg_name], Vcharset_list);
+ Vcharset_ordered_list = nconc2 (Vcharset_ordered_list,
+ Fcons (make_number (id), Qnil));
+ charset_ordered_list_tick++;
+ }
+
+ return Qnil;
+}
+
+
+/* Same as Fdefine_charset_internal but arguments are more convenient
+ to call from C (typically in syms_of_charset). This can define a
+ charset of `offset' method only. Return the ID of the new
+ charset. */
+
+static int
+define_charset_internal (name, dimension, code_space, min_code, max_code,
+ iso_final, iso_revision, emacs_mule_id,
+ ascii_compatible, supplementary,
+ code_offset)
+ Lisp_Object name;
+ int dimension;
+ unsigned char *code_space;
+ unsigned min_code, max_code;
+ int iso_final, iso_revision, emacs_mule_id;
+ int ascii_compatible, supplementary;
+ int code_offset;
+{
+ Lisp_Object args[charset_arg_max];
+ Lisp_Object plist[14];
+ Lisp_Object val;
+ int i;
+
+ args[charset_arg_name] = name;
+ args[charset_arg_dimension] = make_number (dimension);
+ val = Fmake_vector (make_number (8), make_number (0));
+ for (i = 0; i < 8; i++)
+ ASET (val, i, make_number (code_space[i]));
+ args[charset_arg_code_space] = val;
+ args[charset_arg_min_code] = make_number (min_code);
+ args[charset_arg_max_code] = make_number (max_code);
+ args[charset_arg_iso_final]
+ = (iso_final < 0 ? Qnil : make_number (iso_final));
+ args[charset_arg_iso_revision] = make_number (iso_revision);
+ args[charset_arg_emacs_mule_id]
+ = (emacs_mule_id < 0 ? Qnil : make_number (emacs_mule_id));
+ args[charset_arg_ascii_compatible_p] = ascii_compatible ? Qt : Qnil;
+ args[charset_arg_supplementary_p] = supplementary ? Qt : Qnil;
+ args[charset_arg_invalid_code] = Qnil;
+ args[charset_arg_code_offset] = make_number (code_offset);
+ args[charset_arg_map] = Qnil;
+ args[charset_arg_subset] = Qnil;
+ args[charset_arg_superset] = Qnil;
+ args[charset_arg_unify_map] = Qnil;
+
+ plist[0] = intern (":name");
+ plist[1] = args[charset_arg_name];
+ plist[2] = intern (":dimension");
+ plist[3] = args[charset_arg_dimension];
+ plist[4] = intern (":code-space");
+ plist[5] = args[charset_arg_code_space];
+ plist[6] = intern (":iso-final-char");
+ plist[7] = args[charset_arg_iso_final];
+ plist[8] = intern (":emacs-mule-id");
+ plist[9] = args[charset_arg_emacs_mule_id];
+ plist[10] = intern (":ascii-compatible-p");
+ plist[11] = args[charset_arg_ascii_compatible_p];
+ plist[12] = intern (":code-offset");
+ plist[13] = args[charset_arg_code_offset];
+
+ args[charset_arg_plist] = Flist (14, plist);
+ Fdefine_charset_internal (charset_arg_max, args);
+
+ return XINT (CHARSET_SYMBOL_ID (name));
+}
+
+
+DEFUN ("define-charset-alias", Fdefine_charset_alias,
+ Sdefine_charset_alias, 2, 2, 0,
+ doc: /* Define ALIAS as an alias for charset CHARSET. */)
+ (alias, charset)
+ Lisp_Object alias, charset;
+{
+ Lisp_Object attr;
+
+ CHECK_CHARSET_GET_ATTR (charset, attr);
+ Fputhash (alias, attr, Vcharset_hash_table);
+ Vcharset_list = Fcons (alias, Vcharset_list);
+ return Qnil;
+}
+
+
+DEFUN ("unibyte-charset", Funibyte_charset, Sunibyte_charset, 0, 0, 0,
+ doc: /* Return the unibyte charset (set by `set-unibyte-charset'). */)
+ ()
+{
+ return CHARSET_NAME (CHARSET_FROM_ID (charset_unibyte));
+}
+
+
+DEFUN ("set-unibyte-charset", Fset_unibyte_charset, Sset_unibyte_charset,
+ 1, 1, 0,
+ doc: /* Set the unibyte charset to CHARSET.
+This determines how unibyte/multibyte conversion is done. See also
+function `unibyte-charset'. */)
+ (charset)
+ Lisp_Object charset;
+{
+ struct charset *cs;
+ int i, c;
+
+ CHECK_CHARSET_GET_CHARSET (charset, cs);
+ if (! cs->ascii_compatible_p
+ || cs->dimension != 1)
+ error ("Inappropriate unibyte charset: %s", XSYMBOL (charset)->name->data);
+ charset_unibyte = cs->id;
+ for (i = 128; i < 256; i++)
+ {
+ c = DECODE_CHAR (cs, i);
+ unibyte_to_multibyte_table[i] = (c < 0 ? i : c);
+ }
+
+ return Qnil;
+}
+
+
+DEFUN ("charset-plist", Fcharset_plist, Scharset_plist, 1, 1, 0,
+ doc: /* Return the property list of CHARSET. */)
+ (charset)
+ Lisp_Object charset;
+{
+ Lisp_Object attrs;
+
+ CHECK_CHARSET_GET_ATTR (charset, attrs);
+ return CHARSET_ATTR_PLIST (attrs);
+}
+
+
+DEFUN ("set-charset-plist", Fset_charset_plist, Sset_charset_plist, 2, 2, 0,
+ doc: /* Set CHARSET's property list to PLIST. */)
+ (charset, plist)
+ Lisp_Object charset, plist;
+{
+ Lisp_Object attrs;
+
+ CHECK_CHARSET_GET_ATTR (charset, attrs);
+ CHARSET_ATTR_PLIST (attrs) = plist;
+ return plist;
+}
+
+
+DEFUN ("unify-charset", Funify_charset, Sunify_charset, 1, 3, 0,
+ doc: /* Unify characters of CHARSET with Unicode.
+This means reading the relevant file and installing the table defined
+by CHARSET's `:unify-map' property.
+
+Optional second arg UNIFY-MAP is a file name string or a vector. It has
+the same meaning as the `:unify-map' attribute in the function
+`define-charset' (which see).
+
+Optional third argument DEUNIFY, if non-nil, means to de-unify CHARSET. */)
+ (charset, unify_map, deunify)
+ Lisp_Object charset, unify_map, deunify;
+{
+ int id;
+ struct charset *cs;
+
+ CHECK_CHARSET_GET_ID (charset, id);
+ cs = CHARSET_FROM_ID (id);
+ if (CHARSET_METHOD (cs) == CHARSET_METHOD_MAP_DEFERRED)
+ load_charset (cs);
+ if (NILP (deunify)
+ ? CHARSET_UNIFIED_P (cs) && ! NILP (CHARSET_DEUNIFIER (cs))
+ : ! CHARSET_UNIFIED_P (cs))
+ return Qnil;
+
+ CHARSET_UNIFIED_P (cs) = 0;
+ if (NILP (deunify))
+ {
+ if (CHARSET_METHOD (cs) != CHARSET_METHOD_OFFSET)
+ error ("Can't unify charset: %s", XSYMBOL (charset)->name->data);
+ if (NILP (unify_map))
+ unify_map = CHARSET_UNIFY_MAP (cs);
+ if (STRINGP (unify_map))
+ load_charset_map_from_file (cs, unify_map, 2);
+ else if (VECTORP (unify_map))
+ load_charset_map_from_vector (cs, unify_map, 2);
+ else if (NILP (unify_map))
+ error ("No unify-map for charset");
+ else
+ error ("Bad unify-map arg");
+ CHARSET_UNIFIED_P (cs) = 1;
+ }
+ else if (CHAR_TABLE_P (Vchar_unify_table))
+ {
+ int min_code = CHARSET_MIN_CODE (cs);
+ int max_code = CHARSET_MAX_CODE (cs);
+ int min_char = DECODE_CHAR (cs, min_code);
+ int max_char = DECODE_CHAR (cs, max_code);
+
+ char_table_set_range (Vchar_unify_table, min_char, max_char, Qnil);
+ }
+
+ return Qnil;
+}
+
+DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char,
+ Sget_unused_iso_final_char, 2, 2, 0,
+ doc: /*
+Return an unsed ISO final char for a charset of DIMENISION and CHARS.
+DIMENSION is the number of bytes to represent a character: 1 or 2.
+CHARS is the number of characters in a dimension: 94 or 96.
+
+This final char is for private use, thus the range is `0' (48) .. `?' (63).
+If there's no unused final char for the specified kind of charset,
+return nil. */)
+ (dimension, chars)
+ Lisp_Object dimension, chars;
+{
+ int final_char;
+
+ CHECK_NUMBER (dimension);
+ CHECK_NUMBER (chars);
+ if (XINT (dimension) != 1 && XINT (dimension) != 2 && XINT (dimension) != 3)
+ args_out_of_range_3 (dimension, make_number (1), make_number (3));
+ if (XINT (chars) != 94 && XINT (chars) != 96)
+ args_out_of_range_3 (chars, make_number (94), make_number (96));
+ for (final_char = '0'; final_char <= '?'; final_char++)
+ if (ISO_CHARSET_TABLE (XINT (dimension), XINT (chars), final_char) < 0)
+ break;
+ return (final_char <= '?' ? make_number (final_char) : Qnil);
+}
+
+static void
+check_iso_charset_parameter (dimension, chars, final_char)
+ Lisp_Object dimension, chars, final_char;
+{
+ CHECK_NATNUM (dimension);
+ CHECK_NATNUM (chars);
+ CHECK_NATNUM (final_char);
+
+ if (XINT (dimension) > 3)
+ error ("Invalid DIMENSION %d, it should be 1, 2, or 3", XINT (dimension));
+ if (XINT (chars) != 94 && XINT (chars) != 96)
+ error ("Invalid CHARS %d, it should be 94 or 96", XINT (chars));
+ if (XINT (final_char) < '0' || XINT (final_char) > '~')
+ error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", XINT (chars));
+}
+
+
+DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset, Sdeclare_equiv_charset,
+ 4, 4, 0,
+ doc: /*
+Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET.
+CHARSET should be defined by `define-charset' in advance. */)
+ (dimension, chars, final_char, charset)
+ Lisp_Object dimension, chars, final_char, charset;
+{
+ int id;
+
+ CHECK_CHARSET_GET_ID (charset, id);
+ check_iso_charset_parameter (dimension, chars, final_char);
+
+ ISO_CHARSET_TABLE (XINT (dimension), XINT (chars), XINT (final_char)) = id;
+ return Qnil;
+}
+
+
+/* Return information about charsets in the text at PTR of NBYTES
+ bytes, which are NCHARS characters. The value is:
+
+ 0: Each character is represented by one byte. This is always
+ true for a unibyte string. For a multibyte string, true if
+ it contains only ASCII characters.
+
+ 1: No charsets other than ascii, control-1, and latin-1 are
+ found.
+
+ 2: Otherwise.
+*/
+
+int
+string_xstring_p (string)
+ Lisp_Object string;
+{
+ const unsigned char *p = XSTRING (string)->data;
+ const unsigned char *endp = p + STRING_BYTES (XSTRING (string));
+ struct charset *charset;
+
+ if (XSTRING (string)->size == STRING_BYTES (XSTRING (string)))
+ return 0;
+
+ charset = CHARSET_FROM_ID (charset_iso_8859_1);
+ while (p < endp)
+ {
+ int c = STRING_CHAR_ADVANCE (p);
+
+ /* Fixme: comparison of unsigned expression < 0 is always false */
+ if (ENCODE_CHAR (charset, c) < 0)
+ return 2;
+ }
+ return 1;
+}
+
+
+/* Find charsets in the string at PTR of NCHARS and NBYTES.
+
+ CHARSETS is a vector. Each element is a cons of CHARSET and
+ FOUND-FLAG. CHARSET is a charset id, and FOUND-FLAG is nil or t.
+ FOUND-FLAG t (or nil) means that the corresponding charset is
+ already found (or not yet found).
+
+ It may lookup a translation table TABLE if supplied. */
+
+static void
+find_charsets_in_text (ptr, nchars, nbytes, charsets, table)
+ const unsigned char *ptr;
+ int nchars, nbytes;
+ Lisp_Object charsets, table;
+{
+ const unsigned char *pend = ptr + nbytes;
+ int ncharsets = ASIZE (charsets);
+
+ if (nchars == nbytes)
+ return;
+
+ while (ptr < pend)
+ {
+ int c = STRING_CHAR_ADVANCE (ptr);
+ int i;
+ int all_found = 1;
+ Lisp_Object elt;
+
+ if (!NILP (table))
+ c = translate_char (table, c);
+ for (i = 0; i < ncharsets; i++)
+ {
+ elt = AREF (charsets, i);
+ if (NILP (XCDR (elt)))
+ {
+ struct charset *charset = CHARSET_FROM_ID (XINT (XCAR (elt)));
+
+ if (ENCODE_CHAR (charset, c) != CHARSET_INVALID_CODE (charset))
+ XCDR (elt) = Qt;
+ else
+ all_found = 0;
+ }
+ }
+ if (all_found)
+ break;
+ }
+}
+
+/* Fixme: returns nil for unibyte. */
+DEFUN ("find-charset-region", Ffind_charset_region, Sfind_charset_region,
+ 2, 3, 0,
+ doc: /* Return a list of charsets in the region between BEG and END.
+BEG and END are buffer positions.
+Optional arg TABLE if non-nil is a translation table to look up.
+
+If the current buffer is unibyte, the returned list may contain
+only `ascii', `eight-bit-control', and `eight-bit-graphic'. */)
+ (beg, end, table)
+ Lisp_Object beg, end, table;
+{
+ Lisp_Object charsets;
+ int from, from_byte, to, stop, stop_byte, i;
+ Lisp_Object val;
+
+ validate_region (&beg, &end);
+ from = XFASTINT (beg);
+ stop = to = XFASTINT (end);
+
+ if (from < GPT && GPT < to)
+ {
+ stop = GPT;
+ stop_byte = GPT_BYTE;
+ }
+ else
+ stop_byte = CHAR_TO_BYTE (stop);
+
+ from_byte = CHAR_TO_BYTE (from);
+
+ charsets = Fmake_vector (make_number (charset_table_used), Qnil);
+ for (i = 0; i < charset_table_used; i++)
+ ASET (charsets, i, Fcons (make_number (i), Qnil));
+
+ while (1)
+ {
+ find_charsets_in_text (BYTE_POS_ADDR (from_byte), stop - from,
+ stop_byte - from_byte, charsets, table);
+ if (stop < to)
+ {
+ from = stop, from_byte = stop_byte;
+ stop = to, stop_byte = CHAR_TO_BYTE (stop);
+ }
+ else
+ break;
+ }
+
+ val = Qnil;
+ for (i = charset_table_used - 1; i >= 0; i--)
+ if (!NILP (XCDR (AREF (charsets, i))))
+ val = Fcons (CHARSET_NAME (charset_table + i), val);
+ return val;
+}
+
+/* Fixme: returns nil for unibyte. */
+DEFUN ("find-charset-string", Ffind_charset_string, Sfind_charset_string,
+ 1, 2, 0,
+ doc: /* Return a list of charsets in STR.
+Optional arg TABLE if non-nil is a translation table to look up.
+
+If STR is unibyte, the returned list may contain
+only `ascii', `eight-bit-control', and `eight-bit-graphic'. */)
+ (str, table)
+ Lisp_Object str, table;
+{
+ Lisp_Object charsets;
+ int i;
+ Lisp_Object val;
+
+ CHECK_STRING (str);
+
+ charsets = Fmake_vector (make_number (charset_table_used), Qnil);
+ for (i = 0; i < charset_table_used; i++)
+ ASET (charsets, i, Fcons (make_number (i), Qnil));
+ find_charsets_in_text (XSTRING (str)->data, XSTRING (str)->size,
+ STRING_BYTES (XSTRING (str)), charsets, table);
+
+ val = Qnil;
+ for (i = charset_table_used - 1; i >= 0; i--)
+ if (!NILP (XCDR (AREF (charsets, i))))
+ val = Fcons (CHARSET_NAME (charset_table + i), val);
+ return val;
+}
+
+\f
+
+/* Return a character correponding to the code-point CODE of
+ CHARSET. */
+
+int
+decode_char (charset, code)
+ struct charset *charset;
+ unsigned code;
+{
+ int c, char_index;
+ enum charset_method method = CHARSET_METHOD (charset);
+
+ if (code < CHARSET_MIN_CODE (charset) || code > CHARSET_MAX_CODE (charset))
+ return -1;
+
+ if (method == CHARSET_METHOD_MAP_DEFERRED)
+ {
+ load_charset (charset);
+ method = CHARSET_METHOD (charset);
+ }
+
+ if (method == CHARSET_METHOD_SUBSET)
+ {
+ Lisp_Object subset_info;
+
+ subset_info = CHARSET_SUBSET (charset);
+ charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0)));
+ code -= XINT (AREF (subset_info, 3));
+ if (code < XFASTINT (AREF (subset_info, 1))
+ || code > XFASTINT (AREF (subset_info, 2)))
+ c = -1;
+ else
+ c = DECODE_CHAR (charset, code);
+ }
+ else if (method == CHARSET_METHOD_SUPERSET)
+ {
+ Lisp_Object parents;
+
+ parents = CHARSET_SUPERSET (charset);
+ c = -1;
+ for (; CONSP (parents); parents = XCDR (parents))
+ {
+ int id = XINT (XCAR (XCAR (parents)));
+ int code_offset = XINT (XCDR (XCAR (parents)));
+ unsigned this_code = code - code_offset;
+
+ charset = CHARSET_FROM_ID (id);
+ if ((c = DECODE_CHAR (charset, this_code)) >= 0)
+ break;
+ }
+ }
+ else
+ {
+ char_index = CODE_POINT_TO_INDEX (charset, code);
+ if (char_index < 0)
+ return -1;
+
+ if (method == CHARSET_METHOD_MAP)
+ {
+ Lisp_Object decoder;
+
+ decoder = CHARSET_DECODER (charset);
+ if (! VECTORP (decoder))
+ return -1;
+ c = XINT (AREF (decoder, char_index));
+ }
+ else
+ {
+ c = char_index + CHARSET_CODE_OFFSET (charset);
+ }
+ }
+
+ if (CHARSET_UNIFIED_P (charset)
+ && c >= 0)
+ {
+ MAYBE_UNIFY_CHAR (c);
+ }
+
+ return c;
+}
+
+/* Variable used temporarily by the macro ENCODE_CHAR. */
+Lisp_Object charset_work;
+
+/* Return a code-point of CHAR in CHARSET. If CHAR doesn't belong to
+ CHARSET, return CHARSET_INVALID_CODE (CHARSET). If STRICT is true,
+ use CHARSET's strict_max_char instead of max_char. */
+
+unsigned
+encode_char (charset, c)
+ struct charset *charset;
+ int c;
+{
+ unsigned code;
+ enum charset_method method = CHARSET_METHOD (charset);
+
+ if (CHARSET_UNIFIED_P (charset))
+ {
+ Lisp_Object deunifier, deunified;
+
+ deunifier = CHARSET_DEUNIFIER (charset);
+ if (! CHAR_TABLE_P (deunifier))
+ {
+ Funify_charset (CHARSET_NAME (charset), Qnil, Qnil);
+ deunifier = CHARSET_DEUNIFIER (charset);
+ }
+ deunified = CHAR_TABLE_REF (deunifier, c);
+ if (! NILP (deunified))
+ c = XINT (deunified);
+ }
+
+ if (method == CHARSET_METHOD_SUBSET)
+ {
+ Lisp_Object subset_info;
+ struct charset *this_charset;
+
+ subset_info = CHARSET_SUBSET (charset);
+ this_charset = CHARSET_FROM_ID (XFASTINT (AREF (subset_info, 0)));
+ code = ENCODE_CHAR (this_charset, c);
+ if (code == CHARSET_INVALID_CODE (this_charset)
+ || code < XFASTINT (AREF (subset_info, 1))
+ || code > XFASTINT (AREF (subset_info, 2)))
+ return CHARSET_INVALID_CODE (charset);
+ code += XINT (AREF (subset_info, 3));
+ return code;
+ }
+
+ if (method == CHARSET_METHOD_SUPERSET)
+ {
+ Lisp_Object parents;
+
+ parents = CHARSET_SUPERSET (charset);
+ for (; CONSP (parents); parents = XCDR (parents))
+ {
+ int id = XINT (XCAR (XCAR (parents)));
+ int code_offset = XINT (XCDR (XCAR (parents)));
+ struct charset *this_charset = CHARSET_FROM_ID (id);
+
+ code = ENCODE_CHAR (this_charset, c);
+ if (code != CHARSET_INVALID_CODE (this_charset))
+ return code + code_offset;
+ }
+ return CHARSET_INVALID_CODE (charset);
+ }
+
+ if (! CHARSET_FAST_MAP_REF ((c), charset->fast_map)
+ || c < CHARSET_MIN_CHAR (charset) || c > CHARSET_MAX_CHAR (charset))
+ return CHARSET_INVALID_CODE (charset);
+
+ if (method == CHARSET_METHOD_MAP_DEFERRED)
+ {
+ load_charset (charset);
+ method = CHARSET_METHOD (charset);
+ }
+
+ if (method == CHARSET_METHOD_MAP)
+ {
+ Lisp_Object encoder;
+ Lisp_Object val;
+
+ encoder = CHARSET_ENCODER (charset);
+ if (! CHAR_TABLE_P (CHARSET_ENCODER (charset)))
+ return CHARSET_INVALID_CODE (charset);
+ val = CHAR_TABLE_REF (encoder, c);
+ if (NILP (val))
+ return CHARSET_INVALID_CODE (charset);
+ code = XINT (val);
+ if (! CHARSET_COMPACT_CODES_P (charset))
+ code = INDEX_TO_CODE_POINT (charset, code);
+ }
+ else /* method == CHARSET_METHOD_OFFSET */
+ {
+ code = c - CHARSET_CODE_OFFSET (charset);
+ code = INDEX_TO_CODE_POINT (charset, code);
+ }
+
+ return code;
+}
+
+
+DEFUN ("decode-char", Fdecode_char, Sdecode_char, 2, 3, 0,
+ doc: /* Decode the pair of CHARSET and CODE-POINT into a character.
+Return nil if CODE-POINT is not valid in CHARSET.
+
+CODE-POINT may be a cons (HIGHER-16-BIT-VALUE . LOWER-16-BIT-VALUE).
+
+Optional argument RESTRICTION specifies a way to map the pair of CCS
+and CODE-POINT to a chracter. Currently not supported and just ignored. */)
+ (charset, code_point, restriction)
+ Lisp_Object charset, code_point, restriction;
+{
+ int c, id;
+ unsigned code;
+ struct charset *charsetp;
+
+ CHECK_CHARSET_GET_ID (charset, id);
+ if (CONSP (code_point))
+ {
+ CHECK_NATNUM (XCAR (code_point));
+ CHECK_NATNUM (XCDR (code_point));
+ code = (XINT (XCAR (code_point)) << 16) | (XINT (XCDR (code_point)));
+ }
+ else
+ {
+ CHECK_NATNUM (code_point);
+ code = XINT (code_point);
+ }
+ charsetp = CHARSET_FROM_ID (id);
+ c = DECODE_CHAR (charsetp, code);
+ return (c >= 0 ? make_number (c) : Qnil);
+}
+
+
+DEFUN ("encode-char", Fencode_char, Sencode_char, 2, 3, 0,
+ doc: /* Encode the character CH into a code-point of CHARSET.
+Return nil if CHARSET doesn't include CH.
+
+Optional argument RESTRICTION specifies a way to map CHAR to a
+code-point in CCS. Currently not supported and just ignored. */)
+ (ch, charset, restriction)
+ Lisp_Object ch, charset, restriction;
+{
+ int id;
+ unsigned code;
+ struct charset *charsetp;
+
+ CHECK_CHARSET_GET_ID (charset, id);
+ CHECK_NATNUM (ch);
+ charsetp = CHARSET_FROM_ID (id);
+ code = ENCODE_CHAR (charsetp, XINT (ch));
+ if (code == CHARSET_INVALID_CODE (charsetp))
+ return Qnil;
+ if (code > 0x7FFFFFF)
+ return Fcons (make_number (code >> 16), make_number (code & 0xFFFF));
+ return make_number (code);
+}
+
+
+DEFUN ("make-char", Fmake_char, Smake_char, 1, 5, 0,
+ doc:
+ /* Return a character of CHARSET whose position codes are CODEn.
+
+CODE1 through CODE4 are optional, but if you don't supply sufficient
+position codes, it is assumed that the minimum code in each dimension
+is specified. */)
+ (charset, code1, code2, code3, code4)
+ Lisp_Object charset, code1, code2, code3, code4;
+{
+ int id, dimension;
+ struct charset *charsetp;
+ unsigned code;
+ int c;
+
+ CHECK_CHARSET_GET_ID (charset, id);
+ charsetp = CHARSET_FROM_ID (id);
+
+ dimension = CHARSET_DIMENSION (charsetp);
+ if (NILP (code1))
+ code = (CHARSET_ASCII_COMPATIBLE_P (charsetp)
+ ? 0 : CHARSET_MIN_CODE (charsetp));
+ else
+ {
+ CHECK_NATNUM (code1);
+ if (XFASTINT (code1) >= 0x100)
+ args_out_of_range (make_number (0xFF), code1);
+ code = XFASTINT (code1);
+
+ if (dimension > 1)
+ {
+ code <<= 8;
+ if (NILP (code2))
+ code |= charsetp->code_space[(dimension - 2) * 4];
+ else
+ {
+ CHECK_NATNUM (code2);
+ if (XFASTINT (code2) >= 0x100)
+ args_out_of_range (make_number (0xFF), code2);
+ code |= XFASTINT (code2);
+ }
+
+ if (dimension > 2)
+ {
+ code <<= 8;
+ if (NILP (code3))
+ code |= charsetp->code_space[(dimension - 3) * 4];
+ else
+ {
+ CHECK_NATNUM (code3);
+ if (XFASTINT (code3) >= 0x100)
+ args_out_of_range (make_number (0xFF), code3);
+ code |= XFASTINT (code3);
+ }
+
+ if (dimension > 3)
+ {
+ code <<= 8;
+ if (NILP (code4))
+ code |= charsetp->code_space[0];
+ else
+ {
+ CHECK_NATNUM (code4);
+ if (XFASTINT (code4) >= 0x100)
+ args_out_of_range (make_number (0xFF), code4);
+ code |= XFASTINT (code4);
+ }
+ }
+ }
+ }
+ }
+
+ if (CHARSET_ISO_FINAL (charsetp) >= 0)
+ code &= 0x7F7F7F7F;
+ c = DECODE_CHAR (charsetp, code);
+ if (c < 0)
+ error ("Invalid code(s)");
+ return make_number (c);
+}
+
+
+/* Return the first charset in CHARSET_LIST that contains C.
+ CHARSET_LIST is a list of charset IDs. If it is nil, use
+ Vcharset_ordered_list. */
+
+struct charset *
+char_charset (c, charset_list, code_return)
+ int c;
+ Lisp_Object charset_list;
+ unsigned *code_return;
+{
+ if (NILP (charset_list))
+ charset_list = Vcharset_ordered_list;
+
+ while (CONSP (charset_list))
+ {
+ struct charset *charset = CHARSET_FROM_ID (XINT (XCAR (charset_list)));
+ unsigned code = ENCODE_CHAR (charset, c);
+
+ if (code != CHARSET_INVALID_CODE (charset))
+ {
+ if (code_return)
+ *code_return = code;
+ return charset;
+ }
+ charset_list = XCDR (charset_list);
+ }
+ return NULL;
+}
+
+
+/* Fixme: `unknown' can't happen now? */
+DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0,
+ doc: /*Return list of charset and one to three position-codes of CHAR.
+If CHAR is invalid as a character code, return a list `(unknown CHAR)'. */)
+ (ch)
+ Lisp_Object ch;
+{
+ struct charset *charset;
+ int c, dimension;
+ unsigned code;
+ Lisp_Object val;
+
+ CHECK_CHARACTER (ch);
+ c = XFASTINT (ch);
+ charset = CHAR_CHARSET (c);
+ if (! charset)
+ return Fcons (intern ("unknown"), Fcons (ch, Qnil));
+
+ code = ENCODE_CHAR (charset, c);
+ if (code == CHARSET_INVALID_CODE (charset))
+ abort ();
+ dimension = CHARSET_DIMENSION (charset);
+ val = (dimension == 1 ? Fcons (make_number (code), Qnil)
+ : dimension == 2 ? Fcons (make_number (code >> 8),
+ Fcons (make_number (code & 0xFF), Qnil))
+ : Fcons (make_number (code >> 16),
+ Fcons (make_number ((code >> 8) & 0xFF),
+ Fcons (make_number (code & 0xFF), Qnil))));
+ return Fcons (CHARSET_NAME (charset), val);
+}
+
+
+DEFUN ("char-charset", Fchar_charset, Schar_charset, 1, 1, 0,
+ doc: /* Return the charset of highest priority that contains CHAR. */)
+ (ch)
+ Lisp_Object ch;
+{
+ struct charset *charset;
+
+ CHECK_CHARACTER (ch);
+ charset = CHAR_CHARSET (XINT (ch));
+ return (CHARSET_NAME (charset));
+}
+
+
+DEFUN ("charset-after", Fcharset_after, Scharset_after, 0, 1, 0,
+ doc: /*
+Return charset of a character in the current buffer at position POS.
+If POS is nil, it defauls to the current point.
+If POS is out of range, the value is nil. */)
+ (pos)
+ Lisp_Object pos;
+{
+ Lisp_Object ch;
+ struct charset *charset;
+
+ ch = Fchar_after (pos);
+ if (! INTEGERP (ch))
+ return ch;
+ charset = CHAR_CHARSET (XINT (ch));
+ return (CHARSET_NAME (charset));
+}
+
+
+DEFUN ("iso-charset", Fiso_charset, Siso_charset, 3, 3, 0,
+ doc: /*
+Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR.
+
+ISO 2022's designation sequence (escape sequence) distinguishes charsets
+by their DIMENSION, CHARS, and FINAL-CHAR,
+where as Emacs distinguishes them by charset symbol.
+See the documentation of the function `charset-info' for the meanings of
+DIMENSION, CHARS, and FINAL-CHAR. */)
+ (dimension, chars, final_char)
+ Lisp_Object dimension, chars, final_char;
+{
+ int id;
+
+ check_iso_charset_parameter (dimension, chars, final_char);
+ id = ISO_CHARSET_TABLE (XFASTINT (dimension), XFASTINT (chars),
+ XFASTINT (final_char));
+ return (id >= 0 ? CHARSET_NAME (CHARSET_FROM_ID (id)) : Qnil);
+}
+
+
+DEFUN ("clear-charset-maps", Fclear_charset_maps, Sclear_charset_maps,
+ 0, 0, 0,
+ doc: /*
+Clear encoder and decoder of charsets that are loaded from mapfiles. */)
+ ()
+{
+ int i;
+ struct charset *charset;
+ Lisp_Object attrs;
+
+ for (i = 0; i < charset_table_used; i++)
+ {
+ charset = CHARSET_FROM_ID (i);
+ attrs = CHARSET_ATTRIBUTES (charset);
+
+ if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP)
+ {
+ CHARSET_ATTR_DECODER (attrs) = Qnil;
+ CHARSET_ATTR_ENCODER (attrs) = Qnil;
+ CHARSET_METHOD (charset) = CHARSET_METHOD_MAP_DEFERRED;
+ }
+
+ if (CHARSET_UNIFIED_P (charset))
+ CHARSET_ATTR_DEUNIFIER (attrs) = Qnil;
+ }
+
+ if (CHAR_TABLE_P (Vchar_unified_charset_table))
+ {
+ Foptimize_char_table (Vchar_unified_charset_table);
+ Vchar_unify_table = Vchar_unified_charset_table;
+ Vchar_unified_charset_table = Qnil;
+ }
+
+ return Qnil;
+}
+
+DEFUN ("charset-priority-list", Fcharset_priority_list,
+ Scharset_priority_list, 0, 1, 0,
+ doc: /* Return the list of charsets ordered by priority.
+HIGHESTP non-nil means just return the highest priority one. */)
+ (highestp)
+ Lisp_Object highestp;
+{
+ Lisp_Object val = Qnil, list = Vcharset_ordered_list;
+
+ if (!NILP (highestp))
+ return CHARSET_NAME (CHARSET_FROM_ID (XINT (Fcar (list))));
+
+ while (!NILP (list))
+ {
+ val = Fcons (CHARSET_NAME (CHARSET_FROM_ID (XINT (XCAR (list)))), val);
+ list = XCDR (list);
+ }
+ return Fnreverse (val);
+}
+
+DEFUN ("set-charset-priority", Fset_charset_priority, Sset_charset_priority,
+ 1, MANY, 0,
+ doc: /* Assign higher priority to the charsets given as arguments.
+usage: (set-charset-priority &rest charsets) */)
+ (nargs, args)
+ int nargs;
+ Lisp_Object *args;
+{
+ Lisp_Object new_head = Qnil, old_list, arglist[2];
+ int i, id;
+
+ old_list = Fcopy_sequence (Vcharset_ordered_list);
+ for (i = 0; i < nargs; i++)
+ {
+ CHECK_CHARSET_GET_ID (args[i], id);
+ old_list = Fdelq (make_number (id), old_list);
+ new_head = Fcons (make_number (id), new_head);
+ }
+ arglist[0] = Fnreverse (new_head);
+ arglist[1] = old_list;
+ Vcharset_ordered_list = Fnconc (2, arglist);
+ charset_ordered_list_tick++;
+
+ for (old_list = Vcharset_ordered_list, new_head = Qnil;
+ CONSP (old_list); old_list = XCDR (old_list))
+ {
+ if (! NILP (Fmemq (XCAR (old_list), Viso_2022_charset_list)))
+ new_head = Fcons (XCAR (old_list), new_head);
+ }
+ Viso_2022_charset_list = Fnreverse (new_head);
+
+ return Qnil;
+}
+\f
+void
+init_charset ()
+{
+
+}
+
+
+void
+init_charset_once ()
+{
+ int i, j, k;
+
+ for (i = 0; i < ISO_MAX_DIMENSION; i++)
+ for (j = 0; j < ISO_MAX_CHARS; j++)
+ for (k = 0; k < ISO_MAX_FINAL; k++)
+ iso_charset_table[i][j][k] = -1;
+
+ for (i = 0; i < 256; i++)
+ emacs_mule_charset[i] = NULL;
+
+ charset_jisx0201_roman = -1;
+ charset_jisx0208_1978 = -1;
+ charset_jisx0208 = -1;
+
+ for (i = 0; i < 256; i++)
+ unibyte_to_multibyte_table[i] = i;
+}
+
+#ifdef emacs
+
+void
+syms_of_charset ()
+{
+ char *p;
+
+ DEFSYM (Qcharsetp, "charsetp");
+
+ DEFSYM (Qascii, "ascii");
+ DEFSYM (Qunicode, "unicode");
+ DEFSYM (Qeight_bit, "eight-bit");
+ DEFSYM (Qiso_8859_1, "iso-8859-1");
+
+ DEFSYM (Qgl, "gl");
+ DEFSYM (Qgr, "gr");
+
+ p = (char *) xmalloc (30000);
+
+ staticpro (&Vcharset_ordered_list);
+ Vcharset_ordered_list = Qnil;
+
+ staticpro (&Viso_2022_charset_list);
+ Viso_2022_charset_list = Qnil;
+
+ staticpro (&Vemacs_mule_charset_list);
+ Vemacs_mule_charset_list = Qnil;
+
+ staticpro (&Vcharset_hash_table);
+ Vcharset_hash_table = Fmakehash (Qeq);
+
+ charset_table_size = 128;
+ charset_table = ((struct charset *)
+ xmalloc (sizeof (struct charset) * charset_table_size));
+ charset_table_used = 0;
+
+ staticpro (&Vchar_unified_charset_table);
+ Vchar_unified_charset_table = Fmake_char_table (Qnil, make_number (-1));
+
+ defsubr (&Scharsetp);
+ defsubr (&Smap_charset_chars);
+ defsubr (&Sdefine_charset_internal);
+ defsubr (&Sdefine_charset_alias);
+ defsubr (&Sunibyte_charset);
+ defsubr (&Sset_unibyte_charset);
+ defsubr (&Scharset_plist);
+ defsubr (&Sset_charset_plist);
+ defsubr (&Sunify_charset);
+ defsubr (&Sget_unused_iso_final_char);
+ defsubr (&Sdeclare_equiv_charset);
+ defsubr (&Sfind_charset_region);
+ defsubr (&Sfind_charset_string);
+ defsubr (&Sdecode_char);
+ defsubr (&Sencode_char);
+ defsubr (&Ssplit_char);
+ defsubr (&Smake_char);
+ defsubr (&Schar_charset);
+ defsubr (&Scharset_after);
+ defsubr (&Siso_charset);
+ defsubr (&Sclear_charset_maps);
+ defsubr (&Scharset_priority_list);
+ defsubr (&Sset_charset_priority);
+
+ DEFVAR_LISP ("charset-map-directory", &Vcharset_map_directory,
+ doc: /* Directory of charset map files that come with GNU Emacs.
+The default value is sub-directory "charsets" of `data-directory'. */);
+ Vcharset_map_directory = Fexpand_file_name (build_string ("charsets"),
+ Vdata_directory);
+
+ DEFVAR_LISP ("charset-list", &Vcharset_list,
+ doc: /* List of all charsets ever defined. */);
+ Vcharset_list = Qnil;
+
+ charset_ascii
+ = define_charset_internal (Qascii, 1, "\x00\x7F\x00\x00\x00\x00",
+ 0, 127, 'B', -1, 0, 1, 0, 0);
+ charset_iso_8859_1
+ = define_charset_internal (Qiso_8859_1, 1, "\x00\xFF\x00\x00\x00\x00",
+ 0, 255, -1, -1, -1, 1, 0, 0);
+ charset_unicode
+ = define_charset_internal (Qunicode, 3, "\x00\xFF\x00\xFF\x00\x10",
+ 0, MAX_UNICODE_CHAR, -1, 0, -1, 1, 0, 0);
+ charset_eight_bit
+ = define_charset_internal (Qeight_bit, 1, "\x80\xFF\x00\x00\x00\x00",
+ 128, 255, -1, 0, -1, 0, 0,
+ MAX_5_BYTE_CHAR + 1);
+}
+
+#endif /* emacs */