1 ;;; mule.el --- basic commands for mulitilingual environment
3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
4 ;; Licensed to the Free Software Foundation.
5 ;; Copyright (C) 2001 Free Software Foundation, Inc.
7 ;; Keywords: mule, multilingual, character set, coding system
9 ;; This file is part of GNU Emacs.
11 ;; GNU Emacs is free software; you can redistribute it and/or modify
12 ;; it under the terms of the GNU General Public License as published by
13 ;; the Free Software Foundation; either version 2, or (at your option)
16 ;; GNU Emacs is distributed in the hope that it will be useful,
17 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 ;; GNU General Public License for more details.
21 ;; You should have received a copy of the GNU General Public License
22 ;; along with GNU Emacs; see the file COPYING. If not, write to the
23 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
24 ;; Boston, MA 02111-1307, USA.
30 (defconst mule-version "5.0 (SAKAKI)" "\
31 Version number and name of this version of MULE (multilingual environment).")
33 (defconst mule-version-date "1999.12.7" "\
34 Distribution date of this version of MULE (multilingual environment).")
36 (defun load-with-code-conversion (fullname file &optional noerror nomessage)
37 "Execute a file of Lisp code named FILE whose absolute name is FULLNAME.
38 The file contents are decoded before evaluation if necessary.
39 If optional second arg NOERROR is non-nil,
40 report no error if FILE doesn't exist.
41 Print messages at start and end of loading unless
42 optional third arg NOMESSAGE is non-nil.
43 Return t if file exists."
44 (if (null (file-readable-p fullname))
46 (signal 'file-error (list "Cannot open load file" file)))
47 ;; Read file with code conversion, and then eval.
49 ;; To avoid any autoloading, set default-major-mode to
51 ;; So that we don't get completely screwed if the
52 ;; file is encoded in some complicated character set,
53 ;; read it with real decoding, as a multibyte buffer,
54 ;; even if this is a --unibyte Emacs session.
55 (let ((default-major-mode 'fundamental-mode)
56 (default-enable-multibyte-characters t))
57 ;; We can't use `generate-new-buffer' because files.el
59 (get-buffer-create (generate-new-buffer-name " *load*"))))
61 (source (save-match-data (string-match "\\.el\\'" fullname))))
64 (message "Loading %s (source)..." file)
65 (message "Loading %s..." file)))
67 (setq preloaded-file-list (cons file preloaded-file-list)))
69 (let ((load-file-name fullname)
70 (set-auto-coding-for-load t)
71 (inhibit-file-name-operation nil))
74 (insert-file-contents fullname)
75 ;; If the loaded file was inserted with no-conversion or
76 ;; raw-text coding system, make the buffer unibyte.
77 ;; Otherwise, eval-buffer might try to interpret random
78 ;; binary junk as multibyte characters.
79 (if (and enable-multibyte-characters
80 (or (eq (coding-system-type last-coding-system-used) 5)
81 (eq last-coding-system-used 'no-conversion)))
82 (set-buffer-multibyte nil))
83 ;; Make `kill-buffer' quiet.
84 (set-buffer-modified-p nil))
85 ;; Have the original buffer current while we eval.
86 (eval-buffer buffer nil file
87 ;; If this Emacs is running with --unibyte,
88 ;; convert multibyte strings to unibyte
89 ;; after reading them.
90 ;; (not default-enable-multibyte-characters)
93 (let (kill-buffer-hook kill-buffer-query-functions)
94 (kill-buffer buffer)))
95 (let ((hook (assoc file after-load-alist)))
97 (mapcar (function eval) (cdr hook))))
98 (unless (or nomessage noninteractive)
100 (message "Loading %s (source)...done" file)
101 (message "Loading %s...done" file)))
104 ;; API (Application Program Interface) for charsets.
106 (defsubst charset-quoted-standard-p (obj)
107 "Return t if OBJ is a quoted symbol, and is the name of a standard charset."
108 (and (listp obj) (eq (car obj) 'quote)
109 (symbolp (car-safe (cdr obj)))
110 (let ((vector (get (car-safe (cdr obj)) 'charset)))
111 (and (vectorp vector)
112 (< (aref vector 0) 160)))))
114 (defsubst charsetp (object)
115 "T if OBJECT is a charset."
116 (and (symbolp object) (vectorp (get object 'charset))))
118 (defsubst charset-info (charset)
119 "Return a vector of information of CHARSET.
120 The elements of the vector are:
121 CHARSET-ID, BYTES, DIMENSION, CHARS, WIDTH, DIRECTION,
122 LEADING-CODE-BASE, LEADING-CODE-EXT,
123 ISO-FINAL-CHAR, ISO-GRAPHIC-PLANE,
124 REVERSE-CHARSET, SHORT-NAME, LONG-NAME, DESCRIPTION,
127 CHARSET-ID (integer) is the identification number of the charset.
128 BYTES (integer) is the length of multi-byte form of a character in
129 the charset: one of 1, 2, 3, and 4.
130 DIMENSION (integer) is the number of bytes to represent a character of
132 CHARS (integer) is the number of characters in a dimension: 94 or 96.
133 WIDTH (integer) is the number of columns a character in the charset
134 occupies on the screen: one of 0, 1, and 2.
135 DIRECTION (integer) is the rendering direction of characters in the
136 charset when rendering. If 0, render from left to right, else
137 render from right to left.
138 LEADING-CODE-BASE (integer) is the base leading-code for the
140 LEADING-CODE-EXT (integer) is the extended leading-code for the
141 charset. All charsets of less than 0xA0 has the value 0.
142 ISO-FINAL-CHAR (character) is the final character of the
143 corresponding ISO 2022 charset. If the charset is not assigned
144 any final character, the value is -1.
145 ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked
146 while encoding to variants of ISO 2022 coding system, one of the
147 following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).
148 If the charset is not assigned any final character, the value is -1.
149 REVERSE-CHARSET (integer) is the charset which differs only in
150 LEFT-TO-RIGHT value from the charset. If there's no such a
151 charset, the value is -1.
152 SHORT-NAME (string) is the short name to refer to the charset.
153 LONG-NAME (string) is the long name to refer to the charset
154 DESCRIPTION (string) is the description string of the charset.
155 PLIST (property list) may contain any type of information a user
156 want to put and get by functions `put-charset-property' and
157 `get-charset-property' respectively."
158 (get charset 'charset))
160 ;; It is better not to use backquote in this file,
161 ;; because that makes a bootstrapping problem
162 ;; if you need to recompile all the Lisp files using interpreted code.
164 (defmacro charset-id (charset)
165 "Return charset identification number of CHARSET."
166 (if (charset-quoted-standard-p charset)
167 (aref (charset-info (nth 1 charset)) 0)
168 (list 'aref (list 'charset-info charset) 0)))
170 (defmacro charset-bytes (charset)
171 "Return bytes of CHARSET.
172 See the function `charset-info' for more detail."
173 (if (charset-quoted-standard-p charset)
174 (aref (charset-info (nth 1 charset)) 1)
175 (list 'aref (list 'charset-info charset) 1)))
177 (defmacro charset-dimension (charset)
178 "Return dimension of CHARSET.
179 See the function `charset-info' for more detail."
180 (if (charset-quoted-standard-p charset)
181 (aref (charset-info (nth 1 charset)) 2)
182 (list 'aref (list 'charset-info charset) 2)))
184 (defmacro charset-chars (charset)
185 "Return character numbers contained in a dimension of CHARSET.
186 See the function `charset-info' for more detail."
187 (if (charset-quoted-standard-p charset)
188 (aref (charset-info (nth 1 charset)) 3)
189 (list 'aref (list 'charset-info charset) 3)))
191 (defmacro charset-width (charset)
192 "Return width (how many column occupied on a screen) of CHARSET.
193 See the function `charset-info' for more detail."
194 (if (charset-quoted-standard-p charset)
195 (aref (charset-info (nth 1 charset)) 4)
196 (list 'aref (list 'charset-info charset) 4)))
198 (defmacro charset-direction (charset)
199 "Return direction of CHARSET.
200 See the function `charset-info' for more detail."
201 (if (charset-quoted-standard-p charset)
202 (aref (charset-info (nth 1 charset)) 5)
203 (list 'aref (list 'charset-info charset) 5)))
205 (defmacro charset-iso-final-char (charset)
206 "Return final char of CHARSET.
207 See the function `charset-info' for more detail."
208 (if (charset-quoted-standard-p charset)
209 (aref (charset-info (nth 1 charset)) 8)
210 (list 'aref (list 'charset-info charset) 8)))
212 (defmacro charset-iso-graphic-plane (charset)
213 "Return graphic plane of CHARSET.
214 See the function `charset-info' for more detail."
215 (if (charset-quoted-standard-p charset)
216 (aref (charset-info (nth 1 charset)) 9)
217 (list 'aref (list 'charset-info charset) 9)))
219 (defmacro charset-reverse-charset (charset)
220 "Return reverse charset of CHARSET.
221 See the function `charset-info' for more detail."
222 (if (charset-quoted-standard-p charset)
223 (aref (charset-info (nth 1 charset)) 10)
224 (list 'aref (list 'charset-info charset) 10)))
226 (defmacro charset-short-name (charset)
227 "Return short name of CHARSET.
228 See the function `charset-info' for more detail."
229 (if (charset-quoted-standard-p charset)
230 (aref (charset-info (nth 1 charset)) 11)
231 (list 'aref (list 'charset-info charset) 11)))
233 (defmacro charset-long-name (charset)
234 "Return long name of CHARSET.
235 See the function `charset-info' for more detail."
236 (if (charset-quoted-standard-p charset)
237 (aref (charset-info (nth 1 charset)) 12)
238 (list 'aref (list 'charset-info charset) 12)))
240 (defmacro charset-description (charset)
241 "Return description of CHARSET.
242 See the function `charset-info' for more detail."
243 (if (charset-quoted-standard-p charset)
244 (aref (charset-info (nth 1 charset)) 13)
245 (list 'aref (list 'charset-info charset) 13)))
247 (defmacro charset-plist (charset)
248 "Return list charset property of CHARSET.
249 See the function `charset-info' for more detail."
251 (if (charset-quoted-standard-p charset)
252 (charset-info (nth 1 charset))
253 (list 'charset-info charset))
256 (defun set-charset-plist (charset plist)
257 "Set CHARSET's property list to PLIST, and return PLIST."
258 (aset (charset-info charset) 14 plist))
260 (defun make-char (charset &optional code1 code2)
261 "Return a character of CHARSET whose position codes are CODE1 and CODE2.
262 CODE1 and CODE2 are optional, but if you don't supply
263 sufficient position codes, return a generic character which stands for
264 all characters or group of characters in the character set.
265 A generic character can be used to index a char table (e.g. syntax-table).
267 Such character sets as ascii, eight-bit-control, and eight-bit-graphic
268 don't have corresponding generic characters. If CHARSET is one of
269 them and you don't supply CODE1, return the character of the smallest
272 If CODE1 or CODE2 are invalid (out of range), this function signals an
273 error. However, the eighth bit of both CODE1 and CODE2 is zeroed
274 before they are used to index CHARSET. Thus you may use, say, the
275 actual ISO 8859 character code rather than subtracting 128, as you
276 would need to index the corresponding Emacs charset."
277 (make-char-internal (charset-id charset) code1 code2))
279 (put 'make-char 'byte-compile
282 (let ((charset (nth 1 form)))
283 (if (charset-quoted-standard-p charset)
284 (byte-compile-normal-call
285 (cons 'make-char-internal
286 (cons (charset-id (nth 1 charset)) (nthcdr 2 form))))
287 (byte-compile-normal-call
288 (cons 'make-char-internal
289 (cons (list 'charset-id charset) (nthcdr 2 form)))))))))
291 (defun charset-list ()
292 "Return list of charsets ever defined.
294 This function is provided for backward compatibility.
295 Now we have the variable `charset-list'."
298 (defsubst generic-char-p (char)
299 "Return t if and only if CHAR is a generic character.
300 See also the documentation of `make-char'."
302 (let ((l (split-char char)))
303 (and (or (= (nth 1 l) 0) (eq (nth 2 l) 0))
304 (not (eq (car l) 'composition))))))
306 (defun decode-char (ccs code-point &optional restriction)
307 "Return character specified by coded character set CCS and CODE-POINT in it.
308 Return nil if such a character is not supported.
309 Currently the only supported coded character set is `ucs' (ISO/IEC
310 10646: Universal Multi-Octet Coded Character Set).
312 Optional argument RESTRICTION specifies a way to map the pair of CCS
313 and CODE-POINT to a character. Currently not supported and just ignored."
315 (cond ((< code-point 160)
318 (make-char 'latin-iso8859-1 code-point))
319 ((< code-point #x2500)
320 (setq code-point (- code-point #x0100))
321 (make-char 'mule-unicode-0100-24ff
322 (+ (/ code-point 96) 32) (+ (% code-point 96) 32)))
323 ((< code-point #x3400)
324 (setq code-point (- code-point #x2500))
325 (make-char 'mule-unicode-2500-33ff
326 (+ (/ code-point 96) 32) (+ (% code-point 96) 32)))
327 ((and (>= code-point #xe000) (< code-point #x10000))
328 (setq code-point (- code-point #xe000))
329 (make-char 'mule-unicode-e000-ffff
330 (+ (/ code-point 96) 32) (+ (% code-point 96) 32)))
333 (defun encode-char (char ccs &optional restriction)
334 "Return code-point in coded character set CCS that corresponds to CHAR.
335 Return nil if CHAR is not included in CCS.
336 Currently the only supported coded character set is `ucs' (ISO/IEC
337 10646: Universal Multi-Octet Coded Character Set).
339 CHAR should be in one of these charsets:
340 ascii, latin-iso8859-1, mule-unicode-0100-24ff, mule-unicode-2500-33ff,
341 mule-unicode-e000-ffff, eight-bit-control
342 Otherwise, return nil.
344 Optional argument RESTRICTION specifies a way to map CHAR to a
345 code-point in CCS. Currently not supported and just ignored."
346 (let* ((split (split-char char))
347 (charset (car split)))
349 (cond ((eq charset 'ascii)
351 ((eq charset 'latin-iso8859-1)
352 (+ (nth 1 split) 128))
353 ((eq charset 'mule-unicode-0100-24ff)
354 (+ #x0100 (+ (* (- (nth 1 split) 32) 96)
355 (- (nth 2 split) 32))))
356 ((eq charset 'mule-unicode-2500-33ff)
357 (+ #x2500 (+ (* (- (nth 1 split) 32) 96)
358 (- (nth 2 split) 32))))
359 ((eq charset 'mule-unicode-e000-ffff)
360 (+ #xe000 (+ (* (- (nth 1 split) 32) 96)
361 (- (nth 2 split) 32))))
362 ((eq charset 'eight-bit-control)
366 ;; Coding system stuff
368 ;; Coding system is a symbol that has the property `coding-system'.
370 ;; The value of the property `coding-system' is a vector of the
372 ;; [TYPE MNEMONIC DOC-STRING PLIST FLAGS]
373 ;; We call this vector as coding-spec. See comments in src/coding.c
376 (defconst coding-spec-type-idx 0)
377 (defconst coding-spec-mnemonic-idx 1)
378 (defconst coding-spec-doc-string-idx 2)
379 (defconst coding-spec-plist-idx 3)
380 (defconst coding-spec-flags-idx 4)
382 ;; PLIST is a property list of a coding system. To share PLIST among
383 ;; alias coding systems, a coding system has PLIST in coding-spec
384 ;; instead of having it in normal property list of Lisp symbol.
385 ;; Here's a list of coding system properties currently being used.
389 ;; The value is a coding category the coding system belongs to. The
390 ;; function `make-coding-system' sets this value automatically
391 ;; unless its argument PROPERTIES specifies this property.
393 ;; o alias-coding-systems
395 ;; The value is a list of coding systems of the same alias group. The
396 ;; first element is the coding system made at first, which we call as
397 ;; `base coding system'. The function `make-coding-system' sets this
398 ;; value automatically and `define-coding-system-alias' updates it.
400 ;; See the documentation of make-coding-system for the meanings of the
401 ;; following properties.
403 ;; o post-read-conversion
404 ;; o pre-write-conversion
405 ;; o translation-table-for-decode
406 ;; o translation-table-for-encode
410 ;; o valid-codes (meaningful only for a coding system based on CCL)
413 (defsubst coding-system-spec (coding-system)
414 "Return coding-spec of CODING-SYSTEM."
415 (get (check-coding-system coding-system) 'coding-system))
417 (defun coding-system-type (coding-system)
418 "Return the coding type of CODING-SYSTEM.
419 A coding type is an integer value indicating the encoding method
420 of CODING-SYSTEM. See the function `make-coding-system' for more detail."
421 (aref (coding-system-spec coding-system) coding-spec-type-idx))
423 (defun coding-system-mnemonic (coding-system)
424 "Return the mnemonic character of CODING-SYSTEM.
425 The mnemonic character of a coding system is used in mode line
426 to indicate the coding system. If the arg is nil, return ?-."
427 (let ((spec (coding-system-spec coding-system)))
428 (if spec (aref spec coding-spec-mnemonic-idx) ?-)))
430 (defun coding-system-doc-string (coding-system)
431 "Return the documentation string for CODING-SYSTEM."
432 (aref (coding-system-spec coding-system) coding-spec-doc-string-idx))
434 (defun coding-system-plist (coding-system)
435 "Return the property list of CODING-SYSTEM."
436 (aref (coding-system-spec coding-system) coding-spec-plist-idx))
438 (defun coding-system-flags (coding-system)
439 "Return `flags' of CODING-SYSTEM.
440 A `flags' of a coding system is a vector of length 32 indicating detailed
441 information of a coding system. See the function `make-coding-system'
443 (aref (coding-system-spec coding-system) coding-spec-flags-idx))
445 (defun coding-system-get (coding-system prop)
446 "Extract a value from CODING-SYSTEM's property list for property PROP."
447 (plist-get (coding-system-plist coding-system) prop))
449 (defun coding-system-put (coding-system prop val)
450 "Change value in CODING-SYSTEM's property list PROP to VAL."
451 (let ((plist (coding-system-plist coding-system)))
453 (plist-put plist prop val)
454 (aset (coding-system-spec coding-system) coding-spec-plist-idx
457 (defun coding-system-category (coding-system)
458 "Return the coding category of CODING-SYSTEM.
459 See also `coding-category-list'."
460 (coding-system-get coding-system 'coding-category))
462 (defun coding-system-base (coding-system)
463 "Return the base coding system of CODING-SYSTEM.
464 A base coding system is what made by `make-coding-system'.
465 Any alias nor subsidiary coding systems are not base coding system."
466 (car (coding-system-get coding-system 'alias-coding-systems)))
468 (defalias 'coding-system-parent 'coding-system-base)
469 (make-obsolete 'coding-system-parent 'coding-system-base "20.3")
471 ;; Coding system also has a property `eol-type'.
473 ;; This property indicates how the coding system handles end-of-line
474 ;; format. The value is integer 0, 1, 2, or a vector of three coding
475 ;; systems. Each integer value 0, 1, and 2 indicates the format of
476 ;; end-of-line LF, CRLF, and CR respectively. A vector value
477 ;; indicates that the format of end-of-line should be detected
478 ;; automatically. Nth element of the vector is the subsidiary coding
479 ;; system whose `eol-type' property is N.
481 (defun coding-system-eol-type (coding-system)
482 "Return eol-type of CODING-SYSTEM.
483 An eol-type is integer 0, 1, 2, or a vector of coding systems.
485 Integer values 0, 1, and 2 indicate a format of end-of-line; LF,
486 CRLF, and CR respectively.
488 A vector value indicates that a format of end-of-line should be
489 detected automatically. Nth element of the vector is the subsidiary
490 coding system whose eol-type is N."
491 (get coding-system 'eol-type))
493 (defun coding-system-lessp (x y)
494 (cond ((eq x 'no-conversion) t)
495 ((eq y 'no-conversion) nil)
496 ((eq x 'emacs-mule) t)
497 ((eq y 'emacs-mule) nil)
498 ((eq x 'undecided) t)
499 ((eq y 'undecided) nil)
500 (t (let ((c1 (coding-system-mnemonic x))
501 (c2 (coding-system-mnemonic y)))
502 (or (< (downcase c1) (downcase c2))
503 (and (not (> (downcase c1) (downcase c2)))
506 (defun add-to-coding-system-list (coding-system)
507 "Add CODING-SYSTEM to `coding-system-list' while keeping it sorted."
508 (if (or (null coding-system-list)
509 (coding-system-lessp coding-system (car coding-system-list)))
510 (setq coding-system-list (cons coding-system coding-system-list))
511 (let ((len (length coding-system-list))
512 mid (tem coding-system-list))
514 (setq mid (nthcdr (/ len 2) tem))
515 (if (coding-system-lessp (car mid) coding-system)
517 len (- len (/ len 2)))
518 (setq len (/ len 2))))
519 (setcdr tem (cons coding-system (cdr tem))))))
521 (defun coding-system-list (&optional base-only)
522 "Return a list of all existing non-subsidiary coding systems.
523 If optional arg BASE-ONLY is non-nil, only base coding systems are listed.
524 The value doesn't include subsidiary coding systems which are what
525 made from bases and aliases automatically for various end-of-line
526 formats (e.g. iso-latin-1-unix, koi8-r-dos)."
527 (let* ((codings (copy-sequence coding-system-list))
528 (tail (cons nil codings)))
529 ;; Remove subsidiary coding systems (eol variants) and alias
530 ;; coding systems (if necessary).
532 (let* ((coding (car (cdr tail)))
533 (aliases (coding-system-get coding 'alias-coding-systems)))
535 ;; CODING is an eol variant if not in ALIASES.
536 (not (memq coding aliases))
537 ;; CODING is an alias if it is not car of ALIASES.
538 (and base-only (not (eq coding (car aliases)))))
539 (setcdr tail (cdr (cdr tail)))
540 (setq tail (cdr tail)))))
543 (defun map-charset-chars (func charset)
544 "Use FUNC to map over all characters in CHARSET for side effects.
545 FUNC is a function of two args, the start and end (inclusive) of a
546 character code range. Thus FUNC should iterate over [START, END]."
547 (let* ((dim (charset-dimension charset))
548 (chars (charset-chars charset))
549 (start (if (= chars 94)
554 (make-char charset start)
555 (make-char charset (+ start chars -1)))
558 (make-char charset (+ i start) start)
559 (make-char charset (+ i start) (+ start chars -1)))))))
561 (defun register-char-codings (coding-system safe-chars)
562 "Add entries for CODING-SYSTEM to `char-coding-system-table'.
563 If SAFE-CHARS is a char-table, its non-nil entries specify characters
564 which CODING-SYSTEM encodes safely. If SAFE-CHARS is t, register
565 CODING-SYSTEM as a general one which can encode all characters."
566 (let ((general (char-table-extra-slot char-coding-system-table 0))
567 ;; Charsets which have some members in the table, but not all
568 ;; of them (i.e. not just a generic character):
569 (partials (char-table-extra-slot char-coding-system-table 1)))
570 (if (eq safe-chars t)
571 (or (memq coding-system general)
572 (set-char-table-extra-slot char-coding-system-table 0
573 (cons coding-system general)))
576 (if (and (>= key 128) val)
577 (let ((codings (aref char-coding-system-table key))
578 (charset (char-charset key)))
579 (unless (memq coding-system codings)
580 (if (and (generic-char-p key)
581 (memq charset partials))
582 ;; The generic char would clobber individual
583 ;; entries already in the table. First save the
584 ;; separate existing entries for all chars of the
585 ;; charset (with the generic entry added, if
587 (let (entry existing)
590 (while (<= start end)
591 (setq entry (aref char-coding-system-table start))
595 (if (memq coding-system entry)
597 (cons coding-system entry)))
599 (setq start (1+ start))))
601 ;; Update the generic entry.
602 (aset char-coding-system-table key
603 (cons coding-system codings))
604 ;; Override with the saved entries.
605 (dolist (elt existing)
606 (aset char-coding-system-table (car elt) (cdr elt))))
607 (aset char-coding-system-table key
608 (cons coding-system codings))
609 (unless (or (memq charset partials)
610 (generic-char-p key))
611 (push charset partials)))))))
613 (set-char-table-extra-slot char-coding-system-table 1 partials))))
616 (defun make-subsidiary-coding-system (coding-system)
617 "Make subsidiary coding systems (eol-type variants) of CODING-SYSTEM."
618 (let ((coding-spec (coding-system-spec coding-system))
619 (subsidiaries (vector (intern (format "%s-unix" coding-system))
620 (intern (format "%s-dos" coding-system))
621 (intern (format "%s-mac" coding-system))))
625 (put (aref subsidiaries i) 'coding-system coding-spec)
626 (put (aref subsidiaries i) 'eol-type i)
627 (add-to-coding-system-list (aref subsidiaries i))
628 (setq coding-system-alist
629 (cons (list (symbol-name (aref subsidiaries i)))
630 coding-system-alist))
634 (defun transform-make-coding-system-args (name type &optional doc-string props)
635 "For internal use only.
636 Transform XEmacs style args for `make-coding-system' to Emacs style.
637 Value is a list of transformed arguments."
638 (let ((mnemonic (string-to-char (or (plist-get props 'mnemonic) "?")))
639 (eol-type (plist-get props 'eol-type))
642 ((eq eol-type 'lf) (setq eol-type 'unix))
643 ((eq eol-type 'crlf) (setq eol-type 'dos))
644 ((eq eol-type 'cr) (setq eol-type 'mac)))
645 (if (setq tmp (plist-get props 'post-read-conversion))
646 (setq properties (plist-put properties 'post-read-conversion tmp)))
647 (if (setq tmp (plist-get props 'pre-write-conversion))
648 (setq properties (plist-put properties 'pre-write-conversion tmp)))
650 ((eq type 'shift-jis)
651 `(,name 1 ,mnemonic ,doc-string () ,properties ,eol-type))
652 ((eq type 'iso2022) ; This is not perfect.
653 (if (plist-get props 'escape-quoted)
654 (error "escape-quoted is not supported: %S"
655 `(,name ,type ,doc-string ,props)))
656 (let ((g0 (plist-get props 'charset-g0))
657 (g1 (plist-get props 'charset-g1))
658 (g2 (plist-get props 'charset-g2))
659 (g3 (plist-get props 'charset-g3))
662 (eq (cadr (assoc 'latin-jisx0201
663 (plist-get props 'input-charset-conversion)))
665 (eq (cadr (assoc 'ascii
666 (plist-get props 'output-charset-conversion)))
670 (eq (cadr (assoc 'japanese-jisx0208-1978
671 (plist-get props 'input-charset-conversion)))
673 (eq (cadr (assoc 'japanese-jisx0208
674 (plist-get props 'output-charset-conversion)))
675 'japanese-jisx0208-1978))))
677 (if (plist-get props 'force-g0-on-output)
681 (if (plist-get props 'force-g1-on-output)
685 (if (plist-get props 'force-g2-on-output)
689 (if (plist-get props 'force-g3-on-output)
692 `(,name 2 ,mnemonic ,doc-string
694 ,(plist-get props 'short)
695 ,(not (plist-get props 'no-ascii-eol))
696 ,(not (plist-get props 'no-ascii-cntl))
697 ,(plist-get props 'seven)
699 ,(not (plist-get props 'lock-shift))
702 ,(plist-get props 'no-iso6429)
704 ,properties ,eol-type)))
706 `(,name 3 ,mnemonic ,doc-string () ,properties ,eol-type))
708 `(,name 4 ,mnemonic ,doc-string
709 (,(plist-get props 'decode) . ,(plist-get props 'encode))
710 ,properties ,eol-type))
712 (error "unsupported XEmacs style make-coding-style arguments: %S"
713 `(,name ,type ,doc-string ,props))))))
715 (defun make-coding-system (coding-system type mnemonic doc-string
720 "Define a new coding system CODING-SYSTEM (symbol).
721 Remaining arguments are TYPE, MNEMONIC, DOC-STRING, FLAGS (optional),
722 and PROPERTIES (optional) which construct a coding-spec of CODING-SYSTEM
723 in the following format:
724 [TYPE MNEMONIC DOC-STRING PLIST FLAGS]
726 TYPE is an integer value indicating the type of the coding system as follows:
727 0: Emacs internal format,
728 1: Shift-JIS (or MS-Kanji) used mainly on Japanese PC,
729 2: ISO-2022 including many variants,
730 3: Big5 used mainly on Chinese PC,
731 4: private, CCL programs provide encoding/decoding algorithm,
732 5: Raw-text, which means that text contains random 8-bit codes.
734 MNEMONIC is a character to be displayed on mode line for the coding system.
736 DOC-STRING is a documentation string for the coding system.
738 FLAGS specifies more detailed information of the coding system as follows:
740 If TYPE is 2 (ISO-2022), FLAGS is a list of these elements:
741 CHARSET0, CHARSET1, CHARSET2, CHARSET3, SHORT-FORM,
742 ASCII-EOL, ASCII-CNTL, SEVEN, LOCKING-SHIFT, SINGLE-SHIFT,
743 USE-ROMAN, USE-OLDJIS, NO-ISO6429, INIT-BOL, DESIGNATION-BOL,
744 SAFE, ACCEPT-LATIN-EXTRA-CODE.
745 CHARSETn are character sets initially designated to Gn graphic registers.
746 If CHARSETn is nil, Gn is never used.
747 If CHARSETn is t, Gn can be used but nothing designated initially.
748 If CHARSETn is a list of character sets, those character sets are
749 designated to Gn on output, but nothing designated to Gn initially.
750 But, character set `ascii' can be designated only to G0.
751 SHORT-FORM non-nil means use short designation sequence on output.
752 ASCII-EOL non-nil means designate ASCII to g0 at end of line on output.
753 ASCII-CNTL non-nil means designate ASCII to g0 before control codes and
755 SEVEN non-nil means use 7-bit code only on output.
756 LOCKING-SHIFT non-nil means use locking-shift.
757 SINGLE-SHIFT non-nil means use single-shift.
758 USE-ROMAN non-nil means designate JIS0201-1976-Roman instead of ASCII.
759 USE-OLDJIS non-nil means designate JIS0208-1976 instead of JIS0208-1983.
760 NO-ISO6429 non-nil means not use ISO6429's direction specification.
761 INIT-BOL non-nil means any designation state is assumed to be reset
762 to initial at each beginning of line on output.
763 DESIGNATION-BOL non-nil means designation sequences should be placed
764 at beginning of line on output.
765 SAFE non-nil means convert unsafe characters to `?' on output.
766 Characters not specified in the property `safe-charsets' nor
767 `safe-chars' are unsafe.
768 ACCEPT-LATIN-EXTRA-CODE non-nil means code-detection routine accepts
769 a code specified in `latin-extra-code-table' (which see) as a valid
770 code of the coding system.
772 If TYPE is 4 (private), FLAGS should be a cons of CCL programs, for
773 decoding and encoding. CCL programs should be specified by their
776 PROPERTIES is an alist of properties vs the corresponding values. The
777 following properties are recognized:
779 o post-read-conversion
781 The value is a function to call after some text is inserted and
782 decoded by the coding system itself and before any functions in
783 `after-insert-functions' are called. The argument of this
784 function is the same as for a function in
785 `after-insert-file-functions', i.e. LENGTH of the text inserted,
786 with point at the head of the text to be decoded.
788 o pre-write-conversion
790 The value is a function to call after all functions in
791 `write-region-annotate-functions' and `buffer-file-format' are
792 called, and before the text is encoded by the coding system itself.
793 The arguments to this function are the same as those of a function
794 in `write-region-annotate-functions', i.e. FROM and TO, specifying
797 o translation-table-for-decode
799 The value is a translation table to be applied on decoding. See
800 the function `make-translation-table' for the format of translation
801 table. This is not applicable to type 4 (CCL-based) coding systems.
803 o translation-table-for-encode
805 The value is a translation table to be applied on encoding. This is
806 not applicable to type 4 (CCL-based) coding systems.
810 The value is a char table. If a character has non-nil value in it,
811 the character is safely supported by the coding system. This
812 overrides the specification of safe-charsets.
816 The value is a list of charsets safely supported by the coding
817 system. The value t means that all charsets Emacs handles are
818 supported. Even if some charset is not in this list, it doesn't
819 mean that the charset can't be encoded in the coding system;
820 it just means that some other receiver of text encoded
821 in the coding system won't be able to handle that charset.
825 The value is a symbol of which name is `MIME-charset' parameter of
828 o valid-codes (meaningful only for a coding system based on CCL)
830 The value is a list to indicate valid byte ranges of the encoded
831 file. Each element of the list is an integer or a cons of integer.
832 In the former case, the integer value is a valid byte code. In the
833 latter case, the integers specify the range of valid byte codes.
835 These properties are set in PLIST, a property list. This function
836 also sets properties `coding-category' and `alias-coding-systems'
839 EOL-TYPE specifies the EOL type of the coding-system in one of the
842 o symbol (unix, dos, or mac)
844 The symbol `unix' means Unix-like EOL (LF), `dos' means
845 DOS-like EOL (CRLF), and `mac' means MAC-like EOL (CR).
847 o number (0, 1, or 2)
849 The number 0, 1, and 2 mean UNIX, DOS, and MAC-like EOL
852 o vector of coding-systems of length 3
854 The EOL type is detected automatically for the coding system.
855 And, according to the detected EOL type, one of the coding
856 systems in the vector is selected. Elements of the vector
857 corresponds to Unix-like EOL, DOS-like EOL, and Mac-like EOL
860 Kludgy features for backward compatibility:
862 1. If TYPE is 4 and car or cdr of FLAGS is a vector, the vector is
863 treated as a compiled CCL code.
865 2. If PROPERTIES is just a list of character sets, the list is set as
866 a value of `safe-charsets' in PLIST."
868 ;; For compatiblity with XEmacs, we check the type of TYPE. If it
869 ;; is a symbol, perhaps, this function is called with XEmacs-style
870 ;; arguments. Here, try to transform that kind of arguments to
873 (let ((args (transform-make-coding-system-args coding-system type
874 mnemonic doc-string)))
875 (setq coding-system (car args)
877 mnemonic (nth 2 args)
878 doc-string (nth 3 args)
880 properties (nth 5 args)
881 eol-type (nth 6 args))))
883 ;; Set a value of `coding-system' property.
884 (let ((coding-spec (make-vector 5 nil))
885 (no-initial-designation t)
886 (no-alternative-designation t)
887 (accept-latin-extra-code nil)
889 (if (or (not (integerp type)) (< type 0) (> type 5))
890 (error "TYPE argument must be 0..5"))
891 (if (or (not (integerp mnemonic)) (<= mnemonic ? ) (> mnemonic 127))
892 (error "MNEMONIC argument must be an ASCII printable character"))
893 (aset coding-spec coding-spec-type-idx type)
894 (aset coding-spec coding-spec-mnemonic-idx mnemonic)
895 (aset coding-spec coding-spec-doc-string-idx
896 (purecopy (if (stringp doc-string) doc-string "")))
898 (setq coding-category 'coding-category-emacs-mule))
900 (setq coding-category 'coding-category-sjis))
901 ((= type 2) ; ISO2022
903 (vec (make-vector 32 nil))
907 (let ((charset (car fl)))
908 (if (and no-initial-designation
910 (or (charsetp charset)
912 (charsetp (car charset)))))
913 (setq no-initial-designation nil))
914 (if (charsetp charset)
915 (if (= i 1) (setq g1-designation charset))
920 (setq elt (car tail))
922 (setq no-alternative-designation nil)
923 (if (and elt (not (charsetp elt)))
924 (error "Invalid charset: %s" elt)))
925 (setq tail (cdr tail)))
926 (setq g1-designation (car charset)))
929 (setq no-alternative-designation nil)
930 (error "Invalid charset: %s" charset)))))
931 (aset vec i charset))
932 (setq fl (cdr fl) i (1+ i)))
933 (while (and (< i 32) fl)
934 (aset vec i (car fl))
935 (if (and (= i 16) ; ACCEPT-LATIN-EXTRA-CODE
937 (setq accept-latin-extra-code t))
938 (setq fl (cdr fl) i (1+ i)))
939 (aset coding-spec 4 vec)
940 (setq coding-category
941 (if (aref vec 8) ; Use locking-shift.
942 (or (and (aref vec 7) 'coding-category-iso-7-else)
943 'coding-category-iso-8-else)
944 (if (aref vec 7) ; 7-bit only.
945 (if (aref vec 9) ; Use single-shift.
946 'coding-category-iso-7-else
947 (if no-alternative-designation
948 'coding-category-iso-7-tight
949 'coding-category-iso-7))
950 (if (or no-initial-designation
951 (not no-alternative-designation))
952 'coding-category-iso-8-else
953 (if (and (charsetp g1-designation)
954 (= (charset-dimension g1-designation) 2))
955 'coding-category-iso-8-2
956 'coding-category-iso-8-1)))))))
958 (setq coding-category 'coding-category-big5))
959 ((= type 4) ; private
960 (setq coding-category 'coding-category-ccl)
961 (if (not (consp flags))
962 (error "Invalid FLAGS argument for TYPE 4 (CCL)")
963 (let ((decoder (check-ccl-program
965 (intern (format "%s-decoder" coding-system))))
966 (encoder (check-ccl-program
968 (intern (format "%s-encoder" coding-system)))))
969 (if (and decoder encoder)
970 (aset coding-spec 4 (cons decoder encoder))
971 (error "Invalid FLAGS argument for TYPE 4 (CCL)")))))
973 (setq coding-category 'coding-category-raw-text)))
975 (let ((plist (list 'coding-category coding-category
976 'alias-coding-systems (list coding-system))))
977 (if no-initial-designation
978 (plist-put plist 'no-initial-designation t))
980 (or (eq properties t)
981 (not (consp (car properties)))))
982 ;; In the old version, the arg PROPERTIES is a list to be
983 ;; set in PLIST as a value of property `safe-charsets'.
984 (setq properties (list (cons 'safe-charsets properties))))
985 ;; In the current version PROPERTIES is a property list.
986 ;; Reflect it into PLIST one by one while handling safe-chars
988 (let ((safe-charsets (cdr (assq 'safe-charsets properties)))
989 (safe-chars (cdr (assq 'safe-chars properties)))
992 ;; If only safe-charsets is specified, make a char-table from
993 ;; it, and store that char-table as the value of `safe-chars'.
994 (if (and (not safe-chars) safe-charsets)
996 (if (eq safe-charsets t)
998 (setq safe-chars (make-char-table 'safe-chars))
1000 (setq charset (car safe-charsets)
1001 safe-charsets (cdr safe-charsets))
1002 (cond ((eq charset 'ascii)) ; just ignore
1003 ((eq charset 'eight-bit-control)
1006 (aset safe-chars i t)
1008 ((eq charset 'eight-bit-graphic)
1011 (aset safe-chars i t)
1014 (aset safe-chars (make-char charset) t))))
1015 (if accept-latin-extra-code
1018 (if (aref latin-extra-code-table i)
1019 (aset safe-chars i t))
1021 (setq l (cons (cons 'safe-chars safe-chars) l))))
1023 (setq prop (car (car l)) val (cdr (car l)) l (cdr l))
1024 (if (eq prop 'safe-chars)
1026 (if (and (symbolp val)
1027 (get val 'translation-table))
1028 (setq safe-chars (get val 'translation-table)))
1029 (register-char-codings coding-system safe-chars)
1030 (setq val safe-chars)))
1031 (plist-put plist prop val)))
1032 ;; The property `coding-category' may have been set differently
1033 ;; through PROPERTIES.
1034 (setq coding-category (plist-get plist 'coding-category))
1035 (aset coding-spec coding-spec-plist-idx plist))
1036 (put coding-system 'coding-system coding-spec)
1037 (put coding-category 'coding-systems
1038 (cons coding-system (get coding-category 'coding-systems))))
1040 ;; Next, set a value of `eol-type' property.
1042 ;; If EOL-TYPE is nil, set a vector of subsidiary coding
1043 ;; systems, each corresponds to a coding system for the detected
1045 (setq eol-type (make-subsidiary-coding-system coding-system)))
1047 (cond ((or (eq eol-type 'unix) (null eol-type))
1053 ((or (and (vectorp eol-type)
1054 (= (length eol-type) 3))
1055 (and (numberp eol-type)
1056 (and (>= eol-type 0)
1060 (error "Invalid EOL-TYPE spec:%S" eol-type))))
1061 (put coding-system 'eol-type eol-type)
1063 ;; At last, register CODING-SYSTEM in `coding-system-list' and
1064 ;; `coding-system-alist'.
1065 (add-to-coding-system-list coding-system)
1066 (setq coding-system-alist (cons (list (symbol-name coding-system))
1067 coding-system-alist))
1069 ;; For a coding system of cateogory iso-8-1 and iso-8-2, create
1070 ;; XXX-with-esc variants.
1071 (let ((coding-category (coding-system-category coding-system)))
1072 (if (or (eq coding-category 'coding-category-iso-8-1)
1073 (eq coding-category 'coding-category-iso-8-2))
1074 (let ((esc (intern (concat (symbol-name coding-system) "-with-esc")))
1075 (doc (format "Same as %s but can handle any charsets by ISO's escape sequences." coding-system))
1076 (safe-charsets (assq 'safe-charsets properties))
1077 (mime-charset (assq 'mime-charset properties)))
1079 (setcdr safe-charsets t)
1080 (setq properties (cons (cons 'safe-charsets t) properties)))
1082 (setcdr mime-charset nil))
1083 (make-coding-system esc type mnemonic doc
1084 (if (listp (car flags))
1085 (cons (append (car flags) '(t)) (cdr flags))
1086 (cons (list (car flags) t) (cdr flags)))
1091 (defun define-coding-system-alias (alias coding-system)
1092 "Define ALIAS as an alias for coding system CODING-SYSTEM."
1093 (put alias 'coding-system (coding-system-spec coding-system))
1094 (nconc (coding-system-get alias 'alias-coding-systems) (list alias))
1095 (add-to-coding-system-list alias)
1096 (setq coding-system-alist (cons (list (symbol-name alias))
1097 coding-system-alist))
1098 (let ((eol-type (coding-system-eol-type coding-system)))
1099 (if (vectorp eol-type)
1100 (put alias 'eol-type (make-subsidiary-coding-system alias))
1101 (put alias 'eol-type eol-type))))
1103 (defun set-buffer-file-coding-system (coding-system &optional force)
1104 "Set the file coding-system of the current buffer to CODING-SYSTEM.
1105 This means that when you save the buffer, it will be converted
1106 according to CODING-SYSTEM. For a list of possible values of CODING-SYSTEM,
1107 use \\[list-coding-systems].
1109 If the buffer's previous file coding-system value specifies end-of-line
1110 conversion, and CODING-SYSTEM does not specify one, CODING-SYSTEM is
1111 merged with the already-specified end-of-line conversion.
1113 If the buffer's previous file coding-system value specifies text
1114 conversion, and CODING-SYSTEM does not specify one, CODING-SYSTEM is
1115 merged with the already-specified text conversion.
1117 However, if the optional prefix argument FORCE is non-nil, then
1118 CODING-SYSTEM is used exactly as specified.
1120 This marks the buffer modified so that the succeeding \\[save-buffer]
1121 surely saves the buffer with CODING-SYSTEM. From a program, if you
1122 don't want to mark the buffer modified, just set the variable
1123 `buffer-file-coding-system' directly."
1124 (interactive "zCoding system for visited file (default, nil): \nP")
1125 (check-coding-system coding-system)
1126 (if (and coding-system buffer-file-coding-system (null force))
1127 (let ((base (coding-system-base buffer-file-coding-system))
1128 (eol (coding-system-eol-type buffer-file-coding-system)))
1129 ;; If CODING-SYSTEM doesn't specify text conversion, merge
1130 ;; with that of buffer-file-coding-system.
1131 (if (eq (coding-system-base coding-system) 'undecided)
1132 (setq coding-system (coding-system-change-text-conversion
1133 coding-system base)))
1134 ;; If CODING-SYSTEM doesn't specify eol conversion, merge with
1135 ;; that of buffer-file-coding-system.
1136 (if (and (vectorp (coding-system-eol-type coding-system))
1137 (numberp eol) (>= eol 0) (<= eol 2))
1138 (setq coding-system (coding-system-change-eol-conversion
1139 coding-system eol)))))
1140 (setq buffer-file-coding-system coding-system)
1141 (set-buffer-modified-p t)
1142 (force-mode-line-update))
1144 (defvar default-terminal-coding-system nil
1145 "Default value for the terminal coding system.
1146 This is normally set according to the selected language environment.
1147 See also the command `set-terminal-coding-system'.")
1149 (defun set-terminal-coding-system (coding-system)
1150 "Set coding system of your terminal to CODING-SYSTEM.
1151 All text output to the terminal will be encoded
1152 with the specified coding system.
1153 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems].
1154 The default is determined by the selected language environment
1155 or by the previous use of this command."
1157 (list (let ((default (if (and (not (terminal-coding-system))
1158 default-terminal-coding-system)
1159 default-terminal-coding-system)))
1161 (format "Coding system for terminal display (default, %s): "
1164 (if (and (not coding-system)
1165 (not (terminal-coding-system)))
1166 (setq coding-system default-terminal-coding-system))
1168 (setq default-terminal-coding-system coding-system))
1169 (set-terminal-coding-system-internal coding-system)
1170 (redraw-frame (selected-frame)))
1172 (defvar default-keyboard-coding-system nil
1173 "Default value of the keyboard coding system.
1174 This is normally set according to the selected language environment.
1175 See also the command `set-keyboard-coding-system'.")
1177 (defun set-keyboard-coding-system (coding-system)
1178 "Set coding system for keyboard input to CODING-SYSTEM.
1179 In addition, this command enables Encoded-kbd minor mode.
1180 \(If CODING-SYSTEM is nil, Encoded-kbd mode is turned off -- see
1181 `encoded-kbd-mode'.)
1182 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems].
1183 The default is determined by the selected language environment
1184 or by the previous use of this command."
1186 (list (let ((default (if (and (not (keyboard-coding-system))
1187 default-keyboard-coding-system)
1188 default-keyboard-coding-system)))
1190 (format "Coding system for keyboard input (default, %s): "
1193 (if (and (not coding-system)
1194 (not (keyboard-coding-system)))
1195 (setq coding-system default-keyboard-coding-system))
1197 (setq default-keyboard-coding-system coding-system))
1198 (set-keyboard-coding-system-internal coding-system)
1199 (setq keyboard-coding-system coding-system)
1200 (encoded-kbd-mode (if coding-system 1 0)))
1202 (defcustom keyboard-coding-system nil
1203 "Specify coding system for keyboard input.
1204 If you set this on a terminal which can't distinguish Meta keys from
1205 8-bit characters, you will have to use ESC to type Meta characters.
1206 See Info node `Specify Coding' and Info node `Single-Byte Character Support'.
1208 Setting this variable directly does not take effect;
1209 use either M-x customize or \\[set-keyboard-coding-system]."
1210 :type '(coding-system :tag "Coding system")
1211 :link '(info-link "(emacs)Specify Coding")
1212 :link '(info-link "(emacs)Single-Byte Character Support")
1213 :set (lambda (symbol value)
1214 ;; Don't load encoded-kbd-mode unnecessarily.
1215 (if (or value (boundp 'encoded-kbd-mode))
1216 (set-keyboard-coding-system value)
1217 (set-default 'keyboard-coding-system nil))) ; must initialize
1222 (defun set-buffer-process-coding-system (decoding encoding)
1223 "Set coding systems for the process associated with the current buffer.
1224 DECODING is the coding system to be used to decode input from the process,
1225 ENCODING is the coding system to be used to encode output to the process.
1227 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems]."
1229 "zCoding-system for output from the process: \nzCoding-system for input to the process: ")
1230 (let ((proc (get-buffer-process (current-buffer))))
1232 (error "No process")
1233 (check-coding-system decoding)
1234 (check-coding-system encoding)
1235 (set-process-coding-system proc decoding encoding)))
1236 (force-mode-line-update))
1238 (defalias 'set-clipboard-coding-system 'set-selection-coding-system)
1240 (defun set-selection-coding-system (coding-system)
1241 "Make CODING-SYSTEM used for communicating with other X clients .
1242 When sending or receiving text via cut_buffer, selection, and clipboard,
1243 the text is encoded or decoded by CODING-SYSTEM."
1244 (interactive "zCoding system for X selection: ")
1245 (check-coding-system coding-system)
1246 (setq selection-coding-system coding-system))
1248 ;; Coding system lastly specified by the command
1249 ;; set-next-selection-coding-system.
1250 (defvar last-next-selection-coding-system nil)
1252 (defun set-next-selection-coding-system (coding-system)
1253 "Make CODING-SYSTEM used for the next communication with other X clients.
1254 This setting is effective for the next communication only."
1256 (list (read-coding-system
1257 (if last-next-selection-coding-system
1258 (format "Coding system for the next X selection (default, %S): "
1259 last-next-selection-coding-system)
1260 "Coding system for the next X selection: ")
1261 last-next-selection-coding-system)))
1263 (setq last-next-selection-coding-system coding-system)
1264 (setq coding-system last-next-selection-coding-system))
1265 (check-coding-system coding-system)
1267 (setq next-selection-coding-system coding-system))
1269 (defun set-coding-priority (arg)
1270 "Set priority of coding categories according to ARG.
1271 ARG is a list of coding categories ordered by priority."
1273 (current-list (copy-sequence coding-category-list)))
1274 ;; Check the validity of ARG while deleting coding categories in
1275 ;; ARG from CURRENT-LIST. We assume that CODING-CATEGORY-LIST
1276 ;; contains all coding categories.
1278 (if (or (null (get (car l) 'coding-category-index))
1279 (null (memq (car l) current-list)))
1280 (error "Invalid or duplicated element in argument: %s" arg))
1281 (setq current-list (delq (car l) current-list))
1283 ;; Update `coding-category-list' and return it.
1284 (setq coding-category-list (append arg current-list))
1285 (set-coding-priority-internal)))
1289 (defcustom auto-coding-alist
1290 '(("\\.\\(arc\\|zip\\|lzh\\|zoo\\|jar\\|tar\\|tgz\\)\\'" . no-conversion)
1291 ("\\.\\(gz\\|Z\\|bz\\|bz2\\|gpg\\)\\'" . no-conversion))
1292 "Alist of filename patterns vs corresponding coding systems.
1293 Each element looks like (REGEXP . CODING-SYSTEM).
1294 A file whose name matches REGEXP is decoded by CODING-SYSTEM on reading.
1296 The settings in this alist take priority over `coding:' tags
1297 in the file (see the function `set-auto-coding')
1298 and the contents of `file-coding-system-alist'."
1301 :type '(repeat (cons (regexp :tag "File name regexp")
1302 (symbol :tag "Coding system"))))
1304 (defcustom auto-coding-regexp-alist
1305 '(("^BABYL OPTIONS:[ \t]*-\\*-[ \t]*rmail[ \t]*-\\*-" . no-conversion))
1306 "Alist of patterns vs corresponding coding systems.
1307 Each element looks like (REGEXP . CODING-SYSTEM).
1308 A file whose first bytes match REGEXP is decoded by CODING-SYSTEM on reading.
1310 The settings in this alist take priority over `coding:' tags
1311 in the file (see the function `set-auto-coding')
1312 and the contents of `file-coding-system-alist'."
1315 :type '(repeat (cons (regexp :tag "Regexp")
1316 (symbol :tag "Coding system"))))
1318 (defvar set-auto-coding-for-load nil
1319 "Non-nil means look for `load-coding' property instead of `coding'.
1320 This is used for loading and byte-compiling Emacs Lisp files.")
1322 (defun auto-coding-alist-lookup (filename)
1323 "Return the coding system specified by `auto-coding-alist' for FILENAME."
1324 (let ((alist auto-coding-alist)
1325 (case-fold-search (memq system-type '(vax-vms windows-nt ms-dos)))
1327 (while (and alist (not coding-system))
1328 (if (string-match (car (car alist)) filename)
1329 (setq coding-system (cdr (car alist)))
1330 (setq alist (cdr alist))))
1334 (defun auto-coding-from-file-contents (size)
1335 "Determine a coding system from the contents of the current buffer.
1336 The current buffer contains SIZE bytes starting at point.
1337 Value is either a coding system or nil."
1339 (let ((alist auto-coding-regexp-alist)
1341 (while (and alist (not coding-system))
1342 (let ((regexp (car (car alist))))
1343 (when (re-search-forward regexp (+ (point) size) t)
1344 (setq coding-system (cdr (car alist)))))
1345 (setq alist (cdr alist)))
1349 (defun set-auto-coding (filename size)
1350 "Return coding system for a file FILENAME of which SIZE bytes follow point.
1351 These bytes should include at least the first 1k of the file
1352 and the last 3k of the file, but the middle may be omitted.
1354 It checks FILENAME against the variable `auto-coding-alist'. If
1355 FILENAME doesn't match any entries in the variable, it checks the
1356 contents of the current buffer following point against
1357 `auto-coding-regexp-alist'. If no match is found, it checks for a
1358 `coding:' tag in the first one or two lines following point. If no
1359 `coding:' tag is found, it checks for local variables list in the last
1360 3K bytes out of the SIZE bytes.
1362 The return value is the specified coding system,
1363 or nil if nothing specified.
1365 The variable `set-auto-coding-function' (which see) is set to this
1366 function by default."
1367 (or (auto-coding-alist-lookup filename)
1368 (auto-coding-from-file-contents size)
1369 (let* ((case-fold-search t)
1370 (head-start (point))
1371 (head-end (+ head-start (min size 1024)))
1372 (tail-start (+ head-start (max (- size 3072) 0)))
1373 (tail-end (+ head-start size))
1374 coding-system head-found tail-found pos)
1375 ;; Try a short cut by searching for the string "coding:"
1376 ;; and for "unibyte:" at the head and tail of SIZE bytes.
1377 (setq head-found (or (search-forward "coding:" head-end t)
1378 (search-forward "unibyte:" head-end t)))
1379 (if (and head-found (> head-found tail-start))
1380 ;; Head and tail are overlapped.
1381 (setq tail-found head-found)
1382 (goto-char tail-start)
1383 (setq tail-found (or (search-forward "coding:" tail-end t)
1384 (search-forward "unibyte:" tail-end t))))
1386 ;; At first check the head.
1388 (goto-char head-start)
1389 (setq head-end (set-auto-mode-1))
1390 (setq head-start (point))
1391 (when (< head-found head-end)
1392 (goto-char head-start)
1393 (when (and set-auto-coding-for-load
1395 "\\(.*;\\)?[ \t]*unibyte:[ \t]*\\([^ ;]+\\)"
1397 (setq coding-system 'raw-text))
1398 (when (and (not coding-system)
1400 "\\(.*;\\)?[ \t]*coding:[ \t]*\\([^ ;]+\\)"
1402 (setq coding-system (intern (match-string 2)))
1403 (or (coding-system-p coding-system)
1404 (setq coding-system nil)))))
1406 ;; If no coding: tag in the head, check the tail.
1407 (when (and tail-found (not coding-system))
1408 (goto-char tail-start)
1409 (search-forward "\n\^L" nil t)
1410 (if (re-search-forward
1411 "^\\(.*\\)[ \t]*Local Variables:[ \t]*\\(.*\\)$" tail-end t)
1412 ;; The prefix is what comes before "local variables:" in its
1413 ;; line. The suffix is what comes after "local variables:"
1415 (let* ((prefix (regexp-quote (match-string 1)))
1416 (suffix (regexp-quote (match-string 2)))
1420 ;; N.B. without the \n below, the regexp can
1422 "[ \t]*coding[ \t]*:[ \t]*\\([^ \t\n]+\\)[ \t]*"
1427 "[ \t]*unibyte[ \t]*:[ \t]*\\([^ \t\n]+\\)[ \t]*"
1430 (concat "^" prefix "[ \t]*End *:[ \t]*" suffix "$"))
1432 (re-search-forward re-end tail-end 'move)
1433 (setq tail-end (point))
1435 (when (and set-auto-coding-for-load
1436 (re-search-forward re-unibyte tail-end t))
1437 (setq coding-system 'raw-text))
1438 (when (and (not coding-system)
1439 (re-search-forward re-coding tail-end t))
1440 (setq coding-system (intern (match-string 1)))
1441 (or (coding-system-p coding-system)
1442 (setq coding-system nil))))))
1445 (setq set-auto-coding-function 'set-auto-coding)
1447 (defun after-insert-file-set-buffer-file-coding-system (inserted)
1448 "Set `buffer-file-coding-system' of current buffer after text is inserted."
1449 (if last-coding-system-used
1450 (let ((coding-system
1451 (find-new-buffer-file-coding-system last-coding-system-used))
1452 (modified-p (buffer-modified-p)))
1454 (set-buffer-file-coding-system coding-system t)
1455 (if (and enable-multibyte-characters
1456 (or (eq coding-system 'no-conversion)
1457 (eq (coding-system-type coding-system) 5))
1458 ;; If buffer was unmodified and the size is the
1459 ;; same as INSERTED, we must be visiting it.
1461 (= (buffer-size) inserted))
1462 ;; For coding systems no-conversion and raw-text...,
1463 ;; edit the buffer as unibyte.
1464 (let ((pos-byte (position-bytes (+ (point) inserted))))
1465 (set-buffer-multibyte nil)
1466 (setq inserted (- pos-byte (position-bytes (point))))))
1467 (set-buffer-modified-p modified-p))))
1470 (add-hook 'after-insert-file-functions
1471 'after-insert-file-set-buffer-file-coding-system)
1473 ;; The coding-spec and eol-type of coding-system returned is decided
1474 ;; independently in the following order.
1475 ;; 1. That of buffer-file-coding-system locally bound.
1476 ;; 2. That of CODING.
1478 (defun find-new-buffer-file-coding-system (coding)
1479 "Return a coding system for a buffer when a file of CODING is inserted.
1480 The local variable `buffer-file-coding-system' of the current buffer
1481 is set to the returned value.
1482 Return nil if there's no need to set `buffer-file-coding-system'."
1483 (let (local-coding local-eol
1484 found-coding found-eol
1487 ;; Nothing found about coding.
1490 ;; Get information of `buffer-file-coding-system' in LOCAL-EOL
1491 ;; and LOCAL-CODING.
1492 (setq local-eol (coding-system-eol-type buffer-file-coding-system))
1493 (if (null (numberp local-eol))
1494 ;; But eol-type is not yet set.
1495 (setq local-eol nil))
1496 (if (and buffer-file-coding-system
1497 (not (eq (coding-system-type buffer-file-coding-system) t)))
1498 ;; This is not `undecided'.
1499 (setq local-coding (coding-system-base buffer-file-coding-system)))
1501 (if (and (local-variable-p 'buffer-file-coding-system)
1502 local-eol local-coding)
1503 ;; The current buffer has already set full coding-system, we
1504 ;; had better not change it.
1507 (setq found-eol (coding-system-eol-type coding))
1508 (if (null (numberp found-eol))
1509 ;; But eol-type is not found.
1510 ;; If EOL conversions are inhibited, force unix eol-type.
1511 (setq found-eol (if inhibit-eol-conversion 0)))
1512 (if (eq (coding-system-type coding) t)
1513 (setq found-coding 'undecided)
1514 (setq found-coding (coding-system-base coding)))
1516 (if (and (not found-eol) (eq found-coding 'undecided))
1517 ;; No valid coding information found.
1520 ;; Some coding information (eol or text) found.
1522 ;; The local setting takes precedence over the found one.
1523 (setq new-coding (if (local-variable-p 'buffer-file-coding-system)
1524 (or local-coding found-coding)
1525 (or found-coding local-coding)))
1526 (setq new-eol (if (local-variable-p 'buffer-file-coding-system)
1527 (or local-eol found-eol)
1528 (or found-eol local-eol)))
1530 (let ((eol-type (coding-system-eol-type new-coding)))
1531 (if (and (numberp new-eol) (vectorp eol-type))
1532 (aref eol-type new-eol)
1535 (defun modify-coding-system-alist (target-type regexp coding-system)
1536 "Modify one of look up tables for finding a coding system on I/O operation.
1537 There are three of such tables, `file-coding-system-alist',
1538 `process-coding-system-alist', and `network-coding-system-alist'.
1540 TARGET-TYPE specifies which of them to modify.
1541 If it is `file', it affects `file-coding-system-alist' (which see).
1542 If it is `process', it affects `process-coding-system-alist' (which see).
1543 If it is `network', it affects `network-coding-system-alist' (which see).
1545 REGEXP is a regular expression matching a target of I/O operation.
1546 The target is a file name if TARGET-TYPE is `file', a program name if
1547 TARGET-TYPE is `process', or a network service name or a port number
1548 to connect to if TARGET-TYPE is `network'.
1550 CODING-SYSTEM is a coding system to perform code conversion on the I/O
1551 operation, or a cons cell (DECODING . ENCODING) specifying the coding systems
1552 for decoding and encoding respectively,
1553 or a function symbol which, when called, returns such a cons cell."
1554 (or (memq target-type '(file process network))
1555 (error "Invalid target type: %s" target-type))
1556 (or (stringp regexp)
1557 (and (eq target-type 'network) (integerp regexp))
1558 (error "Invalid regular expression: %s" regexp))
1559 (if (symbolp coding-system)
1560 (if (not (fboundp coding-system))
1562 (check-coding-system coding-system)
1563 (setq coding-system (cons coding-system coding-system))))
1564 (check-coding-system (car coding-system))
1565 (check-coding-system (cdr coding-system)))
1566 (cond ((eq target-type 'file)
1567 (let ((slot (assoc regexp file-coding-system-alist)))
1569 (setcdr slot coding-system)
1570 (setq file-coding-system-alist
1571 (cons (cons regexp coding-system)
1572 file-coding-system-alist)))))
1573 ((eq target-type 'process)
1574 (let ((slot (assoc regexp process-coding-system-alist)))
1576 (setcdr slot coding-system)
1577 (setq process-coding-system-alist
1578 (cons (cons regexp coding-system)
1579 process-coding-system-alist)))))
1581 (let ((slot (assoc regexp network-coding-system-alist)))
1583 (setcdr slot coding-system)
1584 (setq network-coding-system-alist
1585 (cons (cons regexp coding-system)
1586 network-coding-system-alist)))))))
1588 (defun make-translation-table (&rest args)
1589 "Make a translation table from arguments.
1590 A translation table is a char table intended for character
1591 translation in CCL programs.
1593 Each argument is a list of elements of the form (FROM . TO), where FROM
1594 is a character to be translated to TO.
1596 FROM can be a generic character (see `make-char'). In this case, TO is
1597 a generic character containing the same number of characters, or an
1598 ordinary character. If FROM and TO are both generic characters, all
1599 characters belonging to FROM are translated to characters belonging to TO
1600 without changing their position code(s).
1602 The arguments and forms in each argument are processed in the given
1603 order, and if a previous form already translates TO to some other
1604 character, say TO-ALT, FROM is also translated to TO-ALT."
1605 (let ((table (make-char-table 'translation-table))
1608 (let ((elts (car args)))
1610 (let* ((from (car (car elts)))
1611 (from-i 0) ; degree of freedom of FROM
1612 (from-rev (nreverse (split-char from)))
1613 (to (cdr (car elts)))
1614 (to-i 0) ; degree of freedom of TO
1615 (to-rev (nreverse (split-char to))))
1616 ;; Check numbers of heading 0s in FROM-REV and TO-REV.
1617 (while (eq (car from-rev) 0)
1618 (setq from-i (1+ from-i) from-rev (cdr from-rev)))
1619 (while (eq (car to-rev) 0)
1620 (setq to-i (1+ to-i) to-rev (cdr to-rev)))
1621 (if (and (/= from-i to-i) (/= to-i 0))
1622 (error "Invalid character pair (%d . %d)" from to))
1623 ;; If we have already translated TO to TO-ALT, FROM should
1624 ;; also be translated to TO-ALT. But, this is only if TO
1625 ;; is a generic character or TO-ALT is not a generic
1627 (let ((to-alt (aref table to)))
1629 (or (> to-i 0) (not (generic-char-p to-alt))))
1632 (set-char-table-default table from to)
1633 (aset table from to))
1634 ;; If we have already translated some chars to FROM, they
1635 ;; should also be translated to TO.
1636 (let ((l (assq from revlist)))
1643 (setq l (cdr l)) ))))
1644 ;; Now update REVLIST.
1645 (let ((l (assq to revlist)))
1647 (setcdr l (cons from (cdr l)))
1648 (setq revlist (cons (list to from) revlist)))))
1649 (setq elts (cdr elts))))
1650 (setq args (cdr args)))
1651 ;; Return TABLE just created.
1654 (defun make-translation-table-from-vector (vec)
1655 "Make translation table from decoding vector VEC.
1656 VEC is an array of 256 elements to map unibyte codes to multibyte
1657 characters. Elements may be nil for undefined code points.
1658 See also the variable `nonascii-translation-table'."
1659 (let ((table (make-char-table 'translation-table))
1660 (rev-table (make-char-table 'translation-table))
1663 (setq ch (aref vec i))
1667 (aset rev-table ch i))))
1668 (set-char-table-extra-slot table 0 rev-table)
1671 (defun define-translation-table (symbol &rest args)
1672 "Define SYMBOL as the name of translation table made by ARGS.
1673 This sets up information so that the table can be used for
1674 translations in a CCL program.
1676 If the first element of ARGS is a char-table whose purpose is
1677 `translation-table', just define SYMBOL to name it. (Note that this
1678 function does not bind SYMBOL.)
1680 Any other ARGS should be suitable as arguments of the function
1681 `make-translation-table' (which see).
1683 This function sets properties `translation-table' and
1684 `translation-table-id' of SYMBOL to the created table itself and the
1685 identification number of the table respectively. It also registers
1686 the table in `translation-table-vector'."
1687 (let ((table (if (and (char-table-p (car args))
1688 (eq (char-table-subtype (car args))
1689 'translation-table))
1691 (apply 'make-translation-table args)))
1692 (len (length translation-table-vector))
1695 (put symbol 'translation-table table)
1698 (setq translation-table-vector
1699 (vconcat translation-table-vector (make-vector len nil))))
1700 (let ((slot (aref translation-table-vector id)))
1702 (eq (car slot) symbol))
1704 (aset translation-table-vector id (cons symbol table))
1706 (setq id (1+ id)))))
1707 (put symbol 'translation-table-id id)
1710 (put 'with-category-table 'lisp-indent-function 1)
1712 (defmacro with-category-table (category-table &rest body)
1713 `(let ((current-category-table (category-table)))
1714 (set-category-table ,category-table)
1717 (set-category-table current-category-table))))
1719 ;;; Initialize some variables.
1721 (put 'use-default-ascent 'char-table-extra-slots 0)
1722 (setq use-default-ascent (make-char-table 'use-default-ascent))
1723 (put 'ignore-relative-composition 'char-table-extra-slots 0)
1724 (setq ignore-relative-composition
1725 (make-char-table 'ignore-relative-composition))
1730 ;;; mule.el ends here