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 translation table ever defined.
78 ID of a translation table is used to index this vector. */
79 Lisp_Object Vtranslation_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_translation_table
;
108 /* List of all possible generic characters. */
109 Lisp_Object Vgeneric_character_list
;
111 #define min(X, Y) ((X) < (Y) ? (X) : (Y))
112 #define max(X, Y) ((X) > (Y) ? (X) : (Y))
115 invalid_character (c
)
118 error ("Invalid character: %o, %d, 0x%x", c
);
122 /* Set STR a pointer to the multi-byte form of the character C. If C
123 is not a composite character, the multi-byte form is set in WORKBUF
124 and STR points WORKBUF. The caller should allocate at least 4-byte
125 area at WORKBUF in advance. Returns the length of the multi-byte
126 form. If C is an invalid character to have a multi-byte form,
129 Use macro `CHAR_STRING (C, WORKBUF, STR)' instead of calling this
130 function directly if C can be an ASCII character. */
133 non_ascii_char_to_string (c
, workbuf
, str
)
135 unsigned char *workbuf
, **str
;
139 if (COMPOSITE_CHAR_P (c
))
141 int cmpchar_id
= COMPOSITE_CHAR_ID (c
);
143 if (cmpchar_id
< n_cmpchars
)
145 *str
= cmpchar_table
[cmpchar_id
]->data
;
146 return cmpchar_table
[cmpchar_id
]->len
;
150 invalid_character (c
);
154 SPLIT_NON_ASCII_CHAR (c
, charset
, c1
, c2
);
156 || ! CHARSET_DEFINED_P (charset
)
157 || c1
>= 0 && c1
< 32
158 || c2
>= 0 && c2
< 32)
159 invalid_character (c
);
162 *workbuf
++ = CHARSET_LEADING_CODE_BASE (charset
);
163 if (*workbuf
= CHARSET_LEADING_CODE_EXT (charset
))
165 *workbuf
++ = c1
| 0x80;
167 *workbuf
++ = c2
| 0x80;
169 return (workbuf
- *str
);
172 /* Return a non-ASCII character of which multi-byte form is at STR of
173 length LEN. If ACTUAL_LEN is not NULL, the actual length of the
174 multibyte form is set to the address ACTUAL_LEN.
176 If exclude_tail_garbage is nonzero, ACTUAL_LEN excludes gabage
177 bytes following the non-ASCII character.
179 Use macro `STRING_CHAR (STR, LEN)' instead of calling this function
180 directly if STR can hold an ASCII character. */
183 string_to_non_ascii_char (str
, len
, actual_len
, exclude_tail_garbage
)
184 const unsigned char *str
;
185 int len
, *actual_len
, exclude_tail_garbage
;
188 unsigned char c1
, c2
;
189 register int c
, bytes
;
194 if (BASE_LEADING_CODE_P (c
))
196 while (bytes
< len
&& ! CHAR_HEAD_P (str
[bytes
])) bytes
++;
198 if (c
== LEADING_CODE_COMPOSITION
)
200 int cmpchar_id
= str_cmpchar_id (str
, bytes
);
204 c
= MAKE_COMPOSITE_CHAR (cmpchar_id
);
205 if (exclude_tail_garbage
)
206 bytes
= cmpchar_table
[cmpchar_id
]->len
;
211 int charset
= c
, c1
, c2
= 0;
212 int char_bytes
= BYTES_BY_CHAR_HEAD (c
);
215 if (c
>= LEADING_CODE_PRIVATE_11
)
217 if (char_bytes
<= bytes
&& CHARSET_DEFINED_P (charset
))
220 if (CHARSET_DIMENSION (charset
) == 2)
222 c
= MAKE_NON_ASCII_CHAR (charset
, c1
, c2
);
223 if (exclude_tail_garbage
)
234 /* Return the length of the multi-byte form at string STR of length LEN. */
236 multibyte_form_length (str
, len
)
237 const unsigned char *str
;
242 if (BASE_LEADING_CODE_P (*str
))
243 while (bytes
< len
&& ! CHAR_HEAD_P (str
[bytes
])) bytes
++;
248 /* Check if string STR of length LEN contains valid multi-byte form of
249 a character. If valid, charset and position codes of the character
250 is set at *CHARSET, *C1, and *C2, and return 0. If not valid,
251 return -1. This should be used only in the macro SPLIT_STRING
252 which checks range of STR in advance. */
255 split_non_ascii_string (str
, len
, charset
, c1
, c2
)
256 register const unsigned char *str
;
257 register unsigned char *c1
, *c2
;
258 register int len
, *charset
;
260 register unsigned int cs
= *str
++;
262 if (cs
== LEADING_CODE_COMPOSITION
)
264 int cmpchar_id
= str_cmpchar_id (str
- 1, len
);
268 *charset
= cs
, *c1
= cmpchar_id
>> 7, *c2
= cmpchar_id
& 0x7F;
270 else if ((cs
< LEADING_CODE_PRIVATE_11
|| (cs
= *str
++) >= 0xA0)
271 && CHARSET_DEFINED_P (cs
))
276 *c1
= (*str
++) & 0x7F;
277 if (CHARSET_DIMENSION (cs
) == 2)
281 *c2
= (*str
++) & 0x7F;
289 /* Translate character C by translation table TABLE. If C
290 is negative, translate a character specified by CHARSET, C1, and C2
291 (C1 and C2 are code points of the character). If no translation is
292 found in TABLE, return C. */
294 translate_char (table
, c
, charset
, c1
, c2
)
296 int c
, charset
, c1
, c2
;
299 int alt_charset
, alt_c1
, alt_c2
, dimension
;
301 if (c
< 0) c
= MAKE_CHAR (charset
, c1
, c2
);
302 if (!CHAR_TABLE_P (table
)
303 || (ch
= Faref (table
, make_number (c
)), !INTEGERP (ch
))
307 SPLIT_CHAR (XFASTINT (ch
), alt_charset
, alt_c1
, alt_c2
);
308 dimension
= CHARSET_DIMENSION (alt_charset
);
309 if (dimension
== 1 && alt_c1
> 0 || dimension
== 2 && alt_c2
> 0)
310 /* CH is not a generic character, just return it. */
311 return XFASTINT (ch
);
313 /* Since CH is a generic character, we must return a specific
314 charater which has the same position codes as C from CH. */
316 SPLIT_CHAR (c
, charset
, c1
, c2
);
317 if (dimension
!= CHARSET_DIMENSION (charset
))
318 /* We can't make such a character because of dimension mismatch. */
320 return MAKE_CHAR (alt_charset
, c1
, c2
);
323 /* Convert the unibyte character C to multibyte based on
324 Vnonascii_translation_table or nonascii_insert_offset. If they can't
325 convert C to a valid multibyte character, convert it based on
326 DEFAULT_NONASCII_INSERT_OFFSET which makes C a Latin-1 character. */
329 unibyte_char_to_multibyte (c
)
332 if (c
>= 0240 && c
< 0400)
336 if (! NILP (Vnonascii_translation_table
))
337 c
= XINT (Faref (Vnonascii_translation_table
, make_number (c
)));
338 else if (nonascii_insert_offset
> 0)
339 c
+= nonascii_insert_offset
;
340 if (c
>= 0240 && (c
< 0400 || ! VALID_MULTIBYTE_CHAR_P (c
)))
341 c
= c_save
+ DEFAULT_NONASCII_INSERT_OFFSET
;
347 /* Convert the multibyte character C to unibyte 8-bit character based
348 on Vnonascii_translation_table or nonascii_insert_offset. If
349 REV_TBL is non-nil, it should be a reverse table of
350 Vnonascii_translation_table, i.e. what given by:
351 Fchar_table_extra_slot (Vnonascii_translation_table, make_number (0)) */
354 multibyte_char_to_unibyte (c
, rev_tbl
)
358 if (!SINGLE_BYTE_CHAR_P (c
))
362 if (! CHAR_TABLE_P (rev_tbl
)
363 && CHAR_TABLE_P (Vnonascii_translation_table
))
364 rev_tbl
= Fchar_table_extra_slot (Vnonascii_translation_table
,
366 if (CHAR_TABLE_P (rev_tbl
))
369 temp
= Faref (rev_tbl
, make_number (c
));
373 else if (nonascii_insert_offset
> 0)
374 c
-= nonascii_insert_offset
;
375 if (c
< 128 || c
>= 256)
376 c
= (c_save
& 0177) + 0200;
383 /* Update the table Vcharset_table with the given arguments (see the
384 document of `define-charset' for the meaning of each argument).
385 Several other table contents are also updated. The caller should
386 check the validity of CHARSET-ID and the remaining arguments in
390 update_charset_table (charset_id
, dimension
, chars
, width
, direction
,
391 iso_final_char
, iso_graphic_plane
,
392 short_name
, long_name
, description
)
393 Lisp_Object charset_id
, dimension
, chars
, width
, direction
;
394 Lisp_Object iso_final_char
, iso_graphic_plane
;
395 Lisp_Object short_name
, long_name
, description
;
397 int charset
= XINT (charset_id
);
399 unsigned char leading_code_base
, leading_code_ext
;
401 if (NILP (CHARSET_TABLE_ENTRY (charset
)))
402 CHARSET_TABLE_ENTRY (charset
)
403 = Fmake_vector (make_number (CHARSET_MAX_IDX
), Qnil
);
405 /* Get byte length of multibyte form, base leading-code, and
406 extended leading-code of the charset. See the comment under the
407 title "GENERAL NOTE on CHARACTER SET (CHARSET)" in charset.h. */
408 bytes
= XINT (dimension
);
409 if (charset
< MIN_CHARSET_PRIVATE_DIMENSION1
)
411 /* Official charset, it doesn't have an extended leading-code. */
412 if (charset
!= CHARSET_ASCII
)
413 bytes
+= 1; /* For a base leading-code. */
414 leading_code_base
= charset
;
415 leading_code_ext
= 0;
419 /* Private charset. */
420 bytes
+= 2; /* For base and extended leading-codes. */
422 = (charset
< LEADING_CODE_EXT_12
423 ? LEADING_CODE_PRIVATE_11
424 : (charset
< LEADING_CODE_EXT_21
425 ? LEADING_CODE_PRIVATE_12
426 : (charset
< LEADING_CODE_EXT_22
427 ? LEADING_CODE_PRIVATE_21
428 : LEADING_CODE_PRIVATE_22
)));
429 leading_code_ext
= charset
;
432 CHARSET_TABLE_INFO (charset
, CHARSET_ID_IDX
) = charset_id
;
433 CHARSET_TABLE_INFO (charset
, CHARSET_BYTES_IDX
) = make_number (bytes
);
434 CHARSET_TABLE_INFO (charset
, CHARSET_DIMENSION_IDX
) = dimension
;
435 CHARSET_TABLE_INFO (charset
, CHARSET_CHARS_IDX
) = chars
;
436 CHARSET_TABLE_INFO (charset
, CHARSET_WIDTH_IDX
) = width
;
437 CHARSET_TABLE_INFO (charset
, CHARSET_DIRECTION_IDX
) = direction
;
438 CHARSET_TABLE_INFO (charset
, CHARSET_LEADING_CODE_BASE_IDX
)
439 = make_number (leading_code_base
);
440 CHARSET_TABLE_INFO (charset
, CHARSET_LEADING_CODE_EXT_IDX
)
441 = make_number (leading_code_ext
);
442 CHARSET_TABLE_INFO (charset
, CHARSET_ISO_FINAL_CHAR_IDX
) = iso_final_char
;
443 CHARSET_TABLE_INFO (charset
, CHARSET_ISO_GRAPHIC_PLANE_IDX
)
445 CHARSET_TABLE_INFO (charset
, CHARSET_SHORT_NAME_IDX
) = short_name
;
446 CHARSET_TABLE_INFO (charset
, CHARSET_LONG_NAME_IDX
) = long_name
;
447 CHARSET_TABLE_INFO (charset
, CHARSET_DESCRIPTION_IDX
) = description
;
448 CHARSET_TABLE_INFO (charset
, CHARSET_PLIST_IDX
) = Qnil
;
451 /* If we have already defined a charset which has the same
452 DIMENSION, CHARS and ISO-FINAL-CHAR but the different
453 DIRECTION, we must update the entry REVERSE-CHARSET of both
454 charsets. If there's no such charset, the value of the entry
458 for (i
= 0; i
<= MAX_CHARSET
; i
++)
459 if (!NILP (CHARSET_TABLE_ENTRY (i
)))
461 if (CHARSET_DIMENSION (i
) == XINT (dimension
)
462 && CHARSET_CHARS (i
) == XINT (chars
)
463 && CHARSET_ISO_FINAL_CHAR (i
) == XINT (iso_final_char
)
464 && CHARSET_DIRECTION (i
) != XINT (direction
))
466 CHARSET_TABLE_INFO (charset
, CHARSET_REVERSE_CHARSET_IDX
)
468 CHARSET_TABLE_INFO (i
, CHARSET_REVERSE_CHARSET_IDX
) = charset_id
;
473 /* No such a charset. */
474 CHARSET_TABLE_INFO (charset
, CHARSET_REVERSE_CHARSET_IDX
)
478 if (charset
!= CHARSET_ASCII
479 && charset
< MIN_CHARSET_PRIVATE_DIMENSION1
)
481 /* Update tables bytes_by_char_head and width_by_char_head. */
482 bytes_by_char_head
[leading_code_base
] = bytes
;
483 width_by_char_head
[leading_code_base
] = XINT (width
);
485 /* Update table emacs_code_class. */
486 emacs_code_class
[charset
] = (bytes
== 2
487 ? EMACS_leading_code_2
489 ? EMACS_leading_code_3
490 : EMACS_leading_code_4
));
493 /* Update table iso_charset_table. */
494 if (ISO_CHARSET_TABLE (dimension
, chars
, iso_final_char
) < 0)
495 ISO_CHARSET_TABLE (dimension
, chars
, iso_final_char
) = charset
;
500 /* Return charset id of CHARSET_SYMBOL, or return -1 if CHARSET_SYMBOL
503 get_charset_id (charset_symbol
)
504 Lisp_Object charset_symbol
;
509 return ((SYMBOLP (charset_symbol
)
510 && (val
= Fget (charset_symbol
, Qcharset
), VECTORP (val
))
511 && (charset
= XINT (XVECTOR (val
)->contents
[CHARSET_ID_IDX
]),
512 CHARSET_VALID_P (charset
)))
516 /* Return an identification number for a new private charset of
517 DIMENSION and WIDTH. If there's no more room for the new charset,
520 get_new_private_charset_id (dimension
, width
)
521 int dimension
, width
;
523 int charset
, from
, to
;
528 from
= LEADING_CODE_EXT_11
, to
= LEADING_CODE_EXT_12
;
530 from
= LEADING_CODE_EXT_12
, to
= LEADING_CODE_EXT_21
;
535 from
= LEADING_CODE_EXT_21
, to
= LEADING_CODE_EXT_22
;
537 from
= LEADING_CODE_EXT_22
, to
= LEADING_CODE_EXT_MAX
+ 1;
540 for (charset
= from
; charset
< to
; charset
++)
541 if (!CHARSET_DEFINED_P (charset
)) break;
543 return make_number (charset
< to
? charset
: 0);
546 DEFUN ("define-charset", Fdefine_charset
, Sdefine_charset
, 3, 3, 0,
547 "Define CHARSET-ID as the identification number of CHARSET with INFO-VECTOR.\n\
548 If CHARSET-ID is nil, it is decided automatically, which means CHARSET is\n\
549 treated as a private charset.\n\
550 INFO-VECTOR is a vector of the format:\n\
551 [DIMENSION CHARS WIDTH DIRECTION ISO-FINAL-CHAR ISO-GRAPHIC-PLANE\n\
552 SHORT-NAME LONG-NAME DESCRIPTION]\n\
553 The meanings of each elements is as follows:\n\
554 DIMENSION (integer) is the number of bytes to represent a character: 1 or 2.\n\
555 CHARS (integer) is the number of characters in a dimension: 94 or 96.\n\
556 WIDTH (integer) is the number of columns a character in the charset\n\
557 occupies on the screen: one of 0, 1, and 2.\n\
559 DIRECTION (integer) is the rendering direction of characters in the\n\
560 charset when rendering. If 0, render from left to right, else\n\
561 render from right to left.\n\
563 ISO-FINAL-CHAR (character) is the final character of the\n\
564 corresponding ISO 2022 charset.\n\
566 ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked\n\
567 while encoding to variants of ISO 2022 coding system, one of the\n\
568 following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).\n\
570 SHORT-NAME (string) is the short name to refer to the charset.\n\
572 LONG-NAME (string) is the long name to refer to the charset.\n\
574 DESCRIPTION (string) is the description string of the charset.")
575 (charset_id
, charset_symbol
, info_vector
)
576 Lisp_Object charset_id
, charset_symbol
, info_vector
;
580 if (!NILP (charset_id
))
581 CHECK_NUMBER (charset_id
, 0);
582 CHECK_SYMBOL (charset_symbol
, 1);
583 CHECK_VECTOR (info_vector
, 2);
585 if (! NILP (charset_id
))
587 if (! CHARSET_VALID_P (XINT (charset_id
)))
588 error ("Invalid CHARSET: %d", XINT (charset_id
));
589 else if (CHARSET_DEFINED_P (XINT (charset_id
)))
590 error ("Already defined charset: %d", XINT (charset_id
));
593 vec
= XVECTOR (info_vector
)->contents
;
594 if (XVECTOR (info_vector
)->size
!= 9
595 || !INTEGERP (vec
[0]) || !(XINT (vec
[0]) == 1 || XINT (vec
[0]) == 2)
596 || !INTEGERP (vec
[1]) || !(XINT (vec
[1]) == 94 || XINT (vec
[1]) == 96)
597 || !INTEGERP (vec
[2]) || !(XINT (vec
[2]) == 1 || XINT (vec
[2]) == 2)
598 || !INTEGERP (vec
[3]) || !(XINT (vec
[3]) == 0 || XINT (vec
[3]) == 1)
599 || !INTEGERP (vec
[4]) || !(XINT (vec
[4]) >= '0' && XINT (vec
[4]) <= '~')
600 || !INTEGERP (vec
[5]) || !(XINT (vec
[5]) == 0 || XINT (vec
[5]) == 1)
603 || !STRINGP (vec
[8]))
604 error ("Invalid info-vector argument for defining charset %s",
605 XSYMBOL (charset_symbol
)->name
->data
);
607 if (NILP (charset_id
))
609 charset_id
= get_new_private_charset_id (XINT (vec
[0]), XINT (vec
[2]));
610 if (XINT (charset_id
) == 0)
611 error ("There's no room for a new private charset %s",
612 XSYMBOL (charset_symbol
)->name
->data
);
615 update_charset_table (charset_id
, vec
[0], vec
[1], vec
[2], vec
[3],
616 vec
[4], vec
[5], vec
[6], vec
[7], vec
[8]);
617 Fput (charset_symbol
, Qcharset
, CHARSET_TABLE_ENTRY (XINT (charset_id
)));
618 CHARSET_SYMBOL (XINT (charset_id
)) = charset_symbol
;
619 Vcharset_list
= Fcons (charset_symbol
, Vcharset_list
);
623 DEFUN ("generic-character-list", Fgeneric_character_list
,
624 Sgeneric_character_list
, 0, 0, 0,
625 "Return a list of all possible generic characters.\n\
626 It includes a generic character for a charset not yet defined.")
629 return Vgeneric_character_list
;
632 DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char
,
633 Sget_unused_iso_final_char
, 2, 2, 0,
634 "Return an unsed ISO's final char for a charset of DIMENISION and CHARS.\n\
635 DIMENSION is the number of bytes to represent a character: 1 or 2.\n\
636 CHARS is the number of characters in a dimension: 94 or 96.\n\
638 This final char is for private use, thus the range is `0' (48) .. `?' (63).\n\
639 If there's no unused final char for the specified kind of charset,\n\
642 Lisp_Object dimension
, chars
;
646 CHECK_NUMBER (dimension
, 0);
647 CHECK_NUMBER (chars
, 1);
648 if (XINT (dimension
) != 1 && XINT (dimension
) != 2)
649 error ("Invalid charset dimension %d, it should be 1 or 2",
651 if (XINT (chars
) != 94 && XINT (chars
) != 96)
652 error ("Invalid charset chars %d, it should be 94 or 96",
654 for (final_char
= '0'; final_char
<= '?'; final_char
++)
656 if (ISO_CHARSET_TABLE (dimension
, chars
, make_number (final_char
)) < 0)
659 return (final_char
<= '?' ? make_number (final_char
) : Qnil
);
662 DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset
, Sdeclare_equiv_charset
,
664 "Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET.\n\
665 CHARSET should be defined by `defined-charset' in advance.")
666 (dimension
, chars
, final_char
, charset_symbol
)
667 Lisp_Object dimension
, chars
, final_char
, charset_symbol
;
671 CHECK_NUMBER (dimension
, 0);
672 CHECK_NUMBER (chars
, 1);
673 CHECK_NUMBER (final_char
, 2);
674 CHECK_SYMBOL (charset_symbol
, 3);
676 if (XINT (dimension
) != 1 && XINT (dimension
) != 2)
677 error ("Invalid DIMENSION %d, it should be 1 or 2", XINT (dimension
));
678 if (XINT (chars
) != 94 && XINT (chars
) != 96)
679 error ("Invalid CHARS %d, it should be 94 or 96", XINT (chars
));
680 if (XINT (final_char
) < '0' || XFASTINT (final_char
) > '~')
681 error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", XINT (chars
));
682 if ((charset
= get_charset_id (charset_symbol
)) < 0)
683 error ("Invalid charset %s", XSYMBOL (charset_symbol
)->name
->data
);
685 ISO_CHARSET_TABLE (dimension
, chars
, final_char
) = charset
;
689 /* Return number of different charsets in STR of length LEN. In
690 addition, for each found charset N, CHARSETS[N] is set 1. The
691 caller should allocate CHARSETS (MAX_CHARSET + 1 elements) in advance.
692 It may lookup a translation table TABLE if supplied.
694 If CMPCHARP is nonzero and some composite character is found,
695 CHARSETS[128] is also set 1 and the returned number is incremented
699 find_charset_in_str (str
, len
, charsets
, table
, cmpcharp
)
705 register int num
= 0, c
;
707 if (! CHAR_TABLE_P (table
))
715 if (c
== LEADING_CODE_COMPOSITION
)
717 int cmpchar_id
= str_cmpchar_id (str
, len
);
722 struct cmpchar_info
*cmpcharp
= cmpchar_table
[cmpchar_id
];
725 for (i
= 0; i
< cmpcharp
->glyph_len
; i
++)
727 c
= cmpcharp
->glyph
[i
];
730 if ((c
= translate_char (table
, c
, 0, 0, 0)) < 0)
731 c
= cmpcharp
->glyph
[i
];
733 if ((charset
= CHAR_CHARSET (c
)) < 0)
734 charset
= CHARSET_ASCII
;
735 if (!charsets
[charset
])
737 charsets
[charset
] = 1;
741 str
+= cmpcharp
->len
;
742 len
-= cmpcharp
->len
;
746 charset
= CHARSET_ASCII
;
751 c
= STRING_CHAR_AND_LENGTH (str
, len
, bytes
);
754 int c1
= translate_char (table
, c
, 0, 0, 0);
758 charset
= CHAR_CHARSET (c
);
761 if (!charsets
[charset
])
763 charsets
[charset
] = 1;
772 DEFUN ("find-charset-region", Ffind_charset_region
, Sfind_charset_region
,
774 "Return a list of charsets in the region between BEG and END.\n\
775 BEG and END are buffer positions.\n\
776 Optional arg TABLE if non-nil is a translation table to look up.")
778 Lisp_Object beg
, end
, table
;
780 int charsets
[MAX_CHARSET
+ 1];
781 int from
, from_byte
, to
, stop
, stop_byte
, i
;
784 validate_region (&beg
, &end
);
785 from
= XFASTINT (beg
);
786 stop
= to
= XFASTINT (end
);
788 if (from
< GPT
&& GPT
< to
)
791 stop_byte
= GPT_BYTE
;
794 stop_byte
= CHAR_TO_BYTE (stop
);
796 from_byte
= CHAR_TO_BYTE (from
);
798 bzero (charsets
, (MAX_CHARSET
+ 1) * sizeof (int));
801 find_charset_in_str (BYTE_POS_ADDR (from_byte
), stop_byte
- from_byte
,
805 from
= stop
, from_byte
= stop_byte
;
806 stop
= to
, stop_byte
= CHAR_TO_BYTE (stop
);
813 for (i
= MAX_CHARSET
; i
>= 0; i
--)
815 val
= Fcons (CHARSET_SYMBOL (i
), val
);
819 DEFUN ("find-charset-string", Ffind_charset_string
, Sfind_charset_string
,
821 "Return a list of charsets in STR.\n\
822 Optional arg TABLE if non-nil is a translation table to look up.")
824 Lisp_Object str
, table
;
826 int charsets
[MAX_CHARSET
+ 1];
830 CHECK_STRING (str
, 0);
832 if (! STRING_MULTIBYTE (str
))
835 bzero (charsets
, (MAX_CHARSET
+ 1) * sizeof (int));
836 find_charset_in_str (XSTRING (str
)->data
, STRING_BYTES (XSTRING (str
)),
839 for (i
= MAX_CHARSET
; i
>= 0; i
--)
841 val
= Fcons (CHARSET_SYMBOL (i
), val
);
845 DEFUN ("make-char-internal", Fmake_char_internal
, Smake_char_internal
, 1, 3, 0,
847 (charset
, code1
, code2
)
848 Lisp_Object charset
, code1
, code2
;
850 CHECK_NUMBER (charset
, 0);
853 XSETFASTINT (code1
, 0);
855 CHECK_NUMBER (code1
, 1);
857 XSETFASTINT (code2
, 0);
859 CHECK_NUMBER (code2
, 2);
861 if (!CHARSET_DEFINED_P (XINT (charset
)))
862 error ("Invalid charset: %d", XINT (charset
));
864 return make_number (MAKE_CHAR (XINT (charset
), XINT (code1
), XINT (code2
)));
867 DEFUN ("split-char", Fsplit_char
, Ssplit_char
, 1, 1, 0,
868 "Return list of charset and one or two position-codes of CHAR.")
875 CHECK_NUMBER (ch
, 0);
876 SPLIT_CHAR (XFASTINT (ch
), charset
, c1
, c2
);
878 ? Fcons (CHARSET_SYMBOL (charset
),
879 Fcons (make_number (c1
), Fcons (make_number (c2
), Qnil
)))
880 : Fcons (CHARSET_SYMBOL (charset
), Fcons (make_number (c1
), Qnil
)));
883 DEFUN ("char-charset", Fchar_charset
, Schar_charset
, 1, 1, 0,
884 "Return charset of CHAR.")
888 CHECK_NUMBER (ch
, 0);
890 return CHARSET_SYMBOL (CHAR_CHARSET (XINT (ch
)));
893 DEFUN ("charset-after", Fcharset_after
, Scharset_after
, 0, 1, 0,
894 "Return charset of a character in current buffer at position POS.\n\
895 If POS is nil, it defauls to the current point.")
899 register int pos_byte
, c
, charset
;
900 register unsigned char *p
;
904 else if (MARKERP (pos
))
905 pos_byte
= marker_byte_position (pos
);
908 CHECK_NUMBER (pos
, 0);
909 pos_byte
= CHAR_TO_BYTE (XINT (pos
));
911 p
= BYTE_POS_ADDR (pos_byte
);
912 c
= STRING_CHAR (p
, Z_BYTE
- pos_byte
);
913 charset
= CHAR_CHARSET (c
);
914 return CHARSET_SYMBOL (charset
);
917 DEFUN ("iso-charset", Fiso_charset
, Siso_charset
, 3, 3, 0,
918 "Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR.\n\
920 ISO 2022's designation sequence (escape sequence) distinguishes charsets\n\
921 by their DIMENSION, CHARS, and FINAL-CHAR,\n\
922 where as Emacs distinguishes them by charset symbol.\n\
923 See the documentation of the function `charset-info' for the meanings of\n\
924 DIMENSION, CHARS, and FINAL-CHAR.")
925 (dimension
, chars
, final_char
)
926 Lisp_Object dimension
, chars
, final_char
;
930 CHECK_NUMBER (dimension
, 0);
931 CHECK_NUMBER (chars
, 1);
932 CHECK_NUMBER (final_char
, 2);
934 if ((charset
= ISO_CHARSET_TABLE (dimension
, chars
, final_char
)) < 0)
936 return CHARSET_SYMBOL (charset
);
939 /* If GENERICP is nonzero, return nonzero iff C is a valid normal or
940 generic character. If GENERICP is zero, return nonzero iff C is a
941 valid normal character. Do not call this function directly,
942 instead use macro CHAR_VALID_P. */
944 char_valid_p (c
, genericp
)
951 if (SINGLE_BYTE_CHAR_P (c
))
953 SPLIT_NON_ASCII_CHAR (c
, charset
, c1
, c2
);
954 if (!CHARSET_VALID_P (charset
))
956 return (c
< MIN_CHAR_COMPOSITION
957 ? ((c
& CHAR_FIELD1_MASK
) /* i.e. dimension of C is two. */
958 ? (genericp
&& c1
== 0 && c2
== 0
959 || c1
>= 32 && c2
>= 32)
960 : (genericp
&& c1
== 0
962 : c
< MIN_CHAR_COMPOSITION
+ n_cmpchars
);
965 DEFUN ("char-valid-p", Fchar_valid_p
, Schar_valid_p
, 1, 2, 0,
966 "Return t if OBJECT is a valid normal character.\n\
967 If optional arg GENERICP is non-nil, also return t if OBJECT is\n\
968 a valid generic character.")
970 Lisp_Object object
, genericp
;
972 if (! NATNUMP (object
))
974 return (CHAR_VALID_P (XFASTINT (object
), !NILP (genericp
)) ? Qt
: Qnil
);
977 DEFUN ("unibyte-char-to-multibyte", Funibyte_char_to_multibyte
,
978 Sunibyte_char_to_multibyte
, 1, 1, 0,
979 "Convert the unibyte character CH to multibyte character.\n\
980 The conversion is done based on `nonascii-translation-table' (which see)\n\
981 or `nonascii-insert-offset' (which see).")
987 CHECK_NUMBER (ch
, 0);
989 if (c
< 0 || c
>= 0400)
990 error ("Invalid unibyte character: %d", c
);
991 c
= unibyte_char_to_multibyte (c
);
993 error ("Can't convert to multibyte character: %d", XINT (ch
));
994 return make_number (c
);
997 DEFUN ("char-bytes", Fchar_bytes
, Schar_bytes
, 1, 1, 0,
998 "Return byte length of multi-byte form of CHAR.")
1005 CHECK_NUMBER (ch
, 0);
1006 if (COMPOSITE_CHAR_P (XFASTINT (ch
)))
1008 unsigned int id
= COMPOSITE_CHAR_ID (XFASTINT (ch
));
1010 bytes
= (id
< n_cmpchars
? cmpchar_table
[id
]->len
: 1);
1014 int charset
= CHAR_CHARSET (XFASTINT (ch
));
1016 bytes
= CHARSET_DEFINED_P (charset
) ? CHARSET_BYTES (charset
) : 1;
1019 XSETFASTINT (val
, bytes
);
1023 /* Return the width of character of which multi-byte form starts with
1024 C. The width is measured by how many columns occupied on the
1025 screen when displayed in the current buffer. */
1027 #define ONE_BYTE_CHAR_WIDTH(c) \
1030 ? XFASTINT (current_buffer->tab_width) \
1031 : (c == '\n' ? 0 : (NILP (current_buffer->ctl_arrow) ? 4 : 2))) \
1035 ? (NILP (current_buffer->ctl_arrow) ? 4 : 2) \
1036 : ((! NILP (current_buffer->enable_multibyte_characters) \
1037 && BASE_LEADING_CODE_P (c)) \
1038 ? WIDTH_BY_CHAR_HEAD (c) \
1041 DEFUN ("char-width", Fchar_width
, Schar_width
, 1, 1, 0,
1042 "Return width of CHAR when displayed in the current buffer.\n\
1043 The width is measured by how many columns it occupies on the screen.")
1047 Lisp_Object val
, disp
;
1049 struct Lisp_Char_Table
*dp
= buffer_display_table ();
1051 CHECK_NUMBER (ch
, 0);
1055 /* Get the way the display table would display it. */
1056 disp
= dp
? DISP_CHAR_VECTOR (dp
, c
) : Qnil
;
1059 XSETINT (val
, XVECTOR (disp
)->size
);
1060 else if (SINGLE_BYTE_CHAR_P (c
))
1061 XSETINT (val
, ONE_BYTE_CHAR_WIDTH (c
));
1062 else if (COMPOSITE_CHAR_P (c
))
1064 int id
= COMPOSITE_CHAR_ID (XFASTINT (ch
));
1065 XSETFASTINT (val
, (id
< n_cmpchars
? cmpchar_table
[id
]->width
: 0));
1069 int charset
= CHAR_CHARSET (c
);
1071 XSETFASTINT (val
, CHARSET_WIDTH (charset
));
1076 /* Return width of string STR of length LEN when displayed in the
1077 current buffer. The width is measured by how many columns it
1078 occupies on the screen. */
1085 unsigned char *endp
= str
+ len
;
1087 struct Lisp_Char_Table
*dp
= buffer_display_table ();
1091 if (*str
== LEADING_CODE_COMPOSITION
)
1093 int id
= str_cmpchar_id (str
, endp
- str
);
1102 width
+= cmpchar_table
[id
]->width
;
1103 str
+= cmpchar_table
[id
]->len
;
1110 int c
= STRING_CHAR_AND_LENGTH (str
, endp
- str
, thislen
);
1112 /* Get the way the display table would display it. */
1114 disp
= DISP_CHAR_VECTOR (dp
, c
);
1119 width
+= XVECTOR (disp
)->size
;
1121 width
+= ONE_BYTE_CHAR_WIDTH (*str
);
1129 DEFUN ("string-width", Fstring_width
, Sstring_width
, 1, 1, 0,
1130 "Return width of STRING when displayed in the current buffer.\n\
1131 Width is measured by how many columns it occupies on the screen.\n\
1132 When calculating width of a multibyte character in STRING,\n\
1133 only the base leading-code is considered; the validity of\n\
1134 the following bytes is not checked.")
1140 CHECK_STRING (str
, 0);
1141 XSETFASTINT (val
, strwidth (XSTRING (str
)->data
,
1142 STRING_BYTES (XSTRING (str
))));
1146 DEFUN ("char-direction", Fchar_direction
, Schar_direction
, 1, 1, 0,
1147 "Return the direction of CHAR.\n\
1148 The returned value is 0 for left-to-right and 1 for right-to-left.")
1154 CHECK_NUMBER (ch
, 0);
1155 charset
= CHAR_CHARSET (XFASTINT (ch
));
1156 if (!CHARSET_DEFINED_P (charset
))
1157 invalid_character (XINT (ch
));
1158 return CHARSET_TABLE_INFO (charset
, CHARSET_DIRECTION_IDX
);
1161 DEFUN ("chars-in-region", Fchars_in_region
, Schars_in_region
, 2, 2, 0,
1162 "Return number of characters between BEG and END.")
1164 Lisp_Object beg
, end
;
1168 CHECK_NUMBER_COERCE_MARKER (beg
, 0);
1169 CHECK_NUMBER_COERCE_MARKER (end
, 1);
1171 from
= min (XFASTINT (beg
), XFASTINT (end
));
1172 to
= max (XFASTINT (beg
), XFASTINT (end
));
1174 return make_number (to
- from
);
1177 /* Return the number of characters in the NBYTES bytes at PTR.
1178 This works by looking at the contents and checking for multibyte sequences.
1179 However, if the current buffer has enable-multibyte-characters = nil,
1180 we treat each byte as a character. */
1183 chars_in_text (ptr
, nbytes
)
1187 unsigned char *endp
, c
;
1190 /* current_buffer is null at early stages of Emacs initialization. */
1191 if (current_buffer
== 0
1192 || NILP (current_buffer
->enable_multibyte_characters
))
1195 endp
= ptr
+ nbytes
;
1202 if (BASE_LEADING_CODE_P (c
))
1203 while (ptr
< endp
&& ! CHAR_HEAD_P (*ptr
)) ptr
++;
1210 /* Return the number of characters in the NBYTES bytes at PTR.
1211 This works by looking at the contents and checking for multibyte sequences.
1212 It ignores enable-multibyte-characters. */
1215 multibyte_chars_in_text (ptr
, nbytes
)
1219 unsigned char *endp
, c
;
1222 endp
= ptr
+ nbytes
;
1229 if (BASE_LEADING_CODE_P (c
))
1230 while (ptr
< endp
&& ! CHAR_HEAD_P (*ptr
)) ptr
++;
1237 DEFUN ("string", Fstring
, Sstring
, 1, MANY
, 0,
1238 "Concatenate all the argument characters and make the result a string.")
1245 = (unsigned char *) alloca (MAX_LENGTH_OF_MULTI_BYTE_FORM
* n
);
1246 unsigned char *p
= buf
;
1249 for (i
= 0; i
< n
; i
++)
1254 if (!INTEGERP (args
[i
]))
1255 CHECK_NUMBER (args
[i
], 0);
1257 len
= CHAR_STRING (c
, p
, str
);
1259 /* C is a composite character. */
1260 bcopy (str
, p
, len
);
1264 val
= make_string_from_bytes (buf
, n
, p
- buf
);
1270 /*** Composite characters staffs ***/
1272 /* Each composite character is identified by CMPCHAR-ID which is
1273 assigned when Emacs needs the character code of the composite
1274 character (e.g. when displaying it on the screen). See the
1275 document "GENERAL NOTE on COMPOSITE CHARACTER" in `charset.h' how a
1276 composite character is represented in Emacs. */
1278 /* If `static' is defined, it means that it is defined to null string. */
1280 /* The following function is copied from lread.c. */
1282 hash_string (ptr
, len
)
1286 register unsigned char *p
= ptr
;
1287 register unsigned char *end
= p
+ len
;
1288 register unsigned char c
;
1289 register int hash
= 0;
1294 if (c
>= 0140) c
-= 40;
1295 hash
= ((hash
<<3) + (hash
>>28) + c
);
1297 return hash
& 07777777777;
1301 #define CMPCHAR_HASH_TABLE_SIZE 0xFFF
1303 static int *cmpchar_hash_table
[CMPCHAR_HASH_TABLE_SIZE
];
1305 /* Each element of `cmpchar_hash_table' is a pointer to an array of
1306 integer, where the 1st element is the size of the array, the 2nd
1307 element is how many elements are actually used in the array, and
1308 the remaining elements are CMPCHAR-IDs of composite characters of
1309 the same hash value. */
1310 #define CMPCHAR_HASH_SIZE(table) table[0]
1311 #define CMPCHAR_HASH_USED(table) table[1]
1312 #define CMPCHAR_HASH_CMPCHAR_ID(table, i) table[i]
1314 /* Return CMPCHAR-ID of the composite character in STR of the length
1315 LEN. If the composite character has not yet been registered,
1316 register it in `cmpchar_table' and assign new CMPCHAR-ID. This
1317 is the sole function for assigning CMPCHAR-ID. */
1319 str_cmpchar_id (str
, len
)
1320 const unsigned char *str
;
1323 int hash_idx
, *hashp
;
1325 int embedded_rule
; /* 1 if composition rule is embedded. */
1326 int chars
; /* number of components. */
1328 struct cmpchar_info
*cmpcharp
;
1330 /* The second byte 0xFF means compostion rule is embedded. */
1331 embedded_rule
= (str
[1] == 0xFF);
1333 /* At first, get the actual length of the composite character. */
1335 const unsigned char *p
, *endp
= str
+ 1, *lastp
= str
+ len
;
1338 while (endp
< lastp
&& ! CHAR_HEAD_P (*endp
)) endp
++;
1340 /* Any composite char have at least 5-byte length. */
1347 if (embedded_rule
) p
++;
1348 /* No need of checking if *P is 0xA0 because
1349 BYTES_BY_CHAR_HEAD (0x80) surely returns 2. */
1350 p
+= BYTES_BY_CHAR_HEAD (*p
- 0x20);
1353 if (p
> endp
|| chars
< 2 || chars
> MAX_COMPONENT_COUNT
)
1354 /* Invalid components. */
1358 hash_idx
= hash_string (str
, len
) % CMPCHAR_HASH_TABLE_SIZE
;
1359 hashp
= cmpchar_hash_table
[hash_idx
];
1361 /* Then, look into the hash table. */
1363 /* Find the correct one among composite characters of the same
1365 for (i
= 2; i
< CMPCHAR_HASH_USED (hashp
); i
++)
1367 cmpcharp
= cmpchar_table
[CMPCHAR_HASH_CMPCHAR_ID (hashp
, i
)];
1368 if (len
== cmpcharp
->len
1369 && ! bcmp (str
, cmpcharp
->data
, len
))
1370 return CMPCHAR_HASH_CMPCHAR_ID (hashp
, i
);
1373 /* We have to register the composite character in cmpchar_table. */
1374 if (n_cmpchars
> (CHAR_FIELD2_MASK
| CHAR_FIELD3_MASK
))
1375 /* No, we have no more room for a new composite character. */
1378 /* Make the entry in hash table. */
1381 /* Make a table for 8 composite characters initially. */
1382 hashp
= (cmpchar_hash_table
[hash_idx
]
1383 = (int *) xmalloc (sizeof (int) * (2 + 8)));
1384 CMPCHAR_HASH_SIZE (hashp
) = 10;
1385 CMPCHAR_HASH_USED (hashp
) = 2;
1387 else if (CMPCHAR_HASH_USED (hashp
) >= CMPCHAR_HASH_SIZE (hashp
))
1389 CMPCHAR_HASH_SIZE (hashp
) += 8;
1390 hashp
= (cmpchar_hash_table
[hash_idx
]
1391 = (int *) xrealloc (hashp
,
1392 sizeof (int) * CMPCHAR_HASH_SIZE (hashp
)));
1394 CMPCHAR_HASH_CMPCHAR_ID (hashp
, CMPCHAR_HASH_USED (hashp
)) = n_cmpchars
;
1395 CMPCHAR_HASH_USED (hashp
)++;
1397 /* Set information of the composite character in cmpchar_table. */
1398 if (cmpchar_table_size
== 0)
1400 /* This is the first composite character to be registered. */
1401 cmpchar_table_size
= 256;
1403 = (struct cmpchar_info
**) xmalloc (sizeof (cmpchar_table
[0])
1404 * cmpchar_table_size
);
1406 else if (cmpchar_table_size
<= n_cmpchars
)
1408 cmpchar_table_size
+= 256;
1410 = (struct cmpchar_info
**) xrealloc (cmpchar_table
,
1411 sizeof (cmpchar_table
[0])
1412 * cmpchar_table_size
);
1415 cmpcharp
= (struct cmpchar_info
*) xmalloc (sizeof (struct cmpchar_info
));
1417 cmpcharp
->len
= len
;
1418 cmpcharp
->data
= (unsigned char *) xmalloc (len
+ 1);
1419 bcopy (str
, cmpcharp
->data
, len
);
1420 cmpcharp
->data
[len
] = 0;
1421 cmpcharp
->glyph_len
= chars
;
1422 cmpcharp
->glyph
= (GLYPH
*) xmalloc (sizeof (GLYPH
) * chars
);
1425 cmpcharp
->cmp_rule
= (unsigned char *) xmalloc (chars
);
1426 cmpcharp
->col_offset
= (float *) xmalloc (sizeof (float) * chars
);
1430 cmpcharp
->cmp_rule
= NULL
;
1431 cmpcharp
->col_offset
= NULL
;
1434 /* Setup GLYPH data and composition rules (if any) so as not to make
1435 them every time on displaying. */
1437 unsigned char *bufp
;
1439 float leftmost
= 0.0, rightmost
= 1.0;
1442 /* At first, col_offset[N] is set to relative to col_offset[0]. */
1443 cmpcharp
->col_offset
[0] = 0;
1445 for (i
= 0, bufp
= cmpcharp
->data
+ 1; i
< chars
; i
++)
1448 cmpcharp
->cmp_rule
[i
] = *bufp
++;
1450 if (*bufp
== 0xA0) /* This is an ASCII character. */
1452 cmpcharp
->glyph
[i
] = FAST_MAKE_GLYPH ((*++bufp
& 0x7F), 0);
1456 else /* Multibyte character. */
1458 /* Make `bufp' point normal multi-byte form temporally. */
1461 = FAST_MAKE_GLYPH (string_to_non_ascii_char (bufp
, 4, 0, 0), 0);
1462 width
= WIDTH_BY_CHAR_HEAD (*bufp
);
1464 bufp
+= BYTES_BY_CHAR_HEAD (*bufp
- 0x20);
1467 if (embedded_rule
&& i
> 0)
1469 /* Reference points (global_ref and new_ref) are
1480 Now, we calculate the column offset of the new glyph
1481 from the left edge of the first glyph. This can avoid
1482 the same calculation everytime displaying this
1483 composite character. */
1485 /* Reference points of global glyph and new glyph. */
1486 int global_ref
= (cmpcharp
->cmp_rule
[i
] - 0xA0) / 9;
1487 int new_ref
= (cmpcharp
->cmp_rule
[i
] - 0xA0) % 9;
1488 /* Column offset relative to the first glyph. */
1489 float left
= (leftmost
1490 + (global_ref
% 3) * (rightmost
- leftmost
) / 2.0
1491 - (new_ref
% 3) * width
/ 2.0);
1493 cmpcharp
->col_offset
[i
] = left
;
1494 if (left
< leftmost
)
1496 if (left
+ width
> rightmost
)
1497 rightmost
= left
+ width
;
1501 if (width
> rightmost
)
1507 /* Now col_offset[N] are relative to the left edge of the
1508 first component. Make them relative to the left edge of
1510 for (i
= 0; i
< chars
; i
++)
1511 cmpcharp
->col_offset
[i
] -= leftmost
;
1512 /* Make rightmost holds width of overall glyph. */
1513 rightmost
-= leftmost
;
1516 cmpcharp
->width
= rightmost
;
1517 if (cmpcharp
->width
< rightmost
)
1518 /* To get a ceiling integer value. */
1522 cmpchar_table
[n_cmpchars
] = cmpcharp
;
1524 return n_cmpchars
++;
1527 /* Return the Nth element of the composite character C. */
1529 cmpchar_component (c
, n
)
1532 int id
= COMPOSITE_CHAR_ID (c
);
1534 if (id
>= n_cmpchars
/* C is not a valid composite character. */
1535 || n
>= cmpchar_table
[id
]->glyph_len
) /* No such component. */
1537 /* No face data is stored in glyph code. */
1538 return ((int) (cmpchar_table
[id
]->glyph
[n
]));
1541 DEFUN ("cmpcharp", Fcmpcharp
, Scmpcharp
, 1, 1, 0,
1542 "T if CHAR is a composite character.")
1546 CHECK_NUMBER (ch
, 0);
1547 return (COMPOSITE_CHAR_P (XINT (ch
)) ? Qt
: Qnil
);
1550 DEFUN ("composite-char-component", Fcmpchar_component
, Scmpchar_component
,
1552 "Return the IDXth component character of composite character CHARACTER.")
1554 Lisp_Object character
, idx
;
1558 CHECK_NUMBER (character
, 0);
1559 CHECK_NUMBER (idx
, 1);
1561 if ((c
= cmpchar_component (XINT (character
), XINT (idx
))) < 0)
1562 args_out_of_range (character
, idx
);
1564 return make_number (c
);
1567 DEFUN ("composite-char-composition-rule", Fcmpchar_cmp_rule
, Scmpchar_cmp_rule
,
1569 "Return the Nth composition rule embedded in composite character CHARACTER.\n\
1570 The returned rule is for composing the Nth component\n\
1571 on the (N-1)th component. If N is 0, the returned value is always 255.")
1573 Lisp_Object character
, n
;
1577 CHECK_NUMBER (character
, 0);
1578 CHECK_NUMBER (n
, 1);
1580 id
= COMPOSITE_CHAR_ID (XINT (character
));
1581 if (id
< 0 || id
>= n_cmpchars
)
1582 error ("Invalid composite character: %d", XINT (character
));
1584 if (i
> cmpchar_table
[id
]->glyph_len
)
1585 args_out_of_range (character
, n
);
1587 return make_number (cmpchar_table
[id
]->cmp_rule
[i
]);
1590 DEFUN ("composite-char-composition-rule-p", Fcmpchar_cmp_rule_p
,
1591 Scmpchar_cmp_rule_p
, 1, 1, 0,
1592 "Return non-nil if composite character CHARACTER contains a embedded rule.")
1594 Lisp_Object character
;
1598 CHECK_NUMBER (character
, 0);
1599 id
= COMPOSITE_CHAR_ID (XINT (character
));
1600 if (id
< 0 || id
>= n_cmpchars
)
1601 error ("Invalid composite character: %d", XINT (character
));
1603 return (cmpchar_table
[id
]->cmp_rule
? Qt
: Qnil
);
1606 DEFUN ("composite-char-component-count", Fcmpchar_cmp_count
,
1607 Scmpchar_cmp_count
, 1, 1, 0,
1608 "Return number of compoents of composite character CHARACTER.")
1610 Lisp_Object character
;
1614 CHECK_NUMBER (character
, 0);
1615 id
= COMPOSITE_CHAR_ID (XINT (character
));
1616 if (id
< 0 || id
>= n_cmpchars
)
1617 error ("Invalid composite character: %d", XINT (character
));
1619 return (make_number (cmpchar_table
[id
]->glyph_len
));
1622 DEFUN ("compose-string", Fcompose_string
, Scompose_string
,
1624 "Return one char string composed from all characters in STRING.")
1628 unsigned char buf
[MAX_LENGTH_OF_MULTI_BYTE_FORM
], *p
, *pend
, *ptemp
;
1631 CHECK_STRING (str
, 0);
1633 buf
[0] = LEADING_CODE_COMPOSITION
;
1634 p
= XSTRING (str
)->data
;
1635 pend
= p
+ STRING_BYTES (XSTRING (str
));
1639 if (*p
< 0x20 || *p
== 127) /* control code */
1640 error ("Invalid component character: %d", *p
);
1641 else if (*p
< 0x80) /* ASCII */
1643 if (i
+ 2 >= MAX_LENGTH_OF_MULTI_BYTE_FORM
)
1644 error ("Too long string to be composed: %s", XSTRING (str
)->data
);
1645 /* Prepend an ASCII charset indicator 0xA0, set MSB of the
1648 buf
[i
++] = *p
++ + 0x80;
1650 else if (*p
== LEADING_CODE_COMPOSITION
) /* composite char */
1652 /* Already composed. Eliminate the heading
1653 LEADING_CODE_COMPOSITION, keep the remaining bytes
1657 while (! CHAR_HEAD_P (*p
)) p
++;
1658 if (i
+ (p
- ptemp
) >= MAX_LENGTH_OF_MULTI_BYTE_FORM
)
1659 error ("Too long string to be composed: %s", XSTRING (str
)->data
);
1660 bcopy (ptemp
, buf
+ i
, p
- ptemp
);
1663 else /* multibyte char */
1665 /* Add 0x20 to the base leading-code, keep the remaining
1667 len
= BYTES_BY_CHAR_HEAD (*p
);
1668 if (i
+ len
>= MAX_LENGTH_OF_MULTI_BYTE_FORM
)
1669 error ("Too long string to be composed: %s", XSTRING (str
)->data
);
1670 bcopy (p
, buf
+ i
, len
);
1677 /* STR contains only one character, which can't be composed. */
1678 error ("Too short string to be composed: %s", XSTRING (str
)->data
);
1680 return make_string_from_bytes (buf
, 1, i
);
1685 charset_id_internal (charset_name
)
1690 val
= Fget (intern (charset_name
), Qcharset
);
1692 error ("Charset %s is not defined", charset_name
);
1694 return (XINT (XVECTOR (val
)->contents
[0]));
1697 DEFUN ("setup-special-charsets", Fsetup_special_charsets
,
1698 Ssetup_special_charsets
, 0, 0, 0, "Internal use only.")
1701 charset_latin_iso8859_1
= charset_id_internal ("latin-iso8859-1");
1702 charset_jisx0208_1978
= charset_id_internal ("japanese-jisx0208-1978");
1703 charset_jisx0208
= charset_id_internal ("japanese-jisx0208");
1704 charset_katakana_jisx0201
= charset_id_internal ("katakana-jisx0201");
1705 charset_latin_jisx0201
= charset_id_internal ("latin-jisx0201");
1706 charset_big5_1
= charset_id_internal ("chinese-big5-1");
1707 charset_big5_2
= charset_id_internal ("chinese-big5-2");
1712 init_charset_once ()
1716 staticpro (&Vcharset_table
);
1717 staticpro (&Vcharset_symbol_table
);
1718 staticpro (&Vgeneric_character_list
);
1720 /* This has to be done here, before we call Fmake_char_table. */
1721 Qcharset_table
= intern ("charset-table");
1722 staticpro (&Qcharset_table
);
1724 /* Intern this now in case it isn't already done.
1725 Setting this variable twice is harmless.
1726 But don't staticpro it here--that is done in alloc.c. */
1727 Qchar_table_extra_slots
= intern ("char-table-extra-slots");
1729 /* Now we are ready to set up this property, so we can
1730 create the charset table. */
1731 Fput (Qcharset_table
, Qchar_table_extra_slots
, make_number (0));
1732 Vcharset_table
= Fmake_char_table (Qcharset_table
, Qnil
);
1734 Vcharset_symbol_table
= Fmake_vector (make_number (MAX_CHARSET
+ 1), Qnil
);
1737 for (i
= 0; i
< 2; i
++)
1738 for (j
= 0; j
< 2; j
++)
1739 for (k
= 0; k
< 128; k
++)
1740 iso_charset_table
[i
][j
][k
] = -1;
1742 bzero (cmpchar_hash_table
, sizeof cmpchar_hash_table
);
1743 cmpchar_table_size
= n_cmpchars
= 0;
1745 for (i
= 0; i
< 256; i
++)
1746 BYTES_BY_CHAR_HEAD (i
) = 1;
1747 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11
) = 3;
1748 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12
) = 3;
1749 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21
) = 4;
1750 BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22
) = 4;
1751 /* The following doesn't reflect the actual bytes, but just to tell
1752 that it is a start of a multibyte character. */
1753 BYTES_BY_CHAR_HEAD (LEADING_CODE_COMPOSITION
) = 2;
1755 for (i
= 0; i
< 128; i
++)
1756 WIDTH_BY_CHAR_HEAD (i
) = 1;
1757 for (; i
< 256; i
++)
1758 WIDTH_BY_CHAR_HEAD (i
) = 4;
1759 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11
) = 1;
1760 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12
) = 2;
1761 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21
) = 1;
1762 WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22
) = 2;
1768 for (i
= 0x81; i
< 0x90; i
++)
1769 val
= Fcons (make_number ((i
- 0x70) << 7), val
);
1770 for (; i
< 0x9A; i
++)
1771 val
= Fcons (make_number ((i
- 0x8F) << 14), val
);
1772 for (i
= 0xA0; i
< 0xF0; i
++)
1773 val
= Fcons (make_number ((i
- 0x70) << 7), val
);
1774 for (; i
< 0xFF; i
++)
1775 val
= Fcons (make_number ((i
- 0xE0) << 14), val
);
1776 val
= Fcons (make_number (GENERIC_COMPOSITION_CHAR
), val
);
1777 Vgeneric_character_list
= Fnreverse (val
);
1786 Qascii
= intern ("ascii");
1787 staticpro (&Qascii
);
1789 Qcharset
= intern ("charset");
1790 staticpro (&Qcharset
);
1792 /* Define ASCII charset now. */
1793 update_charset_table (make_number (CHARSET_ASCII
),
1794 make_number (1), make_number (94),
1799 build_string ("ASCII"),
1800 build_string ("ASCII"),
1801 build_string ("ASCII (ISO646 IRV)"));
1802 CHARSET_SYMBOL (CHARSET_ASCII
) = Qascii
;
1803 Fput (Qascii
, Qcharset
, CHARSET_TABLE_ENTRY (CHARSET_ASCII
));
1805 Qcomposition
= intern ("composition");
1806 staticpro (&Qcomposition
);
1807 CHARSET_SYMBOL (CHARSET_COMPOSITION
) = Qcomposition
;
1809 defsubr (&Sdefine_charset
);
1810 defsubr (&Sgeneric_character_list
);
1811 defsubr (&Sget_unused_iso_final_char
);
1812 defsubr (&Sdeclare_equiv_charset
);
1813 defsubr (&Sfind_charset_region
);
1814 defsubr (&Sfind_charset_string
);
1815 defsubr (&Smake_char_internal
);
1816 defsubr (&Ssplit_char
);
1817 defsubr (&Schar_charset
);
1818 defsubr (&Scharset_after
);
1819 defsubr (&Siso_charset
);
1820 defsubr (&Schar_valid_p
);
1821 defsubr (&Sunibyte_char_to_multibyte
);
1822 defsubr (&Schar_bytes
);
1823 defsubr (&Schar_width
);
1824 defsubr (&Sstring_width
);
1825 defsubr (&Schar_direction
);
1826 defsubr (&Schars_in_region
);
1828 defsubr (&Scmpcharp
);
1829 defsubr (&Scmpchar_component
);
1830 defsubr (&Scmpchar_cmp_rule
);
1831 defsubr (&Scmpchar_cmp_rule_p
);
1832 defsubr (&Scmpchar_cmp_count
);
1833 defsubr (&Scompose_string
);
1834 defsubr (&Ssetup_special_charsets
);
1836 DEFVAR_LISP ("charset-list", &Vcharset_list
,
1837 "List of charsets ever defined.");
1838 Vcharset_list
= Fcons (Qascii
, Qnil
);
1840 DEFVAR_LISP ("translation-table-vector", &Vtranslation_table_vector
,
1841 "Vector of cons cell of a symbol and translation table ever defined.\n\
1842 An ID of a translation table is an index of this vector.");
1843 Vtranslation_table_vector
= Fmake_vector (make_number (16), Qnil
);
1845 DEFVAR_INT ("leading-code-composition", &leading_code_composition
,
1846 "Leading-code of composite characters.");
1847 leading_code_composition
= LEADING_CODE_COMPOSITION
;
1849 DEFVAR_INT ("leading-code-private-11", &leading_code_private_11
,
1850 "Leading-code of private TYPE9N charset of column-width 1.");
1851 leading_code_private_11
= LEADING_CODE_PRIVATE_11
;
1853 DEFVAR_INT ("leading-code-private-12", &leading_code_private_12
,
1854 "Leading-code of private TYPE9N charset of column-width 2.");
1855 leading_code_private_12
= LEADING_CODE_PRIVATE_12
;
1857 DEFVAR_INT ("leading-code-private-21", &leading_code_private_21
,
1858 "Leading-code of private TYPE9Nx9N charset of column-width 1.");
1859 leading_code_private_21
= LEADING_CODE_PRIVATE_21
;
1861 DEFVAR_INT ("leading-code-private-22", &leading_code_private_22
,
1862 "Leading-code of private TYPE9Nx9N charset of column-width 2.");
1863 leading_code_private_22
= LEADING_CODE_PRIVATE_22
;
1865 DEFVAR_INT ("nonascii-insert-offset", &nonascii_insert_offset
,
1866 "Offset for converting non-ASCII unibyte codes 0240...0377 to multibyte.\n\
1867 This is used for converting unibyte text to multibyte,\n\
1868 and for inserting character codes specified by number.\n\n\
1869 This serves to convert a Latin-1 or similar 8-bit character code\n\
1870 to the corresponding Emacs multibyte character code.\n\
1871 Typically the value should be (- (make-char CHARSET 0) 128),\n\
1872 for your choice of character set.\n\
1873 If `nonascii-translation-table' is non-nil, it overrides this variable.");
1874 nonascii_insert_offset
= 0;
1876 DEFVAR_LISP ("nonascii-translation-table", &Vnonascii_translation_table
,
1877 "Translation table to convert non-ASCII unibyte codes to multibyte.\n\
1878 This is used for converting unibyte text to multibyte,\n\
1879 and for inserting character codes specified by number.\n\n\
1880 Conversion is performed only when multibyte characters are enabled,\n\
1881 and it serves to convert a Latin-1 or similar 8-bit character code\n\
1882 to the corresponding Emacs character code.\n\n\
1883 If this is nil, `nonascii-insert-offset' is used instead.\n\
1884 See also the docstring of `make-translation-table'.");
1885 Vnonascii_translation_table
= Qnil
;
1887 DEFVAR_INT ("min-composite-char", &min_composite_char
,
1888 "Minimum character code of a composite character.");
1889 min_composite_char
= MIN_CHAR_COMPOSITION
;