-/* Basic multilingual character support.
- Copyright (C) 1995, 1997, 1998 Electrotechnical Laboratory, JAPAN.
+/* 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.
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>
+#include <unistd.h>
+#include <ctype.h>
#ifdef emacs
#include <sys/types.h>
#include "lisp.h"
-#include "buffer.h"
+#include "character.h"
#include "charset.h"
-#include "composite.h"
#include "coding.h"
#include "disptab.h"
+#include "buffer.h"
#else /* not emacs */
#endif /* emacs */
-Lisp_Object Qcharset, Qascii, Qeight_bit_control, Qeight_bit_graphic;
-Lisp_Object Qunknown;
-
-/* Declaration of special leading-codes. */
-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. The values are set by
- Fsetup_special_charsets. */
-int charset_latin_iso8859_1; /* ISO8859-1 (Latin-1) */
-int charset_jisx0208_1978; /* JISX0208.1978 (Japanese Kanji old set) */
-int charset_jisx0208; /* JISX0208.1983 (Japanese Kanji) */
-int charset_katakana_jisx0201; /* JISX0201.Kana (Japanese Katakana) */
-int charset_latin_jisx0201; /* JISX0201.Roman (Japanese Roman) */
-int charset_big5_1; /* Big5 Level 1 (Chinese Traditional) */
-int charset_big5_2; /* Big5 Level 2 (Chinese Traditional) */
+/*** GENERAL NOTE on CODED CHARACTER SET (CHARSET) ***
-Lisp_Object Qcharset_table;
+ 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 symbol. In C code, a charset is
+ represented by its ID number or by a pointer the struct charset.
-/* A char-table containing information of each character set. */
-Lisp_Object Vcharset_table;
+ 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 struct charset.
-/* 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. */
+/* List of all charsets. This variable is used only from Emacs
+ Lisp. */
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;
+/* 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_control;
+Lisp_Object Qeight_bit_graphic;
+Lisp_Object Qiso_8859_1;
+Lisp_Object Qunicode;
-/* A char-table for characters which may invoke auto-filling. */
-Lisp_Object Vauto_fill_chars;
+/* The corresponding charsets. */
+int charset_ascii;
+int charset_8_bit_control;
+int charset_8_bit_graphic;
+int charset_iso_8859_1;
+int charset_unicode;
-Lisp_Object Qauto_fill_chars;
+/* Value of charset attribute `charset-iso-plane'. */
+Lisp_Object Qgl, Qgr;
-/* 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];
+/* The primary charset. It is a charset of unibyte characters. */
+int charset_primary;
+
+/* List of charsets ordered by the priority. */
+Lisp_Object Vcharset_ordered_list;
+
+/* 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[2][2][128];
+int iso_charset_table[ISO_MAX_DIMENSION][ISO_MAX_CHARS][ISO_MAX_FINAL];
+
+Lisp_Object Vcharset_map_directory;
+
+Lisp_Object Vchar_unified_charset_table;
+
+#define CODE_POINT_TO_INDEX(charset, code) \
+ ((charset)->code_linear_p \
+ ? (code) - (charset)->min_code \
+ : ((((code) >> 24) <= (charset)->code_space[13]) \
+ && ((((code) >> 16) & 0xFF) <= (charset)->code_space[9]) \
+ && ((((code) >> 8) & 0xFF) <= (charset)->code_space[5]) \
+ && (((code) & 0xFF) <= (charset)->code_space[1])) \
+ ? (((((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])) \
+ : -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 \
+ : (((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)))
-/* Variables used locally in the macro FETCH_MULTIBYTE_CHAR. */
-unsigned char *_fetch_multibyte_char_p;
-int _fetch_multibyte_char_len;
+\f
-/* Offset to add to a non-ASCII value when inserting it. */
-int nonascii_insert_offset;
+/* Set to 1 when a charset map is loaded to warn that a buffer text
+ and a string data may be relocated. */
+int charset_map_loaded;
-/* Translation table for converting non-ASCII unibyte characters
- to multibyte codes, or nil. */
-Lisp_Object Vnonascii_translation_table;
+/* Parse the mapping vector MAP which has this form:
+ [CODE0 CHAR0 CODE1 CHAR1 ... ]
-/* List of all possible generic characters. */
-Lisp_Object Vgeneric_character_list;
+ If CONTROL_FLAG is 0, setup CHARSET->min_char and CHARSET->max_char.
-\f
-void
-invalid_character (c)
- int c;
-{
- error ("Invalid character: 0%o, %d, 0x%x", c, c, c);
-}
+ If CONTROL_FLAG is 1, setup CHARSET->min_char, CHARSET->max_char,
+ CHARSET->decoder, and CHARSET->encoder.
-/* Parse string STR of length LENGTH and fetch information of a
- character at STR. Set BYTES to the byte length the character
- occupies, CHARSET, C1, C2 to proper values of the character. */
-
-#define SPLIT_MULTIBYTE_SEQ(str, length, bytes, charset, c1, c2) \
- do { \
- (c1) = *(str); \
- (bytes) = BYTES_BY_CHAR_HEAD (c1); \
- if ((bytes) == 1) \
- (charset) = ASCII_BYTE_P (c1) ? CHARSET_ASCII : CHARSET_8_BIT_GRAPHIC; \
- else if ((bytes) == 2) \
- { \
- if ((c1) == LEADING_CODE_8_BIT_CONTROL) \
- (charset) = CHARSET_8_BIT_CONTROL, (c1) = (str)[1] - 0x20; \
- else \
- (charset) = (c1), (c1) = (str)[1] & 0x7F; \
- } \
- else if ((bytes) == 3) \
- { \
- if ((c1) < LEADING_CODE_PRIVATE_11) \
- (charset) = (c1), (c1) = (str)[1] & 0x7F, (c2) = (str)[2] & 0x7F; \
- else \
- (charset) = (str)[1], (c1) = (str)[2] & 0x7F; \
- } \
- else \
- (charset) = (str)[1], (c1) = (str)[2] & 0x7F, (c2) = (str)[3] & 0x7F; \
- } while (0)
-
-/* 1 if CHARSET, C1, and C2 compose a valid character, else 0. */
-#define CHAR_COMPONENTS_VALID_P(charset, c1, c2) \
- ((charset) == CHARSET_ASCII \
- ? ((c1) >= 0 && (c1) <= 0x7F) \
- : ((charset) == CHARSET_8_BIT_CONTROL \
- ? ((c1) >= 0x80 && (c1) <= 0x9F) \
- : ((charset) == CHARSET_8_BIT_GRAPHIC \
- ? ((c1) >= 0x80 && (c1) <= 0xFF) \
- : (CHARSET_DIMENSION (charset) == 1 \
- ? ((c1) >= 0x20 && (c1) <= 0x7F) \
- : ((c1) >= 0x20 && (c1) <= 0x7F \
- && (c2) >= 0x20 && (c2) <= 0x7F)))))
-
-/* Store multi-byte form of the character C in STR. The caller should
- allocate at least 4-byte area at STR in advance. Returns the
- length of the multi-byte form. If C is an invalid character code,
- return -1. */
+ If CONTROL_FLAG is 2, setup CHARSET->deunifier and
+ Vchar_unify_table. If Vchar_unified_charset_table is non-nil,
+ setup it too. */
-int
-char_to_string_1 (c, str)
- int c;
- unsigned char *str;
+static void
+parse_charset_map (charset, map, control_flag)
+ struct charset *charset;
+ Lisp_Object map;
+ int control_flag;
{
- unsigned char *p = str;
+ Lisp_Object vec, table;
+ unsigned min_code = CHARSET_MIN_CODE (charset);
+ unsigned max_code = CHARSET_MAX_CODE (charset);
+ int ascii_compatible_p = charset->ascii_compatible_p;
+ int min_char, max_char, nonascii_min_char;
+ int size;
+ int i;
+ int first;
+ unsigned char *fast_map = charset->fast_map;
- if (c & CHAR_MODIFIER_MASK) /* This includes the case C is negative. */
+ if (control_flag)
{
- /* Multibyte character can't have a modifier bit. */
- if (! SINGLE_BYTE_CHAR_P ((c & ~CHAR_MODIFIER_MASK)))
- return -1;
+ int n = CODE_POINT_TO_INDEX (charset, max_code) + 1;
+ unsigned invalid_code = CHARSET_INVALID_CODE (charset);
- /* 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));
- }
+ table = Fmake_char_table (Qnil, make_number (invalid_code));
+ 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, make_number (-1));
- /* If C still has any modifier bits, just ignore it. */
- c &= ~CHAR_MODIFIER_MASK;
+ charset_map_loaded = 1;
}
-
- if (SINGLE_BYTE_CHAR_P (c))
+
+ size = ASIZE (map);
+ nonascii_min_char = MAX_CHAR;
+ CHARSET_COMPACT_CODES_P (charset) = 1;
+ for (first = 1, i = 0; i < size; i += 2)
{
- if (ASCII_BYTE_P (c) || c >= 0xA0)
- *p++ = c;
- else
+ Lisp_Object val;
+ unsigned code, temp;
+ int c, char_index;
+
+ val = AREF (map, i);
+ CHECK_NATNUM (val);
+ code = XFASTINT (val);
+ val = AREF (map, i + 1);
+ CHECK_NATNUM (val);
+ c = XFASTINT (val);
+
+ if (code < min_code || code > max_code)
+ continue;
+ char_index = CODE_POINT_TO_INDEX (charset, code);
+ if (char_index < 0
+ || c > MAX_CHAR)
+ continue;
+
+ if (control_flag < 2)
+ {
+ if (first)
+ {
+ min_char = max_char = c;
+ first = 0;
+ }
+ else if (c > max_char)
+ max_char = c;
+ else if (c < min_char)
+ min_char = c;
+ if (ascii_compatible_p && ! ASCII_BYTE_P (c)
+ && c < nonascii_min_char)
+ nonascii_min_char = c;
+
+ CHARSET_FAST_MAP_SET (c, fast_map);
+ }
+
+ if (control_flag)
{
- *p++ = LEADING_CODE_8_BIT_CONTROL;
- *p++ = c + 0x20;
+ if (control_flag == 1)
+ {
+ if (char_index >= ASIZE (vec))
+ abort ();
+ ASET (vec, char_index, make_number (c));
+ if (code > 0x7FFFFFF)
+ {
+ CHAR_TABLE_SET (table, c,
+ Fcons (make_number (code >> 16),
+ make_number (code & 0xFFFF)));
+ CHARSET_COMPACT_CODES_P (charset) = 0;
+ }
+ else
+ CHAR_TABLE_SET (table, c, make_number (code));
+ }
+ else
+ {
+ int c1 = DECODE_CHAR (charset, code);
+ if (c1 >= 0)
+ {
+ CHAR_TABLE_SET (table, c, make_number (c1));
+ CHAR_TABLE_SET (Vchar_unify_table, c1, c);
+ if (CHAR_TABLE_P (Vchar_unified_charset_table))
+ CHAR_TABLE_SET (Vchar_unified_charset_table, c1,
+ CHARSET_NAME (charset));
+ }
+ }
}
}
- else if (CHAR_VALID_P (c, 0))
+
+ if (control_flag < 2)
{
- int charset, c1, c2;
-
- SPLIT_CHAR (c, charset, c1, c2);
-
- if (charset >= LEADING_CODE_EXT_11)
- *p++ = (charset < LEADING_CODE_EXT_12
- ? LEADING_CODE_PRIVATE_11
- : (charset < LEADING_CODE_EXT_21
- ? LEADING_CODE_PRIVATE_12
- : (charset < LEADING_CODE_EXT_22
- ? LEADING_CODE_PRIVATE_21
- : LEADING_CODE_PRIVATE_22)));
- *p++ = charset;
- if ((c1 > 0 && c1 < 32) || (c2 > 0 && c2 < 32))
- return -1;
- if (c1)
+ CHARSET_MIN_CHAR (charset) = (ascii_compatible_p
+ ? nonascii_min_char : min_char);
+ CHARSET_MAX_CHAR (charset) = max_char;
+ if (control_flag)
{
- *p++ = c1 | 0x80;
- if (c2 > 0)
- *p++ = c2 | 0x80;
+ CHARSET_DECODER (charset) = vec;
+ CHARSET_ENCODER (charset) = table;
}
}
else
- return -1;
-
- return (p - str);
+ CHARSET_DEUNIFIER (charset) = table;
}
-/* Store multi-byte form of the character C in STR. The caller should
- allocate at least 4-byte area at STR in advance. Returns the
- length of the multi-byte form. If C is an invalid character code,
- signal an error.
+/* Read a hexadecimal number (preceded by "0x") from the file FP while
+ paying attention to comment charcter '#'. */
- Use macro `CHAR_STRING (C, STR)' instead of calling this function
- directly if C can be an ASCII character. */
-
-int
-char_to_string (c, str)
- int c;
- unsigned char *str;
+static INLINE unsigned
+read_hex (fp, eof)
+ FILE *fp;
+ int *eof;
{
- int len;
- len = char_to_string_1 (c, str);
- if (len == -1)
- invalid_character (c);
- return len;
-}
+ int c;
+ unsigned n;
+ while ((c = getc (fp)) != EOF)
+ {
+ if (c == '#' || 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';
+ return n;
+}
-/* Return the non-ASCII character corresponding to multi-byte form at
- STR of length LEN. If ACTUAL_LEN is not NULL, store the byte
- length of the multibyte form in *ACTUAL_LEN.
- Use macros STRING_CHAR or STRING_CHAR_AND_LENGTH instead of calling
- this function directly if you want ot handle ASCII characters as
- well. */
+/* Return a mapping vector for CHARSET loaded from MAPFILE.
+ Each line of MAPFILE has this form:
+ 0xAAAA 0xBBBB
+ where 0xAAAA is a code-point and 0xBBBB is the corresponding
+ character code.
+ The returned vector has this form:
+ [ CODE1 CHAR1 CODE2 CHAR2 .... ]
+*/
-int
-string_to_char (str, len, actual_len)
- const unsigned char *str;
- int len, *actual_len;
+static Lisp_Object
+load_charset_map (charset, mapfile)
+ struct charset *charset;
+ Lisp_Object mapfile;
{
- int c, bytes, charset, c1, c2;
+ int fd;
+ FILE *fp;
+ int num;
+ unsigned *numbers_table[256];
+ int numbers_table_used;
+ unsigned *numbers;
+ int eof;
+ Lisp_Object suffixes;
+ Lisp_Object vec;
+ int i;
- SPLIT_MULTIBYTE_SEQ (str, len, bytes, charset, c1, c2);
- c = MAKE_CHAR (charset, c1, c2);
- if (actual_len)
- *actual_len = bytes;
- return c;
-}
+ suffixes = Fcons (build_string (".map"),
+ Fcons (build_string (".TXT"), Qnil));
-/* 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;
+ 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 Qnil;
+ }
- PARSE_MULTIBYTE_SEQ (str, len, bytes);
- return bytes;
-}
+ numbers_table_used = 0;
+ num = 0;
+ eof = 0;
+ while (1)
+ {
+ unsigned n = read_hex (fp, &eof);
-/* 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. */
+ if (eof)
+ break;
+ if ((num % 0x10000) == 0)
+ {
+ if (numbers_table_used == 256)
+ break;
+ numbers = (unsigned *) alloca (sizeof (unsigned) * 0x10000);
+ numbers_table[numbers_table_used++] = numbers;
+ }
+ *numbers++ = n;
+ num++;
+ }
+ fclose (fp);
+ close (fd);
-int
-split_string (str, len, charset, c1, c2)
- const unsigned char *str;
- unsigned char *c1, *c2;
- int len, *charset;
-{
- register int bytes, cs, code1, code2 = -1;
+ vec = Fmake_vector (make_number (num), Qnil);
+ for (i = 0; i < num; i++, numbers++)
+ {
+ if ((i % 0x10000) == 0)
+ numbers = numbers_table[i / 0x10000];
+ ASET (vec, i, make_number (*numbers));
+ }
- SPLIT_MULTIBYTE_SEQ (str, len, bytes, cs, code1, code2);
- if (cs == CHARSET_ASCII)
- return -1;
- *charset = cs;
- *c1 = code1;
- *c2 = code2;
- return 0;
+ charset_map_loaded = 1;
+
+ return vec;
}
-/* Return 1 iff character C has valid printable glyph.
- Use the macro CHAR_PRINTABLE_P instead. */
-int
-char_printable_p (c)
- int c;
+static void
+load_charset (charset)
+ struct charset *charset;
{
- int charset, c1, c2;
+ if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP_DEFERRED)
+ {
+ Lisp_Object map;
- if (ASCII_BYTE_P (c))
- return 1;
- else if (SINGLE_BYTE_CHAR_P (c))
- return 0;
- else if (c >= MAX_CHAR)
- return 0;
-
- SPLIT_CHAR (c, charset, c1, c2);
- if (! CHARSET_DEFINED_P (charset))
- return 0;
- if (CHARSET_CHARS (charset) == 94
- ? c1 <= 32 || c1 >= 127
- : c1 < 32)
- return 0;
- if (CHARSET_DIMENSION (charset) == 2
- && (CHARSET_CHARS (charset) == 94
- ? c2 <= 32 || c2 >= 127
- : c2 < 32))
- return 0;
- return 1;
+ map = CHARSET_MAP (charset);
+ if (STRINGP (map))
+ map = load_charset_map (charset, map);
+ parse_charset_map (charset, map, 1);
+ CHARSET_METHOD (charset) = CHARSET_METHOD_MAP;
+ }
}
-/* 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;
+
+DEFUN ("charsetp", Fcharsetp, Scharsetp, 1, 1, 0,
+ doc: /* Return non-nil if and only if OBJECT is a charset.*/)
+ (object)
+ Lisp_Object object;
{
- 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);
+ return (CHARSETP (object) ? Qt : Qnil);
}
-/* 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;
+void
+map_charset_chars (c_function, function, charset_symbol, arg)
+ void (*c_function) (Lisp_Object, Lisp_Object, Lisp_Object);
+ Lisp_Object function, charset_symbol, arg;
{
- if (c < 0400 && c >= 0200)
- {
- int c_save = c;
+ int id;
+ struct charset *charset;
+ Lisp_Object range;
- if (! NILP (Vnonascii_translation_table))
+ CHECK_CHARSET_GET_ID (charset_symbol, id);
+ charset = CHARSET_FROM_ID (id);
+
+ if (CHARSET_METHOD (charset) == CHARSET_METHOD_MAP_DEFERRED)
+ load_charset (charset);
+
+ if (CHARSET_METHOD (charset) == CHARSET_METHOD_OFFSET)
+ {
+ range = Fcons (make_number (CHARSET_MIN_CHAR (charset)),
+ make_number (CHARSET_MAX_CHAR (charset)));
+ if (NILP (function))
+ (*c_function) (arg, range, Qnil);
+ 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))
{
- c = XINT (Faref (Vnonascii_translation_table, make_number (c)));
- if (c >= 0400 && ! char_valid_p (c, 0))
- c = c_save + DEFAULT_NONASCII_INSERT_OFFSET;
+ range = Fcons (make_number (0), make_number (127));
+ if (NILP (function))
+ (*c_function) (arg, range, Qnil);
+ else
+ call2 (function, range, arg);
}
- else if (c >= 0240 && nonascii_insert_offset > 0)
+ map_char_table (c_function, function, CHARSET_ENCODER (charset), arg,
+ 0, NULL);
+ }
+ else /* i.e. CHARSET_METHOD_PARENT */
+ {
+ int from, to, c;
+ unsigned code;
+ int i, j, k, l;
+ int *code_space = CHARSET_CODE_SPACE (charset);
+ Lisp_Object val;
+
+ range = Fcons (Qnil, Qnil);
+ from = to = -2;
+ for (i = code_space[12]; i <= code_space[13]; i++)
+ for (j = code_space[8]; j <= code_space[9]; j++)
+ for (k = code_space[4]; k <= code_space[5]; k++)
+ for (l = code_space[0]; l <= code_space[1]; l++)
+ {
+ code = (i << 24) | (j << 16) | (k << 8) | l;
+ c = DECODE_CHAR (charset, code);
+ if (c == to + 1)
+ {
+ to++;
+ continue;
+ }
+ if (from >= 0)
+ {
+ if (from < to)
+ {
+ XSETCAR (range, make_number (from));
+ XSETCDR (range, make_number (to));
+ val = range;
+ }
+ else
+ val = make_number (from);
+ if (NILP (function))
+ (*c_function) (arg, val, Qnil);
+ else
+ call2 (function, val, arg);
+ }
+ from = to = (c < 0 ? -2 : c);
+ }
+ if (from >= 0)
{
- c += nonascii_insert_offset;
- if (c < 0400 || ! char_valid_p (c, 0))
- c = c_save + DEFAULT_NONASCII_INSERT_OFFSET;
+ if (from < to)
+ {
+ XSETCAR (range, make_number (from));
+ XSETCDR (range, make_number (to));
+ val = range;
+ }
+ else
+ val = make_number (from);
+ if (NILP (function))
+ (*c_function) (arg, val, Qnil);
+ else
+ call2 (function, val, arg);
}
- else if (c >= 0240)
- c = c_save + DEFAULT_NONASCII_INSERT_OFFSET;
}
- return c;
+}
+
+DEFUN ("map-charset-chars", Fmap_charset_chars, Smap_charset_chars, 2, 3, 0,
+ doc: /* Call FUNCTION for each characters in CHARSET.
+FUNCTION is called with three arguments; FROM, TO, and the 3rd optional
+argument ARG.
+FROM and TO indicates a range of character sequence that are contained
+in CHARSET. */)
+ (function, charset, arg)
+ Lisp_Object function, charset, arg;
+{
+ map_charset_chars (NULL, function, charset, arg);
+ return Qnil;
}
-/* 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)) */
+/* 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. */
-int
-multibyte_char_to_unibyte (c, rev_tbl)
- int c;
- Lisp_Object rev_tbl;
+DEFUN ("define-charset-internal", Fdefine_charset_internal,
+ Sdefine_charset_internal, charset_arg_max, MANY, 0,
+ doc: /* For internal use only. */)
+ (nargs, args)
+ int nargs;
+ Lisp_Object *args;
{
- if (!SINGLE_BYTE_CHAR_P (c))
+ /* 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;
+ 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 c_save = c;
+ 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;
+ }
- 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;
- }
+ 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)))));
+
+ 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));
+
+ val = args[charset_arg_invalid_code];
+ if (NILP (val))
+ {
+ if (charset.min_code > 0)
+ charset.invalid_code = 0;
else
{
- if (nonascii_insert_offset > 0)
- c -= nonascii_insert_offset;
- if (c < 128 || c >= 256)
- c = (c_save & 0177) + 0200;
+ 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);
+ }
- 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);
-
- if (NILP (long_name))
- long_name = short_name;
- if (NILP (description))
- description = long_name;
-
- /* Get byte length of multibyte form, base leading-code, and
- extended leading-code of the charset. See the comment under the
- title "GENERAL NOTE on CHARACTER SET (CHARSET)" in charset.h. */
- bytes = XINT (dimension);
- if (charset < MIN_CHARSET_PRIVATE_DIMENSION1)
+ 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
{
- /* Official charset, it doesn't have an extended leading-code. */
- if (charset != CHARSET_ASCII && charset != CHARSET_8_BIT_GRAPHIC)
- bytes += 1; /* For a base leading-code. */
- leading_code_base = charset;
- leading_code_ext = 0;
+ 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
{
- /* 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);
+ 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_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;
+ charset.ascii_compatible_p = ! NILP (args[charset_arg_ascii_compatible_p]);
- {
- /* 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)))
+ 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);
+
+ for (i = charset.min_char; i < 0x10000 && i <= charset.max_char;
+ i += 128)
+ CHARSET_FAST_MAP_SET (i, charset.fast_map);
+ 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))
+ val = load_charset_map (&charset, val);
+ CHECK_VECTOR (val);
+ parse_charset_map (&charset, val, 0);
+ charset.method = CHARSET_METHOD_MAP_DEFERRED;
+ }
+ else if (! NILP (args[charset_arg_parents]))
+ {
+ val = args[charset_arg_parents];
+ CHECK_LIST (val);
+ charset.method = CHARSET_METHOD_INHERIT;
+ val = Fcopy_sequence (val);
+ ASET (attrs, charset_parents, val);
+
+ charset.min_char = MAX_CHAR;
+ charset.max_char = 0;
+ for (; ! NILP (val); val = Fcdr (val))
{
- 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))
+ 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
{
- CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX)
- = make_number (i);
- CHARSET_TABLE_INFO (i, CHARSET_REVERSE_CHARSET_IDX) = charset_id;
- break;
+ 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];
}
- if (i > MAX_CHARSET)
- /* No such a charset. */
- CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX)
- = make_number (-1);
- }
-
- if (charset != CHARSET_ASCII && charset != CHARSET_8_BIT_GRAPHIC
- && charset < MIN_CHARSET_PRIVATE_DIMENSION1)
- {
- bytes_by_char_head[leading_code_base] = bytes;
- 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));
}
+ else
+ error ("None of :code-offset, :map, :parents are specified");
- /* Update table iso_charset_table. */
- if (XINT (iso_final_char) >= 0
- && ISO_CHARSET_TABLE (dimension, chars, iso_final_char) < 0)
- ISO_CHARSET_TABLE (dimension, chars, iso_final_char) = charset;
-}
+ val = args[charset_arg_unify_map];
+ if (! NILP (val) && !STRINGP (val))
+ CHECK_VECTOR (val);
+ ASET (attrs, charset_unify_map, val);
-#ifdef emacs
+ CHECK_LIST (args[charset_arg_plist]);
+ ASET (attrs, charset_plist, args[charset_arg_plist]);
-/* 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;
-
- /* This originally used a ?: operator, but reportedly the HP-UX
- compiler version HP92453-01 A.10.32.22 miscompiles that. */
- if (SYMBOLP (charset_symbol)
- && VECTORP (val = Fget (charset_symbol, Qcharset))
- && CHARSET_VALID_P (charset =
- XINT (XVECTOR (val)->contents[CHARSET_ID_IDX])))
- return charset;
+ charset.hash_index = hash_lookup (hash_table, args[charset_arg_name],
+ &hash_code);
+ if (charset.hash_index >= 0)
+ {
+ new_definition_p = 0;
+ HASH_VALUE (hash_table, charset.hash_index) = attrs;
+ }
else
- return -1;
-}
+ {
+ charset.hash_index = hash_put (hash_table, args[charset_arg_name], attrs,
+ hash_code);
+ if (charset_table_used == charset_table_size)
+ {
+ charset_table_size += 256;
+ charset_table
+ = ((struct charset *)
+ xrealloc (charset_table,
+ sizeof (struct charset) * charset_table_size));
+ }
+ id = charset_table_used++;
+ ASET (attrs, charset_id, make_number (id));
+ new_definition_p = 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)
+ charset.id = id;
+ charset_table[id] = charset;
+
+ if (charset.iso_final >= 0)
{
- from = LEADING_CODE_EXT_11;
- to = LEADING_CODE_EXT_21;
+ 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));
}
- else
+
+ if (charset.emacs_mule_id >= 0)
{
- from = LEADING_CODE_EXT_21;
- to = LEADING_CODE_EXT_MAX + 1;
+ emacs_mule_charset[charset.emacs_mule_id] = CHARSET_FROM_ID (id);
+ if (new_definition_p)
+ Vemacs_mule_charset_list = nconc2 (Vemacs_mule_charset_list,
+ Fcons (make_number (id), Qnil));
}
- for (charset = from; charset < to; charset++)
- if (!CHARSET_DEFINED_P (charset)) break;
+ 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));
+ }
- return make_number (charset < to ? charset : 0);
+ return Qnil;
}
-DEFUN ("define-charset", Fdefine_charset, Sdefine_charset, 3, 3, 0,
- doc: /* Define CHARSET-ID as the identification number of CHARSET with INFO-VECTOR.
-If CHARSET-ID is nil, it is decided automatically, which means CHARSET is
- treated as a private charset.
-INFO-VECTOR is a vector of the format:
- [DIMENSION CHARS WIDTH DIRECTION ISO-FINAL-CHAR ISO-GRAPHIC-PLANE
- SHORT-NAME LONG-NAME DESCRIPTION]
-The meanings of each elements is as follows:
-DIMENSION (integer) is the number of bytes to represent a character: 1 or 2.
-CHARS (integer) is the number of characters in a dimension: 94 or 96.
-WIDTH (integer) is the number of columns a character in the charset
-occupies on the screen: one of 0, 1, and 2.
-
-DIRECTION (integer) is the rendering direction of characters in the
-charset when rendering. If 0, render from left to right, else
-render from right to left.
-
-ISO-FINAL-CHAR (character) is the final character of the
-corresponding ISO 2022 charset.
-It may be -1 if the charset is internal use only.
-
-ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked
-while encoding to variants of ISO 2022 coding system, one of the
-following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).
-It may be -1 if the charset is internal use only.
-
-SHORT-NAME (string) is the short name to refer to the charset.
-
-LONG-NAME (string) is the long name to refer to the charset.
-
-DESCRIPTION (string) is the description string of the charset. */)
- (charset_id, charset_symbol, info_vector)
- Lisp_Object charset_id, charset_symbol, info_vector;
+
+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 *vec;
+ Lisp_Object attr;
+
+ CHECK_CHARSET_GET_ATTR (charset, attr);
+ Fputhash (alias, attr, Vcharset_hash_table);
+ return Qnil;
+}
- if (!NILP (charset_id))
- CHECK_NUMBER (charset_id);
- CHECK_SYMBOL (charset_symbol);
- CHECK_VECTOR (info_vector);
- 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));
- }
+DEFUN ("primary-charset", Fprimary_charset, Sprimary_charset, 0, 0, 0,
+ doc: /* Return the primary charset. */)
+ ()
+{
+ return CHARSET_NAME (CHARSET_FROM_ID (charset_primary));
+}
- 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]) == -1 || (XINT (vec[4]) >= '0' && XINT (vec[4]) <= '~'))
- || !INTEGERP (vec[5])
- || !(XINT (vec[5]) == -1 || 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);
- Fupdate_coding_systems_internal ();
+DEFUN ("set-primary-charset", Fset_primary_charset, Sset_primary_charset,
+ 1, 1, 0,
+ doc: /* Set the primary charset to CHARSET. */)
+ (charset)
+ Lisp_Object charset;
+{
+ int id;
+
+ CHECK_CHARSET_GET_ID (charset, id);
+ charset_primary = id;
return Qnil;
}
-DEFUN ("generic-character-list", Fgeneric_character_list,
- Sgeneric_character_list, 0, 0, 0,
- doc: /* Return a list of all possible generic characters.
-It includes a generic character for a charset not yet defined. */)
- ()
+
+DEFUN ("charset-plist", Fcharset_plist, Scharset_plist, 1, 1, 0,
+ doc: /* Return a 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, 2, 0,
+ doc: /* Unify characters of CHARSET with Unicode. */)
+ (charset, unify_map)
+ Lisp_Object charset, unify_map;
{
- return Vgeneric_character_list;
+ 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 (CHARSET_UNIFIED_P (cs)
+ && CHAR_TABLE_P (CHARSET_DEUNIFIER (cs)))
+ return Qnil;
+ CHARSET_UNIFIED_P (cs) = 0;
+ if (NILP (unify_map))
+ unify_map = CHARSET_UNIFY_MAP (cs);
+ if (STRINGP (unify_map))
+ unify_map = load_charset_map (cs, unify_map);
+ parse_charset_map (cs, unify_map, 2);
+ CHARSET_UNIFIED_P (cs) = 1;
+ 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's final char for a charset of DIMENISION and CHARS.
+ doc: /*
+Return an unsed ISO's 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,
+If there's no unused final char for the attrified kind of charset,
return nil. */)
(dimension, chars)
Lisp_Object dimension, chars;
CHECK_NUMBER (dimension);
CHECK_NUMBER (chars);
- if (XINT (dimension) != 1 && XINT (dimension) != 2)
- error ("Invalid charset dimension %d, it should be 1 or 2",
- XINT (dimension));
+ 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)
- error ("Invalid charset chars %d, it should be 94 or 96",
- XINT (chars));
+ args_out_of_range_3 (chars, make_number (94), make_number (96));
for (final_char = '0'; final_char <= '?'; final_char++)
- {
- if (ISO_CHARSET_TABLE (dimension, chars, make_number (final_char)) < 0)
- break;
- }
+ if (ISO_CHARSET_TABLE (XINT (dimension), XINT (chars), 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,
- doc: /* Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET.
-CHARSET should be defined by `defined-charset' in advance. */)
- (dimension, chars, final_char, charset_symbol)
- Lisp_Object dimension, chars, final_char, charset_symbol;
+static void
+check_iso_charset_parameter (dimension, chars, final_char)
+ Lisp_Object dimension, chars, final_char;
{
- int charset;
+ CHECK_NATNUM (dimension);
+ CHECK_NATNUM (chars);
+ CHECK_NATNUM (final_char);
- CHECK_NUMBER (dimension);
- CHECK_NUMBER (chars);
- CHECK_NUMBER (final_char);
- CHECK_SYMBOL (charset_symbol);
-
- if (XINT (dimension) != 1 && XINT (dimension) != 2)
- error ("Invalid DIMENSION %d, it should be 1 or 2", XINT (dimension));
+ 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' || XFASTINT (final_char) > '~')
+ if (XINT (final_char) < '0' || XINT (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);
+}
+
+
+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 `defined-charset' in advance. */)
+ (dimension, chars, final_char, charset)
+ Lisp_Object dimension, chars, final_char, charset;
+{
+ int id;
- ISO_CHARSET_TABLE (dimension, chars, final_char) = charset;
+ CHECK_CHARSET_GET_ID (charset, id);
+ check_iso_charset_parameter (dimension, chars, final_char);
+
+ ISO_CHARSET_TABLE (dimension, chars, 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 unibyte text.
- 1: No charsets other than ascii eight-bit-control,
- eight-bit-graphic, and latin-1 are found.
- 2: Otherwise.
+ true for a unibyte string. For a multibyte string, true if
+ it contains only ASCII characters.
+
+ 1: No charsets other than ascii, eight-bit-control, and
+ latin-1 are found.
- In addition, if CHARSETS is nonzero, for each found charset N, set
- CHARSETS[N] to 1. For that, callers should allocate CHARSETS
- (MAX_CHARSET + 1 elements) in advance. It may lookup a translation
- table TABLE if supplied. For invalid charsets, set CHARSETS[1] to
- 1 (note that there's no charset whose ID is 1). */
+ 2: Otherwise.
+*/
int
-find_charset_in_text (ptr, nchars, nbytes, charsets, table)
- unsigned char *ptr;
- int nchars, nbytes, *charsets;
- Lisp_Object table;
+string_xstring_p (string)
+ Lisp_Object string;
{
- if (nchars == nbytes)
+ unsigned char *p = XSTRING (string)->data;
+ 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)
{
- if (charsets && nbytes > 0)
- {
- unsigned char *endp = ptr + nbytes;
- int maskbits = 0;
+ int c = STRING_CHAR_ADVANCE (p);
- while (ptr < endp && maskbits != 7)
- {
- maskbits |= (*ptr < 0x80 ? 1 : *ptr < 0xA0 ? 2 : 4);
- ptr++;
- }
-
- if (maskbits & 1)
- charsets[CHARSET_ASCII] = 1;
- if (maskbits & 2)
- charsets[CHARSET_8_BIT_CONTROL] = 1;
- if (maskbits & 4)
- charsets[CHARSET_8_BIT_GRAPHIC] = 1;
- }
- return 0;
+ if (ENCODE_CHAR (charset, c) < 0)
+ return 2;
}
- else
- {
- int return_val = 1;
- int bytes, charset, c1, c2;
+ return 1;
+}
- if (! CHAR_TABLE_P (table))
- table = Qnil;
- while (nchars-- > 0)
- {
- SPLIT_MULTIBYTE_SEQ (ptr, len, bytes, charset, c1, c2);
- ptr += bytes;
+/* Find charsets in the string at PTR of NCHARS and NBYTES.
- if (!CHARSET_DEFINED_P (charset))
- charset = 1;
- else if (! NILP (table))
- {
- int c = translate_char (table, -1, charset, c1, c2);
- if (c >= 0)
- charset = CHAR_CHARSET (c);
- }
+ 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).
- if (return_val == 1
- && charset != CHARSET_ASCII
- && charset != CHARSET_8_BIT_CONTROL
- && charset != CHARSET_8_BIT_GRAPHIC
- && charset != charset_latin_iso8859_1)
- return_val = 2;
+ It may lookup a translation table TABLE if supplied. */
- if (charsets)
- charsets[charset] = 1;
- else if (return_val == 2)
- break;
+static void
+find_charsets_in_text (ptr, nchars, nbytes, charsets, table)
+ unsigned char *ptr;
+ int nchars, nbytes;
+ Lisp_Object charsets, table;
+{
+ 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;
+ }
}
- return return_val;
+ if (all_found)
+ break;
}
}
+
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, end, table)
Lisp_Object beg, end, table;
{
- int charsets[MAX_CHARSET + 1];
+ Lisp_Object charsets;
int from, from_byte, to, stop, stop_byte, i;
Lisp_Object val;
from_byte = CHAR_TO_BYTE (from);
- bzero (charsets, (MAX_CHARSET + 1) * sizeof (int));
+ 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_charset_in_text (BYTE_POS_ADDR (from_byte), stop - from,
- stop_byte - from_byte, charsets, table);
+ 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;
}
val = Qnil;
- if (charsets[1])
- val = Fcons (Qunknown, val);
- for (i = MAX_CHARSET; i >= MIN_CHARSET_OFFICIAL_DIMENSION1; i--)
- if (charsets[i])
- val = Fcons (CHARSET_SYMBOL (i), val);
- if (charsets[0])
- val = Fcons (Qascii, val);
+ 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;
}
`unknown' is included in the returned list.
If STR is unibyte, the returned list may contain
-only `ascii', `eight-bit-control', and `eight-bit-graphic'. */)
+only `ascii', `eight-bit-control', and `eight-bit-graphic'. */)
(str, table)
Lisp_Object str, table;
{
- int charsets[MAX_CHARSET + 1];
+ Lisp_Object charsets;
int i;
Lisp_Object val;
CHECK_STRING (str);
- bzero (charsets, (MAX_CHARSET + 1) * sizeof (int));
- find_charset_in_text (XSTRING (str)->data, XSTRING (str)->size,
- STRING_BYTES (XSTRING (str)), charsets, table);
+ charsets = Fmake_vector (make_number (charset_table_used), Qnil);
+ find_charsets_in_text (XSTRING (str)->data, XSTRING (str)->size,
+ STRING_BYTES (XSTRING (str)), charsets, table);
val = Qnil;
- if (charsets[1])
- val = Fcons (Qunknown, val);
- for (i = MAX_CHARSET; i >= MIN_CHARSET_OFFICIAL_DIMENSION1; i--)
- if (charsets[i])
- val = Fcons (CHARSET_SYMBOL (i), val);
- if (charsets[0])
- val = Fcons (Qascii, val);
+ 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
-DEFUN ("make-char-internal", Fmake_char_internal, Smake_char_internal, 1, 3, 0,
- doc: /* Return a character made from arguments.
-Internal use only. */)
- (charset, code1, code2)
- Lisp_Object charset, code1, code2;
+
+/* Return a character correponding to the code-point CODE of
+ CHARSET. */
+
+int
+decode_char (charset, code)
+ struct charset *charset;
+ unsigned code;
{
- int charset_id, c1, c2;
+ int c, char_index;
+ enum charset_method method = CHARSET_METHOD (charset);
- CHECK_NUMBER (charset);
- charset_id = XINT (charset);
- if (!CHARSET_DEFINED_P (charset_id))
- error ("Invalid charset ID: %d", XINT (charset));
+ if (code < CHARSET_MIN_CODE (charset) || code > CHARSET_MAX_CODE (charset))
+ return -1;
- if (NILP (code1))
- c1 = 0;
- else
+ if (method == CHARSET_METHOD_MAP_DEFERRED)
{
- CHECK_NUMBER (code1);
- c1 = XINT (code1);
- }
- if (NILP (code2))
- c2 = 0;
- else
- {
- CHECK_NUMBER (code2);
- c2 = XINT (code2);
+ load_charset (charset);
+ method = CHARSET_METHOD (charset);
}
- if (charset_id == CHARSET_ASCII)
- {
- if (c1 < 0 || c1 > 0x7F)
- goto invalid_code_posints;
- return make_number (c1);
- }
- else if (charset_id == CHARSET_8_BIT_CONTROL)
- {
- if (NILP (code1))
- c1 = 0x80;
- else if (c1 < 0x80 || c1 > 0x9F)
- goto invalid_code_posints;
- return make_number (c1);
- }
- else if (charset_id == CHARSET_8_BIT_GRAPHIC)
+ if (method == CHARSET_METHOD_INHERIT)
{
- if (NILP (code1))
- c1 = 0xA0;
- else if (c1 < 0xA0 || c1 > 0xFF)
- goto invalid_code_posints;
- return make_number (c1);
- }
- else if (c1 < 0 || c1 > 0xFF || c2 < 0 || c2 > 0xFF)
- goto invalid_code_posints;
- c1 &= 0x7F;
- c2 &= 0x7F;
- if (c1 == 0
- ? c2 != 0
- : (c2 == 0
- ? !CHAR_COMPONENTS_VALID_P (charset_id, c1, 0x20)
- : !CHAR_COMPONENTS_VALID_P (charset_id, c1, c2)))
- goto invalid_code_posints;
- return make_number (MAKE_CHAR (charset_id, c1, c2));
-
- invalid_code_posints:
- error ("Invalid code points for charset ID %d: %d %d", charset_id, c1, c2);
-}
-
-DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0,
- doc: /* Return list of charset and one or two position-codes of CHAR.
-If CHAR is invalid as a character code,
-return a list of symbol `unknown' and CHAR. */)
- (ch)
- Lisp_Object ch;
-{
- int c, charset, c1, c2;
-
- CHECK_NUMBER (ch);
- 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,
- doc: /* Return charset of CHAR. */)
- (ch)
- Lisp_Object ch;
-{
- CHECK_NUMBER (ch);
-
- return CHARSET_SYMBOL (CHAR_CHARSET (XINT (ch)));
-}
-
-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;
- int charset;
-
- ch = Fchar_after (pos);
- if (! INTEGERP (ch))
- return ch;
- charset = CHAR_CHARSET (XINT (ch));
- return CHARSET_SYMBOL (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 charset;
-
- CHECK_NUMBER (dimension);
- CHECK_NUMBER (chars);
- CHECK_NUMBER (final_char);
-
- if ((charset = ISO_CHARSET_TABLE (dimension, chars, final_char)) < 0)
- return Qnil;
- return CHARSET_SYMBOL (charset);
-}
+ Lisp_Object parents;
-/* 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;
+ parents = CHARSET_PARENTS (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;
- if (c < 0 || c >= MAX_CHAR)
- return 0;
- if (SINGLE_BYTE_CHAR_P (c))
- return 1;
- SPLIT_CHAR (c, charset, c1, c2);
- if (genericp)
+ charset = CHARSET_FROM_ID (id);
+ if ((c = DECODE_CHAR (charset, this_code)) >= 0)
+ break;
+ }
+ }
+ else
{
- if (c1)
+ char_index = CODE_POINT_TO_INDEX (charset, code);
+
+ if (method == CHARSET_METHOD_MAP)
{
- if (c2 <= 0) c2 = 0x20;
+ Lisp_Object decoder;
+
+ decoder = CHARSET_DECODER (charset);
+ if (! VECTORP (decoder))
+ return -1;
+ c = XINT (AREF (decoder, char_index));
}
else
{
- if (c2 <= 0) c1 = c2 = 0x20;
+ c = char_index + CHARSET_CODE_OFFSET (charset);
}
}
- 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,
- doc: /* Return t if OBJECT is a valid normal character.
-If optional arg GENERICP is non-nil, also return t if OBJECT is
-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,
- doc: /* Convert the unibyte character CH to multibyte character.
-The conversion is done based on `nonascii-translation-table' (which see)
- or `nonascii-insert-offset' (which see). */)
- (ch)
- Lisp_Object ch;
-{
- int c;
+ if (CHARSET_UNIFIED_P (charset)
+ && c >= 0)
+ MAYBE_UNIFY_CHAR (c);
- CHECK_NUMBER (ch);
- 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);
+ return c;
}
-DEFUN ("multibyte-char-to-unibyte", Fmultibyte_char_to_unibyte,
- Smultibyte_char_to_unibyte, 1, 1, 0,
- doc: /* Convert the multibyte character CH to unibyte character.
-The conversion is done based on `nonascii-translation-table' (which see)
- or `nonascii-insert-offset' (which see). */)
- (ch)
- Lisp_Object ch;
-{
- int c;
- CHECK_NUMBER (ch);
- 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);
-}
+/* Return a code-point of CHAR in CHARSET. If CHAR doesn't belong to
+ CHARSET, return CHARSET_INVALID_CODE (CHARSET). */
-DEFUN ("char-bytes", Fchar_bytes, Schar_bytes, 1, 1, 0,
- doc: /* Return 1 regardless of the argument CHAR.
-This is now an obsolete function. We keep it just for backward compatibility. */)
- (ch)
- Lisp_Object ch;
-{
- CHECK_NUMBER (ch);
- 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)
+unsigned
+encode_char (charset, c)
+ struct charset *charset;
int c;
{
- int charset;
-
- if (ASCII_BYTE_P (c) || (c & ~((1 << CHARACTERBITS) -1)))
- return 1;
- if (SINGLE_BYTE_CHAR_P (c) && c >= 0xA0)
- return 1;
+ unsigned code;
+ enum charset_method method = CHARSET_METHOD (charset);
- charset = CHAR_CHARSET (c);
- return (CHARSET_DEFINED_P (charset) ? CHARSET_BYTES (charset) : 1);
-}
-
-/* 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,
- doc: /* Return width of CHAR when displayed in the current buffer.
-The width is measured by how many columns it occupies on the screen.
-Tab is taken to occupy `tab-width' columns. */)
- (ch)
- Lisp_Object ch;
-{
- Lisp_Object val, disp;
- int c;
- struct Lisp_Char_Table *dp = buffer_display_table ();
-
- CHECK_NUMBER (ch);
-
- 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 (CHARSET_UNIFIED_P (charset))
{
- int charset = CHAR_CHARSET (c);
+ Lisp_Object deunifier;
+ int deunified;
- XSETFASTINT (val, CHARSET_WIDTH (charset));
+ deunifier = CHARSET_DEUNIFIER (charset);
+ if (! CHAR_TABLE_P (deunifier))
+ {
+ Funify_charset (CHARSET_NAME (charset), Qnil);
+ deunifier = CHARSET_DEUNIFIER (charset);
+ }
+ deunified = XINT (CHAR_TABLE_REF (deunifier, c));
+ if (deunified > 0)
+ c = deunified;
}
- 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;
-{
- return c_string_width (str, len, -1, NULL, NULL);
-}
-/* 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. If PRECISION > 0, return the width of
- longest substring that doesn't exceed PRECISION, and set number of
- characters and bytes of the substring in *NCHARS and *NBYTES
- respectively. */
+ if (! CHARSET_FAST_MAP_REF ((c), charset->fast_map)
+ || c < CHARSET_MIN_CHAR (charset) || c > CHARSET_MAX_CHAR (charset))
+ return CHARSET_INVALID_CODE (charset);
-int
-c_string_width (str, len, precision, nchars, nbytes)
- unsigned char *str;
- int precision, *nchars, *nbytes;
-{
- int i = 0, i_byte = 0;
- int width = 0;
- int chars;
- struct Lisp_Char_Table *dp = buffer_display_table ();
-
- while (i_byte < len)
+ if (method == CHARSET_METHOD_INHERIT)
{
- int bytes, thiswidth;
- Lisp_Object val;
+ Lisp_Object parents;
- if (dp)
+ parents = CHARSET_PARENTS (charset);
+ for (; CONSP (parents); parents = XCDR (parents))
{
- int c = STRING_CHAR_AND_LENGTH (str + i_byte, len - i_byte, bytes);
+ int id = XINT (XCAR (XCAR (parents)));
+ int code_offset = XINT (XCDR (XCAR (parents)));
+ struct charset *this_charset = CHARSET_FROM_ID (id);
- chars = 1;
- val = DISP_CHAR_VECTOR (dp, c);
- if (VECTORP (val))
- thiswidth = XVECTOR (val)->size;
- else
- thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]);
- }
- else
- {
- chars = 1;
- PARSE_MULTIBYTE_SEQ (str + i_byte, len - i_byte, bytes);
- thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]);
- }
-
- if (precision > 0
- && (width + thiswidth > precision))
- {
- *nchars = i;
- *nbytes = i_byte;
- return width;
+ code = ENCODE_CHAR (this_charset, c);
+ if (code != CHARSET_INVALID_CODE (this_charset)
+ && (code_offset < 0 || code >= code_offset))
+ {
+ code -= code_offset;
+ if (CODE_POINT_TO_INDEX (charset, code) >= 0)
+ return code;
+ }
}
- i++;
- i_byte += bytes;
- width += thiswidth;
- }
+ return CHARSET_INVALID_CODE (charset);
+ }
- if (precision > 0)
+ if (method == CHARSET_METHOD_MAP_DEFERRED)
{
- *nchars = i;
- *nbytes = i_byte;
+ load_charset (charset);
+ method = CHARSET_METHOD (charset);
}
- return width;
-}
-
-/* Return width of Lisp string STRING when displayed in the current
- buffer. The width is measured by how many columns it occupies on
- the screen while paying attention to compositions. If PRECISION >
- 0, return the width of longest substring that doesn't exceed
- PRECISION, and set number of characters and bytes of the substring
- in *NCHARS and *NBYTES respectively. */
-
-int
-lisp_string_width (string, precision, nchars, nbytes)
- Lisp_Object string;
- int precision, *nchars, *nbytes;
-{
- int len = XSTRING (string)->size;
- int len_byte = STRING_BYTES (XSTRING (string));
- unsigned char *str = XSTRING (string)->data;
- int i = 0, i_byte = 0;
- int width = 0;
- struct Lisp_Char_Table *dp = buffer_display_table ();
-
- while (i < len)
+ if (method == CHARSET_METHOD_MAP)
{
- int chars, bytes, thiswidth;
+ Lisp_Object encoder;
Lisp_Object val;
- int cmp_id;
- int ignore, end;
-
- if (find_composition (i, -1, &ignore, &end, &val, string)
- && ((cmp_id = get_composition_id (i, i_byte, end - i, val, string))
- >= 0))
- {
- thiswidth = composition_table[cmp_id]->width;
- chars = end - i;
- bytes = string_char_to_byte (string, end) - i_byte;
- }
- else if (dp)
- {
- int c = STRING_CHAR_AND_LENGTH (str + i_byte, len - i_byte, bytes);
- chars = 1;
- val = DISP_CHAR_VECTOR (dp, c);
- if (VECTORP (val))
- thiswidth = XVECTOR (val)->size;
- else
- thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]);
- }
+ encoder = CHARSET_ENCODER (charset);
+ if (! CHAR_TABLE_P (CHARSET_ENCODER (charset)))
+ return CHARSET_INVALID_CODE (charset);
+ val = CHAR_TABLE_REF (encoder, c);
+ if (CONSP (val))
+ code = (XINT (XCAR (val)) << 16) | XINT (XCDR (val));
else
- {
- chars = 1;
- PARSE_MULTIBYTE_SEQ (str + i_byte, len_byte - i_byte, bytes);
- thiswidth = ONE_BYTE_CHAR_WIDTH (str[i_byte]);
- }
-
- if (precision > 0
- && (width + thiswidth > precision))
- {
- *nchars = i;
- *nbytes = i_byte;
- return width;
- }
- i += chars;
- i_byte += bytes;
- width += thiswidth;
- }
-
- if (precision > 0)
+ code = XINT (val);
+ }
+ else
{
- *nchars = i;
- *nbytes = i_byte;
+ code = c - CHARSET_CODE_OFFSET (charset);
+ code = INDEX_TO_CODE_POINT (charset, code);
}
- return width;
+ return code;
}
-DEFUN ("string-width", Fstring_width, Sstring_width, 1, 1, 0,
- doc: /* Return width of STRING when displayed in the current buffer.
-Width is measured by how many columns it occupies on the screen.
-When calculating width of a multibyte character in STRING,
-only the base leading-code is considered; the validity of
-the following bytes is not checked. Tabs in STRING are always
-taken to occupy `tab-width' columns. */)
- (str)
- Lisp_Object str;
-{
- Lisp_Object val;
- CHECK_STRING (str);
- XSETFASTINT (val, lisp_string_width (str, -1, NULL, NULL));
- return val;
-}
+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.
-DEFUN ("char-direction", Fchar_direction, Schar_direction, 1, 1, 0,
- doc: /* Return the direction of CHAR.
-The returned value is 0 for left-to-right and 1 for right-to-left. */)
- (ch)
- Lisp_Object ch;
+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 charset;
+ int c, id;
+ unsigned code;
+ struct charset *charsetp;
- CHECK_NUMBER (ch);
- charset = CHAR_CHARSET (XFASTINT (ch));
- if (!CHARSET_DEFINED_P (charset))
- invalid_character (XINT (ch));
- return CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX);
+ 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 (XCAR (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 ("chars-in-region", Fchars_in_region, Schars_in_region, 2, 2, 0,
- doc: /* Return number of characters between BEG and END. */)
- (beg, end)
- Lisp_Object beg, end;
-{
- int from, to;
- CHECK_NUMBER_COERCE_MARKER (beg);
- CHECK_NUMBER_COERCE_MARKER (end);
+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.
- from = min (XFASTINT (beg), XFASTINT (end));
- to = max (XFASTINT (beg), XFASTINT (end));
+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 c, id;
+ unsigned code;
+ struct charset *charsetp;
- return make_number (to - from);
+ CHECK_CHARSET_GET_ID (charset, id);
+ CHECK_NATNUM (ch);
+ c = XINT (ch);
+ charsetp = CHARSET_FROM_ID (id);
+ code = ENCODE_CHAR (charsetp, 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);
}
-/* 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;
+DEFUN ("make-char", Fmake_char, Smake_char, 1, 4, 0,
+ doc: /* Return a character of CHARSET whose position code is CODE.
- return multibyte_chars_in_text (ptr, nbytes);
-}
+If dimension of CHARSET is two, and the third optional arg CODE2 is
+non-nil, CODE actually specifies the first byte of the position code,
+and CODE2 specifies the second byte.
-/* 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;
+If dimension of CHARSET is three, and the third optional arg CODE2 and
+the fourth optional arg CODE3 are both non-nil, CODE actually
+specifies the first byte of the position code, CODE2 the second byte,
+and CODE3 the third byte. */)
+ (charset, code, code2, code3)
+ Lisp_Object charset, code, code2, code3;
{
- unsigned char *endp;
- int chars, bytes;
+ int id, dimension;
+ struct charset *charsetp;
+ unsigned c;
- endp = ptr + nbytes;
- chars = 0;
+ CHECK_CHARSET_GET_ID (charset, id);
+ charsetp = CHARSET_FROM_ID (id);
- while (ptr < endp)
+ if (NILP (code))
+ code = make_number (CHARSET_MIN_CODE (charsetp));
+ else
{
- PARSE_MULTIBYTE_SEQ (ptr, endp - ptr, bytes);
- ptr += bytes;
- chars++;
+ CHECK_NATNUM (code);
+ dimension = CHARSET_DIMENSION (charsetp);
+
+ if (!NILP (code2))
+ {
+ CHECK_NATNUM (code2);
+ if (dimension == 3)
+ CHECK_NATNUM (code3);
+ }
}
- return chars;
+ if (dimension == 1 || NILP (code2))
+ c = XFASTINT (code);
+ else if (dimension == 2)
+ c = (XFASTINT (code) << 8) | XFASTINT (code2);
+ else if (dimension == 3)
+ c = (XFASTINT (code) << 16) | (XFASTINT (code2) << 8) | XFASTINT (code3);
+
+ c = DECODE_CHAR (charsetp, c);
+ return make_number (c);
}
-/* Parse unibyte text at STR of LEN bytes as multibyte text, and
- count the numbers of characters and bytes in it. On counting
- bytes, pay attention to the fact that 8-bit characters in the range
- 0x80..0x9F are represented by 2 bytes in multibyte text. */
-void
-parse_str_as_multibyte (str, len, nchars, nbytes)
- unsigned char *str;
- int len, *nchars, *nbytes;
+
+/* 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;
{
- unsigned char *endp = str + len;
- int n, chars = 0, bytes = 0;
+ if (NILP (charset_list))
+ charset_list = Vcharset_ordered_list;
- while (str < endp)
+ while (CONSP (charset_list))
{
- if (UNIBYTE_STR_AS_MULTIBYTE_P (str, endp - str, n))
- str += n, bytes += n;
- else
- str++, bytes += 2;
- chars++;
+ 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);
}
- *nchars = chars;
- *nbytes = bytes;
- return;
+ return NULL;
}
-/* Arrange unibyte text at STR of NBYTES bytes as multibyte text.
- It actually converts only 8-bit characters in the range 0x80..0x9F
- that don't contruct multibyte characters to multibyte forms. If
- NCHARS is nonzero, set *NCHARS to the number of characters in the
- text. It is assured that we can use LEN bytes at STR as a work
- area and that is enough. Return the number of bytes of the
- resulting text. */
-int
-str_as_multibyte (str, len, nbytes, nchars)
- unsigned char *str;
- int len, nbytes, *nchars;
+DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0,
+ doc: /*Return list of charset and one or two position-codes of CHAR.
+If CHAR is invalid as a character code,
+return a list of symbol `unknown' and CHAR. */)
+ (ch)
+ Lisp_Object ch;
{
- unsigned char *p = str, *endp = str + nbytes;
- unsigned char *to;
- int chars = 0;
- int n;
-
- while (p < endp && UNIBYTE_STR_AS_MULTIBYTE_P (p, endp - p, n))
- p += n, chars++;
- if (nchars)
- *nchars = chars;
- if (p == endp)
- return nbytes;
-
- to = p;
- nbytes = endp - p;
- endp = str + len;
- safe_bcopy (p, endp - nbytes, nbytes);
- p = endp - nbytes;
- while (p < endp)
- {
- if (UNIBYTE_STR_AS_MULTIBYTE_P (p, endp - p, n))
- {
- while (n--)
- *to++ = *p++;
- }
- else
- {
- *to++ = LEADING_CODE_8_BIT_CONTROL;
- *to++ = *p++ + 0x20;
- }
- chars++;
- }
- if (nchars)
- *nchars = chars;
- return (to - str);
+ 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);
}
-/* Parse unibyte string at STR of LEN bytes, and return the number of
- bytes it may ocupy when converted to multibyte string by
- `str_to_multibyte'. */
-int
-parse_str_to_multibyte (str, len)
- unsigned char *str;
- int len;
+DEFUN ("char-charset", Fchar_charset, Schar_charset, 1, 1, 0,
+ doc: /* Return the charset of highest priority that contains CHAR. */)
+ (ch)
+ Lisp_Object ch;
{
- unsigned char *endp = str + len;
- int bytes;
+ struct charset *charset;
- for (bytes = 0; str < endp; str++)
- bytes += (*str < 0x80 || *str >= 0xA0) ? 1 : 2;
- return bytes;
+ CHECK_CHARACTER (ch);
+ charset = CHAR_CHARSET (XINT (ch));
+ return (CHARSET_NAME (charset));
}
-/* Convert unibyte text at STR of NBYTES bytes to multibyte text
- that contains the same single-byte characters. It actually
- converts all 8-bit characters to multibyte forms. It is assured
- that we can use LEN bytes at STR as a work area and that is
- enough. */
-int
-str_to_multibyte (str, len, bytes)
- unsigned char *str;
- int len, bytes;
+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;
{
- unsigned char *p = str, *endp = str + bytes;
- unsigned char *to;
-
- while (p < endp && (*p < 0x80 || *p >= 0xA0)) p++;
- if (p == endp)
- return bytes;
- to = p;
- bytes = endp - p;
- endp = str + len;
- safe_bcopy (p, endp - bytes, bytes);
- p = endp - bytes;
- while (p < endp)
- {
- if (*p < 0x80 || *p >= 0xA0)
- *to++ = *p++;
- else
- *to++ = LEADING_CODE_8_BIT_CONTROL, *to++ = *p++ + 0x20;
- }
- return (to - str);
+ Lisp_Object ch;
+ struct charset *charset;
+
+ ch = Fchar_after (pos);
+ if (! INTEGERP (ch))
+ return ch;
+ charset = CHAR_CHARSET (XINT (ch));
+ return (CHARSET_NAME (charset));
}
-/* Arrange multibyte text at STR of LEN bytes as a unibyte text. It
- actually converts only 8-bit characters in the range 0x80..0x9F to
- unibyte forms. */
-int
-str_as_unibyte (str, bytes)
- unsigned char *str;
- int bytes;
+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;
{
- unsigned char *p = str, *endp = str + bytes;
- unsigned char *to = str;
+ int id;
- while (p < endp && *p != LEADING_CODE_8_BIT_CONTROL) p++;
- to = p;
- while (p < endp)
- {
- if (*p == LEADING_CODE_8_BIT_CONTROL)
- *to++ = *(p + 1) - 0x20, p += 2;
- else
- *to++ = *p++;
- }
- return (to - str);
+ 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);
}
-\f
-DEFUN ("string", Fstring, Sstring, 1, MANY, 0,
- doc: /* Concatenate all the argument characters and make the result a string.
-usage: (string &rest CHARACTERS) */)
- (n, args)
- int n;
- Lisp_Object *args;
+
+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;
- unsigned char *buf = (unsigned char *) alloca (MAX_MULTIBYTE_LENGTH * n);
- unsigned char *p = buf;
- int c;
- int multibyte = 0;
+ struct charset *charset;
+ Lisp_Object attrs;
- for (i = 0; i < n; i++)
+ for (i = 0; i < charset_table_used; i++)
{
- CHECK_NUMBER (args[i]);
- if (!multibyte && !SINGLE_BYTE_CHAR_P (XFASTINT (args[i])))
- multibyte = 1;
+ 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;
}
- for (i = 0; i < n; i++)
+ if (CHAR_TABLE_P (Vchar_unified_charset_table))
{
- c = XINT (args[i]);
- if (multibyte)
- p += CHAR_STRING (c, p);
- else
- *p++ = c;
+ Foptimize_char_table (Vchar_unified_charset_table);
+ Vchar_unify_table = Vchar_unified_charset_table;
+ Vchar_unified_charset_table = Qnil;
}
- return make_string_from_bytes (buf, n, p - buf);
+ return Qnil;
}
-#endif /* emacs */
\f
-int
-charset_id_internal (charset_name)
- char *charset_name;
+void
+init_charset ()
{
- 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, doc: /* 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);
+ 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 < 255; i++)
+ emacs_mule_charset[i] = NULL;
- /* This has to be done here, before we call Fmake_char_table. */
- Qcharset_table = intern ("charset-table");
- staticpro (&Qcharset_table);
+#if 0
+ Vchar_charset_set = Fmake_char_table (Qnil, Qnil);
+ CHAR_TABLE_SET (Vchar_charset_set, make_number (97), Qnil);
+
+ DEFSYM (Qcharset_encode_table, "charset-encode-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;
-
- for (i = 0; i < 256; i++)
- bytes_by_char_head[i] = 1;
- 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;
-
- 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);
- Vgeneric_character_list = Fnreverse (val);
- }
-
- nonascii_insert_offset = 0;
- Vnonascii_translation_table = Qnil;
+ /* Now we are ready to set up this property, so we can create syntax
+ tables. */
+ Fput (Qcharset_encode_table, Qchar_table_extra_slots, make_number (0));
+#endif
}
#ifdef emacs
void
syms_of_charset ()
{
- Qcharset = intern ("charset");
- staticpro (&Qcharset);
-
- Qascii = intern ("ascii");
- staticpro (&Qascii);
-
- Qeight_bit_control = intern ("eight-bit-control");
- staticpro (&Qeight_bit_control);
-
- Qeight_bit_graphic = intern ("eight-bit-graphic");
- staticpro (&Qeight_bit_graphic);
-
- /* Define special charsets ascii, eight-bit-control, and
- eight-bit-graphic. */
- 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"),
- Qnil, /* same as above */
- build_string ("ASCII (ISO646 IRV)"));
- CHARSET_SYMBOL (CHARSET_ASCII) = Qascii;
- Fput (Qascii, Qcharset, CHARSET_TABLE_ENTRY (CHARSET_ASCII));
-
- update_charset_table (make_number (CHARSET_8_BIT_CONTROL),
- make_number (1), make_number (96),
- make_number (4),
- make_number (0),
- make_number (-1),
- make_number (-1),
- build_string ("8-bit control code (0x80..0x9F)"),
- Qnil, /* same as above */
- Qnil); /* same as above */
- CHARSET_SYMBOL (CHARSET_8_BIT_CONTROL) = Qeight_bit_control;
- Fput (Qeight_bit_control, Qcharset,
- CHARSET_TABLE_ENTRY (CHARSET_8_BIT_CONTROL));
-
- update_charset_table (make_number (CHARSET_8_BIT_GRAPHIC),
- make_number (1), make_number (96),
- make_number (4),
- make_number (0),
- make_number (-1),
- make_number (-1),
- build_string ("8-bit graphic char (0xA0..0xFF)"),
- Qnil, /* same as above */
- Qnil); /* same as above */
- CHARSET_SYMBOL (CHARSET_8_BIT_GRAPHIC) = Qeight_bit_graphic;
- Fput (Qeight_bit_graphic, Qcharset,
- CHARSET_TABLE_ENTRY (CHARSET_8_BIT_GRAPHIC));
-
- 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);
+ char *p;
+
+ DEFSYM (Qcharsetp, "charsetp");
+
+ DEFSYM (Qascii, "ascii");
+ DEFSYM (Qunicode, "unicode");
+ DEFSYM (Qeight_bit_control, "eight-bit-control");
+ DEFSYM (Qeight_bit_graphic, "eight-bit-graphic");
+ 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 (&Sprimary_charset);
+ defsubr (&Sset_primary_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 (&Smake_char_internal);
+ defsubr (&Sdecode_char);
+ defsubr (&Sencode_char);
defsubr (&Ssplit_char);
+ defsubr (&Smake_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 (&Ssetup_special_charsets);
+ defsubr (&Sclear_charset_maps);
+
+ DEFVAR_LISP ("charset-map-directory", &Vcharset_map_directory,
+ doc: /* Directory of charset map files that come with GNU Emacs.
+The default value is \"\\[data-directory]/charsets\". */);
+ Vcharset_map_directory = Fexpand_file_name (build_string ("charsets"),
+ Vdata_directory);
DEFVAR_LISP ("charset-list", &Vcharset_list,
doc: /* List of charsets ever defined. */);
- Vcharset_list = Fcons (Qascii, Fcons (Qeight_bit_control,
- Fcons (Qeight_bit_graphic, Qnil)));
-
- DEFVAR_LISP ("translation-table-vector", &Vtranslation_table_vector,
- doc: /* Vector of cons cell of a symbol and translation table ever defined.
-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-private-11", &leading_code_private_11,
- doc: /* 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,
- doc: /* 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,
- doc: /* 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,
- doc: /* 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,
- doc: /* Offset for converting non-ASCII unibyte codes 0240...0377 to multibyte.
-This is used for converting unibyte text to multibyte,
-and for inserting character codes specified by number.
-
-This serves to convert a Latin-1 or similar 8-bit character code
-to the corresponding Emacs multibyte character code.
-Typically the value should be (- (make-char CHARSET 0) 128),
-for your choice of character set.
-If `nonascii-translation-table' is non-nil, it overrides this variable. */);
- nonascii_insert_offset = 0;
-
- DEFVAR_LISP ("nonascii-translation-table", &Vnonascii_translation_table,
- doc: /* Translation table to convert non-ASCII unibyte codes to multibyte.
-This is used for converting unibyte text to multibyte,
-and for inserting character codes specified by number.
-
-Conversion is performed only when multibyte characters are enabled,
-and it serves to convert a Latin-1 or similar 8-bit character code
-to the corresponding Emacs character code.
-
-If this is nil, `nonascii-insert-offset' is used instead.
-See also the docstring of `make-translation-table'. */);
- Vnonascii_translation_table = Qnil;
-
- DEFVAR_LISP ("auto-fill-chars", &Vauto_fill_chars,
- doc: /* A char-table for characters which invoke auto-filling.
-Such characters have 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);
+ Vcharset_list = Qnil;
+
+ /* Make the prerequisite charset `ascii' and `unicode'. */
+ {
+ Lisp_Object args[charset_arg_max];
+ Lisp_Object plist[14];
+ Lisp_Object val;
+
+ plist[0] = intern (":name");
+ plist[1] = args[charset_arg_name] = Qascii;
+ plist[2] = intern (":dimension");
+ plist[3] = args[charset_arg_dimension] = make_number (1);
+ val = Fmake_vector (make_number (8), make_number (0));
+ ASET (val, 1, make_number (127));
+ plist[4] = intern (":code-space");
+ plist[5] = args[charset_arg_code_space] = val;
+ plist[6] = intern (":iso-final-char");
+ plist[7] = args[charset_arg_iso_final] = make_number ('B');
+ args[charset_arg_iso_revision] = Qnil;
+ plist[8] = intern (":emacs-mule-id");
+ plist[9] = args[charset_arg_emacs_mule_id] = make_number (0);
+ plist[10] = intern (":ascii-compatible-p");
+ plist[11] = args[charset_arg_ascii_compatible_p] = Qt;
+ args[charset_arg_supplementary_p] = Qnil;
+ args[charset_arg_invalid_code] = Qnil;
+ plist[12] = intern (":code-offset");
+ plist[13] = args[charset_arg_code_offset] = make_number (0);
+ args[charset_arg_map] = Qnil;
+ args[charset_arg_parents] = Qnil;
+ args[charset_arg_unify_map] = Qnil;
+ /* The actual plist is set by mule-conf.el. */
+ args[charset_arg_plist] = Flist (14, plist);
+ Fdefine_charset_internal (charset_arg_max, args);
+ charset_ascii = CHARSET_SYMBOL_ID (Qascii);
+
+ plist[1] = args[charset_arg_name] = Qunicode;
+ plist[3] = args[charset_arg_dimension] = make_number (3);
+ val = Fmake_vector (make_number (8), make_number (0));
+ ASET (val, 1, make_number (255));
+ ASET (val, 3, make_number (255));
+ ASET (val, 5, make_number (16));
+ plist[5] = args[charset_arg_code_space] = val;
+ plist[7] = args[charset_arg_iso_final] = Qnil;
+ args[charset_arg_iso_revision] = Qnil;
+ plist[9] = args[charset_arg_emacs_mule_id] = Qnil;
+ plist[11] = args[charset_arg_ascii_compatible_p] = Qt;
+ args[charset_arg_supplementary_p] = Qnil;
+ args[charset_arg_invalid_code] = Qnil;
+ plist[13] = args[charset_arg_code_offset] = make_number (0);
+ args[charset_arg_map] = Qnil;
+ args[charset_arg_parents] = Qnil;
+ args[charset_arg_unify_map] = Qnil;
+ /* The actual plist is set by mule-conf.el. */
+ args[charset_arg_plist] = Flist (14, plist);
+ Fdefine_charset_internal (charset_arg_max, args);
+ charset_unicode = CHARSET_SYMBOL_ID (Qunicode);
+ }
}
#endif /* emacs */
-/* Header for multibyte character handler.
+/* Header for charset handler.
Copyright (C) 1995, 1997, 1998 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.
#ifndef EMACS_CHARSET_H
#define EMACS_CHARSET_H
-/* #define BYTE_COMBINING_DEBUG */
-
-/*** GENERAL NOTE on CHARACTER SET (CHARSET) ***
-
- A character set ("charset" hereafter) is a meaningful collection
- (i.e. language, culture, functionality, etc) of characters. Emacs
- handles multiple charsets at once. Each charset corresponds to one
- of the ISO charsets. Emacs identifies a charset by a unique
- identification number, whereas ISO identifies a charset by a triplet
- of DIMENSION, CHARS and FINAL-CHAR. So, hereafter, just saying
- "charset" means an identification number (integer value).
-
- The value range of charsets is 0x00, 0x81..0xFE. There are four
- kinds of charset depending on DIMENSION (1 or 2) and CHARS (94 or
- 96). For instance, a charset of DIMENSION2_CHARS94 contains 94x94
- characters.
-
- Within Emacs Lisp, a charset is treated as a symbol which has a
- property `charset'. The property value is a vector containing
- various information about the charset. For readability of C code,
- we use the following convention for C variable names:
- charset_symbol: Emacs Lisp symbol of a charset
- charset_id: Emacs Lisp integer of an identification number of a charset
- charset: C integer of an identification number of a charset
-
- Each charset (except for ascii) is assigned a base leading-code
- (range 0x80..0x9E). In addition, a charset of greater than 0xA0
- (whose base leading-code is 0x9A..0x9D) is assigned an extended
- leading-code (range 0xA0..0xFE). In this case, each base
- leading-code specifies the allowable range of extended leading-code
- as shown in the table below. A leading-code is used to represent a
- character in Emacs' buffer and string.
-
- We call a charset which has extended leading-code a "private
- charset" because those are mainly for a charset which is not yet
- registered by ISO. On the contrary, we call a charset which does
- not have extended leading-code an "official charset".
-
- ---------------------------------------------------------------------------
- charset dimension base leading-code extended leading-code
- ---------------------------------------------------------------------------
- 0x00 official dim1 -- none -- -- none --
- (ASCII)
- 0x01..0x7F --never used--
- 0x80 official dim1 -- none -- -- none --
- (eight-bit-graphic)
- 0x81..0x8F official dim1 same as charset -- none --
- 0x90..0x99 official dim2 same as charset -- none --
- 0x9A..0x9D --never used--
- 0x9E official dim1 same as charset -- none --
- (eight-bit-control)
- 0x9F --never used--
- 0xA0..0xDF private dim1 0x9A same as charset
- of 1-column width
- 0xE0..0xEF private dim1 0x9B same as charset
- of 2-column width
- 0xF0..0xF4 private dim2 0x9C same as charset
- of 1-column width
- 0xF5..0xFE private dim2 0x9D same as charset
- of 2-column width
- 0xFF --never used--
- ---------------------------------------------------------------------------
-
-*/
-
-/* Definition of special leading-codes. */
-/* Leading-code followed by extended leading-code. */
-#define LEADING_CODE_PRIVATE_11 0x9A /* for private DIMENSION1 of 1-column */
-#define LEADING_CODE_PRIVATE_12 0x9B /* for private DIMENSION1 of 2-column */
-#define LEADING_CODE_PRIVATE_21 0x9C /* for private DIMENSION2 of 1-column */
-#define LEADING_CODE_PRIVATE_22 0x9D /* for private DIMENSION2 of 2-column */
-
-#define LEADING_CODE_8_BIT_CONTROL 0x9E /* for `eight-bit-control' */
-
-/* Extended leading-code. */
-/* Start of each extended leading-codes. */
-#define LEADING_CODE_EXT_11 0xA0 /* follows LEADING_CODE_PRIVATE_11 */
-#define LEADING_CODE_EXT_12 0xE0 /* follows LEADING_CODE_PRIVATE_12 */
-#define LEADING_CODE_EXT_21 0xF0 /* follows LEADING_CODE_PRIVATE_21 */
-#define LEADING_CODE_EXT_22 0xF5 /* follows LEADING_CODE_PRIVATE_22 */
-/* Maximum value of extended leading-codes. */
-#define LEADING_CODE_EXT_MAX 0xFE
-
-/* Definition of minimum/maximum charset of each DIMENSION. */
-#define MIN_CHARSET_OFFICIAL_DIMENSION1 0x80
-#define MAX_CHARSET_OFFICIAL_DIMENSION1 0x8F
-#define MIN_CHARSET_OFFICIAL_DIMENSION2 0x90
-#define MAX_CHARSET_OFFICIAL_DIMENSION2 0x99
-#define MIN_CHARSET_PRIVATE_DIMENSION1 LEADING_CODE_EXT_11
-#define MIN_CHARSET_PRIVATE_DIMENSION2 LEADING_CODE_EXT_21
-
-/* Maximum value of overall charset identification number. */
-#define MAX_CHARSET 0xFE
-
-/* Definition of special charsets. */
-#define CHARSET_ASCII 0 /* 0x00..0x7F */
-#define CHARSET_8_BIT_CONTROL 0x9E /* 0x80..0x9F */
-#define CHARSET_8_BIT_GRAPHIC 0x80 /* 0xA0..0xFF */
-
-extern int charset_latin_iso8859_1; /* ISO8859-1 (Latin-1) */
-extern int charset_jisx0208_1978; /* JISX0208.1978 (Japanese Kanji old set) */
-extern int charset_jisx0208; /* JISX0208.1983 (Japanese Kanji) */
-extern int charset_katakana_jisx0201; /* JISX0201.Kana (Japanese Katakana) */
-extern int charset_latin_jisx0201; /* JISX0201.Roman (Japanese Roman) */
-extern int charset_big5_1; /* Big5 Level 1 (Chinese Traditional) */
-extern int charset_big5_2; /* Big5 Level 2 (Chinese Traditional) */
-
-/* Check if CH is an ASCII character or a base leading-code.
- Nowadays, any byte can be the first byte of a character in a
- multibyte buffer/string. So this macro name is not appropriate. */
-#define CHAR_HEAD_P(ch) ((unsigned char) (ch) < 0xA0)
-
-/*** GENERAL NOTE on CHARACTER REPRESENTATION ***
-
- Firstly, the term "character" or "char" is used for a multilingual
- character (of course, including ASCII characters), not for a byte in
- computer memory. We use the term "code" or "byte" for the latter
- case.
-
- A character is identified by charset and one or two POSITION-CODEs.
- POSITION-CODE is the position of the character in the charset. A
- character of DIMENSION1 charset has one POSITION-CODE: POSITION-CODE-1.
- A character of DIMENSION2 charset has two POSITION-CODE:
- POSITION-CODE-1 and POSITION-CODE-2. The code range of
- POSITION-CODE is 0x20..0x7F.
-
- Emacs has two kinds of representation of a character: multi-byte
- form (for buffers and strings) and single-word form (for character
- objects in Emacs Lisp). The latter is called "character code"
- hereafter. Both representations encode the information of charset
- and POSITION-CODE but in a different way (for instance, the MSB of
- POSITION-CODE is set in multi-byte form).
-
- For details of the multi-byte form, see the section "2. Emacs
- internal format handlers" of `coding.c'.
-
- Emacs uses 19 bits for a character code. The bits are divided into
- 3 fields: FIELD1(5bits):FIELD2(7bits):FIELD3(7bits).
-
- A character code of DIMENSION1 character uses FIELD2 to hold charset
- and FIELD3 to hold POSITION-CODE-1. A character code of DIMENSION2
- character uses FIELD1 to hold charset, FIELD2 and FIELD3 to hold
- POSITION-CODE-1 and POSITION-CODE-2 respectively.
-
- More precisely...
-
- FIELD2 of DIMENSION1 character (except for ascii, eight-bit-control,
- and eight-bit-graphic) is "charset - 0x70". This is to make all
- character codes except for ASCII and 8-bit codes greater than 256.
- So, the range of FIELD2 of DIMENSION1 character is 0, 1, or
- 0x11..0x7F.
-
- FIELD1 of DIMENSION2 character is "charset - 0x8F" for official
- charset and "charset - 0xE0" for private charset. So, the range of
- FIELD1 of DIMENSION2 character is 0x01..0x1E.
-
- -----------------------------------------------------------------------------
- charset FIELD1 (5-bit) FIELD2 (7-bit) FIELD3 (7-bit)
- -----------------------------------------------------------------------------
- ascii 0 0 0x00..0x7F
- eight-bit-control 0 1 0x00..0x1F
- eight-bit-graphic 0 1 0x20..0x7F
- DIMENSION1 0 charset - 0x70 POSITION-CODE-1
- DIMENSION2(o) charset - 0x8F POSITION-CODE-1 POSITION-CODE-2
- DIMENSION2(p) charset - 0xE0 POSITION-CODE-1 POSITION-CODE-2
- -----------------------------------------------------------------------------
- "(o)": official, "(p)": private
- -----------------------------------------------------------------------------
-*/
-
-/* Masks of each field of character code. */
-#define CHAR_FIELD1_MASK (0x1F << 14)
-#define CHAR_FIELD2_MASK (0x7F << 7)
-#define CHAR_FIELD3_MASK 0x7F
-
-/* Macros to access each field of character C. */
-#define CHAR_FIELD1(c) (((c) & CHAR_FIELD1_MASK) >> 14)
-#define CHAR_FIELD2(c) (((c) & CHAR_FIELD2_MASK) >> 7)
-#define CHAR_FIELD3(c) ((c) & CHAR_FIELD3_MASK)
-
-/* Minimum character code of character of each DIMENSION. */
-#define MIN_CHAR_OFFICIAL_DIMENSION1 \
- ((0x81 - 0x70) << 7)
-#define MIN_CHAR_PRIVATE_DIMENSION1 \
- ((MIN_CHARSET_PRIVATE_DIMENSION1 - 0x70) << 7)
-#define MIN_CHAR_OFFICIAL_DIMENSION2 \
- ((MIN_CHARSET_OFFICIAL_DIMENSION2 - 0x8F) << 14)
-#define MIN_CHAR_PRIVATE_DIMENSION2 \
- ((MIN_CHARSET_PRIVATE_DIMENSION2 - 0xE0) << 14)
-/* Maximum character code currently used plus 1. */
-#define MAX_CHAR (0x1F << 14)
-
-/* 1 if C is a single byte character, else 0. */
-#define SINGLE_BYTE_CHAR_P(c) ((unsigned) (c) < 0x100)
-
-/* 1 if BYTE is an ASCII character in itself, in multibyte mode. */
-#define ASCII_BYTE_P(byte) ((byte) < 0x80)
-
-/* A char-table containing information on each character set.
-
- Unlike ordinary char-tables, this doesn't contain any nested tables.
- Only the top level elements are used. Each element is a vector of
- the following information:
- CHARSET-ID, BYTES, DIMENSION, CHARS, WIDTH, DIRECTION,
- LEADING-CODE-BASE, LEADING-CODE-EXT,
- ISO-FINAL-CHAR, ISO-GRAPHIC-PLANE,
- REVERSE-CHARSET, SHORT-NAME, LONG-NAME, DESCRIPTION,
- PLIST.
-
- CHARSET-ID (integer) is the identification number of the charset.
-
- BYTES (integer) is the length of the multi-byte form of a character
- in the charset: one of 1, 2, 3, and 4.
-
- DIMENSION (integer) is the number of bytes to represent a character: 1 or 2.
-
- CHARS (integer) is the number of characters in a dimension: 94 or 96.
-
- WIDTH (integer) is the number of columns a character in the charset
- occupies on the screen: one of 0, 1, and 2..
-
- DIRECTION (integer) is the rendering direction of characters in the
- charset when rendering. If 0, render from left to right, else
- render from right to left.
-
- LEADING-CODE-BASE (integer) is the base leading-code for the
- charset.
-
- LEADING-CODE-EXT (integer) is the extended leading-code for the
- charset. All charsets of less than 0xA0 have the value 0.
-
- ISO-FINAL-CHAR (character) is the final character of the
- corresponding ISO 2022 charset. It is -1 for such a character
- that is used only internally (e.g. `eight-bit-control').
-
- ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked
- while encoding to variants of ISO 2022 coding system, one of the
- following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR). It
- is -1 for such a character that is used only internally
- (e.g. `eight-bit-control').
-
- REVERSE-CHARSET (integer) is the charset which differs only in
- LEFT-TO-RIGHT value from the charset. If there's no such a
- charset, the value is -1.
-
- SHORT-NAME (string) is the short name to refer to the charset.
-
- LONG-NAME (string) is the long name to refer to the charset.
-
- DESCRIPTION (string) is the description string of the charset.
-
- PLIST (property list) may contain any type of information a user
- wants to put and get by functions `put-charset-property' and
- `get-charset-property' respectively. */
-extern Lisp_Object Vcharset_table;
-
-/* Macros to access various information of CHARSET in Vcharset_table.
- We provide these macros for efficiency. No range check of CHARSET. */
-
-/* Return entry of CHARSET (C integer) in Vcharset_table. */
-#define CHARSET_TABLE_ENTRY(charset) \
- XCHAR_TABLE (Vcharset_table)->contents[((charset) == CHARSET_ASCII \
- ? 0 : (charset) + 128)]
-
-/* Return information INFO-IDX of CHARSET. */
-#define CHARSET_TABLE_INFO(charset, info_idx) \
- XVECTOR (CHARSET_TABLE_ENTRY (charset))->contents[info_idx]
-
-#define CHARSET_ID_IDX (0)
-#define CHARSET_BYTES_IDX (1)
-#define CHARSET_DIMENSION_IDX (2)
-#define CHARSET_CHARS_IDX (3)
-#define CHARSET_WIDTH_IDX (4)
-#define CHARSET_DIRECTION_IDX (5)
-#define CHARSET_LEADING_CODE_BASE_IDX (6)
-#define CHARSET_LEADING_CODE_EXT_IDX (7)
-#define CHARSET_ISO_FINAL_CHAR_IDX (8)
-#define CHARSET_ISO_GRAPHIC_PLANE_IDX (9)
-#define CHARSET_REVERSE_CHARSET_IDX (10)
-#define CHARSET_SHORT_NAME_IDX (11)
-#define CHARSET_LONG_NAME_IDX (12)
-#define CHARSET_DESCRIPTION_IDX (13)
-#define CHARSET_PLIST_IDX (14)
-/* Size of a vector of each entry of Vcharset_table. */
-#define CHARSET_MAX_IDX (15)
-
-/* And several more macros to be used frequently. */
-#define CHARSET_BYTES(charset) \
- XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_BYTES_IDX))
-#define CHARSET_DIMENSION(charset) \
- XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_DIMENSION_IDX))
-#define CHARSET_CHARS(charset) \
- XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_CHARS_IDX))
-#define CHARSET_WIDTH(charset) \
- XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_WIDTH_IDX))
-#define CHARSET_DIRECTION(charset) \
- XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX))
-#define CHARSET_LEADING_CODE_BASE(charset) \
- XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_BASE_IDX))
-#define CHARSET_LEADING_CODE_EXT(charset) \
- XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_EXT_IDX))
-#define CHARSET_ISO_FINAL_CHAR(charset) \
- XINT (CHARSET_TABLE_INFO (charset, CHARSET_ISO_FINAL_CHAR_IDX))
-#define CHARSET_ISO_GRAPHIC_PLANE(charset) \
- XINT (CHARSET_TABLE_INFO (charset, CHARSET_ISO_GRAPHIC_PLANE_IDX))
-#define CHARSET_REVERSE_CHARSET(charset) \
- XINT (CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX))
-
-/* Macros to specify direction of a charset. */
-#define CHARSET_DIRECTION_LEFT_TO_RIGHT 0
-#define CHARSET_DIRECTION_RIGHT_TO_LEFT 1
-
-/* A vector of charset symbol indexed by charset-id. This is used
- only for returning charset symbol from C functions. */
-extern Lisp_Object Vcharset_symbol_table;
-
-/* Return symbol of CHARSET. */
-#define CHARSET_SYMBOL(charset) \
- XVECTOR (Vcharset_symbol_table)->contents[charset]
-
-/* 1 if CHARSET is in valid value range, else 0. */
-#define CHARSET_VALID_P(charset) \
- ((charset) == 0 \
- || ((charset) > 0x80 && (charset) <= MAX_CHARSET_OFFICIAL_DIMENSION2) \
- || ((charset) >= MIN_CHARSET_PRIVATE_DIMENSION1 \
- && (charset) <= MAX_CHARSET) \
- || ((charset) == CHARSET_8_BIT_CONTROL) \
- || ((charset) == CHARSET_8_BIT_GRAPHIC))
-
-/* 1 if CHARSET is already defined, else 0. */
-#define CHARSET_DEFINED_P(charset) \
- (((charset) >= 0) && ((charset) <= MAX_CHARSET) \
- && !NILP (CHARSET_TABLE_ENTRY (charset)))
-
-/* Since the information CHARSET-BYTES and CHARSET-WIDTH of
- Vcharset_table can be retrieved only by the first byte of
- multi-byte form (an ASCII code or a base leading-code), we provide
- here tables to be used by macros BYTES_BY_CHAR_HEAD and
- WIDTH_BY_CHAR_HEAD for faster information retrieval. */
-extern int bytes_by_char_head[256];
-extern int width_by_char_head[256];
-
-#define BYTES_BY_CHAR_HEAD(char_head) \
- (ASCII_BYTE_P (char_head) ? 1 : bytes_by_char_head[char_head])
-#define WIDTH_BY_CHAR_HEAD(char_head) \
- (ASCII_BYTE_P (char_head) ? 1 : width_by_char_head[char_head])
-
-/* Charset of the character C. */
-#define CHAR_CHARSET(c) \
- (SINGLE_BYTE_CHAR_P (c) \
- ? (ASCII_BYTE_P (c) \
- ? CHARSET_ASCII \
- : (c) < 0xA0 ? CHARSET_8_BIT_CONTROL : CHARSET_8_BIT_GRAPHIC) \
- : ((c) < MIN_CHAR_OFFICIAL_DIMENSION2 \
- ? CHAR_FIELD2 (c) + 0x70 \
- : ((c) < MIN_CHAR_PRIVATE_DIMENSION2 \
- ? CHAR_FIELD1 (c) + 0x8F \
- : CHAR_FIELD1 (c) + 0xE0)))
+/* Index to arguments of Fdefine_charset_internal. */
+
+enum define_charset_arg_index
+ {
+ charset_arg_name,
+ charset_arg_dimension,
+ charset_arg_code_space,
+ charset_arg_iso_final,
+ charset_arg_iso_revision,
+ charset_arg_emacs_mule_id,
+ charset_arg_ascii_compatible_p,
+ charset_arg_supplementary_p,
+ charset_arg_invalid_code,
+ charset_arg_code_offset,
+ charset_arg_map,
+ charset_arg_parents,
+ charset_arg_unify_map,
+ charset_arg_plist,
+ charset_arg_max
+ };
+
+
+/* Indices to charset attributes vector. */
+
+enum charset_attr_index
+ {
+ /* ID number of the charset. */
+ charset_id,
-/* Check if two characters C1 and C2 belong to the same charset. */
-#define SAME_CHARSET_P(c1, c2) \
- (c1 < MIN_CHAR_OFFICIAL_DIMENSION2 \
- ? (c1 & CHAR_FIELD2_MASK) == (c2 & CHAR_FIELD2_MASK) \
- : (c1 & CHAR_FIELD1_MASK) == (c2 & CHAR_FIELD1_MASK))
-
-/* Return a character of which charset is CHARSET and position-codes
- are C1 and C2. DIMENSION1 character ignores C2. */
-#define MAKE_CHAR(charset, c1, c2) \
- ((charset) == CHARSET_ASCII \
- ? (c1) & 0x7F \
- : (((charset) == CHARSET_8_BIT_CONTROL \
- || (charset) == CHARSET_8_BIT_GRAPHIC) \
- ? ((c1) & 0x7F) | 0x80 \
- : ((CHARSET_DEFINED_P (charset) \
- ? CHARSET_DIMENSION (charset) == 1 \
- : (charset) < MIN_CHARSET_PRIVATE_DIMENSION2) \
- ? (((charset) - 0x70) << 7) | ((c1) <= 0 ? 0 : ((c1) & 0x7F)) \
- : ((((charset) \
- - ((charset) < MIN_CHARSET_PRIVATE_DIMENSION2 ? 0x8F : 0xE0)) \
- << 14) \
- | ((c2) <= 0 ? 0 : ((c2) & 0x7F)) \
- | ((c1) <= 0 ? 0 : (((c1) & 0x7F) << 7))))))
-
-
-/* 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. */
-#define CHAR_VALID_P(c, genericp) \
- ((c) >= 0 \
- && (SINGLE_BYTE_CHAR_P (c) || char_valid_p (c, genericp)))
-
-/* This default value is used when nonascii-translation-table or
- nonascii-insert-offset fail to convert unibyte character to a valid
- multibyte character. This makes a Latin-1 character. */
-
-#define DEFAULT_NONASCII_INSERT_OFFSET 0x800
-
-/* Parse multibyte string STR of length LENGTH and set BYTES to the
- byte length of a character at STR. */
-
-#ifdef BYTE_COMBINING_DEBUG
-
-#define PARSE_MULTIBYTE_SEQ(str, length, bytes) \
+ /* Name of the charset (symbol). */
+ charset_name,
+
+ /* Property list of the charset. */
+ charset_plist,
+
+ /* If the method of the charset is `MAP_DEFERRED', the value is a
+ mappint vector or a file name that contains mapping vector.
+ Otherwise, nil. */
+ charset_map,
+
+ /* If the method of the charset is `MAP', the value is a vector
+ that maps code points of the charset to characters. The vector
+ is indexed by a character index. A character index is
+ calculated from a code point and the code-space table of the
+ charset. */
+ charset_decoder,
+
+ /* If the method of the charset is `MAP', the value is a
+ char-table that maps characters of the charset to code
+ points. */
+ charset_encoder,
+
+ /* If the method of the charset is `INHERIT', the value is a list
+ of the form (PARENT-CHARSET-ID . CODE-OFFSET). */
+ charset_parents,
+
+ /* */
+ charset_unify_map,
+
+ /* */
+ charset_deunifier,
+
+ /* The length of charset attribute vector. */
+ charset_attr_max
+ };
+
+/* Methods for converting code points and characters of charsets. */
+
+enum charset_method
+ {
+ /* For a charset of this method, a character code is calculated
+ from a character index (which is calculated from a code point)
+ simply by adding an offset value. */
+ CHARSET_METHOD_OFFSET,
+
+ /* For a charset of this method, a decoder vector and an encoder
+ char-table is used for code point <-> character code
+ conversion. */
+ CHARSET_METHOD_MAP,
+
+ /* Same as above but decoder and encoder are loaded from a file on
+ demand. Once loaded, the method is changed to
+ CHARSET_METHOD_MAP. */
+ CHARSET_METHOD_MAP_DEFERRED,
+
+ /* A charset of this method inherits characters from the other
+ charsets. */
+ CHARSET_METHOD_INHERIT,
+ };
+
+struct charset
+{
+ int id;
+
+ int hash_index;
+
+ /* Dimension of the charset: 1, 2, 3, or 4. */
+ int dimension;
+
+ /* Minimum byte code in each dimension. */
+ int code_space[16];
+
+ /* 1 if there's no gap in code-points. */
+ int code_linear_p;
+
+ /* If the charset is treated as 94-chars in ISO-2022, the value is 0.
+ If the charset is treated as 96-chars in ISO-2022, the value is 1. */
+ int iso_chars_96;
+
+ /* ISO final character code for the charset: 48..127.
+ It may be 0 if the charset doesn't conform to ISO-2022. */
+ int iso_final;
+
+ int iso_revision;
+
+ /* If the charset is identical to what supported by Emacs 21 and the
+ priors, the identification number of the charset used in those
+ version. Otherwise, -1. */
+ int emacs_mule_id;
+
+ /* Nonzero iff the charset is compatible with ASCII. */
+ int ascii_compatible_p;
+
+ /* Nonzero iff the charset is supplementary. */
+ int supplementary_p;
+
+ /* Nonzero iff all the code points are representable by Lisp_Int. */
+ int compact_codes_p;
+
+ /* The method for encoding/decoding characters of the charset. */
+ enum charset_method method;
+
+ /* Mininum and Maximum code points of the charset. */
+ unsigned min_code, max_code;
+
+ /* Mininum and Maximum character codes of the charset. If the
+ charset is compatible with ASCII, min_char is a minimum non-ASCII
+ character of the charset. */
+ int min_char, max_char;
+
+ /* The code returned by ENCODE_CHAR if a character is not encodable
+ by the charset. */
+ unsigned invalid_code;
+
+ /* If the method of the charset is CHARSET_METHOD_MAP, this is a
+ table of bits used to quickly and roughly guess if a character
+ belongs to the charset.
+
+ The first 64 elements are 512 bits for characters less than
+ 0x10000. Each bit corresponds to 128-character block. The last
+ 126 elements are 1008 bits for the greater characters
+ (0x10000..0x3FFFFF). Each bit corresponds to 4096-character
+ block.
+
+ If a bit is 1, at least one character in the corresponds block is
+ in this charset. */
+ unsigned char fast_map[190];
+
+ /* Offset value to calculate a character code from code-point, and
+ visa versa. */
+ int code_offset;
+
+ int unified_p;
+};
+
+/* Hash table of charset symbols vs. the correponding attribute
+ vectors. */
+extern Lisp_Object Vcharset_hash_table;
+
+/* Table of struct charset. */
+extern struct charset *charset_table;
+extern int charset_table_used;
+
+#define CHARSET_FROM_ID(id) (charset_table + (id))
+
+extern Lisp_Object Vcharset_list;
+extern Lisp_Object Viso_2022_charset_list;
+extern Lisp_Object Vemacs_mule_charset_list;
+
+extern struct charset *emacs_mule_charset[256];
+
+
+/* Macros to access information about charset. */
+
+/* Return the attribute vector of charset whose symbol is SYMBOL. */
+#define CHARSET_SYMBOL_ATTRIBUTES(symbol) \
+ Fgethash ((symbol), Vcharset_hash_table, Qnil)
+
+#define CHARSET_ATTR_ID(attrs) AREF ((attrs), charset_id)
+#define CHARSET_ATTR_NAME(attrs) AREF ((attrs), charset_name)
+#define CHARSET_ATTR_PLIST(attrs) AREF ((attrs), charset_plist)
+#define CHARSET_ATTR_MAP(attrs) AREF ((attrs), charset_map)
+#define CHARSET_ATTR_DECODER(attrs) AREF ((attrs), charset_decoder)
+#define CHARSET_ATTR_ENCODER(attrs) AREF ((attrs), charset_encoder)
+#define CHARSET_ATTR_PARENTS(attrs) AREF ((attrs), charset_parents)
+#define CHARSET_ATTR_UNIFY_MAP(attrs) AREF ((attrs), charset_unify_map)
+#define CHARSET_ATTR_DEUNIFIER(attrs) AREF ((attrs), charset_deunifier)
+
+#define CHARSET_SYMBOL_ID(symbol) \
+ CHARSET_ATTR_ID (CHARSET_SYMBOL_ATTRIBUTES (symbol))
+
+/* Return an index to Vcharset_hash_table of the charset whose symbol
+ is SYMBOL. */
+#define CHARSET_SYMBOL_HASH_INDEX(symbol) \
+ hash_lookup (XHASH_TABLE (Vcharset_hash_table), symbol, NULL)
+
+/* Return the attribute vector of CHARSET. */
+#define CHARSET_ATTRIBUTES(charset) \
+ (HASH_VALUE (XHASH_TABLE (Vcharset_hash_table), (charset)->hash_index))
+
+#define CHARSET_ID(charset) ((charset)->id)
+#define CHARSET_HASH_INDEX(charset) ((charset)->hash_index)
+#define CHARSET_DIMENSION(charset) ((charset)->dimension)
+#define CHARSET_CODE_SPACE(charset) ((charset)->code_space)
+#define CHARSET_CODE_LINEAR_P(charset) ((charset)->code_linear_p)
+#define CHARSET_ISO_CHARS_96(charset) ((charset)->iso_chars_96)
+#define CHARSET_ISO_FINAL(charset) ((charset)->iso_final)
+#define CHARSET_ISO_PLANE(charset) ((charset)->iso_plane)
+#define CHARSET_ISO_REVISION(charset) ((charset)->iso_revision)
+#define CHARSET_EMACS_MULE_ID(charset) ((charset)->emacs_mule_id)
+#define CHARSET_ASCII_COMPATIBLE_P(charset) ((charset)->ascii_compatible_p)
+#define CHARSET_COMPACT_CODES_P(charset) ((charset)->compact_codes_p)
+#define CHARSET_METHOD(charset) ((charset)->method)
+#define CHARSET_MIN_CODE(charset) ((charset)->min_code)
+#define CHARSET_MAX_CODE(charset) ((charset)->max_code)
+#define CHARSET_INVALID_CODE(charset) ((charset)->invalid_code)
+#define CHARSET_MIN_CHAR(charset) ((charset)->min_char)
+#define CHARSET_MAX_CHAR(charset) ((charset)->max_char)
+#define CHARSET_CODE_OFFSET(charset) ((charset)->code_offset)
+#define CHARSET_UNIFIED_P(charset) ((charset)->unified_p)
+
+#define CHARSET_NAME(charset) \
+ (CHARSET_ATTR_NAME (CHARSET_ATTRIBUTES (charset)))
+#define CHARSET_MAP(charset) \
+ (CHARSET_ATTR_MAP (CHARSET_ATTRIBUTES (charset)))
+#define CHARSET_DECODER(charset) \
+ (CHARSET_ATTR_DECODER (CHARSET_ATTRIBUTES (charset)))
+#define CHARSET_ENCODER(charset) \
+ (CHARSET_ATTR_ENCODER (CHARSET_ATTRIBUTES (charset)))
+#define CHARSET_PARENTS(charset) \
+ (CHARSET_ATTR_PARENTS (CHARSET_ATTRIBUTES (charset)))
+#define CHARSET_UNIFY_MAP(charset) \
+ (CHARSET_ATTR_UNIFY_MAP (CHARSET_ATTRIBUTES (charset)))
+#define CHARSET_DEUNIFIER(charset) \
+ (CHARSET_ATTR_DEUNIFIER (CHARSET_ATTRIBUTES (charset)))
+
+
+/* Nonzero iff OBJ is a valid charset symbol. */
+#define CHARSETP(obj) (CHARSET_SYMBOL_HASH_INDEX (obj) >= 0)
+
+/* Check if X is a valid charset symbol. If not, signal an error. */
+#define CHECK_CHARSET(x) \
do { \
- int i = 1; \
- while (i < (length) && ! CHAR_HEAD_P ((str)[i])) i++; \
- (bytes) = BYTES_BY_CHAR_HEAD ((str)[0]); \
- if ((bytes) > i) \
- abort (); \
+ if (! SYMBOLP (x) || CHARSET_SYMBOL_HASH_INDEX (x) < 0) \
+ x = wrong_type_argument (Qcharsetp, (x)); \
} while (0)
-#else /* not BYTE_COMBINING_DEBUG */
-
-#define PARSE_MULTIBYTE_SEQ(str, length, bytes) \
- (bytes) = BYTES_BY_CHAR_HEAD ((str)[0])
-
-#endif /* not BYTE_COMBINING_DEBUG */
-
-/* Return 1 iff the byte sequence at unibyte string STR (LENGTH bytes)
- is valid as a multibyte form. If valid, by a side effect, BYTES is
- set to the byte length of the multibyte form. */
-
-#define UNIBYTE_STR_AS_MULTIBYTE_P(str, length, bytes) \
- (((str)[0] < 0x80 || (str)[0] >= 0xA0) \
- ? ((bytes) = 1) \
- : (((bytes) = BYTES_BY_CHAR_HEAD ((str)[0])), \
- ((bytes) > 1 && (bytes) <= (length) \
- && (str)[0] != LEADING_CODE_8_BIT_CONTROL \
- && !CHAR_HEAD_P ((str)[1]) \
- && ((bytes) == 2 \
- || (!CHAR_HEAD_P ((str)[2]) \
- && ((bytes) == 3 \
- || !CHAR_HEAD_P ((str)[3])))))))
-
-/* Return 1 iff the byte sequence at multibyte string STR is valid as
- a unibyte form. By a side effect, BYTES is set to the byte length
- of one character at STR. */
-
-#define MULTIBYTE_STR_AS_UNIBYTE_P(str, bytes) \
- ((bytes) = BYTES_BY_CHAR_HEAD ((str)[0]), \
- (str)[0] != LEADING_CODE_8_BIT_CONTROL)
-
-/* The charset of character C is stored in CHARSET, and the
- position-codes of C are stored in C1 and C2.
- We store -1 in C2 if the dimension of the charset is 1. */
-
-#define SPLIT_CHAR(c, charset, c1, c2) \
- (SINGLE_BYTE_CHAR_P (c) \
- ? ((charset \
- = (ASCII_BYTE_P (c) \
- ? CHARSET_ASCII \
- : ((c) < 0xA0 ? CHARSET_8_BIT_CONTROL : CHARSET_8_BIT_GRAPHIC))), \
- c1 = (c), c2 = -1) \
- : ((c) & CHAR_FIELD1_MASK \
- ? (charset = (CHAR_FIELD1 (c) \
- + ((c) < MIN_CHAR_PRIVATE_DIMENSION2 ? 0x8F : 0xE0)), \
- c1 = CHAR_FIELD2 (c), \
- c2 = CHAR_FIELD3 (c)) \
- : (charset = CHAR_FIELD2 (c) + 0x70, \
- c1 = CHAR_FIELD3 (c), \
- c2 = -1)))
-
-/* Return 1 iff character C has valid printable glyph. */
-#define CHAR_PRINTABLE_P(c) (ASCII_BYTE_P (c) || char_printable_p (c))
-
-/* The charset of the character at STR is stored in CHARSET, and the
- position-codes are stored in C1 and C2.
- We store -1 in C2 if the character is just 2 bytes. */
-
-#define SPLIT_STRING(str, len, charset, c1, c2) \
- ((BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) < 2 \
- || BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) > len \
- || split_string (str, len, &charset, &c1, &c2) < 0) \
- ? c1 = *(str), charset = CHARSET_ASCII \
- : charset)
-/* Mapping table from ISO2022's charset (specified by DIMENSION,
- CHARS, and FINAL_CHAR) to Emacs' charset. Should be accessed by
- macro ISO_CHARSET_TABLE (DIMENSION, CHARS, FINAL_CHAR). */
-extern int iso_charset_table[2][2][128];
-
-#define ISO_CHARSET_TABLE(dimension, chars, final_char) \
- iso_charset_table[XINT (dimension) - 1][XINT (chars) > 94][XINT (final_char)]
-
-#define BASE_LEADING_CODE_P(c) (BYTES_BY_CHAR_HEAD ((unsigned char) (c)) > 1)
-
-/* Return how many bytes C will occupy in a multibyte buffer. */
-#define CHAR_BYTES(c) \
- (SINGLE_BYTE_CHAR_P (c) \
- ? ((ASCII_BYTE_P (c) || (c) >= 0xA0) ? 1 : 2) \
- : char_bytes (c))
-
-/* The following two macros CHAR_STRING and STRING_CHAR are the main
- entry points to convert between Emacs's two types of character
- representations: multi-byte form and single-word form (character
- code). */
-
-/* Store multi-byte form of the character C in STR. The caller should
- allocate at least MAX_MULTIBYTE_LENGTH bytes area at STR in
- advance. Returns the length of the multi-byte form. If C is an
- invalid character code, signal an error. */
-
-#define CHAR_STRING(c, str) \
- (SINGLE_BYTE_CHAR_P (c) \
- ? ((ASCII_BYTE_P (c) || c >= 0xA0) \
- ? (*(str) = (unsigned char)(c), 1) \
- : (*(str) = LEADING_CODE_8_BIT_CONTROL, *((str)+ 1) = c + 0x20, 2)) \
- : char_to_string (c, (unsigned char *) str))
-
-/* Like CHAR_STRING but don't signal an error if C is invalid.
- Value is -1 in this case. */
-
-#define CHAR_STRING_NO_SIGNAL(c, str) \
- (SINGLE_BYTE_CHAR_P (c) \
- ? ((ASCII_BYTE_P (c) || c >= 0xA0) \
- ? (*(str) = (unsigned char)(c), 1) \
- : (*(str) = LEADING_CODE_8_BIT_CONTROL, *((str)+ 1) = c + 0x20, 2)) \
- : char_to_string_1 (c, (unsigned char *) str))
-
-/* Return a character code of the character of which multi-byte form
- is at STR and the length is LEN. If STR doesn't contain valid
- multi-byte form, only the first byte in STR is returned. */
-
-#define STRING_CHAR(str, len) \
- (BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) == 1 \
- ? (unsigned char) *(str) \
- : string_to_char (str, len, 0))
-
-/* This is like STRING_CHAR but the third arg ACTUAL_LEN is set to the
- length of the multi-byte form. Just to know the length, use
- MULTIBYTE_FORM_LENGTH. */
-
-#define STRING_CHAR_AND_LENGTH(str, len, actual_len) \
- (BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) == 1 \
- ? ((actual_len) = 1), (unsigned char) *(str) \
- : string_to_char (str, len, &(actual_len)))
-
-/* Fetch the "next" character from Lisp string STRING at byte position
- BYTEIDX, character position CHARIDX. Store it into OUTPUT.
-
- All the args must be side-effect-free.
- BYTEIDX and CHARIDX must be lvalues;
- we increment them past the character fetched. */
-
-#define FETCH_STRING_CHAR_ADVANCE(OUTPUT, STRING, CHARIDX, BYTEIDX) \
-if (1) \
- { \
- CHARIDX++; \
- if (STRING_MULTIBYTE (STRING)) \
- { \
- unsigned char *ptr = &XSTRING (STRING)->data[BYTEIDX]; \
- int space_left = XSTRING (STRING)->size_byte - BYTEIDX; \
- int actual_len; \
- \
- OUTPUT = STRING_CHAR_AND_LENGTH (ptr, space_left, actual_len); \
- BYTEIDX += actual_len; \
- } \
- else \
- OUTPUT = XSTRING (STRING)->data[BYTEIDX++]; \
- } \
-else
-
-/* Like FETCH_STRING_CHAR_ADVANCE but assume STRING is multibyte. */
-
-#define FETCH_STRING_CHAR_ADVANCE_NO_CHECK(OUTPUT, STRING, CHARIDX, BYTEIDX) \
-if (1) \
- { \
- unsigned char *fetch_string_char_ptr = &XSTRING (STRING)->data[BYTEIDX]; \
- int fetch_string_char_space_left = XSTRING (STRING)->size_byte - BYTEIDX; \
- int actual_len; \
- \
- OUTPUT \
- = STRING_CHAR_AND_LENGTH (fetch_string_char_ptr, \
- fetch_string_char_space_left, actual_len); \
- \
- BYTEIDX += actual_len; \
- CHARIDX++; \
- } \
-else
-
-/* Like FETCH_STRING_CHAR_ADVANCE but fetch character from the current
- buffer. */
-
-#define FETCH_CHAR_ADVANCE(OUTPUT, CHARIDX, BYTEIDX) \
-if (1) \
- { \
- CHARIDX++; \
- if (!NILP (current_buffer->enable_multibyte_characters)) \
- { \
- unsigned char *ptr = BYTE_POS_ADDR (BYTEIDX); \
- int space_left = ((CHARIDX < GPT ? GPT_BYTE : Z_BYTE) - BYTEIDX); \
- int actual_len; \
- \
- OUTPUT= STRING_CHAR_AND_LENGTH (ptr, space_left, actual_len); \
- BYTEIDX += actual_len; \
- } \
- else \
- { \
- OUTPUT = *(BYTE_POS_ADDR (BYTEIDX)); \
- BYTEIDX++; \
- } \
- } \
-else
-
-/* Return the length of the multi-byte form at string STR of length LEN. */
-
-#define MULTIBYTE_FORM_LENGTH(str, len) \
- (BYTES_BY_CHAR_HEAD (*(unsigned char *)(str)) == 1 \
- ? 1 \
- : multibyte_form_length (str, len))
-
-#ifdef emacs
-
-/* Increase the buffer byte position POS_BYTE of the current buffer to
- the next character boundary. This macro relies on the fact that
- *GPT_ADDR and *Z_ADDR are always accessible and the values are
- '\0'. No range checking of POS. */
-
-#ifdef BYTE_COMBINING_DEBUG
-
-#define INC_POS(pos_byte) \
- do { \
- unsigned char *p = BYTE_POS_ADDR (pos_byte); \
- if (BASE_LEADING_CODE_P (*p)) \
- { \
- int len, bytes; \
- len = Z_BYTE - pos_byte; \
- PARSE_MULTIBYTE_SEQ (p, len, bytes); \
- pos_byte += bytes; \
- } \
- else \
- pos_byte++; \
+/* Check if X is a valid charset symbol. If valid, set ID to the id
+ number of the charset. Otherwise, signal an error. */
+#define CHECK_CHARSET_GET_ID(x, id) \
+ do { \
+ int idx; \
+ \
+ if (! SYMBOLP (x) || (idx = CHARSET_SYMBOL_HASH_INDEX (x)) < 0) \
+ x = wrong_type_argument (Qcharsetp, (x)); \
+ id = AREF (HASH_VALUE (XHASH_TABLE (Vcharset_hash_table), idx), \
+ charset_id); \
} while (0)
-#else /* not BYTE_COMBINING_DEBUG */
-#define INC_POS(pos_byte) \
- do { \
- unsigned char *p = BYTE_POS_ADDR (pos_byte); \
- pos_byte += BYTES_BY_CHAR_HEAD (*p); \
+/* Check if X is a valid charset symbol. If valid, set ATTR to the
+ attr vector of the charset. Otherwise, signal an error. */
+#define CHECK_CHARSET_GET_ATTR(x, attr) \
+ do { \
+ if (!SYMBOLP (x) || NILP (attr = CHARSET_SYMBOL_ATTRIBUTES (x))) \
+ x = wrong_type_argument (Qcharsetp, (x)); \
} while (0)
-#endif /* not BYTE_COMBINING_DEBUG */
-/* Decrease the buffer byte position POS_BYTE of the current buffer to
- the previous character boundary. No range checking of POS. */
-#define DEC_POS(pos_byte) \
- do { \
- unsigned char *p, *p_min; \
- \
- pos_byte--; \
- if (pos_byte < GPT_BYTE) \
- p = BEG_ADDR + pos_byte - 1, p_min = BEG_ADDR; \
- else \
- p = BEG_ADDR + GAP_SIZE + pos_byte - 1, p_min = GAP_END_ADDR; \
- if (p > p_min && !CHAR_HEAD_P (*p)) \
- { \
- unsigned char *pend = p--; \
- int len, bytes; \
- while (p > p_min && !CHAR_HEAD_P (*p)) p--; \
- len = pend + 1 - p; \
- PARSE_MULTIBYTE_SEQ (p, len, bytes); \
- if (bytes == len) \
- pos_byte -= len - 1; \
- } \
+#define CHECK_CHARSET_GET_CHARSET(x, charset) \
+ do { \
+ int id; \
+ CHECK_CHARSET_GET_ID (x, id); \
+ charset = CHARSET_FROM_ID (id); \
} while (0)
-/* Increment both CHARPOS and BYTEPOS, each in the appropriate way. */
-#define INC_BOTH(charpos, bytepos) \
-do \
- { \
- (charpos)++; \
- if (NILP (current_buffer->enable_multibyte_characters)) \
- (bytepos)++; \
- else \
- INC_POS ((bytepos)); \
- } \
-while (0)
-
-/* Decrement both CHARPOS and BYTEPOS, each in the appropriate way. */
-
-#define DEC_BOTH(charpos, bytepos) \
-do \
- { \
- (charpos)--; \
- if (NILP (current_buffer->enable_multibyte_characters)) \
- (bytepos)--; \
- else \
- DEC_POS ((bytepos)); \
- } \
-while (0)
+/* Lookup Vcharset_order_list and return the first charset that
+ contains the character C. */
+#define CHAR_CHARSET(c) \
+ char_charset ((c), Qnil, NULL)
+
+#if 0
+/* Char-table of charset-sets. Each element is a bool vector indexed
+ by a charset ID. */
+extern Lisp_Object Vchar_charset_set;
+
+/* Charset-bag of character C. */
+#define CHAR_CHARSET_SET(c) \
+ CHAR_TABLE_REF (Vchar_charset_set, c)
+
+/* Check if two characters C1 and C2 belong to the same charset. */
+#define SAME_CHARSET_P(c1, c2) \
+ intersection_p (CHAR_CHARSET_SET (c1), CHAR_CHARSET_SET (c2))
+
+#endif
+
+
+/* Return a character correponding to the code-point CODE of CHARSET.
+ Try some optimization before calling decode_char. */
+
+#define DECODE_CHAR(charset, code) \
+ ((ASCII_BYTE_P (code) && (charset)->ascii_compatible_p) \
+ ? (code) \
+ : ((code) < (charset)->min_code || (code) > (charset)->max_code) \
+ ? -1 \
+ : (charset)->unified_p \
+ ? decode_char ((charset), (code)) \
+ : (charset)->method == CHARSET_METHOD_OFFSET \
+ ? ((charset)->code_linear_p \
+ ? (code) - (charset)->min_code + (charset)->code_offset \
+ : decode_char ((charset), (code))) \
+ : (charset)->method == CHARSET_METHOD_MAP \
+ ? ((charset)->code_linear_p \
+ ? XINT (AREF (CHARSET_DECODER (charset), \
+ (code) - (charset)->min_code)) \
+ : decode_char ((charset), (code))) \
+ : decode_char ((charset), (code)))
+
+
+/* Return a code point of CHAR in CHARSET.
+ Try some optimization before calling encode_char. */
+
+#define ENCODE_CHAR(charset, c) \
+ ((ASCII_CHAR_P (c) && (charset)->ascii_compatible_p) \
+ ? (c) \
+ : (charset)->unified_p \
+ ? encode_char ((charset), (c)) \
+ : ((c) < (charset)->min_char || (c) > (charset)->max_char) \
+ ? (charset)->invalid_code \
+ : (charset)->method == CHARSET_METHOD_OFFSET \
+ ? ((charset)->code_linear_p \
+ ? (c) - (charset)->code_offset + (charset)->min_code \
+ : encode_char ((charset), (c))) \
+ : (charset)->method == CHARSET_METHOD_MAP \
+ ? ((charset)->compact_codes_p \
+ ? XFASTINT (CHAR_TABLE_REF (CHARSET_ENCODER (charset), (c))) \
+ : encode_char ((charset), (c))) \
+ : encode_char ((charset), (c)))
+
+
+/* Set to 1 when a charset map is loaded to warn that a buffer text
+ and a string data may be relocated. */
+extern int charset_map_loaded;
+
+
+/* Set CHARSET to the charset highest priority of C, CODE to the
+ code-point of C in CHARSET. */
+#define SPLIT_CHAR(c, charset, code) \
+ ((charset) = char_charset ((c), Qnil, &(code)))
+
+
+#define ISO_MAX_DIMENSION 3
+#define ISO_MAX_CHARS 2
+#define ISO_MAX_FINAL 0x80 /* only 0x30..0xFF are used */
+
+/* Mapping table from ISO2022's charset (specified by DIMENSION,
+ CHARS, and FINAL_CHAR) to Emacs' charset ID. Should be accessed by
+ macro ISO_CHARSET_TABLE (DIMENSION, CHARS, FINAL_CHAR). */
+extern int iso_charset_table[ISO_MAX_DIMENSION][ISO_MAX_CHARS][ISO_MAX_FINAL];
-/* Increase the buffer byte position POS_BYTE of the current buffer to
- the next character boundary. This macro relies on the fact that
- *GPT_ADDR and *Z_ADDR are always accessible and the values are
- '\0'. No range checking of POS_BYTE. */
+/* A charset of type iso2022 who has DIMENSION, CHARS, and FINAL
+ (final character). */
+#define ISO_CHARSET_TABLE(dimension, chars_96, final) \
+ iso_charset_table[(dimension) - 1][(chars_96)][(final)]
-#ifdef BYTE_COMBINING_DEBUG
+/* Nonzero iff the charset who has FAST_MAP may contain C. */
+#define CHARSET_FAST_MAP_REF(c, fast_map) \
+ ((c) < 0x10000 \
+ ? fast_map[(c) >> 10] & (1 << (((c) >> 7) & 7)) \
+ : fast_map[((c) >> 15) + 62] & (1 << (((c) >> 12) & 7)))
-#define BUF_INC_POS(buf, pos_byte) \
+#define CHARSET_FAST_MAP_SET(c, fast_map) \
do { \
- unsigned char *p = BUF_BYTE_ADDRESS (buf, pos_byte); \
- if (BASE_LEADING_CODE_P (*p)) \
- { \
- int len, bytes; \
- len = BUF_Z_BYTE (buf) - pos_byte; \
- PARSE_MULTIBYTE_SEQ (p, len, bytes); \
- pos_byte += bytes; \
- } \
+ if ((c) < 0x10000) \
+ (fast_map)[(c) >> 10] |= 1 << (((c) >> 7) & 7); \
else \
- pos_byte++; \
+ (fast_map)[((c) >> 15) + 62] |= 1 << (((c) >> 12) & 7); \
} while (0)
-#else /* not BYTE_COMBINING_DEBUG */
-#define BUF_INC_POS(buf, pos_byte) \
- do { \
- unsigned char *p = BUF_BYTE_ADDRESS (buf, pos_byte); \
- pos_byte += BYTES_BY_CHAR_HEAD (*p); \
- } while (0)
-#endif /* not BYTE_COMBINING_DEBUG */
+/* 1 iff CHARSET may contain the character C. */
+#define CHAR_CHARSET_P(c, charset) \
+ ((ASCII_CHAR_P (c) && (charset)->ascii_compatible_p) \
+ || (CHARSET_UNIFIED_P (charset) \
+ ? encode_char ((charset), (c)) != (charset)->invalid_code \
+ : (CHARSET_FAST_MAP_REF ((c), (charset)->fast_map) \
+ && ((charset)->method == CHARSET_METHOD_OFFSET \
+ ? (c) >= (charset)->min_char && (c) <= (charset)->max_char \
+ : ((charset)->method == CHARSET_METHOD_MAP \
+ && (charset)->compact_codes_p) \
+ ? (XFASTINT (CHAR_TABLE_REF (CHARSET_ENCODER (charset), (c))) \
+ != (charset)->invalid_code) \
+ : encode_char ((charset), (c)) != (charset)->invalid_code))))
-/* Decrease the buffer byte position POS_BYTE of the current buffer to
- the previous character boundary. No range checking of POS_BYTE. */
-#define BUF_DEC_POS(buf, pos_byte) \
- do { \
- unsigned char *p, *p_min; \
- pos_byte--; \
- if (pos_byte < BUF_GPT_BYTE (buf)) \
- { \
- p = BUF_BEG_ADDR (buf) + pos_byte - 1; \
- p_min = BUF_BEG_ADDR (buf); \
- } \
- else \
- { \
- p = BUF_BEG_ADDR (buf) + BUF_GAP_SIZE (buf) + pos_byte - 1; \
- p_min = BUF_GAP_END_ADDR (buf); \
- } \
- if (p > p_min && !CHAR_HEAD_P (*p)) \
- { \
- unsigned char *pend = p--; \
- int len, bytes; \
- while (p > p_min && !CHAR_HEAD_P (*p)) p--; \
- len = pend + 1 - p; \
- PARSE_MULTIBYTE_SEQ (p, len, bytes); \
- if (bytes == len) \
- pos_byte -= len - 1; \
- } \
- } while (0)
-#endif /* emacs */
-
-/* This is the maximum byte length of multi-byte sequence. */
-#define MAX_MULTIBYTE_LENGTH 4
-
-extern void invalid_character P_ ((int));
-
-extern int translate_char P_ ((Lisp_Object, int, int, int, int));
-extern int split_string P_ ((const unsigned char *, int, int *,
- unsigned char *, unsigned char *));
-extern int char_to_string P_ ((int, unsigned char *));
-extern int char_to_string_1 P_ ((int, unsigned char *));
-extern int string_to_char P_ ((const unsigned char *, int, int *));
-extern int char_printable_p P_ ((int c));
-extern int multibyte_form_length P_ ((const unsigned char *, int));
-extern void parse_str_as_multibyte P_ ((unsigned char *, int, int *, int *));
-extern int str_as_multibyte P_ ((unsigned char *, int, int, int *));
-extern int parse_str_to_multibyte P_ ((unsigned char *, int));
-extern int str_to_multibyte P_ ((unsigned char *, int, int));
-extern int str_as_unibyte P_ ((unsigned char *, int));
-extern int get_charset_id P_ ((Lisp_Object));
-extern int find_charset_in_text P_ ((unsigned char *, int, int, int *,
- Lisp_Object));
-extern int strwidth P_ ((unsigned char *, int));
-extern int c_string_width P_ ((unsigned char *, int, int, int *, int *));
-extern int lisp_string_width P_ ((Lisp_Object, int, int *, int *));
-extern int char_bytes P_ ((int));
-extern int char_valid_p P_ ((int, int));
-
-extern Lisp_Object Vtranslation_table_vector;
-
-/* Return a translation table of id number ID. */
-#define GET_TRANSLATION_TABLE(id) \
- (XCDR(XVECTOR(Vtranslation_table_vector)->contents[(id)]))
-
-/* A char-table for characters which may invoke auto-filling. */
-extern Lisp_Object Vauto_fill_chars;
-
-/* Copy LEN bytes from FROM to TO. This macro should be used only
- when a caller knows that LEN is short and the obvious copy loop is
- faster than calling bcopy which has some overhead. Copying a
- multibyte sequence of a multibyte character is the typical case. */
-
-#define BCOPY_SHORT(from, to, len) \
- do { \
- int i = len; \
- unsigned char *from_p = from, *to_p = to; \
- while (i--) *to_p++ = *from_p++; \
- } while (0)
+extern Lisp_Object Qcharsetp;
+
+extern Lisp_Object Qascii, Qunicode;
+extern int charset_ascii, charset_8_bit_control, charset_8_bit_graphic;
+extern int charset_iso_8859_1;
+extern int charset_primary;
+
+extern struct charset *char_charset P_ ((int, Lisp_Object, unsigned *));
+extern Lisp_Object charset_attributes P_ ((int));
+
+extern int decode_char P_ ((struct charset *, unsigned));
+extern unsigned encode_char P_ ((struct charset *, int));
+extern int string_xstring_p P_ ((Lisp_Object));
+
+EXFUN (Funify_charset, 2);
#endif /* EMACS_CHARSET_H */
projects / emacs.git / commitdiff