1 ;;; mule.el --- basic commands for mulitilingual environment
3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
4 ;; Licensed to the Free Software Foundation.
6 ;; Keywords: mule, multilingual, character set, coding system
8 ;; This file is part of GNU Emacs.
10 ;; GNU Emacs is free software; you can redistribute it and/or modify
11 ;; it under the terms of the GNU General Public License as published by
12 ;; the Free Software Foundation; either version 2, or (at your option)
15 ;; GNU Emacs is distributed in the hope that it will be useful,
16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 ;; GNU General Public License for more details.
20 ;; You should have received a copy of the GNU General Public License
21 ;; along with GNU Emacs; see the file COPYING. If not, write to the
22 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
23 ;; Boston, MA 02111-1307, USA.
27 (defconst mule-version "5.0 (SAKAKI)" "\
28 Version number and name of this version of MULE (multilingual environment).")
30 (defconst mule-version-date "1999.12.7" "\
31 Distribution date of this version of MULE (multilingual environment).")
33 (defun load-with-code-conversion (fullname file &optional noerror nomessage)
34 "Execute a file of Lisp code named FILE whose absolute name is FULLNAME.
35 The file contents are decoded before evaluation if necessary.
36 If optional second arg NOERROR is non-nil,
37 report no error if FILE doesn't exist.
38 Print messages at start and end of loading unless
39 optional third arg NOMESSAGE is non-nil.
40 Return t if file exists."
41 (if (null (file-readable-p fullname))
43 (signal 'file-error (list "Cannot open load file" file)))
44 ;; Read file with code conversion, and then eval.
46 ;; To avoid any autoloading, set default-major-mode to
48 ;; So that we don't get completely screwed if the
49 ;; file is encoded in some complicated character set,
50 ;; read it with real decoding, as a multibyte buffer,
51 ;; even if this is a --unibyte Emacs session.
52 (let ((default-major-mode 'fundamental-mode)
53 (default-enable-multibyte-characters t))
54 ;; We can't use `generate-new-buffer' because files.el
56 (get-buffer-create (generate-new-buffer-name " *load*"))))
58 (source (save-match-data (string-match "\\.el\\'" fullname))))
61 (message "Loading %s (source)..." file)
62 (message "Loading %s..." file)))
64 (setq preloaded-file-list (cons file preloaded-file-list)))
66 (let ((load-file-name fullname)
67 (set-auto-coding-for-load t)
68 (inhibit-file-name-operation nil))
71 (insert-file-contents fullname)
72 ;; If the loaded file was inserted with no-conversion or
73 ;; raw-text coding system, make the buffer unibyte.
74 ;; Otherwise, eval-buffer might try to interpret random
75 ;; binary junk as multibyte characters.
76 (if (and enable-multibyte-characters
77 (or (eq (coding-system-type last-coding-system-used) 5)
78 (eq last-coding-system-used 'no-conversion)))
79 (set-buffer-multibyte nil))
80 ;; Make `kill-buffer' quiet.
81 (set-buffer-modified-p nil))
82 ;; Have the original buffer current while we eval.
83 (eval-buffer buffer nil file
84 ;; If this Emacs is running with --unibyte,
85 ;; convert multibyte strings to unibyte
86 ;; after reading them.
87 ;; (not default-enable-multibyte-characters)
90 (let (kill-buffer-hook kill-buffer-query-functions)
91 (kill-buffer buffer)))
92 (let ((hook (assoc file after-load-alist)))
94 (mapcar (function eval) (cdr hook))))
95 (unless (or nomessage noninteractive)
97 (message "Loading %s (source)...done" file)
98 (message "Loading %s...done" file)))
101 ;; API (Application Program Interface) for charsets.
103 ;; Return t if OBJ is a quoted symbol
104 ;; and the symbol is the name of a standard charset.
105 (defsubst charset-quoted-standard-p (obj)
106 (and (listp obj) (eq (car obj) 'quote)
107 (symbolp (car-safe (cdr obj)))
108 (let ((vector (get (car-safe (cdr obj)) 'charset)))
109 (and (vectorp vector)
110 (< (aref vector 0) 160)))))
112 (defsubst charsetp (object)
113 "T if OBJECT is a charset."
114 (and (symbolp object) (vectorp (get object 'charset))))
116 (defsubst charset-info (charset)
117 "Return a vector of information of CHARSET.
118 The elements of the vector are:
119 CHARSET-ID, BYTES, DIMENSION, CHARS, WIDTH, DIRECTION,
120 LEADING-CODE-BASE, LEADING-CODE-EXT,
121 ISO-FINAL-CHAR, ISO-GRAPHIC-PLANE,
122 REVERSE-CHARSET, SHORT-NAME, LONG-NAME, DESCRIPTION,
125 CHARSET-ID (integer) is the identification number of the charset.
126 BYTES (integer) is the length of multi-byte form of a character in
127 the charset: one of 1, 2, 3, and 4.
128 DIMENSION (integer) is the number of bytes to represent a character of
130 CHARS (integer) is the number of characters in a dimension: 94 or 96.
131 WIDTH (integer) is the number of columns a character in the charset
132 occupies on the screen: one of 0, 1, and 2.
133 DIRECTION (integer) is the rendering direction of characters in the
134 charset when rendering. If 0, render from left to right, else
135 render from right to left.
136 LEADING-CODE-BASE (integer) is the base leading-code for the
138 LEADING-CODE-EXT (integer) is the extended leading-code for the
139 charset. All charsets of less than 0xA0 has the value 0.
140 ISO-FINAL-CHAR (character) is the final character of the
141 corresponding ISO 2022 charset. If the charset is not assigned
142 any final character, the value is -1.
143 ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked
144 while encoding to variants of ISO 2022 coding system, one of the
145 following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).
146 If the charset is not assigned any final character, the value is -1.
147 REVERSE-CHARSET (integer) is the charset which differs only in
148 LEFT-TO-RIGHT value from the charset. If there's no such a
149 charset, the value is -1.
150 SHORT-NAME (string) is the short name to refer to the charset.
151 LONG-NAME (string) is the long name to refer to the charset
152 DESCRIPTION (string) is the description string of the charset.
153 PLIST (property list) may contain any type of information a user
154 want to put and get by functions `put-charset-property' and
155 `get-charset-property' respectively."
156 (get charset 'charset))
158 ;; It is better not to use backquote in this file,
159 ;; because that makes a bootstrapping problem
160 ;; if you need to recompile all the Lisp files using interpreted code.
162 (defmacro charset-id (charset)
163 "Return charset identification number of CHARSET."
164 (if (charset-quoted-standard-p charset)
165 (aref (charset-info (nth 1 charset)) 0)
166 (list 'aref (list 'charset-info charset) 0)))
168 (defmacro charset-bytes (charset)
169 "Return bytes of CHARSET.
170 See the function `charset-info' for more detail."
171 (if (charset-quoted-standard-p charset)
172 (aref (charset-info (nth 1 charset)) 1)
173 (list 'aref (list 'charset-info charset) 1)))
175 (defmacro charset-dimension (charset)
176 "Return dimension of CHARSET.
177 See the function `charset-info' for more detail."
178 (if (charset-quoted-standard-p charset)
179 (aref (charset-info (nth 1 charset)) 2)
180 (list 'aref (list 'charset-info charset) 2)))
182 (defmacro charset-chars (charset)
183 "Return character numbers contained in a dimension of CHARSET.
184 See the function `charset-info' for more detail."
185 (if (charset-quoted-standard-p charset)
186 (aref (charset-info (nth 1 charset)) 3)
187 (list 'aref (list 'charset-info charset) 3)))
189 (defmacro charset-width (charset)
190 "Return width (how many column occupied on a screen) of CHARSET.
191 See the function `charset-info' for more detail."
192 (if (charset-quoted-standard-p charset)
193 (aref (charset-info (nth 1 charset)) 4)
194 (list 'aref (list 'charset-info charset) 4)))
196 (defmacro charset-direction (charset)
197 "Return direction of CHARSET.
198 See the function `charset-info' for more detail."
199 (if (charset-quoted-standard-p charset)
200 (aref (charset-info (nth 1 charset)) 5)
201 (list 'aref (list 'charset-info charset) 5)))
203 (defmacro charset-iso-final-char (charset)
204 "Return final char of CHARSET.
205 See the function `charset-info' for more detail."
206 (if (charset-quoted-standard-p charset)
207 (aref (charset-info (nth 1 charset)) 8)
208 (list 'aref (list 'charset-info charset) 8)))
210 (defmacro charset-iso-graphic-plane (charset)
211 "Return graphic plane of CHARSET.
212 See the function `charset-info' for more detail."
213 (if (charset-quoted-standard-p charset)
214 (aref (charset-info (nth 1 charset)) 9)
215 (list 'aref (list 'charset-info charset) 9)))
217 (defmacro charset-reverse-charset (charset)
218 "Return reverse charset of CHARSET.
219 See the function `charset-info' for more detail."
220 (if (charset-quoted-standard-p charset)
221 (aref (charset-info (nth 1 charset)) 10)
222 (list 'aref (list 'charset-info charset) 10)))
224 (defmacro charset-short-name (charset)
225 "Return short name of CHARSET.
226 See the function `charset-info' for more detail."
227 (if (charset-quoted-standard-p charset)
228 (aref (charset-info (nth 1 charset)) 11)
229 (list 'aref (list 'charset-info charset) 11)))
231 (defmacro charset-long-name (charset)
232 "Return long name of CHARSET.
233 See the function `charset-info' for more detail."
234 (if (charset-quoted-standard-p charset)
235 (aref (charset-info (nth 1 charset)) 12)
236 (list 'aref (list 'charset-info charset) 12)))
238 (defmacro charset-description (charset)
239 "Return description of CHARSET.
240 See the function `charset-info' for more detail."
241 (if (charset-quoted-standard-p charset)
242 (aref (charset-info (nth 1 charset)) 13)
243 (list 'aref (list 'charset-info charset) 13)))
245 (defmacro charset-plist (charset)
246 "Return list charset property of CHARSET.
247 See the function `charset-info' for more detail."
249 (if (charset-quoted-standard-p charset)
250 (charset-info (nth 1 charset))
251 (list 'charset-info charset))
254 (defun set-charset-plist (charset plist)
255 "Set CHARSET's property list to PLIST, and return PLIST."
256 (aset (charset-info charset) 14 plist))
258 (defun make-char (charset &optional code1 code2)
259 "Return a character of CHARSET whose position codes are CODE1 and CODE2.
260 CODE1 and CODE2 are optional, but if you don't supply
261 sufficient position codes, return a generic character which stands for
262 all characters or group of characters in the character set.
263 A generic character can be used to index a char table (e.g. syntax-table).
265 Such character sets as ascii, eight-bit-control, and eight-bit-graphic
266 don't have corresponding generic characters. If CHARSET is one of
267 them and you don't supply CODE1, return the character of the smallest
270 If CODE1 or CODE2 are invalid (out of range), this function signals an error."
271 (make-char-internal (charset-id charset) code1 code2))
273 (put 'make-char 'byte-compile
276 (let ((charset (nth 1 form)))
277 (if (charset-quoted-standard-p charset)
278 (byte-compile-normal-call
279 (cons 'make-char-internal
280 (cons (charset-id (nth 1 charset)) (nthcdr 2 form))))
281 (byte-compile-normal-call
282 (cons 'make-char-internal
283 (cons (list 'charset-id charset) (nthcdr 2 form)))))))))
285 (defun charset-list ()
286 "Return list of charsets ever defined.
288 This function is provided for backward compatibility.
289 Now we have the variable `charset-list'."
292 (defsubst generic-char-p (char)
293 "Return t if and only if CHAR is a generic character.
294 See also the documentation of make-char."
296 (let ((l (split-char char)))
297 (and (or (= (nth 1 l) 0) (eq (nth 2 l) 0))
298 (not (eq (car l) 'composition))))))
300 (defun decode-char (ccs code-point &optional restriction)
301 "Return character specified by coded character set CCS and CODE-POINT in it.
302 Return nil if such a character is not supported.
303 Currently the only supported coded character set is `ucs' (ISO/IEC
304 10646: Universal Multi-Octet Coded Character Set).
306 Optional argument RESTRICTION specifies a way to map the pair of CCS
307 and CODE-POINT to a chracter. Currently not supported and just ignored."
309 (cond ((< code-point 160)
312 (make-char 'latin-iso8859-1 code-point))
313 ((< code-point #x2500)
314 (setq code-point (- code-point #x0100))
315 (make-char 'mule-unicode-0100-24ff
316 (+ (/ code-point 96) 32) (+ (% code-point 96) 32)))
317 ((< code-point #x3400)
318 (setq code-point (- code-point #x2500))
319 (make-char 'mule-unicode-2500-33ff
320 (+ (/ code-point 96) 32) (+ (% code-point 96) 32)))
321 ((and (>= code-point #xe000) (< code-point #x10000))
322 (setq code-point (- code-point #xe000))
323 (make-char 'mule-unicode-e000-ffff
324 (+ (/ code-point 96) 32) (+ (% code-point 96) 32)))
327 (defun encode-char (char ccs &optional restriction)
328 "Return code-point in coded character set CCS that corresponds to CHAR.
329 Return nil if CHAR is not included in CCS.
330 Currently the only supported coded character set is `ucs' (ISO/IEC
331 10646: Universal Multi-Octet Coded Character Set).
333 CHAR should be in one of these charsets:
334 ascii, latin-iso8859-1, mule-unicode-0100-24ff, mule-unicode-2500-33ff,
335 mule-unicode-e000-ffff, eight-bit-control
336 Otherwise, return nil.
338 Optional argument RESTRICTION specifies a way to map CHAR to a
339 code-point in CCS. Currently not supported and just ignored."
340 (let* ((split (split-char char))
341 (charset (car split)))
343 (cond ((eq charset 'ascii)
345 ((eq charset 'latin-iso8859-1)
346 (+ (nth 1 split) 128))
347 ((eq charset 'mule-unicode-0100-24ff)
348 (+ #x0100 (+ (* (- (nth 1 split) 32) 96)
349 (- (nth 2 split) 32))))
350 ((eq charset 'mule-unicode-2500-33ff)
351 (+ #x2500 (+ (* (- (nth 1 split) 32) 96)
352 (- (nth 2 split) 32))))
353 ((eq charset 'mule-unicode-e000-ffff)
354 (+ #xe000 (+ (* (- (nth 1 split) 32) 96)
355 (- (nth 2 split) 32))))
356 ((eq charset 'eight-bit-control)
360 ;; Coding system stuff
362 ;; Coding system is a symbol that has the property `coding-system'.
364 ;; The value of the property `coding-system' is a vector of the
366 ;; [TYPE MNEMONIC DOC-STRING PLIST FLAGS]
367 ;; We call this vector as coding-spec. See comments in src/coding.c
370 (defconst coding-spec-type-idx 0)
371 (defconst coding-spec-mnemonic-idx 1)
372 (defconst coding-spec-doc-string-idx 2)
373 (defconst coding-spec-plist-idx 3)
374 (defconst coding-spec-flags-idx 4)
376 ;; PLIST is a property list of a coding system. To share PLIST among
377 ;; alias coding systems, a coding system has PLIST in coding-spec
378 ;; instead of having it in normal property list of Lisp symbol.
379 ;; Here's a list of coding system properties currently being used.
383 ;; The value is a coding category the coding system belongs to. The
384 ;; function `make-coding-system' sets this value automatically
385 ;; unless its argument PROPERTIES specifies this property.
387 ;; o alias-coding-systems
389 ;; The value is a list of coding systems of the same alias group. The
390 ;; first element is the coding system made at first, which we call as
391 ;; `base coding system'. The function `make-coding-system' sets this
392 ;; value automatically and `define-coding-system-alias' updates it.
394 ;; See the documentation of make-coding-system for the meanings of the
395 ;; following properties.
397 ;; o post-read-conversion
398 ;; o pre-write-conversion
399 ;; o translation-table-for-decode
400 ;; o translation-table-for-encode
404 ;; o valid-codes (meaningful only for a coding system based on CCL)
407 ;; Return coding-spec of CODING-SYSTEM
408 (defsubst coding-system-spec (coding-system)
409 (get (check-coding-system coding-system) 'coding-system))
411 (defun coding-system-type (coding-system)
412 "Return the coding type of CODING-SYSTEM.
413 A coding type is an integer value indicating the encoding method
414 of CODING-SYSTEM. See the function `make-coding-system' for more detail."
415 (aref (coding-system-spec coding-system) coding-spec-type-idx))
417 (defun coding-system-mnemonic (coding-system)
418 "Return the mnemonic character of CODING-SYSTEM.
419 The mnemonic character of a coding system is used in mode line
420 to indicate the coding system. If the arg is nil, return ?-."
421 (let ((spec (coding-system-spec coding-system)))
422 (if spec (aref spec coding-spec-mnemonic-idx) ?-)))
424 (defun coding-system-doc-string (coding-system)
425 "Return the documentation string for CODING-SYSTEM."
426 (aref (coding-system-spec coding-system) coding-spec-doc-string-idx))
428 (defun coding-system-plist (coding-system)
429 "Return the property list of CODING-SYSTEM."
430 (aref (coding-system-spec coding-system) coding-spec-plist-idx))
432 (defun coding-system-flags (coding-system)
433 "Return `flags' of CODING-SYSTEM.
434 A `flags' of a coding system is a vector of length 32 indicating detailed
435 information of a coding system. See the function `make-coding-system'
437 (aref (coding-system-spec coding-system) coding-spec-flags-idx))
439 (defun coding-system-get (coding-system prop)
440 "Extract a value from CODING-SYSTEM's property list for property PROP."
441 (plist-get (coding-system-plist coding-system) prop))
443 (defun coding-system-put (coding-system prop val)
444 "Change value in CODING-SYSTEM's property list PROP to VAL."
445 (let ((plist (coding-system-plist coding-system)))
447 (plist-put plist prop val)
448 (aset (coding-system-spec coding-system) coding-spec-plist-idx
451 (defun coding-system-category (coding-system)
452 "Return the coding category of CODING-SYSTEM."
453 (coding-system-get coding-system 'coding-category))
455 (defun coding-system-base (coding-system)
456 "Return the base coding system of CODING-SYSTEM.
457 A base coding system is what made by `make-coding-system'.
458 Any alias nor subsidiary coding systems are not base coding system."
459 (car (coding-system-get coding-system 'alias-coding-systems)))
461 (defalias 'coding-system-parent 'coding-system-base)
462 (make-obsolete 'coding-system-parent 'coding-system-base "20.3")
464 ;; Coding system also has a property `eol-type'.
466 ;; This property indicates how the coding system handles end-of-line
467 ;; format. The value is integer 0, 1, 2, or a vector of three coding
468 ;; systems. Each integer value 0, 1, and 2 indicates the format of
469 ;; end-of-line LF, CRLF, and CR respectively. A vector value
470 ;; indicates that the format of end-of-line should be detected
471 ;; automatically. Nth element of the vector is the subsidiary coding
472 ;; system whose `eol-type' property is N.
474 (defun coding-system-eol-type (coding-system)
475 "Return eol-type of CODING-SYSTEM.
476 An eol-type is integer 0, 1, 2, or a vector of coding systems.
478 Integer values 0, 1, and 2 indicate a format of end-of-line; LF,
479 CRLF, and CR respectively.
481 A vector value indicates that a format of end-of-line should be
482 detected automatically. Nth element of the vector is the subsidiary
483 coding system whose eol-type is N."
484 (get coding-system 'eol-type))
486 (defun coding-system-lessp (x y)
487 (cond ((eq x 'no-conversion) t)
488 ((eq y 'no-conversion) nil)
489 ((eq x 'emacs-mule) t)
490 ((eq y 'emacs-mule) nil)
491 ((eq x 'undecided) t)
492 ((eq y 'undecided) nil)
493 (t (let ((c1 (coding-system-mnemonic x))
494 (c2 (coding-system-mnemonic y)))
495 (or (< (downcase c1) (downcase c2))
496 (and (not (> (downcase c1) (downcase c2)))
499 ;; Add CODING-SYSTEM to coding-system-list while keeping it sorted.
500 (defun add-to-coding-system-list (coding-system)
501 (if (or (null coding-system-list)
502 (coding-system-lessp coding-system (car coding-system-list)))
503 (setq coding-system-list (cons coding-system coding-system-list))
504 (let ((len (length coding-system-list))
505 mid (tem coding-system-list))
507 (setq mid (nthcdr (/ len 2) tem))
508 (if (coding-system-lessp (car mid) coding-system)
510 len (- len (/ len 2)))
511 (setq len (/ len 2))))
512 (setcdr tem (cons coding-system (cdr tem))))))
514 (defun coding-system-list (&optional base-only)
515 "Return a list of all existing non-subsidiary coding systems.
516 If optional arg BASE-ONLY is non-nil, only base coding systems are listed.
517 The value doesn't include subsidiary coding systems which are what
518 made from bases and aliases automatically for various end-of-line
519 formats (e.g. iso-latin-1-unix, koi8-r-dos)."
520 (let* ((codings (copy-sequence coding-system-list))
521 (tail (cons nil codings)))
522 ;; Remove subsidiary coding systems (eol variants) and alias
523 ;; coding systems (if necessary).
525 (let* ((coding (car (cdr tail)))
526 (aliases (coding-system-get coding 'alias-coding-systems)))
528 ;; CODING is an eol variant if not in ALIASES.
529 (not (memq coding aliases))
530 ;; CODING is an alias if it is not car of ALIASES.
531 (and base-only (not (eq coding (car aliases)))))
532 (setcdr tail (cdr (cdr tail)))
533 (setq tail (cdr tail)))))
536 (defun register-char-codings (coding-system safe-chars)
537 (let ((general (char-table-extra-slot char-coding-system-table 0)))
538 (if (eq safe-chars t)
539 (or (memq coding-system general)
540 (set-char-table-extra-slot char-coding-system-table 0
541 (cons coding-system general)))
545 (if (and (>= key 128) val)
546 (let ((codings (aref char-coding-system-table key)))
547 (or (memq coding-system codings)
548 (aset char-coding-system-table key
549 (cons coding-system codings)))))))
553 ;; Make subsidiary coding systems (eol-type variants) of CODING-SYSTEM.
554 (defun make-subsidiary-coding-system (coding-system)
555 (let ((coding-spec (coding-system-spec coding-system))
556 (subsidiaries (vector (intern (format "%s-unix" coding-system))
557 (intern (format "%s-dos" coding-system))
558 (intern (format "%s-mac" coding-system))))
562 (put (aref subsidiaries i) 'coding-system coding-spec)
563 (put (aref subsidiaries i) 'eol-type i)
564 (add-to-coding-system-list (aref subsidiaries i))
565 (setq coding-system-alist
566 (cons (list (symbol-name (aref subsidiaries i)))
567 coding-system-alist))
571 (defun transform-make-coding-system-args (name type &optional doc-string props)
572 "For internal use only.
573 Transform XEmacs style args for `make-coding-system' to Emacs style.
574 Value is a list of transformed arguments."
575 (let ((mnemonic (string-to-char (or (plist-get props 'mnemonic) "?")))
576 (eol-type (plist-get props 'eol-type))
579 ((eq eol-type 'lf) (setq eol-type 'unix))
580 ((eq eol-type 'crlf) (setq eol-type 'dos))
581 ((eq eol-type 'cr) (setq eol-type 'mac)))
582 (if (setq tmp (plist-get props 'post-read-conversion))
583 (setq properties (plist-put properties 'post-read-conversion tmp)))
584 (if (setq tmp (plist-get props 'pre-write-conversion))
585 (setq properties (plist-put properties 'pre-write-conversion tmp)))
587 ((eq type 'shift-jis)
588 `(,name 1 ,mnemonic ,doc-string () ,properties ,eol-type))
589 ((eq type 'iso2022) ; This is not perfect.
590 (if (plist-get props 'escape-quoted)
591 (error "escape-quoted is not supported: %S"
592 `(,name ,type ,doc-string ,props)))
593 (let ((g0 (plist-get props 'charset-g0))
594 (g1 (plist-get props 'charset-g1))
595 (g2 (plist-get props 'charset-g2))
596 (g3 (plist-get props 'charset-g3))
599 (eq (cadr (assoc 'latin-jisx0201
600 (plist-get props 'input-charset-conversion)))
602 (eq (cadr (assoc 'ascii
603 (plist-get props 'output-charset-conversion)))
607 (eq (cadr (assoc 'japanese-jisx0208-1978
608 (plist-get props 'input-charset-conversion)))
610 (eq (cadr (assoc 'japanese-jisx0208
611 (plist-get props 'output-charset-conversion)))
612 'japanese-jisx0208-1978))))
614 (if (plist-get props 'force-g0-on-output)
618 (if (plist-get props 'force-g1-on-output)
622 (if (plist-get props 'force-g2-on-output)
626 (if (plist-get props 'force-g3-on-output)
629 `(,name 2 ,mnemonic ,doc-string
631 ,(plist-get props 'short)
632 ,(not (plist-get props 'no-ascii-eol))
633 ,(not (plist-get props 'no-ascii-cntl))
634 ,(plist-get props 'seven)
636 ,(not (plist-get props 'lock-shift))
639 ,(plist-get props 'no-iso6429)
641 ,properties ,eol-type)))
643 `(,name 3 ,mnemonic ,doc-string () ,properties ,eol-type))
645 `(,name 4 ,mnemonic ,doc-string
646 (,(plist-get props 'decode) . ,(plist-get props 'encode))
647 ,properties ,eol-type))
649 (error "unsupported XEmacs style make-coding-style arguments: %S"
650 `(,name ,type ,doc-string ,props))))))
652 (defun make-coding-system (coding-system type mnemonic doc-string
657 "Define a new coding system CODING-SYSTEM (symbol).
658 Remaining arguments are TYPE, MNEMONIC, DOC-STRING, FLAGS (optional),
659 and PROPERTIES (optional) which construct a coding-spec of CODING-SYSTEM
660 in the following format:
661 [TYPE MNEMONIC DOC-STRING PLIST FLAGS]
663 TYPE is an integer value indicating the type of the coding system as follows:
664 0: Emacs internal format,
665 1: Shift-JIS (or MS-Kanji) used mainly on Japanese PC,
666 2: ISO-2022 including many variants,
667 3: Big5 used mainly on Chinese PC,
668 4: private, CCL programs provide encoding/decoding algorithm,
669 5: Raw-text, which means that text contains random 8-bit codes.
671 MNEMONIC is a character to be displayed on mode line for the coding system.
673 DOC-STRING is a documentation string for the coding system.
675 FLAGS specifies more detailed information of the coding system as follows:
677 If TYPE is 2 (ISO-2022), FLAGS is a list of these elements:
678 CHARSET0, CHARSET1, CHARSET2, CHARSET3, SHORT-FORM,
679 ASCII-EOL, ASCII-CNTL, SEVEN, LOCKING-SHIFT, SINGLE-SHIFT,
680 USE-ROMAN, USE-OLDJIS, NO-ISO6429, INIT-BOL, DESIGNATION-BOL,
681 SAFE, ACCEPT-LATIN-EXTRA-CODE.
682 CHARSETn are character sets initially designated to Gn graphic registers.
683 If CHARSETn is nil, Gn is never used.
684 If CHARSETn is t, Gn can be used but nothing designated initially.
685 If CHARSETn is a list of character sets, those character sets are
686 designated to Gn on output, but nothing designated to Gn initially.
687 But, character set `ascii' can be designated only to G0.
688 SHORT-FORM non-nil means use short designation sequence on output.
689 ASCII-EOL non-nil means designate ASCII to g0 at end of line on output.
690 ASCII-CNTL non-nil means designate ASCII to g0 before control codes and
692 SEVEN non-nil means use 7-bit code only on output.
693 LOCKING-SHIFT non-nil means use locking-shift.
694 SINGLE-SHIFT non-nil means use single-shift.
695 USE-ROMAN non-nil means designate JIS0201-1976-Roman instead of ASCII.
696 USE-OLDJIS non-nil means designate JIS0208-1976 instead of JIS0208-1983.
697 NO-ISO6429 non-nil means not use ISO6429's direction specification.
698 INIT-BOL non-nil means any designation state is assumed to be reset
699 to initial at each beginning of line on output.
700 DESIGNATION-BOL non-nil means designation sequences should be placed
701 at beginning of line on output.
702 SAFE non-nil means convert unsafe characters to `?' on output.
703 Characters not specified in the property `safe-charsets' nor
704 `safe-chars' are unsafe.
705 ACCEPT-LATIN-EXTRA-CODE non-nil means code-detection routine accepts
706 a code specified in `latin-extra-code-table' (which see) as a valid
707 code of the coding system.
709 If TYPE is 4 (private), FLAGS should be a cons of CCL programs, for
710 decoding and encoding. CCL programs should be specified by their
713 PROPERTIES is an alist of properties vs the corresponding values. The
714 following properties are recognized:
716 o post-read-conversion
718 The value is a function to call after some text is inserted and
719 decoded by the coding system itself and before any functions in
720 `after-insert-functions' are called. The arguments to this
721 function is the same as those of a function in
722 `after-insert-functions', i.e. LENGTH of a text while putting point
723 at the head of the text to be decoded
725 o pre-write-conversion
727 The value is a function to call after all functions in
728 `write-region-annotate-functions' and `buffer-file-format' are
729 called, and before the text is encoded by the coding system itself.
730 The arguments to this function is the same as those of a function
731 in `write-region-annotate-functions', i.e. FROM and TO specifying
734 o translation-table-for-decode
736 The value is a translation table to be applied on decoding. See
737 the function `make-translation-table' for the format of translation
740 o translation-table-for-encode
742 The value is a translation table to be applied on encoding.
746 The value is a char table. If a character has non-nil value in it,
747 the character is safely supported by the coding system. This
748 overrides the specification of safe-charsets.
752 The value is a list of charsets safely supported by the coding
753 system. The value t means that all charsets Emacs handles are
754 supported. Even if some charset is not in this list, it doesn't
755 mean that the charset can't be encoded in the coding system,
756 instead, it just means that some other receiver of a text encoded
757 in the coding system won't be able to handle that charset.
761 The value is a symbol of which name is `MIME-charset' parameter of
764 o valid-codes (meaningful only for a coding system based on CCL)
766 The value is a list to indicate valid byte ranges of the encoded
767 file. Each element of the list is an integer or a cons of integer.
768 In the former case, the integer value is a valid byte code. In the
769 latter case, the integers specifies the range of valid byte codes.
771 These properties are set in PLIST, a property list. This function
772 also sets properties `coding-category' and `alias-coding-systems'
775 EOL-TYPE specifies the EOL type of the coding-system in one of the
778 o symbol (unix, dos, or mac)
780 The symbol `unix' means Unix-like EOL (LF), `dos' means
781 DOS-like EOL (CRLF), and `mac' means MAC-like EOL (CR).
783 o number (0, 1, or 2)
785 The number 0, 1, and 2 mean UNIX, DOS, and MAC-like EOL
788 o vector of coding-systems of length 3
790 The EOL type is detected automatically for the coding system.
791 And, according to the detected EOL type, one of the coding
792 systems in the vector is selected. Elements of the vector
793 corresponds to Unix-like EOL, DOS-like EOL, and Mac-like EOL
796 Kludgy features for backward compatibility:
798 1. If TYPE is 4 and car or cdr of FLAGS is a vector, the vector is
799 treated as a compiled CCL code.
801 2. If PROPERTIES is just a list of character sets, the list is set as
802 a value of `safe-charsets' in PLIST."
804 ;; For compatiblity with XEmacs, we check the type of TYPE. If it
805 ;; is a symbol, perhaps, this function is called with XEmacs-style
806 ;; arguments. Here, try to transform that kind of arguments to
809 (let ((args (transform-make-coding-system-args coding-system type
810 mnemonic doc-string)))
811 (setq coding-system (car args)
813 mnemonic (nth 2 args)
814 doc-string (nth 3 args)
816 properties (nth 5 args)
817 eol-type (nth 6 args))))
819 ;; Set a value of `coding-system' property.
820 (let ((coding-spec (make-vector 5 nil))
821 (no-initial-designation t)
822 (no-alternative-designation t)
823 (accept-latin-extra-code nil)
825 (if (or (not (integerp type)) (< type 0) (> type 5))
826 (error "TYPE argument must be 0..5"))
827 (if (or (not (integerp mnemonic)) (<= mnemonic ? ) (> mnemonic 127))
828 (error "MNEMONIC argument must be an ASCII printable character."))
829 (aset coding-spec coding-spec-type-idx type)
830 (aset coding-spec coding-spec-mnemonic-idx mnemonic)
831 (aset coding-spec coding-spec-doc-string-idx
832 (purecopy (if (stringp doc-string) doc-string "")))
834 (setq coding-category 'coding-category-emacs-mule))
836 (setq coding-category 'coding-category-sjis))
837 ((= type 2) ; ISO2022
839 (vec (make-vector 32 nil))
843 (let ((charset (car fl)))
844 (if (and no-initial-designation
846 (or (charsetp charset)
848 (charsetp (car charset)))))
849 (setq no-initial-designation nil))
850 (if (charsetp charset)
851 (if (= i 1) (setq g1-designation charset))
856 (setq elt (car tail))
858 (setq no-alternative-designation nil)
859 (if (and elt (not (charsetp elt)))
860 (error "Invalid charset: %s" elt)))
861 (setq tail (cdr tail)))
862 (setq g1-designation (car charset)))
865 (setq no-alternative-designation nil)
866 (error "Invalid charset: %s" charset)))))
867 (aset vec i charset))
868 (setq fl (cdr fl) i (1+ i)))
869 (while (and (< i 32) fl)
870 (aset vec i (car fl))
871 (if (and (= i 16) ; ACCEPT-LATIN-EXTRA-CODE
873 (setq accept-latin-extra-code t))
874 (setq fl (cdr fl) i (1+ i)))
875 (aset coding-spec 4 vec)
876 (setq coding-category
877 (if (aref vec 8) ; Use locking-shift.
878 (or (and (aref vec 7) 'coding-category-iso-7-else)
879 'coding-category-iso-8-else)
880 (if (aref vec 7) ; 7-bit only.
881 (if (aref vec 9) ; Use single-shift.
882 'coding-category-iso-7-else
883 (if no-alternative-designation
884 'coding-category-iso-7-tight
885 'coding-category-iso-7))
886 (if (or no-initial-designation
887 (not no-alternative-designation))
888 'coding-category-iso-8-else
889 (if (and (charsetp g1-designation)
890 (= (charset-dimension g1-designation) 2))
891 'coding-category-iso-8-2
892 'coding-category-iso-8-1)))))))
894 (setq coding-category 'coding-category-big5))
895 ((= type 4) ; private
896 (setq coding-category 'coding-category-ccl)
897 (if (not (consp flags))
898 (error "Invalid FLAGS argument for TYPE 4 (CCL)")
899 (let ((decoder (check-ccl-program
901 (intern (format "%s-decoder" coding-system))))
902 (encoder (check-ccl-program
904 (intern (format "%s-encoder" coding-system)))))
905 (if (and decoder encoder)
906 (aset coding-spec 4 (cons decoder encoder))
907 (error "Invalid FLAGS argument for TYPE 4 (CCL)")))))
909 (setq coding-category 'coding-category-raw-text)))
911 (let ((plist (list 'coding-category coding-category
912 'alias-coding-systems (list coding-system))))
913 (if no-initial-designation
914 (plist-put plist 'no-initial-designation t))
916 (or (eq properties t)
917 (not (consp (car properties)))))
918 ;; In the old version, the arg PROPERTIES is a list to be
919 ;; set in PLIST as a value of property `safe-charsets'.
920 (setq properties (list (cons 'safe-charsets properties))))
921 ;; In the current version PROPERTIES is a property list.
922 ;; Reflect it into PLIST one by one while handling safe-chars
924 (let ((safe-charsets (cdr (assq 'safe-charsets properties)))
925 (safe-chars (cdr (assq 'safe-chars properties)))
928 ;; If only safe-charsets is specified, make a char-table from
929 ;; it, and store that char-table as the value of `safe-chars'.
930 (if (and (not safe-chars) safe-charsets)
932 (if (eq safe-charsets t)
934 (setq safe-chars (make-char-table 'safe-chars))
936 (setq charset (car safe-charsets)
937 safe-charsets (cdr safe-charsets))
938 (cond ((eq charset 'ascii)) ; just ignore
939 ((eq charset 'eight-bit-control)
942 (aset safe-chars i t)
944 ((eq charset 'eight-bit-graphic)
947 (aset safe-chars i t)
950 (aset safe-chars (make-char charset) t))))
951 (if accept-latin-extra-code
954 (if (aref latin-extra-code-table i)
955 (aset safe-chars i t))
957 (setq l (cons (cons 'safe-chars safe-chars) l))))
959 (setq prop (car (car l)) val (cdr (car l)) l (cdr l))
960 (if (eq prop 'safe-chars)
962 (if (and (symbolp val)
963 (get val 'translation-table))
964 (setq safe-chars (get val 'translation-table)))
965 (register-char-codings coding-system safe-chars)
966 (setq val safe-chars)))
967 (plist-put plist prop val)))
968 ;; The property `coding-category' may have been set differently
969 ;; through PROPERTIES.
970 (setq coding-category (plist-get plist 'coding-category))
971 (aset coding-spec coding-spec-plist-idx plist))
972 (put coding-system 'coding-system coding-spec)
973 (put coding-category 'coding-systems
974 (cons coding-system (get coding-category 'coding-systems))))
976 ;; Next, set a value of `eol-type' property.
978 ;; If EOL-TYPE is nil, set a vector of subsidiary coding
979 ;; systems, each corresponds to a coding system for the detected
981 (setq eol-type (make-subsidiary-coding-system coding-system)))
983 (cond ((or (eq eol-type 'unix) (null eol-type))
989 ((or (and (vectorp eol-type)
990 (= (length eol-type) 3))
991 (and (numberp eol-type)
996 (error "Invalid EOL-TYPE spec:%S" eol-type))))
997 (put coding-system 'eol-type eol-type)
999 ;; At last, register CODING-SYSTEM in `coding-system-list' and
1000 ;; `coding-system-alist'.
1001 (add-to-coding-system-list coding-system)
1002 (setq coding-system-alist (cons (list (symbol-name coding-system))
1003 coding-system-alist))
1005 ;; For a coding system of cateogory iso-8-1 and iso-8-2, create
1006 ;; XXX-with-esc variants.
1007 (let ((coding-category (coding-system-category coding-system)))
1008 (if (or (eq coding-category 'coding-category-iso-8-1)
1009 (eq coding-category 'coding-category-iso-8-2))
1010 (let ((esc (intern (concat (symbol-name coding-system) "-with-esc")))
1011 (doc (format "Same as %s but can handle any charsets by ISO's escape sequences." coding-system))
1012 (safe-charsets (assq 'safe-charsets properties))
1013 (mime-charset (assq 'mime-charset properties)))
1015 (setcdr safe-charsets t)
1016 (setq properties (cons (cons 'safe-charsets t) properties)))
1018 (setcdr mime-charset nil))
1019 (make-coding-system esc type mnemonic doc
1020 (if (listp (car flags))
1021 (cons (append (car flags) '(t)) (cdr flags))
1022 (cons (list (car flags) t) (cdr flags)))
1027 (defun define-coding-system-alias (alias coding-system)
1028 "Define ALIAS as an alias for coding system CODING-SYSTEM."
1029 (put alias 'coding-system (coding-system-spec coding-system))
1030 (nconc (coding-system-get alias 'alias-coding-systems) (list alias))
1031 (add-to-coding-system-list alias)
1032 (setq coding-system-alist (cons (list (symbol-name alias))
1033 coding-system-alist))
1034 (let ((eol-type (coding-system-eol-type coding-system)))
1035 (if (vectorp eol-type)
1036 (put alias 'eol-type (make-subsidiary-coding-system alias))
1037 (put alias 'eol-type eol-type))))
1039 (defun set-buffer-file-coding-system (coding-system &optional force)
1040 "Set the file coding-system of the current buffer to CODING-SYSTEM.
1041 This means that when you save the buffer, it will be converted
1042 according to CODING-SYSTEM. For a list of possible values of CODING-SYSTEM,
1043 use \\[list-coding-systems].
1045 If the buffer's previous file coding-system value specifies end-of-line
1046 conversion, and CODING-SYSTEM does not specify one, CODING-SYSTEM is
1047 merged with the already-specified end-of-line conversion.
1049 If the buffer's previous file coding-system value specifies text
1050 conversion, and CODING-SYSTEM does not specify one, CODING-SYSTEM is
1051 merged with the already-specified text conversion.
1053 However, if the optional prefix argument FORCE is non-nil, then
1054 CODING-SYSTEM is used exactly as specified.
1056 This marks the buffer modified so that the succeeding \\[save-buffer]
1057 surely saves the buffer with CODING-SYSTEM. From a program, if you
1058 don't want to mark the buffer modified, just set the variable
1059 `buffer-file-coding-system' directly."
1060 (interactive "zCoding system for visited file (default, nil): \nP")
1061 (check-coding-system coding-system)
1062 (if (and coding-system buffer-file-coding-system (null force))
1063 (let ((base (coding-system-base buffer-file-coding-system))
1064 (eol (coding-system-eol-type buffer-file-coding-system)))
1065 ;; If CODING-SYSTEM doesn't specify text conversion, merge
1066 ;; with that of buffer-file-coding-system.
1067 (if (eq (coding-system-base coding-system) 'undecided)
1068 (setq coding-system (coding-system-change-text-conversion
1069 coding-system base)))
1070 ;; If CODING-SYSTEM doesn't specify eol conversion, merge with
1071 ;; that of buffer-file-coding-system.
1072 (if (and (vectorp (coding-system-eol-type coding-system))
1073 (numberp eol) (>= eol 0) (<= eol 2))
1074 (setq coding-system (coding-system-change-eol-conversion
1075 coding-system eol)))))
1076 (setq buffer-file-coding-system coding-system)
1077 (set-buffer-modified-p t)
1078 (force-mode-line-update))
1080 (defvar default-terminal-coding-system nil
1081 "Default value for the terminal coding system.
1082 This is normally set according to the selected language environment.
1083 See also the command `set-terminal-coding-system'.")
1085 (defun set-terminal-coding-system (coding-system)
1086 "Set coding system of your terminal to CODING-SYSTEM.
1087 All text output to the terminal will be encoded
1088 with the specified coding system.
1089 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems].
1090 The default is determined by the selected language environment
1091 or by the previous use of this command."
1093 (list (let ((default (if (and (not (terminal-coding-system))
1094 default-terminal-coding-system)
1095 default-terminal-coding-system)))
1097 (format "Coding system for terminal display (default, %s): "
1100 (if (and (not coding-system)
1101 (not (terminal-coding-system)))
1102 (setq coding-system default-terminal-coding-system))
1104 (setq default-terminal-coding-system coding-system))
1105 (set-terminal-coding-system-internal coding-system)
1106 (redraw-frame (selected-frame)))
1108 (defvar default-keyboard-coding-system nil
1109 "Default value of the keyboard coding system.
1110 This is normally set according to the selected language environment.
1111 See also the command `set-keyboard-coding-system'.")
1113 (defun set-keyboard-coding-system (coding-system)
1114 "Set coding system for keyboard input to CODING-SYSTEM.
1115 In addition, this command enables Encoded-kbd minor mode.
1116 \(If CODING-SYSTEM is nil, Encoded-kbd mode is turned off -- see
1117 `encoded-kbd-mode'.)
1118 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems].
1119 The default is determined by the selected language environment
1120 or by the previous use of this command."
1122 (list (let ((default (if (and (not (keyboard-coding-system))
1123 default-keyboard-coding-system)
1124 default-keyboard-coding-system)))
1126 (format "Coding system for keyboard input (default, %s): "
1129 (if (and (not coding-system)
1130 (not (keyboard-coding-system)))
1131 (setq coding-system default-keyboard-coding-system))
1133 (setq default-keyboard-coding-system coding-system))
1134 (set-keyboard-coding-system-internal coding-system)
1135 (encoded-kbd-mode (if coding-system 1 0)))
1137 (defcustom keyboard-coding-system nil
1138 "Specify coding system for keyboard input.
1139 If you set this on a terminal which can't distinguish Meta keys from
1140 8-bit characters, you will have to use ESC to type Meta characters.
1141 See Info node `Specify Coding' and Info node `Single-Byte Character Support'.
1143 Setting this variable directly does not take effect;
1144 use either M-x customize or \\[set-keyboard-coding-system]."
1145 :type '(coding-system :tag "Coding system")
1146 :link '(info-link "(emacs)Specify Coding")
1147 :link '(info-link "(emacs)Single-Byte Character Support")
1148 :set (lambda (symbol value)
1149 ;; Don't load encoded-kbd-mode unnecessarily.
1150 (if (or value (boundp 'encoded-kbd-mode))
1151 (set-keyboard-coding-system value)
1152 (set-default 'keyboard-coding-system nil))) ; must initialize
1157 (defun set-buffer-process-coding-system (decoding encoding)
1158 "Set coding systems for the process associated with the current buffer.
1159 DECODING is the coding system to be used to decode input from the process,
1160 ENCODING is the coding system to be used to encode output to the process.
1162 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems]."
1164 "zCoding-system for output from the process: \nzCoding-system for input to the process: ")
1165 (let ((proc (get-buffer-process (current-buffer))))
1167 (error "no process")
1168 (check-coding-system decoding)
1169 (check-coding-system encoding)
1170 (set-process-coding-system proc decoding encoding)))
1171 (force-mode-line-update))
1173 (defalias 'set-clipboard-coding-system 'set-selection-coding-system)
1175 (defun set-selection-coding-system (coding-system)
1176 "Make CODING-SYSTEM used for communicating with other X clients .
1177 When sending or receiving text via cut_buffer, selection, and clipboard,
1178 the text is encoded or decoded by CODING-SYSTEM."
1179 (interactive "zCoding system for X selection: ")
1180 (check-coding-system coding-system)
1181 (setq selection-coding-system coding-system))
1183 ;; Coding system lastly specified by the command
1184 ;; set-next-selection-coding-system.
1185 (defvar last-next-selection-coding-system nil)
1187 (defun set-next-selection-coding-system (coding-system)
1188 "Make CODING-SYSTEM used for the next communication with other X clients.
1189 This setting is effective for the next communication only."
1191 (list (read-coding-system
1192 (if last-next-selection-coding-system
1193 (format "Coding system for the next X selection (default, %S): "
1194 last-next-selection-coding-system)
1195 "Coding system for the next X selection: ")
1196 last-next-selection-coding-system)))
1198 (setq last-next-selection-coding-system coding-system)
1199 (setq coding-system last-next-selection-coding-system))
1200 (check-coding-system coding-system)
1202 (setq next-selection-coding-system coding-system))
1204 (defun set-coding-priority (arg)
1205 "Set priority of coding categories according to LIST.
1206 LIST is a list of coding categories ordered by priority."
1208 (current-list (copy-sequence coding-category-list)))
1209 ;; Check the validity of ARG while deleting coding categories in
1210 ;; ARG from CURRENT-LIST. We assume that CODING-CATEGORY-LIST
1211 ;; contains all coding categories.
1213 (if (or (null (get (car l) 'coding-category-index))
1214 (null (memq (car l) current-list)))
1215 (error "Invalid or duplicated element in argument: %s" arg))
1216 (setq current-list (delq (car l) current-list))
1218 ;; Update `coding-category-list' and return it.
1219 (setq coding-category-list (append arg current-list))
1220 (set-coding-priority-internal)))
1224 (defcustom auto-coding-alist
1225 '(("\\.\\(arc\\|zip\\|lzh\\|zoo\\|jar\\|tar\\|tgz\\)\\'" . no-conversion)
1226 ("\\.\\(ARC\\|ZIP\\|LZH\\|ZOO\\|JAR\\|TAR\\|TGZ\\)\\'" . no-conversion))
1227 "Alist of filename patterns vs corresponding coding systems.
1228 Each element looks like (REGEXP . CODING-SYSTEM).
1229 A file whose name matches REGEXP is decoded by CODING-SYSTEM on reading.
1231 The settings in this alist take priority over `coding:' tags
1232 in the file (see the function `set-auto-coding')
1233 and the contents of `file-coding-system-alist'."
1236 :type '(repeat (cons (regexp :tag "File name regexp")
1237 (symbol :tag "Coding system"))))
1239 (defvar set-auto-coding-for-load nil
1240 "Non-nil means look for `load-coding' property instead of `coding'.
1241 This is used for loading and byte-compiling Emacs Lisp files.")
1243 (defun auto-coding-alist-lookup (filename)
1244 "Return the coding system specified by `auto-coding-alist' for FILENAME."
1245 (let ((alist auto-coding-alist)
1246 (case-fold-search (memq system-type '(vax-vms windows-nt ms-dos)))
1248 (while (and alist (not coding-system))
1249 (if (string-match (car (car alist)) filename)
1250 (setq coding-system (cdr (car alist)))
1251 (setq alist (cdr alist))))
1254 (defun set-auto-coding (filename size)
1255 "Return coding system for a file FILENAME of which SIZE bytes follow point.
1256 These bytes should include at least the first 1k of the file
1257 and the last 3k of the file, but the middle may be omitted.
1259 It checks FILENAME against the variable `auto-coding-alist'.
1260 If FILENAME doesn't match any entries in the variable,
1261 it checks for a `coding:' tag in the first one or two lines following
1262 point. If no `coding:' tag is found, it checks for local variables
1263 list in the last 3K bytes out of the SIZE bytes.
1265 The return value is the specified coding system,
1266 or nil if nothing specified.
1268 The variable `set-auto-coding-function' (which see) is set to this
1269 function by default."
1270 (let ((coding-system (auto-coding-alist-lookup filename)))
1273 (let* ((case-fold-search t)
1274 (head-start (point))
1275 (head-end (+ head-start (min size 1024)))
1276 (tail-start (+ head-start (max (- size 3072) 0)))
1277 (tail-end (+ head-start size))
1278 coding-system head-found tail-found pos)
1279 ;; Try a short cut by searching for the string "coding:"
1280 ;; and for "unibyte:" at the head and tail of SIZE bytes.
1281 (setq head-found (or (search-forward "coding:" head-end t)
1282 (search-forward "unibyte:" head-end t)))
1283 (if (and head-found (> head-found tail-start))
1284 ;; Head and tail are overlapped.
1285 (setq tail-found head-found)
1286 (goto-char tail-start)
1287 (setq tail-found (or (search-forward "coding:" tail-end t)
1288 (search-forward "unibyte:" tail-end t))))
1290 ;; At first check the head.
1292 (goto-char head-start)
1293 (setq pos (re-search-forward "[\n\r]" head-end t))
1295 (= (char-after head-start) ?#)
1296 (= (char-after (1+ head-start)) ?!))
1297 ;; If the file begins with "#!" (exec interpreter magic),
1298 ;; look for coding frobs in the first two lines. You cannot
1299 ;; necessarily put them in the first line of such a file
1300 ;; without screwing up the interpreter invocation.
1301 (setq pos (search-forward "\n" head-end t)))
1302 (if pos (setq head-end pos))
1303 (when (< head-found head-end)
1304 (goto-char head-start)
1305 (when (and set-auto-coding-for-load
1307 "-\\*-\\(.*;\\)?[ \t]*unibyte:[ \t]*\\([^ ;]+\\)"
1309 (setq coding-system 'raw-text))
1310 (when (and (not coding-system)
1312 "-\\*-\\(.*;\\)?[ \t]*coding:[ \t]*\\([^ ;]+\\)"
1314 (setq coding-system (intern (match-string 2)))
1315 (or (coding-system-p coding-system)
1316 (setq coding-system nil)))))
1318 ;; If no coding: tag in the head, check the tail.
1319 (when (and tail-found (not coding-system))
1320 (goto-char tail-start)
1321 (search-forward "\n\^L" nil t)
1322 (if (re-search-forward
1323 "^\\(.*\\)[ \t]*Local Variables:[ \t]*\\(.*\\)$" tail-end t)
1324 ;; The prefix is what comes before "local variables:" in its
1325 ;; line. The suffix is what comes after "local variables:"
1327 (let* ((prefix (regexp-quote (match-string 1)))
1328 (suffix (regexp-quote (match-string 2)))
1332 "[ \t]*coding[ \t]*:[ \t]*\\([^ \t]+\\)[ \t]*"
1337 "[ \t]*unibyte[ \t]*:[ \t]*\\([^ \t]+\\)[ \t]*"
1340 (concat "^" prefix "[ \t]*end *:[ \t]*" suffix "$"))
1342 (re-search-forward re-end tail-end 'move)
1343 (setq tail-end (point))
1345 (when (and set-auto-coding-for-load
1346 (re-search-forward re-unibyte tail-end t))
1347 (setq coding-system 'raw-text))
1348 (when (and (not coding-system)
1349 (re-search-forward re-coding tail-end t))
1350 (setq coding-system (intern (match-string 1)))
1351 (or (coding-system-p coding-system)
1352 (setq coding-system nil))))))
1355 (setq set-auto-coding-function 'set-auto-coding)
1357 ;; Set buffer-file-coding-system of the current buffer after some text
1359 (defun after-insert-file-set-buffer-file-coding-system (inserted)
1360 (if last-coding-system-used
1361 (let ((coding-system
1362 (find-new-buffer-file-coding-system last-coding-system-used))
1363 (modified-p (buffer-modified-p)))
1365 (set-buffer-file-coding-system coding-system t)
1366 (if (and enable-multibyte-characters
1367 (or (eq coding-system 'no-conversion)
1368 (eq (coding-system-type coding-system) 5))
1369 ;; If buffer was unmodified and the size is the
1370 ;; same as INSERTED, we must be visiting it.
1372 (= (buffer-size) inserted))
1373 ;; For coding systems no-conversion and raw-text...,
1374 ;; edit the buffer as unibyte.
1375 (let ((pos-byte (position-bytes (+ (point) inserted))))
1376 (set-buffer-multibyte nil)
1377 (setq inserted (- pos-byte (position-bytes (point))))))
1378 (set-buffer-modified-p modified-p))))
1381 (add-hook 'after-insert-file-functions
1382 'after-insert-file-set-buffer-file-coding-system)
1384 ;; The coding-spec and eol-type of coding-system returned is decided
1385 ;; independently in the following order.
1386 ;; 1. That of buffer-file-coding-system locally bound.
1387 ;; 2. That of CODING.
1389 (defun find-new-buffer-file-coding-system (coding)
1390 "Return a coding system for a buffer when a file of CODING is inserted.
1391 The local variable `buffer-file-coding-system' of the current buffer
1392 is set to the returned value.
1393 Return nil if there's no need to set `buffer-file-coding-system'."
1394 (let (local-coding local-eol
1395 found-coding found-eol
1398 ;; Nothing found about coding.
1401 ;; Get information of `buffer-file-coding-system' in LOCAL-EOL
1402 ;; and LOCAL-CODING.
1403 (setq local-eol (coding-system-eol-type buffer-file-coding-system))
1404 (if (null (numberp local-eol))
1405 ;; But eol-type is not yet set.
1406 (setq local-eol nil))
1407 (if (and buffer-file-coding-system
1408 (not (eq (coding-system-type buffer-file-coding-system) t)))
1409 ;; This is not `undecided'.
1410 (setq local-coding (coding-system-base buffer-file-coding-system)))
1412 (if (and (local-variable-p 'buffer-file-coding-system)
1413 local-eol local-coding)
1414 ;; The current buffer has already set full coding-system, we
1415 ;; had better not change it.
1418 (setq found-eol (coding-system-eol-type coding))
1419 (if (null (numberp found-eol))
1420 ;; But eol-type is not found.
1421 ;; If EOL conversions are inhibited, force unix eol-type.
1422 (setq found-eol (if inhibit-eol-conversion 0)))
1423 (if (eq (coding-system-type coding) t)
1424 (setq found-coding 'undecided)
1425 (setq found-coding (coding-system-base coding)))
1427 (if (and (not found-eol) (eq found-coding 'undecided))
1428 ;; No valid coding information found.
1431 ;; Some coding information (eol or text) found.
1433 ;; The local setting takes precedence over the found one.
1434 (setq new-coding (if (local-variable-p 'buffer-file-coding-system)
1435 (or local-coding found-coding)
1436 (or found-coding local-coding)))
1437 (setq new-eol (if (local-variable-p 'buffer-file-coding-system)
1438 (or local-eol found-eol)
1439 (or found-eol local-eol)))
1441 (let ((eol-type (coding-system-eol-type new-coding)))
1442 (if (and (numberp new-eol) (vectorp eol-type))
1443 (aref eol-type new-eol)
1446 (defun modify-coding-system-alist (target-type regexp coding-system)
1447 "Modify one of look up tables for finding a coding system on I/O operation.
1448 There are three of such tables, `file-coding-system-alist',
1449 `process-coding-system-alist', and `network-coding-system-alist'.
1451 TARGET-TYPE specifies which of them to modify.
1452 If it is `file', it affects `file-coding-system-alist' (which see).
1453 If it is `process', it affects `process-coding-system-alist' (which see).
1454 If it is `network', it affects `network-coding-system-alist' (which see).
1456 REGEXP is a regular expression matching a target of I/O operation.
1457 The target is a file name if TARGET-TYPE is `file', a program name if
1458 TARGET-TYPE is `process', or a network service name or a port number
1459 to connect to if TARGET-TYPE is `network'.
1461 CODING-SYSTEM is a coding system to perform code conversion on the I/O
1462 operation, or a cons cell (DECODING . ENCODING) specifying the coding systems
1463 for decoding and encoding respectively,
1464 or a function symbol which, when called, returns such a cons cell."
1465 (or (memq target-type '(file process network))
1466 (error "Invalid target type: %s" target-type))
1467 (or (stringp regexp)
1468 (and (eq target-type 'network) (integerp regexp))
1469 (error "Invalid regular expression: %s" regexp))
1470 (if (symbolp coding-system)
1471 (if (not (fboundp coding-system))
1473 (check-coding-system coding-system)
1474 (setq coding-system (cons coding-system coding-system))))
1475 (check-coding-system (car coding-system))
1476 (check-coding-system (cdr coding-system)))
1477 (cond ((eq target-type 'file)
1478 (let ((slot (assoc regexp file-coding-system-alist)))
1480 (setcdr slot coding-system)
1481 (setq file-coding-system-alist
1482 (cons (cons regexp coding-system)
1483 file-coding-system-alist)))))
1484 ((eq target-type 'process)
1485 (let ((slot (assoc regexp process-coding-system-alist)))
1487 (setcdr slot coding-system)
1488 (setq process-coding-system-alist
1489 (cons (cons regexp coding-system)
1490 process-coding-system-alist)))))
1492 (let ((slot (assoc regexp network-coding-system-alist)))
1494 (setcdr slot coding-system)
1495 (setq network-coding-system-alist
1496 (cons (cons regexp coding-system)
1497 network-coding-system-alist)))))))
1499 (defun make-translation-table (&rest args)
1500 "Make a translation table (char table) from arguments.
1501 Each argument is a list of the form (FROM . TO),
1502 where FROM is a character to be translated to TO.
1504 FROM can be a generic character (see `make-char'). In this case, TO is
1505 a generic character containing the same number of characters, or a
1506 ordinary character. If FROM and TO are both generic characters, all
1507 characters belonging to FROM are translated to characters belonging to TO
1508 without changing their position code(s).
1510 The arguments and forms in each argument are processed in the given
1511 order, and if a previous form already translates TO to some other
1512 character, say TO-ALT, FROM is also translated to TO-ALT."
1513 (let ((table (make-char-table 'translation-table))
1516 (let ((elts (car args)))
1518 (let* ((from (car (car elts)))
1519 (from-i 0) ; degree of freedom of FROM
1520 (from-rev (nreverse (split-char from)))
1521 (to (cdr (car elts)))
1522 (to-i 0) ; degree of freedom of TO
1523 (to-rev (nreverse (split-char to))))
1524 ;; Check numbers of heading 0s in FROM-REV and TO-REV.
1525 (while (eq (car from-rev) 0)
1526 (setq from-i (1+ from-i) from-rev (cdr from-rev)))
1527 (while (eq (car to-rev) 0)
1528 (setq to-i (1+ to-i) to-rev (cdr to-rev)))
1529 (if (and (/= from-i to-i) (/= to-i 0))
1530 (error "Invalid character pair (%d . %d)" from to))
1531 ;; If we have already translated TO to TO-ALT, FROM should
1532 ;; also be translated to TO-ALT. But, this is only if TO
1533 ;; is a generic character or TO-ALT is not a generic
1535 (let ((to-alt (aref table to)))
1537 (or (> to-i 0) (not (generic-char-p to-alt))))
1540 (set-char-table-default table from to)
1541 (aset table from to))
1542 ;; If we have already translated some chars to FROM, they
1543 ;; should also be translated to TO.
1544 (let ((l (assq from revlist)))
1551 (setq l (cdr l)) ))))
1552 ;; Now update REVLIST.
1553 (let ((l (assq to revlist)))
1555 (setcdr l (cons from (cdr l)))
1556 (setq revlist (cons (list to from) revlist)))))
1557 (setq elts (cdr elts))))
1558 (setq args (cdr args)))
1559 ;; Return TABLE just created.
1562 (defun make-translation-table-from-vector (vec)
1563 "Make translation table from decoding vector VEC.
1564 VEC is an array of 256 elements to map unibyte codes to multibyte characters.
1565 See also the variable `nonascii-translation-table'."
1566 (let ((table (make-char-table 'translation-table))
1567 (rev-table (make-char-table 'translation-table))
1571 (setq ch (aref vec i))
1574 (aset rev-table ch i))
1576 (set-char-table-extra-slot table 0 rev-table)
1579 (defun define-translation-table (symbol &rest args)
1580 "Define SYMBOL as a name of translation table made by ARGS.
1582 If the first element of ARGS is a char-table of which purpose is
1583 translation-table, just define SYMBOL as the name of it.
1585 In the other case, ARGS are the same as arguments to the function
1586 `make-translation-table' (which see).
1588 This function sets properties `translation-table' and
1589 `translation-table-id' of SYMBOL to the created table itself and
1590 identification number of the table respectively."
1591 (let ((table (if (and (char-table-p (car args))
1592 (eq (char-table-subtype (car args))
1593 'translation-table))
1595 (apply 'make-translation-table args)))
1596 (len (length translation-table-vector))
1599 (put symbol 'translation-table table)
1602 (setq translation-table-vector
1603 (vconcat translation-table-vector (make-vector len nil))))
1604 (let ((slot (aref translation-table-vector id)))
1606 (eq (car slot) symbol))
1608 (aset translation-table-vector id (cons symbol table))
1610 (setq id (1+ id)))))
1611 (put symbol 'translation-table-id id)
1614 (put 'with-category-table 'lisp-indent-function 1)
1616 (defmacro with-category-table (category-table &rest body)
1617 `(let ((current-category-table (category-table)))
1618 (set-category-table ,category-table)
1621 (set-category-table current-category-table))))
1623 ;;; Initialize some variables.
1625 (put 'use-default-ascent 'char-table-extra-slots 0)
1626 (setq use-default-ascent (make-char-table 'use-default-ascent))
1627 (put 'ignore-relative-composition 'char-table-extra-slots 0)
1628 (setq ignore-relative-composition
1629 (make-char-table 'ignore-relative-composition))
1634 ;;; mule.el ends here