1 /* Basic multilingual character support.
2 Copyright (C) 1995, 1997, 1998 Electrotechnical Laboratory, JAPAN.
3 Licensed to the Free Software Foundation.
5 This file is part of GNU Emacs.
7 GNU Emacs is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 2, or (at your option)
12 GNU Emacs is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
17 You should have received a copy of the GNU General Public License
18 along with GNU Emacs; see the file COPYING. If not, write to
19 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
20 Boston, MA 02111-1307, USA. */
22 /* At first, see the document in `charset.h' to understand the code in
29 #include <sys/types.h>
43 Lisp_Object Qcharset
, Qascii
, Qcomposition
;
45 /* Declaration of special leading-codes. */
46 int leading_code_composition
; /* for composite characters */
47 int leading_code_private_11
; /* for private DIMENSION1 of 1-column */
48 int leading_code_private_12
; /* for private DIMENSION1 of 2-column */
49 int leading_code_private_21
; /* for private DIMENSION2 of 1-column */
50 int leading_code_private_22
; /* for private DIMENSION2 of 2-column */
52 /* Declaration of special charsets. */
53 int charset_ascii
; /* ASCII */
54 int charset_composition
; /* for a composite character */
55 int charset_latin_iso8859_1
; /* ISO8859-1 (Latin-1) */
56 int charset_jisx0208_1978
; /* JISX0208.1978 (Japanese Kanji old set) */
57 int charset_jisx0208
; /* JISX0208.1983 (Japanese Kanji) */
58 int charset_katakana_jisx0201
; /* JISX0201.Kana (Japanese Katakana) */
59 int charset_latin_jisx0201
; /* JISX0201.Roman (Japanese Roman) */
60 int charset_big5_1
; /* Big5 Level 1 (Chinese Traditional) */
61 int charset_big5_2
; /* Big5 Level 2 (Chinese Traditional) */
63 int min_composite_char
;
65 Lisp_Object Qcharset_table
;
67 /* A char-table containing information of each character set. */
68 Lisp_Object Vcharset_table
;
70 /* A vector of charset symbol indexed by charset-id. This is used
71 only for returning charset symbol from C functions. */
72 Lisp_Object Vcharset_symbol_table
;
74 /* A list of charset symbols ever defined. */
75 Lisp_Object Vcharset_list
;
77 /* Vector of unification table ever defined.
78 An ID of a unification table is an index of this vector. */
79 Lisp_Object Vcharacter_unification_table_vector
;
81 /* Tables used by macros BYTES_BY_CHAR_HEAD and WIDTH_BY_CHAR_HEAD. */
82 int bytes_by_char_head
[256];
83 int width_by_char_head
[256];
85 /* Mapping table from ISO2022's charset (specified by DIMENSION,
86 CHARS, and FINAL-CHAR) to Emacs' charset. */
87 int iso_charset_table
[2][2][128];
89 /* Table of pointers to the structure `cmpchar_info' indexed by
91 struct cmpchar_info
**cmpchar_table
;
92 /* The current size of `cmpchar_table'. */
93 static int cmpchar_table_size
;
94 /* Number of the current composite characters. */
97 /* Variables used locally in the macro FETCH_MULTIBYTE_CHAR. */
98 unsigned char *_fetch_multibyte_char_p
;
99 int _fetch_multibyte_char_len
;
101 /* Offset to add to a non-ASCII value when inserting it. */
102 int nonascii_insert_offset
;
104 /* Translation table for converting non-ASCII unibyte characters
105 to multibyte codes, or nil. */
106 Lisp_Object Vnonascii_translate_table
;
108 #define min(X, Y) ((X) < (Y) ? (X) : (Y))
109 #define max(X, Y) ((X) > (Y) ? (X) : (Y))
112 invalid_character (c
)
115 error ("Invalid character: %o, %d, 0x%x", c
);
119 /* Set STR a pointer to the multi-byte form of the character C. If C
120 is not a composite character, the multi-byte form is set in WORKBUF
121 and STR points WORKBUF. The caller should allocate at least 4-byte
122 area at WORKBUF in advance. Returns the length of the multi-byte
123 form. If C is an invalid character to have a multi-byte form,
126 Use macro `CHAR_STRING (C, WORKBUF, STR)' instead of calling this
127 function directly if C can be an ASCII character. */
130 non_ascii_char_to_string (c
, workbuf
, str
)
132 unsigned char *workbuf
, **str
;
136 if (COMPOSITE_CHAR_P (c
))
138 int cmpchar_id
= COMPOSITE_CHAR_ID (c
);
140 if (cmpchar_id
< n_cmpchars
)
142 *str
= cmpchar_table
[cmpchar_id
]->data
;
143 return cmpchar_table
[cmpchar_id
]->len
;
147 invalid_character (c
);
151 SPLIT_NON_ASCII_CHAR (c
, charset
, c1
, c2
);
153 || ! CHARSET_DEFINED_P (charset
)
154 || c1
>= 0 && c1
< 32
155 || c2
>= 0 && c2
< 32)
156 invalid_character (c
);
159 *workbuf
++ = CHARSET_LEADING_CODE_BASE (charset
);
160 if (*workbuf
= CHARSET_LEADING_CODE_EXT (charset
))
162 *workbuf
++ = c1
| 0x80;
164 *workbuf
++ = c2
| 0x80;
166 return (workbuf
- *str
);
169 /* Return a non-ASCII character of which multi-byte form is at STR of
170 length LEN. If ACTUAL_LEN is not NULL, the actual length of the
171 character is set to the address ACTUAL_LEN.
173 Use macro `STRING_CHAR (STR, LEN)' instead of calling this function
174 directly if STR can hold an ASCII character. */
177 string_to_non_ascii_char (str
, len
, actual_len
)
178 const unsigned char *str
;
179 int len
, *actual_len
;
182 unsigned char c1
, c2
;
183 register int c
, bytes
;
188 if (BASE_LEADING_CODE_P (c
))
190 while (bytes
< len
&& ! CHAR_HEAD_P (str
[bytes
])) bytes
++;
192 if (c
== LEADING_CODE_COMPOSITION
)
194 int cmpchar_id
= str_cmpchar_id (str
, bytes
);
197 c
= MAKE_COMPOSITE_CHAR (cmpchar_id
);
201 int charset
= c
, c1
, c2
= 0;
204 if (c
>= LEADING_CODE_PRIVATE_11
)
206 if (BYTES_BY_CHAR_HEAD (c
) <= bytes
&& CHARSET_DEFINED_P (charset
))
209 if (CHARSET_DIMENSION (charset
) == 2)
211 c
= MAKE_NON_ASCII_CHAR (charset
, c1
, c2
);
221 /* Return the length of the multi-byte form at string STR of length LEN. */
223 multibyte_form_length (str
, len
)
224 const unsigned char *str
;
229 if (BASE_LEADING_CODE_P (*str
))
230 while (bytes
< len
&& ! CHAR_HEAD_P (str
[bytes
])) bytes
++;
235 /* Check if string STR of length LEN contains valid multi-byte form of
236 a character. If valid, charset and position codes of the character
237 is set at *CHARSET, *C1, and *C2, and return 0. If not valid,
238 return -1. This should be used only in the macro SPLIT_STRING
239 which checks range of STR in advance. */
242 split_non_ascii_string (str
, len
, charset
, c1
, c2
)
243 register const unsigned char *str
;
244 register unsigned char *c1
, *c2
;
245 register int len
, *charset
;
247 register unsigned int cs
= *str
++;
249 if (cs
== LEADING_CODE_COMPOSITION
)
251 int cmpchar_id
= str_cmpchar_id (str
- 1, len
);
255 *charset
= cs
, *c1
= cmpchar_id
>> 7, *c2
= cmpchar_id
& 0x7F;
257 else if ((cs
< LEADING_CODE_PRIVATE_11
|| (cs
= *str
++) >= 0xA0)
258 && CHARSET_DEFINED_P (cs
))
263 *c1
= (*str
++) & 0x7F;
264 if (CHARSET_DIMENSION (cs
) == 2)
268 *c2
= (*str
++) & 0x7F;
276 /* Return a character unified with C (or a character made of CHARSET,
277 C1, and C2) in unification table TABLE. If no unification is found
278 in TABLE, return C. */
280 unify_char (table
, c
, charset
, c1
, c2
)
282 int c
, charset
, c1
, c2
;
285 int alt_charset
, alt_c1
, alt_c2
, dimension
;
287 if (c
< 0) c
= MAKE_CHAR (charset
, c1
, c2
);
288 if (!CHAR_TABLE_P (table
)
289 || (ch
= Faref (table
, make_number (c
)), !INTEGERP (ch
))
293 SPLIT_CHAR (XFASTINT (ch
), alt_charset
, alt_c1
, alt_c2
);
294 dimension
= CHARSET_DIMENSION (alt_charset
);
295 if (dimension
== 1 && alt_c1
> 0 || dimension
== 2 && alt_c2
> 0)
296 /* CH is not a generic character, just return it. */
297 return XFASTINT (ch
);
299 /* Since CH is a generic character, we must return a specific
300 charater which has the same position codes as C from CH. */
302 SPLIT_CHAR (c
, charset
, c1
, c2
);
303 if (dimension
!= CHARSET_DIMENSION (charset
))
304 /* We can't make such a character because of dimension mismatch. */
306 return MAKE_CHAR (alt_charset
, c1
, c2
);
309 /* Convert the unibyte character C to multibyte based on
310 Vnonascii_translate_table or nonascii_insert_offset. If they can't
311 convert C to a valid multibyte character, convert it based on
312 DEFAULT_NONASCII_INSERT_OFFSET which makes C a Latin-1 character. */
315 unibyte_char_to_multibyte (c
)
318 if (c
>= 0240 && c
< 0400)
322 if (! NILP (Vnonascii_translate_table
))
323 c
= XINT (Faref (Vnonascii_translate_table
, make_number (c
)));
324 else if (nonascii_insert_offset
> 0)
325 c
+= nonascii_insert_offset
;
326 if (c
>= 0240 && (c
< 0400 || ! VALID_MULTIBYTE_CHAR_P (c
)))
327 c
= c_save
+ DEFAULT_NONASCII_INSERT_OFFSET
;
332 /* Update the table Vcharset_table with the given arguments (see the
333 document of `define-charset' for the meaning of each argument).
334 Several other table contents are also updated. The caller should
335 check the validity of CHARSET-ID and the remaining arguments in
339 update_charset_table (charset_id
, dimension
, chars
, width
, direction
,
340 iso_final_char
, iso_graphic_plane
,
341 short_name
, long_name
, description
)
342 Lisp_Object charset_id
, dimension
, chars
, width
, direction
;
343 Lisp_Object iso_final_char
, iso_graphic_plane
;
344 Lisp_Object short_name
, long_name
, description
;
346 int charset
= XINT (charset_id
);
348 unsigned char leading_code_base
, leading_code_ext
;
350 if (NILP (CHARSET_TABLE_ENTRY (charset
)))
351 CHARSET_TABLE_ENTRY (charset
)
352 = Fmake_vector (make_number (CHARSET_MAX_IDX
), Qnil
);
354 /* Get byte length of multibyte form, base leading-code, and
355 extended leading-code of the charset. See the comment under the
356 title "GENERAL NOTE on CHARACTER SET (CHARSET)" in charset.h. */
357 bytes
= XINT (dimension
);
358 if (charset
< MIN_CHARSET_PRIVATE_DIMENSION1
)
360 /* Official charset, it doesn't have an extended leading-code. */
361 if (charset
!= CHARSET_ASCII
)
362 bytes
+= 1; /* For a base leading-code. */
363 leading_code_base
= charset
;
364 leading_code_ext
= 0;
368 /* Private charset. */
369 bytes
+= 2; /* For base and extended leading-codes. */
371 = (charset
< LEADING_CODE_EXT_12
372 ? LEADING_CODE_PRIVATE_11
373 : (charset
< LEADING_CODE_EXT_21
374 ? LEADING_CODE_PRIVATE_12
375 : (charset
< LEADING_CODE_EXT_22
376 ? LEADING_CODE_PRIVATE_21
377 : LEADING_CODE_PRIVATE_22
)));
378 leading_code_ext
= charset
;
381 CHARSET_TABLE_INFO (charset
, CHARSET_ID_IDX
) = charset_id
;
382 CHARSET_TABLE_INFO (charset
, CHARSET_BYTES_IDX
) = make_number (bytes
);
383 CHARSET_TABLE_INFO (charset
, CHARSET_DIMENSION_IDX
) = dimension
;
384 CHARSET_TABLE_INFO (charset
, CHARSET_CHARS_IDX
) = chars
;
385 CHARSET_TABLE_INFO (charset
, CHARSET_WIDTH_IDX
) = width
;
386 CHARSET_TABLE_INFO (charset
, CHARSET_DIRECTION_IDX
) = direction
;
387 CHARSET_TABLE_INFO (charset
, CHARSET_LEADING_CODE_BASE_IDX
)
388 = make_number (leading_code_base
);
389 CHARSET_TABLE_INFO (charset
, CHARSET_LEADING_CODE_EXT_IDX
)
390 = make_number (leading_code_ext
);
391 CHARSET_TABLE_INFO (charset
, CHARSET_ISO_FINAL_CHAR_IDX
) = iso_final_char
;
392 CHARSET_TABLE_INFO (charset
, CHARSET_ISO_GRAPHIC_PLANE_IDX
)
394 CHARSET_TABLE_INFO (charset
, CHARSET_SHORT_NAME_IDX
) = short_name
;
395 CHARSET_TABLE_INFO (charset
, CHARSET_LONG_NAME_IDX
) = long_name
;
396 CHARSET_TABLE_INFO (charset
, CHARSET_DESCRIPTION_IDX
) = description
;
397 CHARSET_TABLE_INFO (charset
, CHARSET_PLIST_IDX
) = Qnil
;
400 /* If we have already defined a charset which has the same
401 DIMENSION, CHARS and ISO-FINAL-CHAR but the different
402 DIRECTION, we must update the entry REVERSE-CHARSET of both
403 charsets. If there's no such charset, the value of the entry
407 for (i
= 0; i
<= MAX_CHARSET
; i
++)
408 if (!NILP (CHARSET_TABLE_ENTRY (i
)))
410 if (CHARSET_DIMENSION (i
) == XINT (dimension
)
411 && CHARSET_CHARS (i
) == XINT (chars
)
412 && CHARSET_ISO_FINAL_CHAR (i
) == XINT (iso_final_char
)
413 && CHARSET_DIRECTION (i
) != XINT (direction
))
415 CHARSET_TABLE_INFO (charset
, CHARSET_REVERSE_CHARSET_IDX
)
417 CHARSET_TABLE_INFO (i
, CHARSET_REVERSE_CHARSET_IDX
) = charset_id
;
422 /* No such a charset. */
423 CHARSET_TABLE_INFO (charset
, CHARSET_REVERSE_CHARSET_IDX
)
427 if (charset
!= CHARSET_ASCII
428 && charset
< MIN_CHARSET_PRIVATE_DIMENSION1
)
430 /* Update tables bytes_by_char_head and width_by_char_head. */
431 bytes_by_char_head
[leading_code_base
] = bytes
;
432 width_by_char_head
[leading_code_base
] = XINT (width
);
434 /* Update table emacs_code_class. */
435 emacs_code_class
[charset
] = (bytes
== 2
436 ? EMACS_leading_code_2
438 ? EMACS_leading_code_3
439 : EMACS_leading_code_4
));
442 /* Update table iso_charset_table. */
443 if (ISO_CHARSET_TABLE (dimension
, chars
, iso_final_char
) < 0)
444 ISO_CHARSET_TABLE (dimension
, chars
, iso_final_char
) = charset
;
449 /* Return charset id of CHARSET_SYMBOL, or return -1 if CHARSET_SYMBOL
452 get_charset_id (charset_symbol
)
453 Lisp_Object charset_symbol
;
458 return ((SYMBOLP (charset_symbol
)
459 && (val
= Fget (charset_symbol
, Qcharset
), VECTORP (val
))
460 && (charset
= XINT (XVECTOR (val
)->contents
[CHARSET_ID_IDX
]),
461 CHARSET_VALID_P (charset
)))
465 /* Return an identification number for a new private charset of
466 DIMENSION and WIDTH. If there's no more room for the new charset,
469 get_new_private_charset_id (dimension
, width
)
470 int dimension
, width
;
472 int charset
, from
, to
;
477 from
= LEADING_CODE_EXT_11
, to
= LEADING_CODE_EXT_12
;
479 from
= LEADING_CODE_EXT_12
, to
= LEADING_CODE_EXT_21
;
484 from
= LEADING_CODE_EXT_21
, to
= LEADING_CODE_EXT_22
;
486 from
= LEADING_CODE_EXT_22
, to
= LEADING_CODE_EXT_MAX
+ 1;
489 for (charset
= from
; charset
< to
; charset
++)
490 if (!CHARSET_DEFINED_P (charset
)) break;
492 return make_number (charset
< to
? charset
: 0);
495 DEFUN ("define-charset", Fdefine_charset
, Sdefine_charset
, 3, 3, 0,
496 "Define CHARSET-ID as the identification number of CHARSET with INFO-VECTOR.\n\
497 If CHARSET-ID is nil, it is decided automatically, which means CHARSET is\n\
498 treated as a private charset.\n\
499 INFO-VECTOR is a vector of the format:\n\
500 [DIMENSION CHARS WIDTH DIRECTION ISO-FINAL-CHAR ISO-GRAPHIC-PLANE\n\
501 SHORT-NAME LONG-NAME DESCRIPTION]\n\
502 The meanings of each elements is as follows:\n\
503 DIMENSION (integer) is the number of bytes to represent a character: 1 or 2.\n\
504 CHARS (integer) is the number of characters in a dimension: 94 or 96.\n\
505 WIDTH (integer) is the number of columns a character in the charset\n\
506 occupies on the screen: one of 0, 1, and 2.\n\
508 DIRECTION (integer) is the rendering direction of characters in the\n\
509 charset when rendering. If 0, render from right to left, else\n\
510 render from left to right.\n\
512 ISO-FINAL-CHAR (character) is the final character of the\n\
513 corresponding ISO 2022 charset.\n\
515 ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked\n\
516 while encoding to variants of ISO 2022 coding system, one of the\n\
517 following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).\n\
519 SHORT-NAME (string) is the short name to refer to the charset.\n\
521 LONG-NAME (string) is the long name to refer to the charset.\n\
523 DESCRIPTION (string) is the description string of the charset.")
524 (charset_id
, charset_symbol
, info_vector
)
525 Lisp_Object charset_id
, charset_symbol
, info_vector
;
529 if (!NILP (charset_id
))
530 CHECK_NUMBER (charset_id
, 0);
531 CHECK_SYMBOL (charset_symbol
, 1);
532 CHECK_VECTOR (info_vector
, 2);
534 if (! NILP (charset_id
))
536 if (! CHARSET_VALID_P (XINT (charset_id
)))
537 error ("Invalid CHARSET: %d", XINT (charset_id
));
538 else if (CHARSET_DEFINED_P (XINT (charset_id
)))
539 error ("Already defined charset: %d", XINT (charset_id
));
542 vec
= XVECTOR (info_vector
)->contents
;
543 if (XVECTOR (info_vector
)->size
!= 9
544 || !INTEGERP (vec
[0]) || !(XINT (vec
[0]) == 1 || XINT (vec
[0]) == 2)
545 || !INTEGERP (vec
[1]) || !(XINT (vec
[1]) == 94 || XINT (vec
[1]) == 96)
546 || !INTEGERP (vec
[2]) || !(XINT (vec
[2]) == 1 || XINT (vec
[2]) == 2)
547 || !INTEGERP (vec
[3]) || !(XINT (vec
[3]) == 0 || XINT (vec
[3]) == 1)
548 || !INTEGERP (vec
[4]) || !(XINT (vec
[4]) >= '0' && XINT (vec
[4]) <= '~')
549 || !INTEGERP (vec
[5]) || !(XINT (vec
[5]) == 0 || XINT (vec
[5]) == 1)
552 || !STRINGP (vec
[8]))
553 error ("Invalid info-vector argument for defining charset %s",
554 XSYMBOL (charset_symbol
)->name
->data
);
556 if (NILP (charset_id
))
558 charset_id
= get_new_private_charset_id (XINT (vec
[0]), XINT (vec
[2]));
559 if (XINT (charset_id
) == 0)
560 error ("There's no room for a new private charset %s",
561 XSYMBOL (charset_symbol
)->name
->data
);
564 update_charset_table (charset_id
, vec
[0], vec
[1], vec
[2], vec
[3],
565 vec
[4], vec
[5], vec
[6], vec
[7], vec
[8]);
566 Fput (charset_symbol
, Qcharset
, CHARSET_TABLE_ENTRY (XINT (charset_id
)));
567 CHARSET_SYMBOL (XINT (charset_id
)) = charset_symbol
;
568 Vcharset_list
= Fcons (charset_symbol
, Vcharset_list
);
572 DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char
,
573 Sget_unused_iso_final_char
, 2, 2, 0,
574 "Return an unsed ISO's final char for a charset of DIMENISION and CHARS.\n\
575 DIMENSION is the number of bytes to represent a character: 1 or 2.\n\
576 CHARS is the number of characters in a dimension: 94 or 96.\n\
578 This final char is for private use, thus the range is `0' (48) .. `?' (63).\n\
579 If there's no unused final char for the specified kind of charset,\n\
582 Lisp_Object dimension
, chars
;
586 CHECK_NUMBER (dimension
, 0);
587 CHECK_NUMBER (chars
, 1);
588 if (XINT (dimension
) != 1 && XINT (dimension
) != 2)
589 error ("Invalid charset dimension %d, it should be 1 or 2",
591 if (XINT (chars
) != 94 && XINT (chars
) != 96)
592 error ("Invalid charset chars %d, it should be 94 or 96",
594 for (final_char
= '0'; final_char
<= '?'; final_char
++)
596 if (ISO_CHARSET_TABLE (dimension
, chars
, make_number (final_char
)) < 0)
599 return (final_char
<= '?' ? make_number (final_char
) : Qnil
);
602 DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset
, Sdeclare_equiv_charset
,
604 "Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET.\n\
605 CHARSET should be defined by `defined-charset' in advance.")
606 (dimension
, chars
, final_char
, charset_symbol
)
607 Lisp_Object dimension
, chars
, final_char
, charset_symbol
;
611 CHECK_NUMBER (dimension
, 0);
612 CHECK_NUMBER (chars
, 1);
613 CHECK_NUMBER (final_char
, 2);
614 CHECK_SYMBOL (charset_symbol
, 3);
616 if (XINT (dimension
) != 1 && XINT (dimension
) != 2)
617 error ("Invalid DIMENSION %d, it should be 1 or 2", XINT (dimension
));
618 if (XINT (chars
) != 94 && XINT (chars
) != 96)
619 error ("Invalid CHARS %d, it should be 94 or 96", XINT (chars
));
620 if (XINT (final_char
) < '0' || XFASTINT (final_char
) > '~')
621 error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", XINT (chars
));
622 if ((charset
= get_charset_id (charset_symbol
)) < 0)
623 error ("Invalid charset %s", XSYMBOL (charset_symbol
)->name
->data
);
625 ISO_CHARSET_TABLE (dimension
, chars
, final_char
) = charset
;
629 /* Return number of different charsets in STR of length LEN. In
630 addition, for each found charset N, CHARSETS[N] is set 1. The
631 caller should allocate CHARSETS (MAX_CHARSET + 1 elements) in advance.
632 It may lookup a unification table TABLE if supplied. */
635 find_charset_in_str (str
, len
, charsets
, table
)
640 register int num
= 0, c
;
642 if (! CHAR_TABLE_P (table
))
650 if (c
== LEADING_CODE_COMPOSITION
)
652 int cmpchar_id
= str_cmpchar_id (str
, len
);
657 struct cmpchar_info
*cmpcharp
= cmpchar_table
[cmpchar_id
];
660 for (i
= 0; i
< cmpcharp
->glyph_len
; i
++)
662 c
= cmpcharp
->glyph
[i
];
665 if ((c
= unify_char (table
, c
, 0, 0, 0)) < 0)
666 c
= cmpcharp
->glyph
[i
];
668 if ((charset
= CHAR_CHARSET (c
)) < 0)
669 charset
= CHARSET_ASCII
;
670 if (!charsets
[charset
])
672 charsets
[charset
] = 1;
676 str
+= cmpcharp
->len
;
677 len
-= cmpcharp
->len
;
681 charset
= CHARSET_ASCII
;
686 c
= STRING_CHAR_AND_LENGTH (str
, len
, bytes
);
689 int c1
= unify_char (table
, c
, 0, 0, 0);
693 charset
= CHAR_CHARSET (c
);
696 if (!charsets
[charset
])
698 charsets
[charset
] = 1;
707 DEFUN ("find-charset-region", Ffind_charset_region
, Sfind_charset_region
,
709 "Return a list of charsets in the region between BEG and END.\n\
710 BEG and END are buffer positions.\n\
711 Optional arg TABLE if non-nil is a unification table to look up.")
713 Lisp_Object beg
, end
, table
;
715 int charsets
[MAX_CHARSET
+ 1];
716 int from
, from_byte
, to
, stop
, stop_byte
, i
;
719 validate_region (&beg
, &end
);
720 from
= XFASTINT (beg
);
721 stop
= to
= XFASTINT (end
);
723 if (from
< GPT
&& GPT
< to
)
726 stop_byte
= GPT_BYTE
;
729 stop_byte
= CHAR_TO_BYTE (stop
);
731 from_byte
= CHAR_TO_BYTE (from
);
733 bzero (charsets
, (MAX_CHARSET
+ 1) * sizeof (int));
736 find_charset_in_str (BYTE_POS_ADDR (from_byte
), stop_byte
- from_byte
,
740 from
= stop
, from_byte
= stop_byte
;
741 stop
= to
, stop_byte
= CHAR_TO_BYTE (stop
);
748 for (i
= MAX_CHARSET
; i
>= 0; i
--)
750 val
= Fcons (CHARSET_SYMBOL (i
), val
);
754 DEFUN ("find-charset-string", Ffind_charset_string
, Sfind_charset_string
,
756 "Return a list of charsets in STR.\n\
757 Optional arg TABLE if non-nil is a unification table to look up.")
759 Lisp_Object str
, table
;
761 int charsets
[MAX_CHARSET
+ 1];
765 CHECK_STRING (str
, 0);
767 if (! STRING_MULTIBYTE (str
))
770 bzero (charsets
, (MAX_CHARSET
+ 1) * sizeof (int));
771 find_charset_in_str (XSTRING (str
)->data
, STRING_BYTES (XSTRING (str
)),
774 for (i
= MAX_CHARSET
; i
>= 0; i
--)
776 val
= Fcons (CHARSET_SYMBOL (i
), val
);
780 DEFUN ("make-char-internal", Fmake_char_internal
, Smake_char_internal
, 1, 3, 0,
782 (charset
, code1
, code2
)
783 Lisp_Object charset
, code1
, code2
;
785 CHECK_NUMBER (charset
, 0);
788 XSETFASTINT (code1
, 0);
790 CHECK_NUMBER (code1
, 1);
792 XSETFASTINT (code2
, 0);
794 CHECK_NUMBER (code2
, 2);
796 if (!CHARSET_DEFINED_P (XINT (charset
)))
797 error ("Invalid charset: %d", XINT (charset
));
799 return make_number (MAKE_CHAR (XINT (charset
), XINT (code1
), XINT (code2
)));
802 DEFUN ("split-char", Fsplit_char
, Ssplit_char
, 1, 1, 0,
803 "Return list of charset and one or two position-codes of CHAR.")
810 CHECK_NUMBER (ch
, 0);
811 SPLIT_CHAR (XFASTINT (ch
), charset
, c1
, c2
);
813 ? Fcons (CHARSET_SYMBOL (charset
),
814 Fcons (make_number (c1
), Fcons (make_number (c2
), Qnil
)))
815 : Fcons (CHARSET_SYMBOL (charset
), Fcons (make_number (c1
), Qnil
)));
818 DEFUN ("char-charset", Fchar_charset
, Schar_charset
, 1, 1, 0,
819 "Return charset of CHAR.")
823 CHECK_NUMBER (ch
, 0);
825 return CHARSET_SYMBOL (CHAR_CHARSET (XINT (ch
)));
828 DEFUN ("charset-after", Fcharset_after
, Scharset_after
, 0, 1, 0,
829 "Return charset of a character in current buffer at position POS.\n\
830 If POS is nil, it defauls to the current point.")
834 register int pos_byte
, c
, charset
;
835 register unsigned char *p
;
839 else if (MARKERP (pos
))
840 pos_byte
= marker_byte_position (pos
);
843 CHECK_NUMBER (pos
, 0);
844 pos_byte
= CHAR_TO_BYTE (XINT (pos
));
846 p
= BYTE_POS_ADDR (pos_byte
);
847 c
= STRING_CHAR (p
, Z_BYTE
- pos_byte
);
848 charset
= CHAR_CHARSET (c
);
849 return CHARSET_SYMBOL (charset
);
852 DEFUN ("iso-charset", Fiso_charset
, Siso_charset
, 3, 3, 0,
853 "Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR.\n\
855 ISO 2022's designation sequence (escape sequence) distinguishes charsets\n\
856 by their DIMENSION, CHARS, and FINAL-CHAR,\n\
857 where as Emacs distinguishes them by charset symbol.\n\
858 See the documentation of the function `charset-info' for the meanings of\n\
859 DIMENSION, CHARS, and FINAL-CHAR.")
860 (dimension
, chars
, final_char
)
861 Lisp_Object dimension
, chars
, final_char
;
865 CHECK_NUMBER (dimension
, 0);
866 CHECK_NUMBER (chars
, 1);
867 CHECK_NUMBER (final_char
, 2);
869 if ((charset
= ISO_CHARSET_TABLE (dimension
, chars
, final_char
)) < 0)
871 return CHARSET_SYMBOL (charset
);
874 /* If GENERICP is nonzero, return nonzero iff C is a valid normal or
875 generic character. If GENERICP is zero, return nonzero iff C is a
876 valid normal character. Do not call this function directly,
877 instead use macro CHAR_VALID_P. */
879 char_valid_p (c
, genericp
)
886 if (SINGLE_BYTE_CHAR_P (c
))
888 SPLIT_NON_ASCII_CHAR (c
, charset
, c1
, c2
);
889 if (!CHARSET_VALID_P (charset
))
891 return (c
< MIN_CHAR_COMPOSITION
892 ? ((c
& CHAR_FIELD1_MASK
) /* i.e. dimension of C is two. */
893 ? (genericp
&& c1
== 0 && c2
== 0
894 || c1
>= 32 && c2
>= 32)
895 : (genericp
&& c1
== 0
897 : c
< MIN_CHAR_COMPOSITION
+ n_cmpchars
);
900 DEFUN ("char-valid-p", Fchar_valid_p
, Schar_valid_p
, 1, 2, 0,
901 "Return t if OBJECT is a valid normal character.\n\
902 If optional arg GENERICP is non-nil, also return t if OBJECT is\n\
903 a valid generic character.")
905 Lisp_Object object
, genericp
;
907 if (! NATNUMP (object
))
909 return (CHAR_VALID_P (XFASTINT (object
), !NILP (genericp
)) ? Qt
: Qnil
);
912 DEFUN ("unibyte-char-to-multibyte", Funibyte_char_to_multibyte
,
913 Sunibyte_char_to_multibyte
, 1, 1, 0,
914 "Convert the unibyte character CH to multibyte character.\n\
915 The conversion is done based on nonascii-translate-table (which see)\n\
916 or nonascii-insert-offset (which see).")
922 CHECK_NUMBER (ch
, 0);
924 if (c
< 0 || c
>= 0400)
925 error ("Invalid unibyte character: %d", c
);
926 c
= unibyte_char_to_multibyte (c
);
928 error ("Can't convert to multibyte character: %d", XINT (ch
));
929 return make_number (c
);
932 DEFUN ("char-bytes", Fchar_bytes
, Schar_bytes
, 1, 1, 0,
933 "Return byte length of multi-byte form of CHAR.")
940 CHECK_NUMBER (ch
, 0);
941 if (COMPOSITE_CHAR_P (XFASTINT (ch
)))
943 unsigned int id
= COMPOSITE_CHAR_ID (XFASTINT (ch
));
945 bytes
= (id
< n_cmpchars
? cmpchar_table
[id
]->len
: 1);
949 int charset
= CHAR_CHARSET (XFASTINT (ch
));
951 bytes
= CHARSET_DEFINED_P (charset
) ? CHARSET_BYTES (charset
) : 1;
954 XSETFASTINT (val
, bytes
);
958 /* Return the width of character of which multi-byte form starts with
959 C. The width is measured by how many columns occupied on the
960 screen when displayed in the current buffer. */
962 #define ONE_BYTE_CHAR_WIDTH(c) \
965 ? XFASTINT (current_buffer->tab_width) \
966 : (c == '\n' ? 0 : (NILP (current_buffer->ctl_arrow) ? 4 : 2))) \
970 ? (NILP (current_buffer->ctl_arrow) ? 4 : 2) \
971 : ((! NILP (current_buffer->enable_multibyte_characters) \
972 && BASE_LEADING_CODE_P (c)) \
973 ? WIDTH_BY_CHAR_HEAD (c) \
977 DEFUN ("char-width", Fchar_width, Schar_width, 1, 1, 0,
978 "Return width of CHAR when displayed in the current buffer.\n\
979 The width is measured by how many columns it occupies on the screen.")
983 Lisp_Object val
, disp
;
985 struct Lisp_Char_Table
*dp
= buffer_display_table ();
987 CHECK_NUMBER (ch
, 0);
991 /* Get the way the display table would display it. */
992 disp
= dp
? DISP_CHAR_VECTOR (dp
, c
) : Qnil
;
995 XSETINT (val
, XVECTOR (disp
)->size
);
996 else if (SINGLE_BYTE_CHAR_P (c
))
997 XSETINT (val
, ONE_BYTE_CHAR_WIDTH (c
));
998 else if (COMPOSITE_CHAR_P (c
))
1000 int id
= COMPOSITE_CHAR_ID (XFASTINT (ch
));
1001 XSETFASTINT (val
, (id
< n_cmpchars
? cmpchar_table
[id
]->width
: 0));
1005 int charset
= CHAR_CHARSET (c
);
1007 XSETFASTINT (val
, CHARSET_WIDTH (charset
));
1012 /* Return width of string STR of length LEN when displayed in the
1013 current buffer. The width is measured by how many columns it
1014 occupies on the screen. */
1021 unsigned char *endp
= str
+ len
;
1023 struct Lisp_Char_Table
*dp
= buffer_display_table ();
1027 if (*str
== LEADING_CODE_COMPOSITION
)
1029 int id
= str_cmpchar_id (str
, endp
- str
);
1038 width
+= cmpchar_table
[id
]->width
;
1039 str
+= cmpchar_table
[id
]->len
;
1046 int c
= STRING_CHAR_AND_LENGTH (str
, endp
- str
, thislen
);
1048 /* Get the way the display table would display it. */
1050 disp
= DISP_CHAR_VECTOR (dp
, c
);
1055 width
+= XVECTOR (disp
)->size
;
1057 width
+= ONE_BYTE_CHAR_WIDTH (*str
);
1065 DEFUN ("string-width", Fstring_width
, Sstring_width
, 1, 1, 0,
1066 "Return width of STRING when displayed in the current buffer.\n\
1067 Width is measured by how many columns it occupies on the screen.\n\
1068 When calculating width of a multibyte character in STRING,\n\
1069 only the base leading-code is considered; the validity of\n\
1070 the following bytes is not checked.")
1076 CHECK_STRING (str
, 0);
1077 XSETFASTINT (val
, strwidth (XSTRING (str
)->data
,
1078 STRING_BYTES (XSTRING (str
))));
1082 DEFUN ("char-direction", Fchar_direction
, Schar_direction
, 1, 1, 0,
1083 "Return the direction of CHAR.\n\
1084 The returned value is 0 for left-to-right and 1 for right-to-left.")
1090 CHECK_NUMBER (ch
, 0);
1091 charset
= CHAR_CHARSET (XFASTINT (ch
));
1092 if (!CHARSET_DEFINED_P (charset
))
1093 invalid_character (XINT (ch
));
1094 return CHARSET_TABLE_INFO (charset
, CHARSET_DIRECTION_IDX
);
1097 DEFUN ("chars-in-region", Fchars_in_region
, Schars_in_region
, 2, 2, 0,
1098 "Return number of characters between BEG and END.")
1100 Lisp_Object beg
, end
;
1104 from
= min (XFASTINT (beg
), XFASTINT (end
));
1105 to
= max (XFASTINT (beg
), XFASTINT (end
));
1107 return make_number (to
- from
);
1110 /* Return the number of characters in the NBYTES bytes at PTR.
1111 This works by looking at the contents and checking for multibyte sequences.
1112 However, if the current buffer has enable-multibyte-characters = nil,
1113 we treat each byte as a character. */
1116 chars_in_text (ptr
, nbytes
)
1120 unsigned char *endp
, c
;
1123 /* current_buffer is null at early stages of Emacs initialization. */
1124 if (current_buffer
== 0
1125 || NILP (current_buffer
->enable_multibyte_characters
))
1128 endp
= ptr
+ nbytes
;
1135 if (BASE_LEADING_CODE_P (c
))
1136 while (ptr
< endp
&& ! CHAR_HEAD_P (*ptr
)) ptr
++;
1143 /* Return the number of characters in the NBYTES bytes at PTR.
1144 This works by looking at the contents and checking for multibyte sequences.
1145 It ignores enable-multibyte-characters. */
1148 multibyte_chars_in_text (ptr
, nbytes
)
1152 unsigned char *endp
, c
;
1155 endp
= ptr
+ nbytes
;
1162 if (BASE_LEADING_CODE_P (c
))
1163 while (ptr
< endp
&& ! CHAR_HEAD_P (*ptr
)) ptr
++;
1170 DEFUN ("string", Fstring
, Sstring
, 1, MANY
, 0,
1171 "Concatenate all the argument characters and make the result a string.")
1178 = (unsigned char *) alloca (MAX_LENGTH_OF_MULTI_BYTE_FORM
* n
);
1179 unsigned char *p
= buf
;
1182 for (i
= 0; i
< n
; i
++)
1187 if (!INTEGERP (args
[i
]))
1188 CHECK_NUMBER (args
[i
], 0);
1190 len
= CHAR_STRING (c
, p
, str
);
1192 /* C is a composite character. */
1193 bcopy (str
, p
, len
);
1197 val
= make_string_from_bytes (buf
, n
, p
- buf
);
1203 /*** Composite characters staffs ***/
1205 /* Each composite character is identified by CMPCHAR-ID which is
1206 assigned when Emacs needs the character code of the composite
1207 character (e.g. when displaying it on the screen). See the
1208 document "GENERAL NOTE on COMPOSITE CHARACTER" in `charset.h' how a
1209 composite character is represented in Emacs. */
1211 /* If `static' is defined, it means that it is defined to null string. */
1213 /* The following function is copied from lread.c. */
1215 hash_string (ptr
, len
)
1219 register unsigned char *p
= ptr
;
1220 register unsigned char *end
= p
+ len
;
1221 register unsigned char c
;
1222 register int hash
= 0;
1227 if (c
>= 0140) c
-= 40;
1228 hash
= ((hash
<<3) + (hash
>>28) + c
);
1230 return hash
& 07777777777;
1234 #define CMPCHAR_HASH_TABLE_SIZE 0xFFF
1236 static int *cmpchar_hash_table
[CMPCHAR_HASH_TABLE_SIZE
];
1238 /* Each element of `cmpchar_hash_table' is a pointer to an array of
1239 integer, where the 1st element is the size of the array, the 2nd
1240 element is how many elements are actually used in the array, and
1241 the remaining elements are CMPCHAR-IDs of composite characters of
1242 the same hash value. */
1243 #define CMPCHAR_HASH_SIZE(table) table[0]
1244 #define CMPCHAR_HASH_USED(table) table[1]
1245 #define CMPCHAR_HASH_CMPCHAR_ID(table, i) table[i]
1247 /* Return CMPCHAR-ID of the composite character in STR of the length
1248 LEN. If the composite character has not yet been registered,
1249 register it in `cmpchar_table' and assign new CMPCHAR-ID. This
1250 is the sole function for assigning CMPCHAR-ID. */
1252 str_cmpchar_id (str
, len
)
1253 const unsigned char *str
;
1256 int hash_idx
, *hashp
;
1258 int embedded_rule
; /* 1 if composition rule is embedded. */
1259 int chars
; /* number of components. */
1261 struct cmpchar_info
*cmpcharp
;
1263 /* The second byte 0xFF means compostion rule is embedded. */
1264 embedded_rule
= (str
[1] == 0xFF);
1266 /* At first, get the actual length of the composite character. */
1268 const unsigned char *p
, *endp
= str
+ 1, *lastp
= str
+ len
;
1271 while (endp
< lastp
&& ! CHAR_HEAD_P (*endp
)) endp
++;
1273 /* Any composite char have at least 5-byte length. */
1280 if (embedded_rule
) p
++;
1281 /* No need of checking if *P is 0xA0 because
1282 BYTES_BY_CHAR_HEAD (0x80) surely returns 2. */
1283 p
+= BYTES_BY_CHAR_HEAD (*p
- 0x20);
1286 if (p
> endp
|| chars
< 2 || chars
> MAX_COMPONENT_COUNT
)
1287 /* Invalid components. */
1291 hash_idx
= hash_string (str
, len
) % CMPCHAR_HASH_TABLE_SIZE
;
1292 hashp
= cmpchar_hash_table
[hash_idx
];
1294 /* Then, look into the hash table. */
1296 /* Find the correct one among composite characters of the same
1298 for (i
= 2; i
< CMPCHAR_HASH_USED (hashp
); i
++)
1300 cmpcharp
= cmpchar_table
[CMPCHAR_HASH_CMPCHAR_ID (hashp
, i
)];
1301 if (len
== cmpcharp
->len
1302 && ! bcmp (str
, cmpcharp
->data
, len
))
1303 return CMPCHAR_HASH_CMPCHAR_ID (hashp
, i
);
1306 /* We have to register the composite character in cmpchar_table. */
1307 if (n_cmpchars
> (CHAR_FIELD2_MASK
| CHAR_FIELD3_MASK
))
1308 /* No, we have no more room for a new composite character. */
1311 /* Make the entry in hash table. */
1314 /* Make a table for 8 composite characters initially. */
1315 hashp
= (cmpchar_hash_table
[hash_idx
]
1316 = (int *) xmalloc (sizeof (int) * (2 + 8)));
1317 CMPCHAR_HASH_SIZE (hashp
) = 10;
1318 CMPCHAR_HASH_USED (hashp
) = 2;
1320 else if (CMPCHAR_HASH_USED (hashp
) >= CMPCHAR_HASH_SIZE (hashp
))
1322 CMPCHAR_HASH_SIZE (hashp
) += 8;
1323 hashp
= (cmpchar_hash_table
[hash_idx
]
1324 = (int *) xrealloc (hashp
,
1325 sizeof (int) * CMPCHAR_HASH_SIZE (hashp
)));
1327 CMPCHAR_HASH_CMPCHAR_ID (hashp
, CMPCHAR_HASH_USED (hashp
)) = n_cmpchars
;
1328 CMPCHAR_HASH_USED (hashp
)++;
1330 /* Set information of the composite character in cmpchar_table. */
1331 if (cmpchar_table_size
== 0)
1333 /* This is the first composite character to be registered. */
1334 cmpchar_table_size
= 256;
1336 = (struct cmpchar_info
**) xmalloc (sizeof (cmpchar_table
[0])
1337 * cmpchar_table_size
);
1339 else if (cmpchar_table_size
<= n_cmpchars
)
1341 cmpchar_table_size
+= 256;
1343 = (struct cmpchar_info
**) xrealloc (cmpchar_table
,
1344 sizeof (cmpchar_table
[0])
1345 * cmpchar_table_size
);
1348 cmpcharp
= (struct cmpchar_info
*) xmalloc (sizeof (struct cmpchar_info
));
1350 cmpcharp
->len
= len
;
1351 cmpcharp
->data
= (unsigned char *) xmalloc (len
+ 1);
1352 bcopy (str
, cmpcharp
->data
, len
);
1353 cmpcharp
->data
[len
] = 0;
1354 cmpcharp
->glyph_len
= chars
;
1355 cmpcharp
->glyph
= (GLYPH
*) xmalloc (sizeof (GLYPH
) * chars
);
1358 cmpcharp
->cmp_rule
= (unsigned char *) xmalloc (chars
);
1359 cmpcharp
->col_offset
= (float *) xmalloc (sizeof (float) * chars
);
1363 cmpcharp
->cmp_rule
= NULL
;
1364 cmpcharp
->col_offset
= NULL
;
1367 /* Setup GLYPH data and composition rules (if any) so as not to make
1368 them every time on displaying. */
1370 unsigned char *bufp
;
1372 float leftmost
= 0.0, rightmost
= 1.0;
1375 /* At first, col_offset[N] is set to relative to col_offset[0]. */
1376 cmpcharp
->col_offset
[0] = 0;
1378 for (i
= 0, bufp
= cmpcharp
->data
+ 1; i
< chars
; i
++)
1381 cmpcharp
->cmp_rule
[i
] = *bufp
++;
1383 if (*bufp
== 0xA0) /* This is an ASCII character. */
1385 cmpcharp
->glyph
[i
] = FAST_MAKE_GLYPH ((*++bufp
& 0x7F), 0);
1389 else /* Multibyte character. */
1391 /* Make `bufp' point normal multi-byte form temporally. */
1394 = FAST_MAKE_GLYPH (string_to_non_ascii_char (bufp
, 4, 0), 0);
1395 width
= WIDTH_BY_CHAR_HEAD (*bufp
);
1397 bufp
+= BYTES_BY_CHAR_HEAD (*bufp
- 0x20);
1400 if (embedded_rule
&& i
> 0)
1402 /* Reference points (global_ref and new_ref) are
1413 Now, we calculate the column offset of the new glyph
1414 from the left edge of the first glyph. This can avoid
1415 the same calculation everytime displaying this
1416 composite character. */
1418 /* Reference points of global glyph and new glyph. */
1419 int global_ref
= (cmpcharp
->cmp_rule
[i
] - 0xA0) / 9;
1420 int new_ref
= (cmpcharp
->cmp_rule
[i
] - 0xA0) % 9;
1421 /* Column offset relative to the first glyph. */
1422 float left
= (leftmost
1423 + (global_ref
% 3) * (rightmost
- leftmost
) / 2.0
1424 - (new_ref
% 3) * width
/ 2.0);
1426 cmpcharp
->col_offset
[i
] = left
;
1427 if (left
< leftmost
)
1429 if (left
+ width
> rightmost
)
1430 rightmost
= left
+ width
;
1434 if (width
> rightmost
)
1440 /* Now col_offset[N] are relative to the left edge of the
1441 first component. Make them relative to the left edge of
1443 for (i
= 0; i
< chars
; i
++)
1444 cmpcharp
->col_offset
[i
] -= leftmost
;
1445 /* Make rightmost holds width of overall glyph. */
1446 rightmost
-= leftmost
;
1449 cmpcharp
->width
= rightmost
;
1450 if (cmpcharp
->width
< rightmost
)
1451 /* To get a ceiling integer value. */
1455 cmpchar_table
[n_cmpchars
] = cmpcharp
;
1457 return n_cmpchars
++;
1460 /* Return the Nth element of the composite character C. */
1462 cmpchar_component (c
, n
)
1465 int id
= COMPOSITE_CHAR_ID (c
);
1467 if (id
>= n_cmpchars
/* C is not a valid composite character. */
1468 || n
>= cmpchar_table
[id
]->glyph_len
) /* No such component. */
1470 /* No face data is stored in glyph code. */
1471 return ((int) (cmpchar_table
[id
]->glyph
[n
]));
1474 DEFUN ("cmpcharp", Fcmpcharp
, Scmpcharp
, 1, 1, 0,
1475 "T if CHAR is a composite character.")
1479 CHECK_NUMBER (ch
, 0);
1480 return (COMPOSITE_CHAR_P (XINT (ch
)) ? Qt
: Qnil
);
1483 DEFUN ("composite-char-component", Fcmpchar_component
, Scmpchar_component
,
1485 "Return the IDXth component character of composite character CHARACTER.")
1487 Lisp_Object character
, idx
;
1491 CHECK_NUMBER (character
, 0);
1492 CHECK_NUMBER (idx
, 1);
1494 if ((c
= cmpchar_component (XINT (character
), XINT (idx
))) < 0)
1495 args_out_of_range (character
, idx
);
1497 return make_number (c
);
1500 DEFUN ("composite-char-composition-rule", Fcmpchar_cmp_rule
, Scmpchar_cmp_rule
,
1502 "Return the Nth composition rule embedded in composite character CHARACTER.\n\
1503 The returned rule is for composing the Nth component\n\
1504 on the (N-1)th component. If N is 0, the returned value is always 255.")
1506 Lisp_Object character
, n
;
1510 CHECK_NUMBER (character
, 0);
1511 CHECK_NUMBER (n
, 1);
1513 id
= COMPOSITE_CHAR_ID (XINT (character
));
1514 if (id
< 0 || id
>= n_cmpchars
)
1515 error ("Invalid composite character: %d", XINT (character
));
1517 if (i
> cmpchar_table
[id
]->glyph_len
)
1518 args_out_of_range (character
, n
);
1520 return make_number (cmpchar_table
[id
]->cmp_rule
[i
]);
1523 DEFUN ("composite-char-composition-rule-p", Fcmpchar_cmp_rule_p
,
1524 Scmpchar_cmp_rule_p
, 1, 1, 0,
1525 "Return non-nil if composite character CHARACTER contains a embedded rule.")
1527 Lisp_Object character
;
1531 CHECK_NUMBER (character
, 0);
1532 id
= COMPOSITE_CHAR_ID (XINT (character
));
1533 if (id
< 0 || id
>= n_cmpchars
)
1534 error ("Invalid composite character: %d", XINT (character
));
1536 return (cmpchar_table
[id
]->cmp_rule
? Qt
: Qnil
);
1539 DEFUN ("composite-char-component-count", Fcmpchar_cmp_count
,
1540 Scmpchar_cmp_count
, 1, 1, 0,
1541 "Return number of compoents of composite character CHARACTER.")
1543 Lisp_Object character
;
1547 CHECK_NUMBER (character
, 0);
1548 id
= COMPOSITE_CHAR_ID (XINT (character
));
1549 if (id
< 0 || id
>= n_cmpchars
)
1550 error ("Invalid composite character: %d", XINT (character
));
1552 return (make_number (cmpchar_table
[id
]->glyph_len
));
1555 DEFUN ("compose-string", Fcompose_string
, Scompose_string
,
1557 "Return one char string composed from all characters in STRING.")
1561 unsigned char buf
[MAX_LENGTH_OF_MULTI_BYTE_FORM
], *p
, *pend
, *ptemp
;
1564 CHECK_STRING (str
, 0);
1566 buf
[0] = LEADING_CODE_COMPOSITION
;
1567 p
= XSTRING (str
)->data
;
1568 pend
= p
+ STRING_BYTES (XSTRING (str
));
1572 if (*p
< 0x20 || *p
== 127) /* control code */
1573 error ("Invalid component character: %d", *p
);
1574 else if (*p
< 0x80) /* ASCII */
1576 if (i
+ 2 >= MAX_LENGTH_OF_MULTI_BYTE_FORM
)
1577 error ("Too long string to be composed: %s", XSTRING (str
)->data
);
1578 /* Prepend an ASCII charset indicator 0xA0, set MSB of the
1581 buf
[i
++] = *p
++ + 0x80;
1583 else if (*p
== LEADING_CODE_COMPOSITION
) /* composite char */
1585 /* Already composed. Eliminate the heading
1586 LEADING_CODE_COMPOSITION, keep the remaining bytes
1590 while (! CHAR_HEAD_P (*p
)) p
++;
1591 if (i
+ (p
- ptemp
) >= MAX_LENGTH_OF_MULTI_BYTE_FORM
)
1592 error ("Too long string to be composed: %s", XSTRING (str
)->data
);
1593 bcopy (ptemp
, buf
+ i
, p
- ptemp
);
1596 else /* multibyte char */
1598 /* Add 0x20 to the base leading-code, keep the remaining
1600 len
= BYTES_BY_CHAR_HEAD (*p
);
1601 if (i
+ len
>= MAX_LENGTH_OF_MULTI_BYTE_FORM
)
1602 error ("Too long string to be composed: %s", XSTRING (str
)->data
);
1603 bcopy (p
, buf
+ i
, len
);
1610 /* STR contains only one character, which can't be composed. */
1611 error ("Too short string to be composed: %s", XSTRING (str
)->data
);
1613 return make_string_from_bytes (buf
, 1, i
);
1618 charset_id_internal (charset_name
)
1621 Lisp_Object val
= Fget (intern (charset_name
), Qcharset
);
1624 error ("Charset %s is not defined", charset_name
);
1626 return (XINT (XVECTOR (val
)->contents
[0]));
1629 DEFUN ("setup-special-charsets", Fsetup_special_charsets
,
1630 Ssetup_special_charsets
, 0, 0, 0, "Internal use only.")
1633 charset_latin_iso8859_1
= charset_id_internal ("latin-iso8859-1");
1634 charset_jisx0208_1978
= charset_id_internal ("japanese-jisx0208-1978");
1635 charset_jisx0208
= charset_id_internal ("japanese-jisx0208");
1636 charset_katakana_jisx0201
= charset_id_internal ("katakana-jisx0201");
1637 charset_latin_jisx0201
= charset_id_internal ("latin-jisx0201");
1638 charset_big5_1
= charset_id_internal ("chinese-big5-1");
1639 charset_big5_2
= charset_id_internal ("chinese-big5-2");
1644 init_charset_once ()
1648 staticpro (&Vcharset_table
);
1649 staticpro (&Vcharset_symbol_table
);
1651 /* This has to be done here, before we call Fmake_char_table. */
1652 Qcharset_table
= intern ("charset-table");
1653 staticpro (&Qcharset_table
);
1655 /* Intern this now in case it isn't already done.
1656 Setting this variable twice is harmless.
1657 But don't staticpro it here--that is done in alloc.c. */
1658 Qchar_table_extra_slots
= intern ("char-table-extra-slots");
1660 /* Now we are ready to set up this property, so we can
1661 create the charset table. */
1662 Fput (Qcharset_table
, Qchar_table_extra_slots
, make_number (0));
1663 Vcharset_table
= Fmake_char_table (Qcharset_table
, Qnil
);
1665 Vcharset_symbol_table
= Fmake_vector (make_number (MAX_CHARSET
+ 1), Qnil
);
1668 for (i
= 0; i
< 2; i
++)
1669 for (j
= 0; j
< 2; j
++)
1670 for (k
= 0; k
< 128; k
++)
1671 iso_charset_table
[i
][j
][k
] = -1;
1673 bzero (cmpchar_hash_table
, sizeof cmpchar_hash_table
);
1674 cmpchar_table_size
= n_cmpchars
= 0;
1676 for (i
= 0; i
< 256; i
++)
1677 BYTES_BY_CHAR_HEAD (i
) = 1;
1678 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11
) = 3;
1679 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12
) = 3;
1680 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21
) = 4;
1681 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22
) = 4;
1682 /* The following doesn't reflect the actual bytes, but just to tell
1683 that it is a start of a multibyte character. */
1684 BYTES_BY_CHAR_HEAD (LEADING_CODE_COMPOSITION
) = 2;
1686 for (i
= 0; i
< 128; i
++)
1687 WIDTH_BY_CHAR_HEAD (i
) = 1;
1688 for (; i
< 256; i
++)
1689 WIDTH_BY_CHAR_HEAD (i
) = 4;
1690 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11
) = 1;
1691 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12
) = 2;
1692 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21
) = 1;
1693 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22
) = 2;
1701 Qascii
= intern ("ascii");
1702 staticpro (&Qascii
);
1704 Qcharset
= intern ("charset");
1705 staticpro (&Qcharset
);
1707 /* Define ASCII charset now. */
1708 update_charset_table (make_number (CHARSET_ASCII
),
1709 make_number (1), make_number (94),
1714 build_string ("ASCII"),
1715 build_string ("ASCII"),
1716 build_string ("ASCII (ISO646 IRV)"));
1717 CHARSET_SYMBOL (CHARSET_ASCII
) = Qascii
;
1718 Fput (Qascii
, Qcharset
, CHARSET_TABLE_ENTRY (CHARSET_ASCII
));
1720 Qcomposition
= intern ("composition");
1721 staticpro (&Qcomposition
);
1722 CHARSET_SYMBOL (CHARSET_COMPOSITION
) = Qcomposition
;
1724 defsubr (&Sdefine_charset
);
1725 defsubr (&Sget_unused_iso_final_char
);
1726 defsubr (&Sdeclare_equiv_charset
);
1727 defsubr (&Sfind_charset_region
);
1728 defsubr (&Sfind_charset_string
);
1729 defsubr (&Smake_char_internal
);
1730 defsubr (&Ssplit_char
);
1731 defsubr (&Schar_charset
);
1732 defsubr (&Scharset_after
);
1733 defsubr (&Siso_charset
);
1734 defsubr (&Schar_valid_p
);
1735 defsubr (&Sunibyte_char_to_multibyte
);
1736 defsubr (&Schar_bytes
);
1737 defsubr (&Schar_width
);
1738 defsubr (&Sstring_width
);
1739 defsubr (&Schar_direction
);
1740 defsubr (&Schars_in_region
);
1742 defsubr (&Scmpcharp
);
1743 defsubr (&Scmpchar_component
);
1744 defsubr (&Scmpchar_cmp_rule
);
1745 defsubr (&Scmpchar_cmp_rule_p
);
1746 defsubr (&Scmpchar_cmp_count
);
1747 defsubr (&Scompose_string
);
1748 defsubr (&Ssetup_special_charsets
);
1750 DEFVAR_LISP ("charset-list", &Vcharset_list
,
1751 "List of charsets ever defined.");
1752 Vcharset_list
= Fcons (Qascii
, Qnil
);
1754 DEFVAR_LISP ("character-unification-table-vector",
1755 &Vcharacter_unification_table_vector
,
1756 "Vector of cons cell of a symbol and unification table ever defined.\n\
1757 An ID of a unification table is an index of this vector.");
1758 Vcharacter_unification_table_vector
= Fmake_vector (make_number (16), Qnil
);
1760 DEFVAR_INT ("leading-code-composition", &leading_code_composition
,
1761 "Leading-code of composite characters.");
1762 leading_code_composition
= LEADING_CODE_COMPOSITION
;
1764 DEFVAR_INT ("leading-code-private-11", &leading_code_private_11
,
1765 "Leading-code of private TYPE9N charset of column-width 1.");
1766 leading_code_private_11
= LEADING_CODE_PRIVATE_11
;
1768 DEFVAR_INT ("leading-code-private-12", &leading_code_private_12
,
1769 "Leading-code of private TYPE9N charset of column-width 2.");
1770 leading_code_private_12
= LEADING_CODE_PRIVATE_12
;
1772 DEFVAR_INT ("leading-code-private-21", &leading_code_private_21
,
1773 "Leading-code of private TYPE9Nx9N charset of column-width 1.");
1774 leading_code_private_21
= LEADING_CODE_PRIVATE_21
;
1776 DEFVAR_INT ("leading-code-private-22", &leading_code_private_22
,
1777 "Leading-code of private TYPE9Nx9N charset of column-width 2.");
1778 leading_code_private_22
= LEADING_CODE_PRIVATE_22
;
1780 DEFVAR_INT ("nonascii-insert-offset", &nonascii_insert_offset
,
1781 "Offset for converting non-ASCII unibyte codes 0240...0377 to multibyte.\n\
1782 This is used for converting unibyte text to multibyte,\n\
1783 and for inserting character codes specified by number.\n\n\
1784 Conversion is performed only when multibyte characters are enabled,\n\
1785 and it serves to convert a Latin-1 or similar 8-bit character code\n\
1786 to the corresponding Emacs character code.\n\
1787 If `nonascii-translate-table' is non-nil, it overrides this variable.");
1788 nonascii_insert_offset
= 0;
1790 DEFVAR_LISP ("nonascii-translate-table", &Vnonascii_translate_table
,
1791 "Translate table for converting non-ASCII unibyte codes to multibyte.\n\
1792 This is used for converting unibyte text to multibyte,\n\
1793 and for inserting character codes specified by number.\n\n\
1794 Conversion is performed only when multibyte characters are enabled,\n\
1795 and it serves to convert a Latin-1 or similar 8-bit character code\n\
1796 to the corresponding Emacs character code.\n\n\
1797 If this is nil, `nonascii-insert-offset' is used instead.");
1798 Vnonascii_translate_table
= Qnil
;
1800 DEFVAR_INT ("min-composite-char", &min_composite_char
,
1801 "Minimum character code of a composite character.");
1802 min_composite_char
= MIN_CHAR_COMPOSITION
;