1 ;;; mule-cmds.el --- commands for multilingual environment -*- lexical-binding:t -*-
3 ;; Copyright (C) 1997-2015 Free Software Foundation, Inc.
4 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
5 ;; 2005, 2006, 2007, 2008, 2009, 2010, 2011
6 ;; National Institute of Advanced Industrial Science and Technology (AIST)
7 ;; Registration Number H14PRO021
9 ;; National Institute of Advanced Industrial Science and Technology (AIST)
10 ;; Registration Number H13PRO009
12 ;; Keywords: mule, i18n
14 ;; This file is part of GNU Emacs.
16 ;; GNU Emacs is free software: you can redistribute it and/or modify
17 ;; it under the terms of the GNU General Public License as published by
18 ;; the Free Software Foundation, either version 3 of the License, or
19 ;; (at your option) any later version.
21 ;; GNU Emacs is distributed in the hope that it will be useful,
22 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
23 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
24 ;; GNU General Public License for more details.
26 ;; You should have received a copy of the GNU General Public License
27 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
33 (eval-when-compile (require 'cl-lib))
36 (autoload 'widget-value "wid-edit")
38 (defvar mac-system-coding-system)
40 ;;; MULE related key bindings and menus.
43 (let ((map (make-sparse-keymap)))
44 (define-key map "f" 'set-buffer-file-coding-system)
45 (define-key map "r" 'revert-buffer-with-coding-system)
46 (define-key map "F" 'set-file-name-coding-system)
47 (define-key map "t" 'set-terminal-coding-system)
48 (define-key map "k" 'set-keyboard-coding-system)
49 (define-key map "p" 'set-buffer-process-coding-system)
50 (define-key map "x" 'set-selection-coding-system)
51 (define-key map "X" 'set-next-selection-coding-system)
52 (define-key map "\C-\\" 'set-input-method)
53 (define-key map "c" 'universal-coding-system-argument)
54 (define-key map "l" 'set-language-environment)
56 "Keymap for Mule (Multilingual environment) specific commands.")
58 ;; Keep "C-x C-m ..." for mule specific commands.
59 (define-key ctl-x-map "\C-m" mule-keymap)
61 (defvar describe-language-environment-map
62 (let ((map (make-sparse-keymap "Describe Language Environment")))
63 (bindings--define-key map
64 [Default] '(menu-item "Default" describe-specified-language-support))
67 (defvar setup-language-environment-map
68 (let ((map (make-sparse-keymap "Set Language Environment")))
69 (bindings--define-key map
70 [Default] '(menu-item "Default" setup-specified-language-environment))
73 (defvar set-coding-system-map
74 (let ((map (make-sparse-keymap "Set Coding System")))
75 (bindings--define-key map [set-buffer-process-coding-system]
76 '(menu-item "For I/O with Subprocess" set-buffer-process-coding-system
77 :visible (fboundp 'start-process)
78 :enable (get-buffer-process (current-buffer))
79 :help "How to en/decode I/O from/to subprocess connected to this buffer"))
80 (bindings--define-key map [set-next-selection-coding-system]
81 '(menu-item "For Next X Selection" set-next-selection-coding-system
82 :visible (display-selections-p)
83 :help "How to en/decode next selection/clipboard operation"))
84 (bindings--define-key map [set-selection-coding-system]
85 '(menu-item "For X Selections/Clipboard" set-selection-coding-system
86 :visible (display-selections-p)
87 :help "How to en/decode data to/from selection/clipboard"))
89 (bindings--define-key map [separator-3] menu-bar-separator)
90 (bindings--define-key map [set-terminal-coding-system]
91 '(menu-item "For Terminal" set-terminal-coding-system
92 :enable (null (memq initial-window-system '(x w32 ns)))
93 :help "How to encode terminal output"))
94 (bindings--define-key map [set-keyboard-coding-system]
95 '(menu-item "For Keyboard" set-keyboard-coding-system
96 :help "How to decode keyboard input"))
98 (bindings--define-key map [separator-2] menu-bar-separator)
99 (bindings--define-key map [set-file-name-coding-system]
100 '(menu-item "For File Name" set-file-name-coding-system
101 :help "How to decode/encode file names"))
102 (bindings--define-key map [revert-buffer-with-coding-system]
103 '(menu-item "For Reverting This File Now"
104 revert-buffer-with-coding-system
105 :enable buffer-file-name
106 :help "Revisit this file immediately using specified coding system"))
107 (bindings--define-key map [set-buffer-file-coding-system]
108 '(menu-item "For Saving This Buffer" set-buffer-file-coding-system
109 :help "How to encode this buffer when saved"))
110 (bindings--define-key map [separator-1] menu-bar-separator)
111 (bindings--define-key map [universal-coding-system-argument]
112 '(menu-item "For Next Command" universal-coding-system-argument
113 :help "Coding system to be used by next command"))
116 (defvar mule-menu-keymap
117 (let ((map (make-sparse-keymap "Mule (Multilingual Environment)")))
118 (bindings--define-key map [mule-diag]
119 '(menu-item "Show All Multilingual Settings" mule-diag
120 :help "Display multilingual environment settings"))
121 (bindings--define-key map [list-character-sets]
122 '(menu-item "List Character Sets" list-character-sets
123 :help "Show table of available character sets"))
124 (bindings--define-key map [describe-coding-system]
125 '(menu-item "Describe Coding System..." describe-coding-system))
126 (bindings--define-key map [describe-input-method]
127 '(menu-item "Describe Input Method..." describe-input-method
128 :help "Keyboard layout for a specific input method"))
129 (bindings--define-key map [describe-language-environment]
130 `(menu-item "Describe Language Environment"
131 ,describe-language-environment-map
132 :help "Show multilingual settings for a specific language"))
134 (bindings--define-key map [separator-coding-system] menu-bar-separator)
135 (bindings--define-key map [view-hello-file]
136 '(menu-item "Show Multilingual Sample Text" view-hello-file
137 :enable (file-readable-p
138 (expand-file-name "HELLO" data-directory))
139 :help "Demonstrate various character sets"))
140 (bindings--define-key map [set-various-coding-system]
141 `(menu-item "Set Coding Systems" ,set-coding-system-map
142 :enable (default-value 'enable-multibyte-characters)))
144 (bindings--define-key map [separator-input-method] menu-bar-separator)
145 (bindings--define-key map [describe-input-method]
146 '(menu-item "Describe Input Method" describe-input-method))
147 (bindings--define-key map [set-input-method]
148 '(menu-item "Select Input Method..." set-input-method))
149 (bindings--define-key map [toggle-input-method]
150 '(menu-item "Toggle Input Method" toggle-input-method))
152 (bindings--define-key map [separator-mule] menu-bar-separator)
153 (bindings--define-key map [set-language-environment]
154 `(menu-item "Set Language Environment" ,setup-language-environment-map))
156 "Keymap for Mule (Multilingual environment) menu specific commands.")
158 ;; This should be a single character key binding because users use it
159 ;; very frequently while editing multilingual text. Now we can use
160 ;; only two such keys: "\C-\\" and "\C-^", but the latter is not
161 ;; convenient because it requires shifting on most keyboards. An
162 ;; alternative is "\C-\]" which is now bound to `abort-recursive-edit'
163 ;; but it won't be used that frequently.
164 (define-key global-map "\C-\\" 'toggle-input-method)
166 ;; This is no good because people often type Shift-SPC
167 ;; meaning to type SPC. -- rms.
168 ;; ;; Here's an alternative key binding for X users (Shift-SPACE).
169 ;; (define-key global-map [?\S- ] 'toggle-input-method)
171 ;;; Mule related hyperlinks.
172 (defconst help-xref-mule-regexp-template
173 (purecopy (concat "\\(\\<\\("
174 "\\(coding system\\)\\|"
175 "\\(input method\\)\\|"
176 "\\(character set\\)\\|"
179 ;; Note starting with word-syntax character:
180 "['`‘]\\(\\sw\\(\\sw\\|\\s_\\)+\\)['’]")))
182 (defun coding-system-change-eol-conversion (coding-system eol-type)
183 "Return a coding system which differs from CODING-SYSTEM in EOL conversion.
184 The returned coding system converts end-of-line by EOL-TYPE
185 but text as the same way as CODING-SYSTEM.
186 EOL-TYPE should be `unix', `dos', `mac', or nil.
187 If EOL-TYPE is nil, the returned coding system detects
188 how end-of-line is formatted automatically while decoding.
190 EOL-TYPE can be specified by an integer 0, 1, or 2.
191 They means `unix', `dos', and `mac' respectively."
192 (if (symbolp eol-type)
193 (setq eol-type (cond ((eq eol-type 'unix) 0)
194 ((eq eol-type 'dos) 1)
195 ((eq eol-type 'mac) 2)
197 ;; We call `coding-system-base' before `coding-system-eol-type',
198 ;; because the coding-system may not be initialized until then.
199 (let* ((base (coding-system-base coding-system))
200 (orig-eol-type (coding-system-eol-type coding-system)))
201 (cond ((vectorp orig-eol-type)
204 (aref orig-eol-type eol-type)))
207 ((= eol-type orig-eol-type)
209 ((progn (setq orig-eol-type (coding-system-eol-type base))
210 (vectorp orig-eol-type))
211 (aref orig-eol-type eol-type)))))
213 (defun coding-system-change-text-conversion (coding-system coding)
214 "Return a coding system which differs from CODING-SYSTEM in text conversion.
215 The returned coding system converts text by CODING
216 but end-of-line as the same way as CODING-SYSTEM.
217 If CODING is nil, the returned coding system detects
218 how text is formatted automatically while decoding."
219 (let ((eol-type (coding-system-eol-type coding-system)))
220 (coding-system-change-eol-conversion
221 (if coding coding 'undecided)
222 (if (numberp eol-type) (aref [unix dos mac] eol-type)))))
224 ;; Canonicalize the coding system name NAME by removing some prefixes
225 ;; and delimiter characters. Support function of
226 ;; coding-system-from-name.
227 (defun canonicalize-coding-system-name (name)
228 (if (string-match "^\\(ms\\|ibm\\|windows-\\)\\([0-9]+\\)$" name)
229 ;; "ms950", "ibm950", "windows-950" -> "cp950"
230 (concat "cp" (match-string 2 name))
231 (if (string-match "^iso[-_ ]?[0-9]" name)
232 ;; "iso-8859-1" -> "8859-1", "iso-2022-jp" ->"2022-jp"
233 (setq name (substring name (1- (match-end 0)))))
234 (let ((idx (string-match "[-_ /]" name)))
235 ;; Delete "-", "_", " ", "/" but do distinguish "16-be" and "16be".
238 (eq (string-match "16-[lb]e$" name (- idx 2))
240 (setq idx (string-match "[-_ /]" name (match-end 0)))
241 (setq name (concat (substring name 0 idx) (substring name (1+ idx)))
242 idx (string-match "[-_ /]" name idx))))
245 (defun coding-system-from-name (name)
246 "Return a coding system whose name matches with NAME (string or symbol)."
248 (if (stringp name) (setq sym (intern name))
249 (setq sym name name (symbol-name name)))
250 (if (coding-system-p sym)
253 (if (string-match "-\\(unix\\|dos\\|mac\\)$" name)
254 (prog1 (intern (match-string 1 name))
255 (setq name (substring name 0 (match-beginning 0)))))))
256 (setq name (canonicalize-coding-system-name (downcase name)))
258 (dolist (elt (coding-system-list))
259 (if (string= (canonicalize-coding-system-name (symbol-name elt))
261 (throw 'tag (if eol-type (coding-system-change-eol-conversion
265 (defun toggle-enable-multibyte-characters (&optional arg)
266 "Change whether this buffer uses multibyte characters.
267 With ARG, use multibyte characters if the ARG is positive.
269 Note that this command does not convert the byte contents of
270 the buffer; it only changes the way those bytes are interpreted.
271 In general, therefore, this command *changes* the sequence of
272 characters that the current buffer contains.
274 We suggest you avoid using this command unless you know what you are
275 doing. If you use it by mistake, and the buffer is now displayed
276 wrong, use this command again to toggle back to the right mode."
279 (if (null arg) (null enable-multibyte-characters)
280 (> (prefix-numeric-value arg) 0))))
281 (set-buffer-multibyte new-flag))
282 (force-mode-line-update))
284 (defun view-hello-file ()
285 "Display the HELLO file, which lists many languages and characters."
287 ;; We have to decode the file in any environment.
288 (let ((coding-system-for-read 'iso-2022-7bit))
289 (view-file (expand-file-name "HELLO" data-directory))))
291 (defun universal-coding-system-argument (coding-system)
292 "Execute an I/O command using the specified coding system."
294 (let ((default (and buffer-file-coding-system
295 (not (eq (coding-system-type buffer-file-coding-system)
297 buffer-file-coding-system)))
298 (list (read-coding-system
300 (format "Coding system for following command (default %s): " default)
301 "Coding system for following command: ")
303 (let* ((keyseq (read-key-sequence
304 (format "Command to execute with %s:" coding-system)))
305 (cmd (key-binding keyseq))
307 ;; read-key-sequence ignores quit, so make an explicit check.
308 ;; Like many places, this assumes quit == C-g, but it need not be.
309 (if (equal last-input-event ?\C-g)
311 (when (memq cmd '(universal-argument digit-argument))
312 (call-interactively cmd)
314 ;; Process keys bound in `universal-argument-map'.
316 (setq keyseq (read-key-sequence nil t)
317 cmd (key-binding keyseq t))
318 (not (eq cmd 'universal-argument-other-key)))
319 (let ((current-prefix-arg prefix-arg)
320 ;; Have to bind `last-command-event' here so that
321 ;; `digit-argument', for instance, can compute the
323 (last-command-event (aref keyseq 0)))
324 (call-interactively cmd)))
326 ;; This is the final call to `universal-argument-other-key', which
327 ;; set's the final `prefix-arg.
328 (let ((current-prefix-arg prefix-arg))
329 (call-interactively cmd))
331 ;; Read the command to execute with the given prefix arg.
332 (setq prefix prefix-arg
333 keyseq (read-key-sequence nil t)
334 cmd (key-binding keyseq)))
336 (let ((coding-system-for-read coding-system)
337 (coding-system-for-write coding-system)
338 (coding-system-require-warning t)
339 (current-prefix-arg prefix))
341 (call-interactively cmd))))
343 (defun set-default-coding-systems (coding-system)
344 "Set default value of various coding systems to CODING-SYSTEM.
345 This sets the following coding systems:
346 o coding system of a newly created buffer
347 o default coding system for subprocess I/O
348 This also sets the following values:
349 o default value used as `file-name-coding-system' for converting file names
350 if CODING-SYSTEM is ASCII-compatible
351 o default value for the command `set-terminal-coding-system'
352 o default value for the command `set-keyboard-coding-system'
353 if CODING-SYSTEM is ASCII-compatible"
354 (check-coding-system coding-system)
355 (setq-default buffer-file-coding-system coding-system)
357 (if (eq system-type 'darwin)
358 ;; The file-name coding system on Darwin systems is always utf-8.
359 (setq default-file-name-coding-system 'utf-8)
360 (if (and (default-value 'enable-multibyte-characters)
361 (or (not coding-system)
362 (coding-system-get coding-system 'ascii-compatible-p)))
363 (setq default-file-name-coding-system coding-system)))
364 (setq default-terminal-coding-system coding-system)
365 ;; Prevent default-terminal-coding-system from converting ^M to ^J.
366 (setq default-keyboard-coding-system
367 (coding-system-change-eol-conversion coding-system 'unix))
368 ;; Preserve eol-type from existing default-process-coding-systems.
369 ;; On non-unix-like systems in particular, these may have been set
370 ;; carefully by the user, or by the startup code, to deal with the
371 ;; users shell appropriately, so should not be altered by changing
372 ;; language environment.
374 (coding-system-change-text-conversion
375 (car default-process-coding-system) coding-system))
377 (coding-system-change-text-conversion
378 (cdr default-process-coding-system) coding-system)))
379 (setq default-process-coding-system
380 (cons output-coding input-coding))))
382 (defun prefer-coding-system (coding-system)
383 "Add CODING-SYSTEM at the front of the priority list for automatic detection.
384 This also sets the following coding systems:
385 o coding system of a newly created buffer
386 o default coding system for subprocess I/O
387 This also sets the following values:
388 o default value used as `file-name-coding-system' for converting file names
389 o default value for the command `set-terminal-coding-system'
390 o default value for the command `set-keyboard-coding-system'
392 If CODING-SYSTEM specifies a certain type of EOL conversion, the coding
393 systems set by this function will use that type of EOL conversion.
395 A coding system that requires automatic detection of text+encoding
396 \(e.g. undecided, unix) can't be preferred.
398 To prefer, for instance, utf-8, say the following:
400 \(prefer-coding-system \\='utf-8)"
401 (interactive "zPrefer coding system: ")
402 (if (not (and coding-system (coding-system-p coding-system)))
403 (error "Invalid coding system `%s'" coding-system))
404 (if (memq (coding-system-type coding-system) '(raw-text undecided))
405 (error "Can't prefer the coding system `%s'" coding-system))
406 (let ((base (coding-system-base coding-system))
407 (eol-type (coding-system-eol-type coding-system)))
408 (set-coding-system-priority base)
409 (and (called-interactively-p 'interactive)
410 (or (eq base coding-system)
411 (message "Highest priority is set to %s (base of %s)"
412 base coding-system)))
413 ;; If they asked for specific EOL conversion, honor that.
414 (if (memq eol-type '(0 1 2))
416 (coding-system-change-eol-conversion base eol-type)))
417 (set-default-coding-systems base)
418 (if (called-interactively-p 'interactive)
419 (or (eq base default-file-name-coding-system)
420 (message "The default value of `file-name-coding-system' was not changed because the specified coding system is not suitable for file names.")))))
422 (defvar sort-coding-systems-predicate nil
423 "If non-nil, a predicate function to sort coding systems.
425 It is called with two coding systems, and should return t if the first
426 one is \"less\" than the second.
428 The function `sort-coding-systems' use it.")
430 (defun sort-coding-systems (codings)
431 "Sort coding system list CODINGS by a priority of each coding system.
432 Return the sorted list. CODINGS is modified by side effects.
434 If a coding system is most preferred, it has the highest priority.
435 Otherwise, coding systems that correspond to MIME charsets have
436 higher priorities. Among them, a coding system included in the
437 `coding-system' key of the current language environment has higher
438 priority. See also the documentation of `language-info-alist'.
440 If the variable `sort-coding-systems-predicate' (which see) is
441 non-nil, it is used to sort CODINGS instead."
442 (if sort-coding-systems-predicate
443 (sort codings sort-coding-systems-predicate)
444 (let* ((from-priority (coding-system-priority-list))
445 (most-preferred (car from-priority))
446 (lang-preferred (get-language-info current-language-environment
450 (let ((base (coding-system-base x)))
451 ;; We calculate the priority number 0..255 by
452 ;; using the 8 bits PMMLCEII as this:
453 ;; P: 1 if most preferred.
454 ;; MM: greater than 0 if mime-charset.
455 ;; L: 1 if one of the current lang. env.'s codings.
456 ;; C: 1 if one of codings listed in the category list.
457 ;; E: 1 if not XXX-with-esc
458 ;; II: if iso-2022 based, 0..3, else 1.
460 (lsh (if (eq base most-preferred) 1 0) 7)
462 (let ((mime (coding-system-get base :mime-charset)))
463 ;; Prefer coding systems corresponding to a
466 ;; Lower utf-16 priority so that we
467 ;; normally prefer utf-8 to it, and put
468 ;; x-ctext below that.
469 (cond ((string-match-p "utf-16"
472 ((string-match-p "^x-" (symbol-name mime))
477 (lsh (if (memq base lang-preferred) 1 0) 4)
478 (lsh (if (memq base from-priority) 1 0) 3)
479 (lsh (if (string-match-p "-with-esc\\'"
482 (if (eq (coding-system-type base) 'iso-2022)
483 (let ((category (coding-system-category base)))
484 ;; For ISO based coding systems, prefer
485 ;; one that doesn't use designation nor
486 ;; locking/single shifting.
488 ((or (eq category 'coding-category-iso-8-1)
489 (eq category 'coding-category-iso-8-2))
491 ((or (eq category 'coding-category-iso-7-tight)
492 (eq category 'coding-category-iso-7))
498 (sort codings (function (lambda (x y)
499 (> (funcall func x) (funcall func y))))))))
501 (defun find-coding-systems-region (from to)
502 "Return a list of proper coding systems to encode a text between FROM and TO.
504 If FROM is a string, find coding systems in that instead of the buffer.
505 All coding systems in the list can safely encode any multibyte characters
508 If the text contains no multibyte characters, return a list of a single
509 element `undecided'."
510 (let ((codings (find-coding-systems-region-internal from to)))
512 ;; The text contains only ASCII characters. Any coding
515 ;; We need copy-sequence because sorting will alter the argument.
516 (sort-coding-systems (copy-sequence codings)))))
518 (defun find-coding-systems-string (string)
519 "Return a list of proper coding systems to encode STRING.
520 All coding systems in the list can safely encode any multibyte characters
523 If STRING contains no multibyte characters, return a list of a single
524 element `undecided'."
525 (find-coding-systems-region string nil))
527 (defun find-coding-systems-for-charsets (charsets)
528 "Return a list of proper coding systems to encode characters of CHARSETS.
529 CHARSETS is a list of character sets.
531 This only finds coding systems of type `charset', whose
532 `:charset-list' property includes all of CHARSETS (plus `ascii' for
533 ASCII-compatible coding systems). It was used in older versions of
534 Emacs, but is unlikely to be what you really want now."
535 ;; Deal with aliases.
536 (setq charsets (mapcar (lambda (c)
537 (get-charset-property c :name))
539 (cond ((or (null charsets)
540 (and (= (length charsets) 1)
541 (eq 'ascii (car charsets))))
543 ((or (memq 'eight-bit-control charsets)
544 (memq 'eight-bit-graphic charsets))
545 '(raw-text utf-8-emacs))
548 (dolist (cs (coding-system-list t))
549 (let ((cs-charsets (and (eq (coding-system-type cs) 'charset)
550 (coding-system-charset-list cs)))
552 (if (coding-system-get cs :ascii-compatible-p)
553 (cl-pushnew 'ascii cs-charsets))
557 (unless (memq (pop charsets) cs-charsets)
561 (nreverse codings)))))
563 (defun find-multibyte-characters (from to &optional maxcount excludes)
564 "Find multibyte characters in the region specified by FROM and TO.
565 If FROM is a string, find multibyte characters in the string.
566 The return value is an alist of the following format:
567 ((CHARSET COUNT CHAR ...) ...)
569 CHARSET is a character set,
570 COUNT is a number of characters,
571 CHARs are the characters found from the character set.
572 Optional 3rd arg MAXCOUNT limits how many CHARs are put in the above list.
573 Optional 4th arg EXCLUDES is a list of character sets to be ignored."
577 (if (multibyte-string-p from)
579 (while (setq idx (string-match-p "[^\000-\177]" from idx))
580 (setq char (aref from idx)
581 charset (char-charset char))
582 (unless (memq charset excludes)
583 (let ((slot (assq charset chars)))
585 (if (not (memq char (nthcdr 2 slot)))
586 (let ((count (nth 1 slot)))
587 (setcar (cdr slot) (1+ count))
588 (if (or (not maxcount) (< count maxcount))
589 (nconc slot (list char)))))
590 (setq chars (cons (list charset 1 char) chars)))))
591 (setq idx (1+ idx)))))
592 (if enable-multibyte-characters
595 (while (re-search-forward "[^\000-\177]" to t)
596 (setq char (preceding-char)
597 charset (char-charset char))
598 (unless (memq charset excludes)
599 (let ((slot (assq charset chars)))
601 (if (not (member char (nthcdr 2 slot)))
602 (let ((count (nth 1 slot)))
603 (setcar (cdr slot) (1+ count))
604 (if (or (not maxcount) (< count maxcount))
605 (nconc slot (list char)))))
606 (setq chars (cons (list charset 1 char) chars)))))))))
609 (defun search-unencodable-char (coding-system)
610 "Search forward from point for a character that is not encodable.
611 It asks which coding system to check.
612 If such a character is found, set point after that character.
613 Otherwise, don't move point.
615 When called from a program, the value is the position of the unencodable
616 character found, or nil if all characters are encodable."
618 (list (let ((default (or buffer-file-coding-system 'us-ascii)))
620 (format "Coding-system (default %s): " default)
622 (let ((pos (unencodable-char-position (point) (point-max) coding-system)))
625 (message "All following characters are encodable by %s" coding-system))
628 (defvar last-coding-system-specified nil
629 "Most recent coding system explicitly specified by the user when asked.
630 This variable is set whenever Emacs asks the user which coding system
631 to use in order to write a file. If you set it to nil explicitly,
632 then call `write-region', then afterward this variable will be non-nil
633 only if the user was explicitly asked and specified a coding system.")
635 (defvar select-safe-coding-system-accept-default-p nil
636 "If non-nil, a function to control the behavior of coding system selection.
637 The meaning is the same as the argument ACCEPT-DEFAULT-P of the
638 function `select-safe-coding-system' (which see). This variable
639 overrides that argument.")
641 (defun sanitize-coding-system-list (codings)
642 "Return a list of coding systems presumably more user-friendly than CODINGS."
643 ;; Change each safe coding system to the corresponding
644 ;; mime-charset name if it is also a coding system. Such a name
645 ;; is more friendly to users.
648 (let ((mime-charset (coding-system-get cs 'mime-charset)))
649 (if (and mime-charset (coding-system-p mime-charset)
650 (coding-system-equal cs mime-charset))
654 ;; Don't offer variations with locking shift, which you
655 ;; basically never want.
657 (dolist (elt codings (setq codings (nreverse l)))
658 (unless (or (eq 'coding-category-iso-7-else
659 (coding-system-category elt))
660 (eq 'coding-category-iso-8-else
661 (coding-system-category elt)))
664 ;; Remove raw-text, emacs-mule and no-conversion unless nothing
665 ;; else is available.
668 (delq 'no-conversion (copy-sequence codings))))
671 (defun select-safe-coding-system-interactively (from to codings unsafe
672 &optional rejected default)
673 "Select interactively a coding system for the region FROM ... TO.
674 FROM can be a string, as in `write-region'.
675 CODINGS is the list of base coding systems known to be safe for this region,
676 typically obtained with `find-coding-systems-region'.
677 UNSAFE is a list of coding systems known to be unsafe for this region.
678 REJECTED is a list of coding systems which were safe but for some reason
679 were not recommended in the particular context.
680 DEFAULT is the coding system to use by default in the query."
681 ;; At first, if some defaults are unsafe, record at most 11
682 ;; problematic characters and their positions for them by turning
685 ;; ((CODING (POS . CHAR) (POS . CHAR) ...) ...)
688 (mapcar #'(lambda (coding)
691 (mapcar #'(lambda (pos)
692 (cons pos (aref from pos)))
693 (unencodable-char-position
694 0 (length from) coding
696 (mapcar #'(lambda (pos)
697 (cons pos (char-after pos)))
698 (unencodable-char-position
699 from to coding 11)))))
702 (setq codings (sanitize-coding-system-list codings))
704 (let ((window-configuration (current-window-configuration))
705 (bufname (buffer-name))
708 ;; If some defaults are unsafe, make sure the offending
709 ;; buffer is displayed.
710 (when (and unsafe (not (stringp from)))
711 (pop-to-buffer bufname)
712 (goto-char (apply 'min (mapcar #'(lambda (x) (car (cadr x)))
714 ;; Then ask users to select one from CODINGS while showing
715 ;; the reason why none of the defaults are not used.
716 (with-output-to-temp-buffer "*Warning*"
717 (with-current-buffer standard-output
718 (if (and (null rejected) (null unsafe))
719 (insert "No default coding systems to try for "
721 (format "string \"%s\"." from)
722 (format-message "buffer `%s'." bufname)))
724 "These default coding systems were tried to encode"
726 (concat " \"" (if (> (length from) 10)
727 (concat (substring from 0 10) "...\"")
729 (format-message " text\nin the buffer `%s'" bufname))
733 (dolist (x (append rejected unsafe))
734 (princ " ") (princ x))
736 (fill-region-as-paragraph pos (point)))
738 (insert "These safely encode the text in the buffer,
739 but are not recommended for encoding text in this context,
740 e.g., for sending an email message.\n ")
742 (princ " ") (princ x))
745 (insert (if rejected "The other coding systems"
746 "However, each of them")
747 (substitute-command-keys
748 " encountered characters it couldn't encode:\n"))
749 (dolist (coding unsafe)
750 (insert (format " %s cannot encode these:" (car coding)))
753 #'(lambda (bufname pos)
754 (when (buffer-live-p (get-buffer bufname))
755 (pop-to-buffer bufname)
758 #'(lambda (bufname pos coding)
759 (when (buffer-live-p (get-buffer bufname))
760 (pop-to-buffer bufname)
764 (search-unencodable-char coding)
765 (forward-char -1))))))
766 (dolist (elt (cdr coding))
769 (insert (if (< i 10) (cdr elt) "..."))
776 "mouse-2, RET: jump to this character"
778 'help-args (list bufname (car elt)))
784 "mouse-2, RET: next unencodable character"
786 'help-args (list bufname (car elt)
790 (insert (substitute-command-keys "\
792 Click on a character (or switch to this window by `\\[other-window]'\n\
793 and select the characters by RET) to jump to the place it appears,\n\
794 where `\\[universal-argument] \\[what-cursor-position]' will give information about it.\n"))))
795 (insert (substitute-command-keys "\nSelect \
796 one of the safe coding systems listed below,\n\
797 or cancel the writing with \\[keyboard-quit] and edit the buffer\n\
798 to remove or modify the problematic characters,\n\
799 or specify any other coding system (and risk losing\n\
800 the problematic characters).\n\n"))
804 (princ " ") (princ x))
806 (fill-region-as-paragraph pos (point)))))
808 ;; Read a coding system.
811 (format "Select coding system (default %s): " default)
813 (setq last-coding-system-specified coding-system))
815 (kill-buffer "*Warning*")
816 (set-window-configuration window-configuration)
819 (defun select-safe-coding-system (from to &optional default-coding-system
820 accept-default-p file)
821 "Ask a user to select a safe coding system from candidates.
822 The candidates of coding systems which can safely encode a text
823 between FROM and TO are shown in a popup window. Among them, the most
824 proper one is suggested as the default.
826 The list of `buffer-file-coding-system' of the current buffer, the
827 default `buffer-file-coding-system', and the most preferred coding
828 system (if it corresponds to a MIME charset) is treated as the
829 default coding system list. Among them, the first one that safely
830 encodes the text is normally selected silently and returned without
831 any user interaction. See also the command `prefer-coding-system'.
833 However, the user is queried if the chosen coding system is
834 inconsistent with what would be selected by `find-auto-coding' from
835 coding cookies &c. if the contents of the region were read from a
836 file. (That could lead to data corruption in a file subsequently
837 re-visited and edited.)
839 Optional 3rd arg DEFAULT-CODING-SYSTEM specifies a coding system or a
840 list of coding systems to be prepended to the default coding system
841 list. However, if DEFAULT-CODING-SYSTEM is a list and the first
842 element is t, the cdr part is used as the default coding system list,
843 i.e. current `buffer-file-coding-system', default `buffer-file-coding-system',
844 and the most preferred coding system are not used.
846 Optional 4th arg ACCEPT-DEFAULT-P, if non-nil, is a function to
847 determine the acceptability of the silently selected coding system.
848 It is called with that coding system, and should return nil if it
849 should not be silently selected and thus user interaction is required.
851 Optional 5th arg FILE is the file name to use for this purpose.
852 That is different from `buffer-file-name' when handling `write-region'
855 The variable `select-safe-coding-system-accept-default-p', if non-nil,
856 overrides ACCEPT-DEFAULT-P.
858 Kludgy feature: if FROM is a string, the string is the target text,
860 (if (not (listp default-coding-system))
861 (setq default-coding-system (list default-coding-system)))
863 (let ((no-other-defaults nil)
865 (unless (or (stringp from) find-file-literally)
866 ;; Find an auto-coding that is specified for the current
867 ;; buffer and file from the region FROM and TO.
872 (setq auto-cs (find-auto-coding (or file buffer-file-name "")
875 (if (coding-system-p (car auto-cs))
876 (setq auto-cs (car auto-cs))
880 Invalid coding system `%s' is specified
881 for the current buffer/file by the %s.
882 It is highly recommended to fix it before writing to a file."
884 (if (eq (cdr auto-cs) :coding) ":coding tag"
885 (format-message "variable `%s'" (cdr auto-cs))))
887 (or (yes-or-no-p "Really proceed with writing? ")
888 (error "Save aborted"))
889 (setq auto-cs nil))))))
891 (if (eq (car default-coding-system) t)
892 (setq no-other-defaults t
893 default-coding-system (cdr default-coding-system)))
895 ;; Change elements of the list to (coding . base-coding).
896 (setq default-coding-system
897 (mapcar (function (lambda (x) (cons x (coding-system-base x))))
898 default-coding-system))
900 (if (and auto-cs (not no-other-defaults))
901 ;; If the file has a coding cookie, use it regardless of any
903 (let ((base (coding-system-base auto-cs)))
904 (unless (memq base '(nil undecided))
905 (setq default-coding-system (list (cons auto-cs base)))
906 (setq no-other-defaults t))))
908 (unless no-other-defaults
909 ;; If buffer-file-coding-system is not nil nor undecided, append it
911 (if buffer-file-coding-system
912 (let ((base (coding-system-base buffer-file-coding-system)))
913 (or (eq base 'undecided)
914 (rassq base default-coding-system)
915 (setq default-coding-system
916 (append default-coding-system
917 (list (cons buffer-file-coding-system base)))))))
919 (unless (and buffer-file-coding-system-explicit
920 (cdr buffer-file-coding-system-explicit))
921 ;; If default buffer-file-coding-system is not nil nor undecided,
922 ;; append it to the defaults.
923 (when (default-value 'buffer-file-coding-system)
924 (let ((base (coding-system-base
925 (default-value 'buffer-file-coding-system))))
926 (or (eq base 'undecided)
927 (rassq base default-coding-system)
928 (setq default-coding-system
929 (append default-coding-system
930 (list (cons (default-value
931 'buffer-file-coding-system)
934 ;; If the most preferred coding system has the property mime-charset,
935 ;; append it to the defaults.
936 (let ((preferred (coding-system-priority-list t))
938 (and (coding-system-p preferred)
939 (setq base (coding-system-base preferred))
940 (coding-system-get preferred :mime-charset)
941 (not (rassq base default-coding-system))
942 (setq default-coding-system
943 (append default-coding-system
944 (list (cons preferred base))))))))
946 (if select-safe-coding-system-accept-default-p
947 (setq accept-default-p select-safe-coding-system-accept-default-p))
949 ;; Decide the eol-type from the top of the default codings,
950 ;; current buffer-file-coding-system, or default buffer-file-coding-system.
951 (if default-coding-system
952 (let ((default-eol-type (coding-system-eol-type
953 (caar default-coding-system))))
954 (if (and (vectorp default-eol-type) buffer-file-coding-system)
955 (setq default-eol-type (coding-system-eol-type
956 buffer-file-coding-system)))
957 (if (and (vectorp default-eol-type)
958 (default-value 'buffer-file-coding-system))
959 (setq default-eol-type
960 (coding-system-eol-type
961 (default-value 'buffer-file-coding-system))))
962 (if (and default-eol-type (not (vectorp default-eol-type)))
963 (dolist (elt default-coding-system)
964 (setcar elt (coding-system-change-eol-conversion
965 (car elt) default-eol-type))))))
967 (let ((codings (find-coding-systems-region from to))
969 (tick (if (not (stringp from)) (buffer-chars-modified-tick)))
970 safe rejected unsafe)
971 (if (eq (car codings) 'undecided)
972 ;; Any coding system is ok.
973 (setq coding-system (caar default-coding-system))
974 ;; Reverse the list so that elements are accumulated in safe,
975 ;; rejected, and unsafe in the correct order.
976 (setq default-coding-system (nreverse default-coding-system))
978 ;; Classify the defaults into safe, rejected, and unsafe.
979 (dolist (elt default-coding-system)
980 (if (memq (cdr elt) codings)
981 ;; This is safe. Is it acceptable?
982 (if (and (functionp accept-default-p)
983 (not (funcall accept-default-p (cdr elt))))
984 ;; No, not acceptable.
985 (push (car elt) rejected)
987 (push (car elt) safe))
989 (push (car elt) unsafe)))
990 ;; If there are safe ones, the first one is what we want.
992 (setq coding-system (car safe))))
994 ;; If all the defaults failed, ask a user.
995 (when (not coding-system)
996 (setq coding-system (select-safe-coding-system-interactively
997 from to codings unsafe rejected (car codings))))
999 ;; Check we're not inconsistent with what `coding:' spec &c would
1000 ;; give when file is re-read.
1001 ;; But don't do this if we explicitly ignored the cookie
1002 ;; by using `find-file-literally'.
1006 (memq (coding-system-type coding-system) '(0 5)))))
1007 ;; Merge coding-system and auto-cs as far as possible.
1008 (if (not coding-system)
1009 (setq coding-system auto-cs)
1011 (setq auto-cs coding-system)
1012 (let ((eol-type-1 (coding-system-eol-type coding-system))
1013 (eol-type-2 (coding-system-eol-type auto-cs)))
1014 (if (eq (coding-system-base coding-system) 'undecided)
1015 (setq coding-system (coding-system-change-text-conversion
1016 coding-system auto-cs))
1017 (if (eq (coding-system-base auto-cs) 'undecided)
1018 (setq auto-cs (coding-system-change-text-conversion
1019 auto-cs coding-system))))
1020 (if (vectorp eol-type-1)
1021 (or (vectorp eol-type-2)
1022 (setq coding-system (coding-system-change-eol-conversion
1023 coding-system eol-type-2)))
1024 (if (vectorp eol-type-2)
1025 (setq auto-cs (coding-system-change-eol-conversion
1026 auto-cs eol-type-1)))))))
1029 ;; Don't barf if writing a compressed file, say.
1030 ;; This check perhaps isn't ideal, but is probably
1031 ;; the best thing to do.
1032 (not (auto-coding-alist-lookup (or file buffer-file-name "")))
1033 (not (coding-system-equal coding-system auto-cs)))
1034 (unless (yes-or-no-p
1035 (format "Selected encoding %s disagrees with \
1036 %s specified by file contents. Really save (else edit coding cookies \
1037 and try again)? " coding-system auto-cs))
1038 (error "Save aborted"))))
1039 (when (and tick (/= tick (buffer-chars-modified-tick)))
1040 (error "Canceled because the buffer was modified"))
1041 (if (and (eq (coding-system-type coding-system) 'undecided)
1042 (coding-system-get coding-system :prefer-utf-8)
1043 (or (multibyte-string-p from)
1044 (and (number-or-marker-p from)
1046 (- (position-bytes to) (position-bytes from))))))
1048 (coding-system-change-text-conversion coding-system 'utf-8)))
1051 (setq select-safe-coding-system-function 'select-safe-coding-system)
1053 (defun select-message-coding-system ()
1054 "Return a coding system to encode the outgoing message of the current buffer.
1055 It at first tries the first coding system found in these variables
1057 (1) local value of `buffer-file-coding-system'
1058 (2) value of `sendmail-coding-system'
1059 (3) value of `default-sendmail-coding-system'
1060 (4) default value of `buffer-file-coding-system'
1061 If the found coding system can't encode the current buffer,
1062 or none of them are bound to a coding system,
1063 it asks the user to select a proper coding system."
1064 (let ((coding (or (and (local-variable-p 'buffer-file-coding-system)
1065 buffer-file-coding-system)
1066 sendmail-coding-system
1067 default-sendmail-coding-system
1068 (default-value 'buffer-file-coding-system))))
1069 (if (eq coding 'no-conversion)
1070 ;; We should never use no-conversion for outgoing mail.
1072 (if (fboundp select-safe-coding-system-function)
1073 (funcall select-safe-coding-system-function
1074 (point-min) (point-max) coding
1075 (function (lambda (x) (coding-system-get x :mime-charset))))
1078 ;;; Language support stuff.
1080 (defvar language-info-alist nil
1081 "Alist of language environment definitions.
1082 Each element looks like:
1083 (LANGUAGE-NAME . ((KEY . INFO) ...))
1084 where LANGUAGE-NAME is a string, the name of the language environment,
1085 KEY is a symbol denoting the kind of information, and
1086 INFO is the data associated with KEY.
1087 Meaningful values for KEY include
1089 documentation value is documentation of what this language environment
1090 is meant for, and how to use it.
1091 charset value is a list of the character sets mainly used
1092 by this language environment.
1093 sample-text value is an expression which is evalled to generate
1094 a line of text written using characters appropriate
1095 for this language environment.
1096 setup-function value is a function to call to switch to this
1097 language environment.
1098 exit-function value is a function to call to leave this
1099 language environment.
1100 coding-system value is a list of coding systems that are good for
1101 saving text written in this language environment.
1102 This list serves as suggestions to the user;
1103 in effect, as a kind of documentation.
1104 coding-priority value is a list of coding systems for this language
1105 environment, in order of decreasing priority.
1106 This is used to set up the coding system priority
1107 list when you switch to this language environment.
1108 nonascii-translation
1109 value is a charset of dimension one to use for
1110 converting a unibyte character to multibyte
1112 input-method value is a default input method for this language
1114 features value is a list of features requested in this
1115 language environment.
1116 ctext-non-standard-encodings
1117 value is a list of non-standard encoding names used
1118 in extended segments of CTEXT. See the variable
1119 `ctext-non-standard-encodings' for more detail.
1121 The following key takes effect only when multibyte characters are
1122 globally disabled, i.e. the default value of `enable-multibyte-characters'
1123 is nil (which is an obsolete and deprecated use):
1125 unibyte-display value is a coding system to encode characters for
1126 the terminal. Characters in the range of 160 to
1127 255 display not as octal escapes, but as non-ASCII
1128 characters in this language environment.")
1130 (defun get-language-info (lang-env key)
1131 "Return information listed under KEY for language environment LANG-ENV.
1132 KEY is a symbol denoting the kind of information.
1133 For a list of useful values for KEY and their meanings,
1134 see `language-info-alist'."
1135 (if (symbolp lang-env)
1136 (setq lang-env (symbol-name lang-env)))
1137 (let ((lang-slot (assoc-string lang-env language-info-alist t)))
1139 (cdr (assq key (cdr lang-slot))))))
1141 (defun set-language-info (lang-env key info)
1142 "Modify part of the definition of language environment LANG-ENV.
1143 Specifically, this stores the information INFO under KEY
1144 in the definition of this language environment.
1145 KEY is a symbol denoting the kind of information.
1146 INFO is the value for that information.
1148 For a list of useful values for KEY and their meanings,
1149 see `language-info-alist'."
1150 (if (symbolp lang-env)
1151 (setq lang-env (symbol-name lang-env)))
1152 (set-language-info-internal lang-env key info)
1153 (if (equal lang-env current-language-environment)
1154 (cond ((eq key 'coding-priority)
1155 (set-language-environment-coding-systems lang-env)
1156 (set-language-environment-charset lang-env))
1157 ((eq key 'input-method)
1158 (set-language-environment-input-method lang-env))
1159 ((eq key 'nonascii-translation)
1160 (set-language-environment-nonascii-translation lang-env))
1162 (set-language-environment-charset lang-env))
1163 ((and (not (default-value 'enable-multibyte-characters))
1164 (or (eq key 'unibyte-syntax) (eq key 'unibyte-display)))
1165 (set-language-environment-unibyte lang-env)))))
1167 (defun set-language-info-internal (lang-env key info)
1169 Arguments are the same as `set-language-info'."
1170 (let (lang-slot key-slot)
1171 (setq lang-slot (assoc lang-env language-info-alist))
1172 (if (null lang-slot) ; If no slot for the language, add it.
1173 (setq lang-slot (list lang-env)
1174 language-info-alist (cons lang-slot language-info-alist)))
1175 (setq key-slot (assq key lang-slot))
1176 (if (null key-slot) ; If no slot for the key, add it.
1178 (setq key-slot (list key))
1179 (setcdr lang-slot (cons key-slot (cdr lang-slot)))))
1180 (setcdr key-slot (purecopy info))
1181 ;; Update the custom-type of `current-language-environment'.
1182 (put 'current-language-environment 'custom-type
1183 (cons 'choice (mapcar
1186 (sort (mapcar 'car language-info-alist) 'string<))))))
1188 (defun set-language-info-alist (lang-env alist &optional parents)
1189 "Store ALIST as the definition of language environment LANG-ENV.
1190 ALIST is an alist of KEY and INFO values. See the documentation of
1191 `language-info-alist' for the meanings of KEY and INFO.
1193 Optional arg PARENTS is a list of parent menu names; it specifies
1194 where to put this language environment in the
1195 Describe Language Environment and Set Language Environment menus.
1196 For example, (\"European\") means to put this language environment
1197 in the European submenu in each of those two menus."
1198 (cond ((symbolp lang-env)
1199 (setq lang-env (symbol-name lang-env)))
1201 (setq lang-env (purecopy lang-env))))
1202 (let ((describe-map describe-language-environment-map)
1203 (setup-map setup-language-environment-map))
1206 map parent-symbol parent prompt)
1208 (if (symbolp (setq parent-symbol (car l)))
1209 (setq parent (symbol-name parent))
1210 (setq parent parent-symbol parent-symbol (intern parent)))
1211 (setq map (lookup-key describe-map (vector parent-symbol)))
1212 ;; This prompt string is for define-prefix-command, so
1213 ;; that the map it creates will be suitable for a menu.
1214 (or map (setq prompt (format "%s Environment" parent)))
1217 (setq map (intern (format "describe-%s-environment-map"
1218 (downcase parent))))
1219 (define-prefix-command map nil prompt)
1220 (define-key-after describe-map (vector parent-symbol)
1221 (cons parent map))))
1222 (setq describe-map (symbol-value map))
1223 (setq map (lookup-key setup-map (vector parent-symbol)))
1226 (setq map (intern (format "setup-%s-environment-map"
1227 (downcase parent))))
1228 (define-prefix-command map nil prompt)
1229 (define-key-after setup-map (vector parent-symbol)
1230 (cons parent map))))
1231 (setq setup-map (symbol-value map))
1234 ;; Set up menu items for this language env.
1235 (let ((doc (assq 'documentation alist)))
1237 (define-key-after describe-map (vector (intern lang-env))
1238 (cons lang-env 'describe-specified-language-support))))
1239 (define-key-after setup-map (vector (intern lang-env))
1240 (cons lang-env 'setup-specified-language-environment))
1243 (set-language-info-internal lang-env (car elt) (cdr elt)))
1245 (if (equal lang-env current-language-environment)
1246 (set-language-environment lang-env))))
1248 (defun read-language-name (key prompt &optional default)
1249 "Read a language environment name which has information for KEY.
1250 If KEY is nil, read any language environment.
1251 Prompt with PROMPT. DEFAULT is the default choice of language environment.
1252 This returns a language environment name as a string."
1253 (let* ((completion-ignore-case t)
1254 (name (completing-read prompt
1257 (function (lambda (elm) (and (listp elm) (assq key elm)))))
1258 t nil nil default)))
1259 (if (and (> (length name) 0)
1261 (get-language-info name key)))
1264 ;;; Multilingual input methods.
1266 "LEIM: Libraries of Emacs Input Methods."
1269 (defconst leim-list-file-name "leim-list.el"
1270 "Name of LEIM list file.
1271 This file contains a list of libraries of Emacs input methods (LEIM)
1272 in the format of Lisp expression for registering each input method.
1273 Emacs loads this file at startup time.")
1275 (defconst leim-list-header (format
1276 ";;; %s -- list of LEIM (Library of Emacs Input Method) -*-coding: utf-8;-*-
1278 ;; This file is automatically generated.
1280 ;; This file contains a list of LEIM (Library of Emacs Input Method)
1281 ;; methods in the same directory as this file. Loading this file
1282 ;; registers all the input methods in Emacs.
1284 ;; Each entry has the form:
1285 ;; (register-input-method
1286 ;; INPUT-METHOD LANGUAGE-NAME ACTIVATE-FUNC
1287 ;; TITLE DESCRIPTION
1289 ;; See the function ‘register-input-method’ for the meanings of the arguments.
1291 ;; If this directory is included in ‘load-path’, Emacs automatically
1292 ;; loads this file at startup time.
1295 leim-list-file-name)
1296 "Header to be inserted in LEIM list file.")
1298 (defconst leim-list-entry-regexp "^(register-input-method"
1299 "Regexp matching head of each entry in LEIM list file.
1300 See also the variable `leim-list-header'.")
1302 (defvar update-leim-list-functions
1303 '(quail-update-leim-list-file)
1304 "List of functions to call to update LEIM list file.
1305 Each function is called with one arg, LEIM directory name.")
1307 (defun update-leim-list-file (&rest dirs)
1308 "Update LEIM list file in directories DIRS."
1309 (dolist (function update-leim-list-functions)
1310 (apply function dirs)))
1312 (defvar current-input-method nil
1313 "The current input method for multilingual text.
1314 If nil, that means no input method is activated now.")
1315 (make-variable-buffer-local 'current-input-method)
1316 (put 'current-input-method 'permanent-local t)
1318 (defvar current-input-method-title nil
1319 "Title string of the current input method shown in mode line.")
1320 (make-variable-buffer-local 'current-input-method-title)
1321 (put 'current-input-method-title 'permanent-local t)
1323 (define-widget 'mule-input-method-string 'string
1324 "String widget with completion for input method."
1326 (lambda (string pred action)
1327 (let ((completion-ignore-case t))
1328 (complete-with-action action input-method-alist string pred)))
1329 :prompt-history 'input-method-history)
1331 (defcustom default-input-method nil
1332 "Default input method for multilingual text (a string).
1333 This is the input method activated automatically by the command
1334 `toggle-input-method' (\\[toggle-input-method])."
1335 :link '(custom-manual "(emacs)Input Methods")
1337 :type `(choice (const nil)
1338 mule-input-method-string)
1339 :set-after '(current-language-environment))
1341 (put 'input-method-function 'permanent-local t)
1343 (defvar input-method-history nil
1344 "History list of input methods read from the minibuffer.
1346 Maximum length of the history list is determined by the value
1347 of `history-length', which see.")
1348 (make-variable-buffer-local 'input-method-history)
1349 (put 'input-method-history 'permanent-local t)
1351 (define-obsolete-variable-alias
1352 'inactivate-current-input-method-function
1353 'deactivate-current-input-method-function "24.3")
1354 (defvar deactivate-current-input-method-function nil
1355 "Function to call for deactivating the current input method.
1356 Every input method should set this to an appropriate value when activated.
1357 This function is called with no argument.
1359 This function should never change the value of `current-input-method'.
1360 It is set to nil by the function `deactivate-input-method'.")
1361 (make-variable-buffer-local 'deactivate-current-input-method-function)
1362 (put 'deactivate-current-input-method-function 'permanent-local t)
1364 (defvar describe-current-input-method-function nil
1365 "Function to call for describing the current input method.
1366 This function is called with no argument.")
1367 (make-variable-buffer-local 'describe-current-input-method-function)
1368 (put 'describe-current-input-method-function 'permanent-local t)
1370 (defvar input-method-alist nil
1371 "Alist of input method names vs how to use them.
1372 Each element has the form:
1373 (INPUT-METHOD LANGUAGE-ENV ACTIVATE-FUNC TITLE DESCRIPTION ARGS...)
1374 See the function `register-input-method' for the meanings of the elements.")
1376 (put 'input-method-alist 'risky-local-variable t)
1378 (defun register-input-method (input-method lang-env &rest args)
1379 "Register INPUT-METHOD as an input method for language environment LANG-ENV.
1381 INPUT-METHOD and LANG-ENV are symbols or strings.
1382 ACTIVATE-FUNC is a function to call to activate this method.
1383 TITLE is a string to show in the mode line when this method is active.
1384 DESCRIPTION is a string describing this method and what it is good for.
1385 The ARGS, if any, are passed as arguments to ACTIVATE-FUNC.
1386 All told, the arguments to ACTIVATE-FUNC are INPUT-METHOD and the ARGS.
1388 This function is mainly used in the file \"leim-list.el\" which is
1389 created at Emacs build time, registering all Quail input methods
1390 contained in the Emacs distribution.
1392 In case you want to register a new Quail input method by yourself, be
1393 careful to use the same input method title as given in the third
1394 parameter of `quail-define-package'. (If the values are different, the
1395 string specified in this function takes precedence.)
1397 The commands `describe-input-method' and `list-input-methods' need
1398 these duplicated values to show some information about input methods
1399 without loading the relevant Quail packages.
1400 \n(fn INPUT-METHOD LANG-ENV ACTIVATE-FUNC TITLE DESCRIPTION &rest ARGS)"
1401 (if (symbolp lang-env)
1402 (setq lang-env (symbol-name lang-env))
1403 (setq lang-env (purecopy lang-env)))
1404 (if (symbolp input-method)
1405 (setq input-method (symbol-name input-method))
1406 (setq input-method (purecopy input-method)))
1407 (setq args (mapcar 'purecopy args))
1408 (let ((info (cons lang-env args))
1409 (slot (assoc input-method input-method-alist)))
1412 (setq slot (cons input-method info))
1413 (setq input-method-alist (cons slot input-method-alist)))))
1415 (defun read-input-method-name (prompt &optional default inhibit-null)
1416 "Read a name of input method from a minibuffer prompting with PROMPT.
1417 If DEFAULT is non-nil, use that as the default,
1418 and substitute it into PROMPT at the first `%s'.
1419 If INHIBIT-NULL is non-nil, null input signals an error.
1421 The return value is a string."
1423 (setq prompt (format prompt default)))
1424 (let* ((completion-ignore-case t)
1425 ;; As it is quite normal to change input method in the
1426 ;; minibuffer, we must enable it even if
1427 ;; enable-recursive-minibuffers is currently nil.
1428 (enable-recursive-minibuffers t)
1429 ;; This binding is necessary because input-method-history is
1431 (input-method (completing-read prompt input-method-alist
1432 nil t nil 'input-method-history
1433 (if (and default (symbolp default))
1434 (symbol-name default)
1436 (if (and input-method (symbolp input-method))
1437 (setq input-method (symbol-name input-method)))
1438 (if (> (length input-method) 0)
1441 (error "No valid input method is specified")))))
1443 (defun activate-input-method (input-method)
1444 "Switch to input method INPUT-METHOD for the current buffer.
1445 If some other input method is already active, turn it off first.
1446 If INPUT-METHOD is nil, deactivate any current input method."
1447 (if (and input-method (symbolp input-method))
1448 (setq input-method (symbol-name input-method)))
1449 (if (and current-input-method
1450 (not (string= current-input-method input-method)))
1451 (deactivate-input-method))
1452 (unless (or current-input-method (null input-method))
1453 (let ((slot (assoc input-method input-method-alist)))
1455 (error "Can't activate input method `%s'" input-method))
1456 (setq current-input-method-title nil)
1457 (let ((func (nth 2 slot)))
1458 (if (functionp func)
1459 (apply (nth 2 slot) input-method (nthcdr 5 slot))
1460 (if (and (consp func) (symbolp (car func)) (symbolp (cdr func)))
1462 (require (cdr func))
1463 (apply (car func) input-method (nthcdr 5 slot)))
1464 (error "Can't activate input method `%s'" input-method))))
1465 (setq current-input-method input-method)
1466 (or (stringp current-input-method-title)
1467 (setq current-input-method-title (nth 3 slot)))
1469 (run-hooks 'input-method-activate-hook)
1470 (force-mode-line-update)))))
1472 (defun deactivate-input-method ()
1473 "Turn off the current input method."
1474 (when current-input-method
1475 (if input-method-history
1476 (unless (string= current-input-method (car input-method-history))
1477 (setq input-method-history
1478 (cons current-input-method
1479 (delete current-input-method input-method-history))))
1480 (setq input-method-history (list current-input-method)))
1483 (setq input-method-function nil
1484 current-input-method-title nil)
1485 (funcall deactivate-current-input-method-function))
1488 'input-method-inactivate-hook ; for backward compatibility
1489 'input-method-deactivate-hook)
1490 (setq current-input-method nil)
1491 (force-mode-line-update)))))
1493 (define-obsolete-function-alias
1494 'inactivate-input-method
1495 'deactivate-input-method "24.3")
1497 (defun set-input-method (input-method &optional interactive)
1498 "Select and activate input method INPUT-METHOD for the current buffer.
1499 This also sets the default input method to the one you specify.
1500 If INPUT-METHOD is nil, this function turns off the input method, and
1501 also causes you to be prompted for a name of an input method the next
1502 time you invoke \\[toggle-input-method].
1503 When called interactively, the optional arg INTERACTIVE is non-nil,
1504 which marks the variable `default-input-method' as set for Custom buffers.
1506 To deactivate the input method interactively, use \\[toggle-input-method].
1507 To deactivate it programmatically, use `deactivate-input-method'."
1509 (let* ((default (or (car input-method-history) default-input-method)))
1510 (list (read-input-method-name
1511 (if default "Select input method (default %s): " "Select input method: ")
1514 (activate-input-method input-method)
1515 (setq default-input-method input-method)
1517 (customize-mark-as-set 'default-input-method))
1518 default-input-method)
1520 (defvar toggle-input-method-active nil
1521 "Non-nil inside `toggle-input-method'.")
1523 (defun toggle-input-method (&optional arg interactive)
1524 "Enable or disable multilingual text input method for the current buffer.
1525 Only one input method can be enabled at any time in a given buffer.
1527 The normal action is to enable an input method if none was enabled,
1528 and disable the current one otherwise. Which input method to enable
1529 can be determined in various ways--either the one most recently used,
1530 or the one specified by `default-input-method', or as a last resort
1531 by reading the name of an input method in the minibuffer.
1533 With a prefix argument ARG, read an input method name with the minibuffer
1534 and enable that one. The default is the most recent input method specified
1535 \(not including the currently active input method, if any).
1537 When called interactively, the optional argument INTERACTIVE is non-nil,
1538 which marks the variable `default-input-method' as set for Custom buffers."
1540 (interactive "P\np")
1541 (if toggle-input-method-active
1542 (error "Recursive use of `toggle-input-method'"))
1543 (if (and current-input-method (not arg))
1544 (deactivate-input-method)
1545 (let ((toggle-input-method-active t)
1546 (default (or (car input-method-history) default-input-method)))
1547 (if (and arg default (equal current-input-method default)
1548 (> (length input-method-history) 1))
1549 (setq default (nth 1 input-method-history)))
1550 (activate-input-method
1551 (if (or arg (not default))
1553 (read-input-method-name
1554 (if default "Input method (default %s): " "Input method: " )
1557 (unless default-input-method
1559 (setq default-input-method current-input-method)
1561 (customize-mark-as-set 'default-input-method)))))))
1563 (autoload 'help-buffer "help-mode")
1565 (defun describe-input-method (input-method)
1566 "Describe input method INPUT-METHOD."
1568 (list (read-input-method-name
1569 "Describe input method (default current choice): ")))
1570 (if (and input-method (symbolp input-method))
1571 (setq input-method (symbol-name input-method)))
1572 (help-setup-xref (list #'describe-input-method
1573 (or input-method current-input-method))
1574 (called-interactively-p 'interactive))
1576 (if (null input-method)
1577 (describe-current-input-method)
1578 (let ((current current-input-method))
1582 (activate-input-method input-method)
1583 (describe-current-input-method))
1584 (activate-input-method current))
1586 (activate-input-method current)
1587 (help-setup-xref (list #'describe-input-method input-method)
1588 (called-interactively-p 'interactive))
1589 (with-output-to-temp-buffer (help-buffer)
1590 (let ((elt (assoc input-method input-method-alist)))
1591 (princ (format-message
1592 "Input method: %s (`%s' in mode line) for %s\n %s\n"
1593 input-method (nth 3 elt) (nth 1 elt) (nth 4 elt))))))))))
1595 (defun describe-current-input-method ()
1596 "Describe the input method currently in use.
1597 This is a subroutine for `describe-input-method'."
1598 (if current-input-method
1599 (if (and (symbolp describe-current-input-method-function)
1600 (fboundp describe-current-input-method-function))
1601 (funcall describe-current-input-method-function)
1602 (message "No way to describe the current input method `%s'"
1603 current-input-method)
1605 (error "No input method is activated now")))
1607 (defun read-multilingual-string (prompt &optional initial-input input-method)
1608 "Read a multilingual string from minibuffer, prompting with string PROMPT.
1609 The input method selected last time is activated in minibuffer.
1610 If optional second argument INITIAL-INPUT is non-nil, insert it in the
1611 minibuffer initially.
1612 Optional 3rd argument INPUT-METHOD specifies the input method to be activated
1613 instead of the one selected last time. It is a symbol or a string."
1616 current-input-method
1617 default-input-method
1618 (read-input-method-name "Input method: " nil t)))
1619 (if (and input-method (symbolp input-method))
1620 (setq input-method (symbol-name input-method)))
1621 (let ((prev-input-method current-input-method))
1624 (activate-input-method input-method)
1625 (read-string prompt initial-input nil nil t))
1626 (activate-input-method prev-input-method))))
1628 ;; Variables to control behavior of input methods. All input methods
1629 ;; should react to these variables.
1631 (defcustom input-method-verbose-flag 'default
1632 "A flag to control extra guidance given by input methods.
1633 The value should be nil, t, `complex-only', or `default'.
1635 The extra guidance is done by showing list of available keys in echo
1636 area. When you use the input method in the minibuffer, the guidance
1637 is shown at the bottom short window (split from the existing window).
1639 If the value is t, extra guidance is always given, if the value is
1640 nil, extra guidance is always suppressed.
1642 If the value is `complex-only', only complex input methods such as
1643 `chinese-py' and `japanese' give extra guidance.
1645 If the value is `default', complex input methods always give extra
1646 guidance, but simple input methods give it only when you are not in
1649 See also the variable `input-method-highlight-flag'."
1650 :type '(choice (const :tag "Always" t) (const :tag "Never" nil)
1651 (const complex-only) (const default))
1654 (defcustom input-method-highlight-flag t
1655 "If this flag is non-nil, input methods highlight partially-entered text.
1656 For instance, while you are in the middle of a Quail input method sequence,
1657 the text inserted so far is temporarily underlined.
1658 The underlining goes away when you finish or abort the input method sequence.
1659 See also the variable `input-method-verbose-flag'."
1663 (defcustom input-method-activate-hook nil
1664 "Normal hook run just after an input method is activated.
1666 The variable `current-input-method' keeps the input method name
1671 (define-obsolete-variable-alias
1672 'input-method-inactivate-hook
1673 'input-method-deactivate-hook "24.3")
1675 (defcustom input-method-deactivate-hook nil
1676 "Normal hook run just after an input method is deactivated.
1678 The variable `current-input-method' still keeps the input method name
1684 (defcustom input-method-after-insert-chunk-hook nil
1685 "Normal hook run just after an input method insert some chunk of text."
1689 (defvar input-method-exit-on-first-char nil
1690 "This flag controls when an input method returns.
1691 Usually, the input method does not return while there's a possibility
1692 that it may find a different translation if a user types another key.
1693 But, if this flag is non-nil, the input method returns as soon as the
1694 current key sequence gets long enough to have some valid translation.")
1696 (defcustom input-method-use-echo-area nil
1697 "This flag controls how an input method shows an intermediate key sequence.
1698 Usually, the input method inserts the intermediate key sequence,
1699 or candidate translations corresponding to the sequence,
1700 at point in the current buffer.
1701 But, if this flag is non-nil, it displays them in echo area instead."
1705 (defvar input-method-exit-on-invalid-key nil
1706 "This flag controls the behavior of an input method on invalid key input.
1707 Usually, when a user types a key which doesn't start any character
1708 handled by the input method, the key is handled by turning off the
1709 input method temporarily. After that key, the input method is re-enabled.
1710 But, if this flag is non-nil, the input method is never back on.")
1713 (defcustom set-language-environment-hook nil
1714 "Normal hook run after some language environment is set.
1716 When you set some hook function here, that effect usually should not
1717 be inherited to another language environment. So, you had better set
1718 another function in `exit-language-environment-hook' (which see) to
1723 (defcustom exit-language-environment-hook nil
1724 "Normal hook run after exiting from some language environment.
1725 When this hook is run, the variable `current-language-environment'
1726 is still bound to the language environment being exited.
1728 This hook is mainly used for canceling the effect of
1729 `set-language-environment-hook' (which see)."
1733 (put 'setup-specified-language-environment 'apropos-inhibit t)
1735 (defun setup-specified-language-environment ()
1736 "Switch to a specified language environment."
1738 (let (language-name)
1739 (if (and (symbolp last-command-event)
1740 (or (not (eq last-command-event 'Default))
1741 (setq last-command-event 'English))
1742 (setq language-name (symbol-name last-command-event)))
1744 (set-language-environment language-name)
1745 (customize-mark-as-set 'current-language-environment))
1746 (error "Bogus calling sequence"))))
1748 (defcustom current-language-environment "English"
1749 "The last language environment specified with `set-language-environment'.
1750 This variable should be set only with \\[customize], which is equivalent
1751 to using the function `set-language-environment'."
1752 :link '(custom-manual "(emacs)Language Environments")
1753 :set (lambda (_symbol value) (set-language-environment value))
1755 (or (car-safe (assoc-string
1756 (if (symbolp current-language-environment)
1757 (symbol-name current-language-environment)
1758 current-language-environment)
1759 language-info-alist t))
1761 ;; custom type will be updated with `set-language-info'.
1762 :type (if language-info-alist
1763 (cons 'choice (mapcar
1766 (sort (mapcar 'car language-info-alist) 'string<)))
1768 :initialize 'custom-initialize-default
1771 (defun reset-language-environment ()
1772 "Reset multilingual environment of Emacs to the default status.
1774 The default status is as follows:
1776 The default value of `buffer-file-coding-system' is nil.
1777 The default coding system for process I/O is nil.
1778 The default value for the command `set-terminal-coding-system' is nil.
1779 The default value for the command `set-keyboard-coding-system' is nil.
1781 The order of priorities of coding systems are as follows:
1790 ;; This function formerly set default-enable-multibyte-characters to t,
1791 ;; but that is incorrect. It should not alter the unibyte/multibyte choice.
1793 (set-coding-system-priority
1802 (set-default-coding-systems nil)
1803 (setq default-sendmail-coding-system 'iso-latin-1)
1804 ;; On Darwin systems, this should be utf-8, but when this file is loaded
1805 ;; utf-8 is not yet defined, so we set it in set-locale-environment instead.
1806 (setq default-file-name-coding-system 'iso-latin-1)
1807 ;; Preserve eol-type from existing default-process-coding-systems.
1808 ;; On non-unix-like systems in particular, these may have been set
1809 ;; carefully by the user, or by the startup code, to deal with the
1810 ;; users shell appropriately, so should not be altered by changing
1811 ;; language environment.
1812 (let ((output-coding
1813 ;; When bootstrapping, coding-systems are not defined yet, so
1814 ;; we need to catch the error from check-coding-system.
1816 (coding-system-change-text-conversion
1817 (car default-process-coding-system) 'undecided)
1818 (coding-system-error 'undecided)))
1821 (coding-system-change-text-conversion
1822 (cdr default-process-coding-system) 'iso-latin-1)
1823 (coding-system-error 'iso-latin-1))))
1824 (setq default-process-coding-system
1825 (cons output-coding input-coding)))
1827 ;; Put the highest priority to the charset iso-8859-1 to prefer the
1828 ;; registry iso8859-1 over iso8859-2 in font selection. It also
1829 ;; makes unibyte-display-via-language-environment to use iso-8859-1
1830 ;; as the unibyte charset.
1831 (set-charset-priority 'iso-8859-1)
1833 ;; Don't alter the terminal and keyboard coding systems here.
1834 ;; The terminal still supports the same coding system
1835 ;; that it supported a minute ago.
1836 ;; (set-terminal-coding-system-internal nil)
1837 ;; (set-keyboard-coding-system-internal nil)
1839 ;; Back in Emacs-20, it was necessary to provide some fallback implicit
1840 ;; conversion, because almost no packages handled coding-system issues.
1841 ;; Nowadays it'd just paper over bugs.
1842 ;; (set-unibyte-charset 'iso-8859-1)
1845 (reset-language-environment)
1847 (defun set-display-table-and-terminal-coding-system (language-name &optional coding-system display)
1848 "Set up the display table and terminal coding system for LANGUAGE-NAME."
1849 (let ((coding (get-language-info language-name 'unibyte-display)))
1851 (or (not coding-system)
1852 (coding-system-equal coding coding-system)))
1853 (standard-display-european-internal)
1854 ;; The following 2 lines undo the 8-bit display that we set up
1855 ;; in standard-display-european-internal, which see. This is in
1856 ;; case the user has used standard-display-european earlier in
1858 (when standard-display-table
1860 (aset standard-display-table (+ i 128) nil))))
1861 (set-terminal-coding-system (or coding-system coding) display)))
1863 (defun set-language-environment (language-name)
1864 "Set up multilingual environment for using LANGUAGE-NAME.
1865 This sets the coding system priority and the default input method
1866 and sometimes other things. LANGUAGE-NAME should be a string
1867 which is the name of a language environment. For example, \"Latin-1\"
1868 specifies the character set for the major languages of Western Europe.
1870 If there is a prior value for `current-language-environment', this
1871 runs the hook `exit-language-environment-hook'. After setting up
1872 the new language environment, it runs `set-language-environment-hook'."
1873 (interactive (list (read-language-name
1875 "Set language environment (default English): ")))
1877 (if (symbolp language-name)
1878 (setq language-name (symbol-name language-name)))
1879 (setq language-name "English"))
1880 (let ((slot (assoc-string language-name language-info-alist t)))
1882 (error "Language environment not defined: %S" language-name))
1883 (setq language-name (car slot)))
1884 (if current-language-environment
1885 (let ((func (get-language-info current-language-environment
1887 (run-hooks 'exit-language-environment-hook)
1888 (if (functionp func) (funcall func))))
1890 (reset-language-environment)
1891 ;; The features might set up coding systems.
1892 (let ((required-features (get-language-info language-name 'features)))
1893 (while required-features
1894 (require (car required-features))
1895 (setq required-features (cdr required-features))))
1897 (setq current-language-environment language-name)
1899 (set-language-environment-coding-systems language-name)
1900 (set-language-environment-input-method language-name)
1901 (set-language-environment-nonascii-translation language-name)
1902 (set-language-environment-charset language-name)
1903 ;; Unibyte setups if necessary.
1904 (unless (default-value 'enable-multibyte-characters)
1905 (set-language-environment-unibyte language-name))
1907 (let ((func (get-language-info language-name 'setup-function)))
1908 (if (functionp func)
1911 (setq current-iso639-language
1912 (or (get-language-info language-name 'iso639-language)
1913 current-iso639-language))
1915 (run-hooks 'set-language-environment-hook)
1916 (force-mode-line-update t))
1918 (define-widget 'charset 'symbol
1922 (lambda (string pred action)
1923 (let ((completion-ignore-case t))
1924 (completion-table-with-predicate
1925 obarray #'charsetp 'strict string pred action)))
1927 :validate (lambda (widget)
1928 (unless (charsetp (widget-value widget))
1929 (widget-put widget :error (format "Invalid charset: %S"
1930 (widget-value widget)))
1932 :prompt-history 'charset-history)
1934 (defcustom language-info-custom-alist nil
1935 "Customizations of language environment parameters.
1936 Value is an alist with elements like those of `language-info-alist'.
1937 These are used to set values in `language-info-alist' which replace
1938 the defaults. A typical use is replacing the default input method for
1939 the environment. Use \\[describe-language-environment] to find the environment's settings.
1941 This option is intended for use at startup. Removing items doesn't
1942 remove them from the language info until you next restart Emacs.
1944 Setting this variable directly does not take effect.
1945 See `set-language-info-alist' for use in programs."
1949 (custom-set-default s v)
1950 ;; Can't do this before language environments are set up.
1952 ;; modify language-info-alist
1954 (set-language-info-alist (car elt) (cdr elt)))
1955 ;; re-set the environment in case its parameters changed
1956 (set-language-environment current-language-environment)))
1958 :key-type (string :tag "Language environment"
1960 (lambda (string pred action)
1961 (let ((completion-ignore-case t))
1962 (complete-with-action
1963 action language-info-alist string pred))))
1965 (alist :key-type symbol
1966 :options ((documentation string)
1967 (charset (repeat charset))
1968 (sample-text string)
1969 (setup-function function)
1970 (exit-function function)
1971 (coding-system (repeat coding-system))
1972 (coding-priority (repeat coding-system))
1973 (nonascii-translation charset)
1974 (input-method mule-input-method-string)
1975 (features (repeat symbol))
1976 (unibyte-display coding-system)))))
1978 (declare-function x-server-vendor "xfns.c" (&optional terminal))
1979 (declare-function x-server-version "xfns.c" (&optional terminal))
1981 (defun standard-display-european-internal ()
1982 ;; Actually set up direct output of non-ASCII characters.
1983 (standard-display-8bit (if (eq window-system 'pc) 128 160) 255)
1984 ;; Unibyte Emacs on MS-DOS wants to display all 8-bit characters with
1985 ;; the native font, and codes 160 and 146 stand for something very
1987 (or (and (eq window-system 'pc) (not (default-value
1988 'enable-multibyte-characters)))
1990 ;; Most X fonts used to do the wrong thing for latin-1 code 160.
1991 (unless (and (eq window-system 'x)
1992 ;; XFree86 4 has fixed the fonts.
1993 (string= "The XFree86 Project, Inc" (x-server-vendor))
1994 (> (aref (number-to-string (nth 2 (x-server-version))) 0)
1996 ;; Make non-line-break space display as a plain space.
1997 (aset standard-display-table (unibyte-char-to-multibyte 160) [32]))
1998 ;; Most Windows programs send out apostrophes as \222. Most X fonts
1999 ;; don't contain a character at that position. Map it to the ASCII
2000 ;; apostrophe. [This is actually RIGHT SINGLE QUOTATION MARK,
2001 ;; U+2019, normally from the windows-1252 character set. XFree 4
2002 ;; fonts probably have the appropriate glyph at this position,
2003 ;; so they could use standard-display-8bit. It's better to use a
2004 ;; proper windows-1252 coding system. --fx]
2005 (aset standard-display-table (unibyte-char-to-multibyte 146) [39]))))
2007 (defun set-language-environment-coding-systems (language-name)
2008 "Do various coding system setups for language environment LANGUAGE-NAME."
2009 (let* ((priority (get-language-info language-name 'coding-priority))
2010 (default-coding (car priority))
2011 ;; If the default buffer-file-coding-system is nil, don't use
2012 ;; coding-system-eol-type, because it treats nil as
2013 ;; `no-conversion'. The default buffer-file-coding-system is set
2014 ;; to nil by reset-language-environment, and in that case we
2015 ;; want to have here the native EOL type for each platform.
2016 ;; FIXME: there should be a common code that runs both on
2017 ;; startup and here to set the default EOL type correctly.
2018 ;; Right now, DOS/Windows platforms set this on dos-w32.el,
2019 ;; which works only as long as the order of loading files at
2020 ;; dump time and calling functions at startup is not modified
2021 ;; significantly, i.e. as long as this function is called
2022 ;; _after_ the default buffer-file-coding-system was set by
2025 (coding-system-eol-type
2026 (or (default-value 'buffer-file-coding-system)
2027 (if (memq system-type '(windows-nt ms-dos)) 'dos 'unix)))))
2029 (set-default-coding-systems
2030 (if (memq eol-type '(0 1 2 unix dos mac))
2031 (coding-system-change-eol-conversion default-coding eol-type)
2033 (setq default-sendmail-coding-system default-coding)
2034 (apply 'set-coding-system-priority priority))))
2036 (defun set-language-environment-input-method (language-name)
2037 "Do various input method setups for language environment LANGUAGE-NAME."
2038 (let ((input-method (get-language-info language-name 'input-method)))
2040 (setq default-input-method input-method)
2041 (if input-method-history
2042 (setq input-method-history
2044 (delete input-method input-method-history)))))))
2046 (defun set-language-environment-nonascii-translation (language-name)
2047 "Do unibyte/multibyte translation setup for language environment LANGUAGE-NAME."
2048 ;; Note: For DOS, we assumed that the charset cpXXX is already
2050 (let ((nonascii (get-language-info language-name 'nonascii-translation)))
2051 (if (eq window-system 'pc)
2052 (setq nonascii (intern (format "cp%d" dos-codepage))))
2053 (or (and (charsetp nonascii)
2054 (get-charset-property nonascii :ascii-compatible-p))
2055 (setq nonascii 'iso-8859-1))
2056 ;; Back in Emacs-20, it was necessary to provide some fallback implicit
2057 ;; conversion, because almost no packages handled coding-system issues.
2058 ;; Nowadays it'd just paper over bugs.
2059 ;; (set-unibyte-charset nonascii)
2062 (defun set-language-environment-charset (language-name)
2063 "Do various charset setups for language environment LANGUAGE-NAME."
2064 ;; Put higher priorities to such charsets that are supported by the
2065 ;; coding systems of higher priorities in this environment.
2066 (let ((charsets (get-language-info language-name 'charset)))
2067 (dolist (coding (get-language-info language-name 'coding-priority))
2068 (let ((list (coding-system-charset-list coding)))
2070 (setq charsets (append charsets list)))))
2072 (apply 'set-charset-priority charsets))))
2074 (defun set-language-environment-unibyte (language-name)
2075 "Do various unibyte-mode setups for language environment LANGUAGE-NAME."
2076 (set-display-table-and-terminal-coding-system language-name))
2078 (defun princ-list (&rest args)
2079 "Print all arguments with `princ', then print \"\\n\"."
2080 (declare (obsolete "use mapc and princ instead." "23.3"))
2084 (put 'describe-specified-language-support 'apropos-inhibit t)
2086 ;; Print language-specific information such as input methods,
2087 ;; charsets, and coding systems. This function is intended to be
2088 ;; called from the menu:
2089 ;; [menu-bar mule describe-language-environment LANGUAGE]
2090 ;; and should not run it by `M-x describe-current-input-method-function'.
2091 (defun describe-specified-language-support ()
2092 "Describe how Emacs supports the specified language environment."
2094 (let (language-name)
2095 (if (not (and (symbolp last-command-event)
2096 (or (not (eq last-command-event 'Default))
2097 (setq last-command-event 'English))
2098 (setq language-name (symbol-name last-command-event))))
2099 (error "This command should only be called from the menu bar"))
2100 (describe-language-environment language-name)))
2102 (defun describe-language-environment (language-name)
2103 "Describe how Emacs supports language environment LANGUAGE-NAME."
2105 (list (read-language-name
2107 "Describe language environment (default current choice): ")))
2108 (if (null language-name)
2109 (setq language-name current-language-environment))
2110 (if (or (null language-name)
2111 (null (get-language-info language-name 'documentation)))
2112 (error "No documentation for the specified language"))
2113 (if (symbolp language-name)
2114 (setq language-name (symbol-name language-name)))
2115 (dolist (feature (get-language-info language-name 'features))
2117 (let ((doc (get-language-info language-name 'documentation)))
2118 (help-setup-xref (list #'describe-language-environment language-name)
2119 (called-interactively-p 'interactive))
2120 (with-output-to-temp-buffer (help-buffer)
2121 (with-current-buffer standard-output
2122 (insert language-name " language environment\n\n")
2124 (insert doc "\n\n"))
2126 (let ((str (eval (get-language-info language-name 'sample-text))))
2128 (insert "Sample text:\n "
2129 (replace-regexp-in-string "\n" "\n " str)
2132 (let ((input-method (get-language-info language-name 'input-method))
2133 (l (copy-sequence input-method-alist))
2135 (when (and input-method
2136 (setq input-method (assoc input-method l)))
2137 (insert "Input methods (default " (car input-method) ")\n")
2138 (setq l (cons input-method (delete input-method l))
2141 (when (or (eq input-method elt)
2142 (eq t (compare-strings language-name nil nil
2143 (nth 1 elt) nil nil t)))
2145 (insert "Input methods:\n")
2147 (insert " " (car elt))
2148 (search-backward (car elt))
2149 (help-xref-button 0 'help-input-method (car elt))
2150 (goto-char (point-max))
2152 (if (stringp (nth 3 elt)) (nth 3 elt) (car (nth 3 elt)))
2153 "\" in mode line)\n")))
2156 (insert "Character sets:\n")
2157 (let ((l (get-language-info language-name 'charset)))
2159 (insert " nothing specific to " language-name "\n")
2161 (insert " " (symbol-name (car l)))
2162 (search-backward (symbol-name (car l)))
2163 (help-xref-button 0 'help-character-set (car l))
2164 (goto-char (point-max))
2165 (insert ": " (charset-description (car l)) "\n")
2168 (insert "Coding systems:\n")
2169 (let ((l (get-language-info language-name 'coding-system)))
2171 (insert " nothing specific to " language-name "\n")
2173 (insert " " (symbol-name (car l)))
2174 (search-backward (symbol-name (car l)))
2175 (help-xref-button 0 'help-coding-system (car l))
2176 (goto-char (point-max))
2177 (insert (substitute-command-keys " (`")
2178 (coding-system-mnemonic (car l))
2179 (substitute-command-keys "' in mode line):\n\t")
2180 (substitute-command-keys
2181 (coding-system-doc-string (car l)))
2183 (let ((aliases (coding-system-aliases (car l))))
2185 (insert "\t(alias:")
2187 (insert " " (symbol-name (car aliases)))
2188 (setq aliases (cdr aliases)))
2190 (setq l (cdr l)))))))))
2194 (defvar locale-translation-file-name nil
2195 "File name for the system's file of locale-name aliases, or nil if none.")
2197 ;; The following definitions might as well be marked as constants and
2198 ;; purecopied, since they're normally used on startup, and probably
2199 ;; should reflect the facilities of the base Emacs.
2200 (defconst locale-language-names
2203 ;; Locale names of the form LANGUAGE[_TERRITORY][.CODESET][@MODIFIER]
2204 ;; as specified in the Single Unix Spec, Version 2.
2205 ;; LANGUAGE is a language code taken from ISO 639:1988 (E/F)
2206 ;; with additions from ISO 639/RA Newsletter No.1/1989;
2207 ;; see Internet RFC 2165 (1997-06) and
2208 ;; http://www.evertype.com/standards/iso639/iso639-en.html
2209 ;; TERRITORY is a country code taken from ISO 3166
2210 ;; http://www.din.de/gremien/nas/nabd/iso3166ma/codlstp1/en_listp1.html.
2211 ;; CODESET and MODIFIER are implementation-dependent.
2213 ;; jasonr comments: MS Windows uses three letter codes for
2214 ;; languages instead of the two letter ISO codes that POSIX
2215 ;; uses. In most cases the first two letters are the same, so
2216 ;; most of the regexps in locale-language-names work. Japanese
2217 ;; and Chinese are exceptions, which are listed in the
2218 ;; non-standard section at the bottom of locale-language-names.
2220 ("aa_DJ" . "Latin-1") ; Afar
2223 ("af" . "Latin-1") ; Afrikaans
2224 ("am" "Ethiopic" utf-8) ; Amharic
2225 ("an" . "Latin-9") ; Aragonese
2229 ("az" . "UTF-8") ; Azerbaijani
2231 ("be" "Belarusian" cp1251) ; Belarusian [Byelorussian until early 1990s]
2232 ("bg" "Bulgarian" cp1251) ; Bulgarian
2235 ("bn" . "UTF-8") ; Bengali, Bangla
2237 ("br" . "Latin-1") ; Breton
2238 ("bs" . "Latin-2") ; Bosnian
2239 ("byn" . "UTF-8") ; Bilin; Blin
2240 ("ca" . "Latin-1") ; Catalan
2242 ("cs" "Czech" iso-8859-2)
2243 ("cy" "Welsh" iso-8859-14)
2244 ("da" . "Latin-1") ; Danish
2245 ("de" "German" iso-8859-1)
2248 ("el" "Greek" iso-8859-7)
2249 ;; Users who specify "en" explicitly typically want Latin-1, not ASCII.
2250 ;; That's actually what the GNU locales define, modulo things like
2252 ("en_IN" "English" utf-8) ; glibc uses utf-8 for English in India
2253 ("en" "English" iso-8859-1) ; English
2254 ("eo" . "Esperanto") ; Esperanto
2255 ("es" "Spanish" iso-8859-1)
2256 ("et" . "Latin-1") ; Estonian
2257 ("eu" . "Latin-1") ; Basque
2258 ("fa" . "UTF-8") ; Persian
2259 ("fi" . "Latin-1") ; Finnish
2260 ("fj" . "Latin-1") ; Fiji
2261 ("fo" . "Latin-1") ; Faroese
2262 ("fr" "French" iso-8859-1) ; French
2263 ("fy" . "Latin-1") ; Frisian
2264 ("ga" . "Latin-1") ; Irish Gaelic (new orthography)
2265 ("gd" . "Latin-9") ; Scots Gaelic
2266 ("gez" "Ethiopic" utf-8) ; Geez
2267 ("gl" . "Latin-1") ; Gallegan; Galician
2269 ("gu" . "UTF-8") ; Gujarati
2270 ("gv" . "Latin-1") ; Manx Gaelic
2272 ("he" "Hebrew" iso-8859-8)
2273 ("hi" "Devanagari" utf-8) ; Hindi
2274 ("hr" "Croatian" iso-8859-2) ; Croatian
2275 ("hu" . "Latin-2") ; Hungarian
2278 ("id" . "Latin-1") ; Indonesian
2281 ("is" . "Latin-1") ; Icelandic
2282 ("it" "Italian" iso-8859-1) ; Italian
2284 ("iw" "Hebrew" iso-8859-8)
2285 ("ja" "Japanese" euc-jp)
2287 ("ka" "Georgian" georgian-ps) ; Georgian
2289 ("kl" . "Latin-1") ; Greenlandic
2291 ("kn" "Kannada" utf-8)
2292 ("ko" "Korean" euc-kr)
2295 ("kw" . "Latin-1") ; Cornish
2297 ("la" . "Latin-1") ; Latin
2298 ("lb" . "Latin-1") ; Luxemburgish
2299 ("lg" . "Laint-6") ; Ganda
2301 ("lo" "Lao" utf-8) ; Laothian
2302 ("lt" "Lithuanian" iso-8859-13)
2303 ("lv" . "Latvian") ; Latvian, Lettish
2305 ("mi" . "Latin-7") ; Maori
2306 ("mk" "Cyrillic-ISO" iso-8859-5) ; Macedonian
2307 ("ml" "Malayalam" utf-8)
2308 ("mn" . "UTF-8") ; Mongolian
2310 ("mr" "Devanagari" utf-8) ; Marathi
2311 ("ms" . "Latin-1") ; Malay
2312 ("mt" . "Latin-3") ; Maltese
2315 ("nb" . "Latin-1") ; Norwegian
2316 ("ne" "Devanagari" utf-8) ; Nepali
2317 ("nl" "Dutch" iso-8859-1)
2318 ("no" . "Latin-1") ; Norwegian
2319 ("oc" . "Latin-1") ; Occitan
2320 ("om_ET" . "UTF-8") ; (Afan) Oromo
2321 ("om" . "Latin-1") ; (Afan) Oromo
2323 ("pa" . "UTF-8") ; Punjabi
2324 ("pl" . "Latin-2") ; Polish
2326 ("pt" . "Latin-1") ; Portuguese
2328 ("rm" . "Latin-1") ; Rhaeto-Romanic
2330 ("ro" "Romanian" iso-8859-2)
2331 ("ru_RU" "Russian" iso-8859-5)
2332 ("ru_UA" "Russian" koi8-u)
2334 ("sa" . "Devanagari") ; Sanskrit
2336 ("se" . "UTF-8") ; Northern Sami
2338 ("sh" . "Latin-2") ; Serbo-Croatian
2340 ("sid" . "UTF-8") ; Sidamo
2341 ("sk" "Slovak" iso-8859-2)
2342 ("sl" "Slovenian" iso-8859-2)
2345 ("so_ET" "UTF-8") ; Somali
2346 ("so" "Latin-1") ; Somali
2347 ("sq" . "Latin-1") ; Albanian
2348 ("sr" . "Latin-2") ; Serbian (Latin alphabet)
2350 ("st" . "Latin-1") ; Sesotho
2352 ("sv" "Swedish" iso-8859-1) ; Swedish
2353 ("sw" . "Latin-1") ; Swahili
2354 ("ta" "Tamil" utf-8)
2355 ("te" . "UTF-8") ; Telugu
2356 ("tg" "Tajik" koi8-t)
2357 ("th" "Thai" tis-620)
2358 ("ti" "Ethiopic" utf-8) ; Tigrinya
2359 ("tig_ER" . "UTF-8") ; Tigre
2361 ("tl" . "Latin-1") ; Tagalog
2364 ("tr" "Turkish" iso-8859-9)
2366 ("tt" . "UTF-8") ; Tatar
2369 ("uk" "Ukrainian" koi8-u)
2370 ("ur" . "UTF-8") ; Urdu
2371 ("uz_UZ@cyrillic" . "UTF-8"); Uzbek
2372 ("uz" . "Latin-1") ; Uzbek
2373 ("vi" "Vietnamese" utf-8)
2375 ("wa" . "Latin-1") ; Walloon
2377 ("xh" . "Latin-1") ; Xhosa
2378 ("yi" . "Windows-1255") ; Yiddish
2381 ("zh_HK" . "Chinese-Big5")
2382 ; zh_HK/BIG5-HKSCS \
2383 ("zh_TW" . "Chinese-Big5")
2384 ("zh_CN.GB2312" "Chinese-GB")
2385 ("zh_CN.GBK" "Chinese-GBK")
2386 ("zh_CN.GB18030" "Chinese-GB18030")
2387 ("zh_CN.UTF-8" . "Chinese-GBK")
2388 ("zh_CN" . "Chinese-GB")
2389 ("zh" . "Chinese-GB")
2390 ("zu" . "Latin-1") ; Zulu
2392 ;; ISO standard locales
2394 ("posix$" . "ASCII")
2396 ;; The "IPA" Emacs language environment does not correspond
2397 ;; to any ISO 639 code, so let it stand for itself.
2400 ;; Nonstandard or obsolete language codes
2401 ("cz" . "Czech") ; e.g. Solaris 2.6
2402 ("ee" . "Latin-4") ; Estonian, e.g. X11R6.4
2403 ("iw" . "Hebrew") ; e.g. X11R6.4
2404 ("sp" . "Cyrillic-ISO") ; Serbian (Cyrillic alphabet), e.g. X11R6.4
2405 ("su" . "Latin-1") ; Finnish, e.g. Solaris 2.6
2406 ("jp" . "Japanese") ; e.g. MS Windows
2407 ("chs" . "Chinese-GBK") ; MS Windows Chinese Simplified
2408 ("cht" . "Chinese-BIG5") ; MS Windows Chinese Traditional
2409 ("gbz" . "UTF-8") ; MS Windows Dari Persian
2410 ("div" . "UTF-8") ; MS Windows Divehi (Maldives)
2411 ("wee" . "Latin-2") ; MS Windows Lower Sorbian
2412 ("wen" . "Latin-2") ; MS Windows Upper Sorbian
2414 "Alist of locale regexps vs the corresponding languages and coding systems.
2415 Each element has this form:
2416 \(LOCALE-REGEXP LANG-ENV CODING-SYSTEM)
2417 The first element whose LOCALE-REGEXP matches the start of a
2418 downcased locale specifies the LANG-ENV \(language environment)
2419 and CODING-SYSTEM corresponding to that locale. If there is no
2420 appropriate language environment, the element may have this form:
2421 \(LOCALE-REGEXP . LANG-ENV)
2422 In this case, LANG-ENV is one of generic language environments for an
2423 specific encoding such as \"Latin-1\" and \"UTF-8\".")
2425 (defconst locale-charset-language-names
2427 '((".*8859[-_]?1\\>" . "Latin-1")
2428 (".*8859[-_]?2\\>" . "Latin-2")
2429 (".*8859[-_]?3\\>" . "Latin-3")
2430 (".*8859[-_]?4\\>" . "Latin-4")
2431 (".*8859[-_]?9\\>" . "Latin-5")
2432 (".*8859[-_]?14\\>" . "Latin-8")
2433 (".*8859[-_]?15\\>" . "Latin-9")
2434 (".*utf\\(?:-?8\\)?\\>" . "UTF-8")
2435 ;; utf-8@euro exists, so put this last. (@euro really specifies
2436 ;; the currency, rather than the charset.)
2437 (".*@euro\\>" . "Latin-9")))
2438 "List of pairs of locale regexps and charset language names.
2439 The first element whose locale regexp matches the start of a downcased locale
2440 specifies the language name whose charset corresponds to that locale.
2441 This language name is used if the locale is not listed in
2442 `locale-language-names'.")
2444 (defconst locale-preferred-coding-systems
2446 '((".*8859[-_]?1\\>" . iso-8859-1)
2447 (".*8859[-_]?2\\>" . iso-8859-2)
2448 (".*8859[-_]?3\\>" . iso-8859-3)
2449 (".*8859[-_]?4\\>" . iso-8859-4)
2450 (".*8859[-_]?9\\>" . iso-8859-9)
2451 (".*8859[-_]?14\\>" . iso-8859-14)
2452 (".*8859[-_]?15\\>" . iso-8859-15)
2453 (".*utf\\(?:-?8\\)?" . utf-8)
2454 ;; utf-8@euro exists, so put this after utf-8. (@euro really
2455 ;; specifies the currency, rather than the charset.)
2456 (".*@euro" . iso-8859-15)
2457 ("koi8-?r" . koi8-r)
2458 ("koi8-?u" . koi8-u)
2460 ("big5[-_]?hkscs" . big5-hkscs)
2462 ("euc-?tw" . euc-tw)
2463 ("euc-?cn" . euc-cn)
2466 ("gb18030" . gb18030)
2467 ("ja.*[._]euc" . japanese-iso-8bit)
2468 ("ja.*[._]jis7" . iso-2022-jp)
2469 ("ja.*[._]pck" . japanese-shift-jis)
2470 ("ja.*[._]sjis" . japanese-shift-jis)
2471 ("jpn" . japanese-shift-jis) ; MS-Windows uses this.
2473 "List of pairs of locale regexps and preferred coding systems.
2474 The first element whose locale regexp matches the start of a downcased locale
2475 specifies the coding system to prefer when using that locale.
2476 This coding system is used if the locale specifies a specific charset.")
2478 (defun locale-name-match (key alist)
2479 "Search for KEY in ALIST, which should be a list of regexp-value pairs.
2480 Return the value corresponding to the first regexp that matches the
2481 start of KEY, or nil if there is no match."
2483 (while (and alist (not element))
2484 (if (string-match-p (concat "\\`\\(?:" (car (car alist)) "\\)") key)
2485 (setq element (car alist)))
2486 (setq alist (cdr alist)))
2489 (defun locale-charset-match-p (charset1 charset2)
2490 "Whether charset names (strings) CHARSET1 and CHARSET2 are equivalent.
2491 Matching is done ignoring case and any hyphens and underscores in the
2492 names. E.g. `ISO_8859-1' and `iso88591' both match `iso-8859-1'."
2493 (setq charset1 (replace-regexp-in-string "[-_]" "" charset1))
2494 (setq charset2 (replace-regexp-in-string "[-_]" "" charset2))
2495 (eq t (compare-strings charset1 nil nil charset2 nil nil t)))
2497 (defvar locale-charset-alist nil
2498 "Coding system alist keyed on locale-style charset name.
2499 Used by `locale-charset-to-coding-system'.")
2501 (defun locale-charset-to-coding-system (charset)
2502 "Find coding system corresponding to CHARSET.
2503 CHARSET is any sort of non-Emacs charset name, such as might be used
2504 in a locale codeset, or elsewhere. It is matched to a coding system
2505 first by case-insensitive lookup in `locale-charset-alist'. Then
2506 matches are looked for in the coding system list, treating case and
2507 the characters `-' and `_' as insignificant. The coding system base
2508 is returned. Thus, for instance, if charset \"ISO8859-2\",
2509 `iso-latin-2' is returned."
2510 (or (car (assoc-string charset locale-charset-alist t))
2511 (let ((cs coding-system-alist)
2513 (while (and (not c) cs)
2514 (if (locale-charset-match-p charset (caar cs))
2515 (setq c (intern (caar cs)))
2517 (if c (coding-system-base c)))))
2519 ;; Fixme: This ought to deal with the territory part of the locale
2520 ;; too, for setting things such as calendar holidays, ps-print paper
2521 ;; size, spelling dictionary.
2523 (declare-function w32-get-console-codepage "w32proc.c" ())
2524 (declare-function w32-get-console-output-codepage "w32proc.c" ())
2526 (defun locale-translate (locale)
2527 "Expand LOCALE according to `locale-translation-file-name', if possible.
2528 For example, translate \"swedish\" into \"sv_SE.ISO8859-1\"."
2529 (if locale-translation-file-name
2531 (set-buffer-multibyte nil)
2532 (insert-file-contents locale-translation-file-name)
2533 (if (re-search-forward
2534 (concat "^" (regexp-quote locale) ":?[ \t]+") nil t)
2535 (buffer-substring (point) (line-end-position))
2539 (defun set-locale-environment (&optional locale-name frame)
2540 "Set up multilingual environment for using LOCALE-NAME.
2541 This sets the language environment, the coding system priority,
2542 the default input method and sometimes other things.
2544 LOCALE-NAME should be a string which is the name of a locale supported
2545 by the system. Often it is of the form xx_XX.CODE, where xx is a
2546 language, XX is a country, and CODE specifies a character set and
2547 coding system. For example, the locale name \"ja_JP.EUC\" might name
2548 a locale for Japanese in Japan using the `japanese-iso-8bit'
2549 coding-system. The name may also have a modifier suffix, e.g. `@euro'
2552 If LOCALE-NAME is nil, its value is taken from the environment
2553 variables LC_ALL, LC_CTYPE and LANG (the first one that is set).
2555 The locale names supported by your system can typically be found in a
2556 directory named `/usr/share/locale' or `/usr/lib/locale'. LOCALE-NAME
2557 will be translated according to the table specified by
2558 `locale-translation-file-name'.
2560 If FRAME is non-nil, only set the keyboard coding system and the
2561 terminal coding system for the terminal of that frame, and don't
2562 touch session-global parameters like the language environment.
2564 See also `locale-charset-language-names', `locale-language-names',
2565 `locale-preferred-coding-systems' and `locale-coding-system'."
2566 (interactive "sSet environment for locale: ")
2568 ;; Do this at runtime for the sake of binaries possibly transported
2569 ;; to a system without X.
2570 (setq locale-translation-file-name
2572 '("/usr/share/X11/locale/locale.alias" ; e.g. X11R7
2573 "/usr/lib/X11/locale/locale.alias" ; e.g. X11R6.4
2574 "/usr/X11R6/lib/X11/locale/locale.alias" ; XFree86, e.g. RedHat 4.2
2575 "/usr/openwin/lib/locale/locale.alias" ; e.g. Solaris 2.6
2577 ;; The following name appears after the X-related names above,
2578 ;; since the X-related names are what X actually uses.
2579 "/usr/share/locale/locale.alias" ; GNU/Linux sans X
2581 (while (and files (not (file-exists-p (car files))))
2582 (setq files (cdr files)))
2585 (let ((locale locale-name))
2588 ;; Use the first of these three environment variables
2589 ;; that has a nonempty value.
2590 (let ((vars '("LC_ALL" "LC_CTYPE" "LANG")))
2592 (= 0 (length locale))) ; nil or empty string
2593 (setq locale (getenv (pop vars) frame)))))
2596 (setq locale (locale-translate locale))
2598 ;; Leave the system locales alone if the caller did not specify
2599 ;; an explicit locale name, as their defaults are set from
2600 ;; LC_MESSAGES and LC_TIME, not LC_CTYPE, and the user might not
2601 ;; want to set them to the same value as LC_CTYPE.
2603 (setq system-messages-locale locale)
2604 (setq system-time-locale locale))
2606 (if (string-match "^[a-z][a-z]" locale)
2607 ;; The value of 'current-iso639-language' is matched against
2608 ;; the ':lang' property of font-spec objects when selecting
2609 ;; and prioritizing available fonts for displaying
2610 ;; characters; see fontset.c.
2611 (setq current-iso639-language
2612 ;; The call to 'downcase' is for w32, where the
2613 ;; MS-Windows locale names are in caps, as in "ENU",
2614 ;; the equivalent of the Posix "en_US". Since the
2615 ;; match mentioned above uses memq, and ':lang'
2616 ;; properties have lower-case values, the letter-case
2617 ;; must match exactly.
2618 (intern (downcase (match-string 0 locale))))))
2621 (or system-messages-locale
2622 (let ((msglocale (getenv "LC_MESSAGES" frame)))
2623 (if (zerop (length msglocale))
2625 (locale-translate msglocale)))))
2628 (setq locale (downcase locale))
2630 (let ((language-name
2631 (locale-name-match locale locale-language-names))
2632 (charset-language-name
2633 (locale-name-match locale locale-charset-language-names))
2634 (default-eol-type (coding-system-eol-type
2635 (default-value 'buffer-file-coding-system)))
2637 (or (locale-name-match locale locale-preferred-coding-systems)
2639 (if (string-match "\\.\\([^@]+\\)" locale)
2640 (locale-charset-to-coding-system
2641 (match-string 1 locale)))))))
2643 (if (consp language-name)
2644 ;; locale-language-names specify both lang-env and coding.
2645 ;; But, what specified in locale-preferred-coding-systems
2646 ;; has higher priority.
2647 (setq coding-system (or coding-system
2648 (nth 1 language-name))
2649 language-name (car language-name))
2650 ;; Otherwise, if locale is not listed in locale-language-names,
2651 ;; use what listed in locale-charset-language-names.
2652 (if (not language-name)
2653 (setq language-name charset-language-name)))
2655 ;; If a specific EOL conversion was specified in the default
2656 ;; buffer-file-coding-system, preserve it in the coding system
2657 ;; we will be using from now on.
2658 (if (and (memq default-eol-type '(0 1 2 unix dos mac))
2660 (coding-system-p coding-system))
2661 (setq coding-system (coding-system-change-eol-conversion
2662 coding-system default-eol-type)))
2666 ;; Set up for this character set. This is now the right way
2667 ;; to do it for both unibyte and multibyte modes.
2669 (set-language-environment language-name))
2671 ;; If the default enable-multibyte-characters is nil,
2672 ;; we are using single-byte characters,
2673 ;; so the display table and terminal coding system are irrelevant.
2674 (when (default-value 'enable-multibyte-characters)
2675 (set-display-table-and-terminal-coding-system
2676 language-name coding-system frame))
2678 ;; Set the `keyboard-coding-system' if appropriate (tty
2679 ;; only). At least X and MS Windows can generate
2680 ;; multilingual input.
2681 ;; XXX This was disabled unless `window-system', but that
2682 ;; leads to buggy behavior when a tty frame is opened
2683 ;; later. Setting the keyboard coding system has no adverse
2684 ;; effect on X, so let's do it anyway. -- Lorentey
2685 (let ((kcs (or coding-system
2686 (car (get-language-info language-name
2688 (if kcs (set-keyboard-coding-system kcs frame)))
2691 (setq locale-coding-system
2692 (car (get-language-info language-name 'coding-priority)))))
2694 (when (and (not frame)
2696 (not (coding-system-equal coding-system
2697 locale-coding-system)))
2698 (prefer-coding-system coding-system)
2699 ;; Fixme: perhaps prefer-coding-system should set this too.
2700 ;; But it's not the time to do such a fundamental change.
2701 (setq default-sendmail-coding-system coding-system)
2702 (setq locale-coding-system coding-system))))
2704 ;; On Windows, override locale-coding-system,
2705 ;; default-file-name-coding-system, keyboard-coding-system,
2706 ;; terminal-coding-system with the ANSI or console codepage.
2707 (when (and (eq system-type 'windows-nt)
2708 (boundp 'w32-ansi-code-page))
2709 (let* ((code-page-coding
2710 (intern (format "cp%d" (if noninteractive
2711 (w32-get-console-codepage)
2712 w32-ansi-code-page))))
2715 (intern (format "cp%d" (w32-get-console-output-codepage)))
2717 (when (coding-system-p code-page-coding)
2718 (or output-coding (setq output-coding code-page-coding))
2719 (unless frame (setq locale-coding-system code-page-coding))
2720 (set-keyboard-coding-system code-page-coding frame)
2721 (set-terminal-coding-system output-coding frame)
2722 (setq default-file-name-coding-system code-page-coding))))
2724 (when (eq system-type 'darwin)
2725 ;; On Darwin, file names are always encoded in utf-8, no matter
2727 (setq default-file-name-coding-system 'utf-8)
2728 ;; Mac OS X's Terminal.app by default uses utf-8 regardless of
2730 (when (and (null window-system)
2731 (equal (getenv "TERM_PROGRAM" frame) "Apple_Terminal"))
2732 (set-terminal-coding-system 'utf-8)
2733 (set-keyboard-coding-system 'utf-8)))
2735 ;; Default to A4 paper if we're not in a C, POSIX or US locale.
2736 ;; (See comments in Flocale_info.)
2738 (let ((locale locale)
2739 (paper (locale-info 'paper)))
2741 ;; This will always be null at the time of writing.
2743 ((equal paper '(216 279))
2744 (setq ps-paper-type 'letter))
2745 ((equal paper '(210 297))
2746 (setq ps-paper-type 'a4)))
2747 (let ((vars '("LC_ALL" "LC_PAPER" "LANG")))
2748 (while (and vars (= 0 (length locale)))
2749 (setq locale (getenv (pop vars) frame))))
2751 ;; As of glibc 2.2.5, these are the only US Letter locales,
2752 ;; and the rest are A4.
2754 (or (locale-name-match locale '(("c$" . letter)
2759 ("enu$" . letter) ; Windows
2766 ;;; Character property
2768 (put 'char-code-property-table 'char-table-extra-slots 5)
2770 (defun define-char-code-property (name table &optional docstring)
2771 "Define NAME as a character code property given by TABLE.
2772 TABLE is a char-table of purpose `char-code-property-table' with
2775 2nd: Function to call to get a property value of a character.
2776 It is called with three arguments CHAR, VAL, and TABLE, where
2777 CHAR is a character, VAL is the value of (aref TABLE CHAR).
2778 3rd: Function to call to put a property value of a character.
2779 It is called with the same arguments as above.
2780 4th: Function to call to get a description string of a property value.
2781 It is called with one argument VALUE, a property value.
2782 5th: Data used by the above functions.
2784 TABLE may be a name of file to load to build a char-table. The
2785 file should contain a call of `define-char-code-property' with a
2786 char-table of the above format as the argument TABLE.
2788 TABLE may also be nil, in which case no property value is pre-assigned.
2790 Optional 3rd argument DOCSTRING is a documentation string of the property.
2792 See also the documentation of `get-char-code-property' and
2793 `put-char-code-property'."
2795 (error "Not a symbol: %s" name))
2796 (if (char-table-p table)
2797 (or (and (eq (char-table-subtype table) 'char-code-property-table)
2798 (eq (char-table-extra-slot table 0) name))
2799 (error "Invalid char-table: %s" table))
2801 (error "Not a char-table nor a file name: %s" table)))
2802 (if (stringp table) (setq table (purecopy table)))
2803 (setf (alist-get name char-code-property-alist) table)
2804 (put name 'char-code-property-documentation (purecopy docstring)))
2806 (defvar char-code-property-table
2807 (make-char-table 'char-code-property-table)
2808 "Char-table containing a property list of each character code.
2809 This table is used for properties not listed in `char-code-property-alist'.
2810 See also the documentation of `get-char-code-property' and
2811 `put-char-code-property'.")
2813 (defun get-char-code-property (char propname)
2814 "Return the value of CHAR's PROPNAME property."
2815 (let ((table (unicode-property-table-internal propname)))
2817 (let ((func (char-table-extra-slot table 1)))
2818 (if (functionp func)
2819 (funcall func char (aref table char) table)
2820 (get-unicode-property-internal table char)))
2821 (plist-get (aref char-code-property-table char) propname))))
2823 (defun put-char-code-property (char propname value)
2824 "Store CHAR's PROPNAME property with VALUE.
2825 It can be retrieved with `(get-char-code-property CHAR PROPNAME)'."
2826 (let ((table (unicode-property-table-internal propname)))
2828 (let ((func (char-table-extra-slot table 2)))
2829 (if (functionp func)
2830 (funcall func char value table)
2831 (put-unicode-property-internal table char value)))
2832 (let* ((plist (aref char-code-property-table char))
2833 (x (plist-put plist propname value)))
2835 (aset char-code-property-table char x))))
2838 (defun char-code-property-description (prop value)
2839 "Return a description string of character property PROP's value VALUE.
2840 If there's no description string for VALUE, return nil."
2841 (let ((table (unicode-property-table-internal prop)))
2843 (let ((func (char-table-extra-slot table 3)))
2844 (if (functionp func)
2845 (funcall func value))))))
2848 ;; Pretty description of encoded string
2850 ;; Alist of ISO 2022 control code vs the corresponding mnemonic string.
2851 (defconst iso-2022-control-alist
2859 (defun encoded-string-description (str coding-system)
2860 "Return a pretty description of STR that is encoded by CODING-SYSTEM."
2861 (setq str (string-as-unibyte str))
2863 (if (and coding-system (eq (coding-system-type coding-system) 'iso-2022))
2864 ;; Try to get a pretty description for ISO 2022 escape sequences.
2865 (function (lambda (x) (or (cdr (assq x iso-2022-control-alist))
2866 (format "#x%02X" x))))
2867 (function (lambda (x) (format "#x%02X" x))))
2870 (defun encode-coding-char (char coding-system &optional charset)
2871 "Encode CHAR by CODING-SYSTEM and return the resulting string.
2872 If CODING-SYSTEM can't safely encode CHAR, return nil.
2873 The 3rd optional argument CHARSET, if non-nil, is a charset preferred
2875 (let* ((str1 (string-as-multibyte (string char)))
2876 (str2 (string-as-multibyte (string char char)))
2877 (found (find-coding-systems-string str1))
2879 (if (and (consp found)
2880 (eq (car found) 'undecided))
2882 (when (memq (coding-system-base coding-system) found)
2883 ;; We must find the encoded string of CHAR. But, just encoding
2884 ;; CHAR will put extra control sequences (usually to designate
2885 ;; ASCII charset) at the tail if type of CODING is ISO 2022.
2886 ;; To exclude such tailing bytes, we at first encode one-char
2887 ;; string and two-char string, then check how many bytes at the
2888 ;; tail of both encoded strings are the same.
2891 (put-text-property 0 1 'charset charset str1)
2892 (put-text-property 0 2 'charset charset str2))
2893 (setq enc1 (encode-coding-string str1 coding-system)
2895 enc2 (encode-coding-string str2 coding-system)
2897 (while (and (> i1 0) (= (aref enc1 (1- i1)) (aref enc2 (1- i2))))
2898 (setq i1 (1- i1) i2 (1- i2)))
2900 ;; Now (substring enc1 i1) and (substring enc2 i2) are the same,
2901 ;; and they are the extra control sequences at the tail to
2903 (substring enc2 0 i2)))))
2905 ;; Backwards compatibility. These might be better with :init-value t,
2906 ;; but that breaks loadup.
2907 (define-minor-mode unify-8859-on-encoding-mode
2908 "Exists only for backwards compatibility."
2911 ;; Doc said "obsolete" in 23.1, this statement only added in 24.1.
2912 (make-obsolete 'unify-8859-on-encoding-mode "don't use it." "23.1")
2914 (define-minor-mode unify-8859-on-decoding-mode
2915 "Exists only for backwards compatibility."
2918 ;; Doc said "obsolete" in 23.1, this statement only added in 24.1.
2919 (make-obsolete 'unify-8859-on-decoding-mode "don't use it." "23.1")
2921 (defvar nonascii-insert-offset 0)
2922 (make-obsolete-variable 'nonascii-insert-offset "do not use it." "23.1")
2923 (defvar nonascii-translation-table nil)
2924 (make-obsolete-variable 'nonascii-translation-table "do not use it." "23.1")
2926 (defvar ucs-names nil
2927 "Alist of cached (CHAR-NAME . CHAR-CODE) pairs.")
2930 "Return alist of (CHAR-NAME . CHAR-CODE) pairs cached in `ucs-names'."
2934 ;; (#x3400 . #x4DBF) CJK Ideographs Extension A
2936 ;; (#x4E00 . #x9FFF) CJK Unified Ideographs
2938 ;; (#xD800 . #xFAFF) Surrogate/Private
2940 ;; (#x13500 . #x167FF) unused
2942 ;; (#x16A40 . #x1AFFF) unused
2944 ;; (#x1B100 . #x1CFFF) unused
2946 ;; (#x20000 . #xDFFFF) CJK Ideograph Extension A, B, etc, unused
2947 (#xE0000 . #xE01FF)))
2948 (gc-cons-threshold 10000000)
2950 (dolist (range ranges)
2951 (let ((c (car range))
2954 (let ((new-name (get-char-code-property c 'name))
2955 (old-name (get-char-code-property c 'old-name)))
2956 ;; In theory this code could end up pushing an "old-name" that
2957 ;; shadows a "new-name" but in practice every time an
2958 ;; `old-name' conflicts with a `new-name', the newer one has a
2959 ;; higher code, so it gets pushed later!
2960 (if new-name (push (cons new-name c) names))
2961 (if old-name (push (cons old-name c) names))
2963 ;; Special case for "BELL" which is apparently the only char which
2964 ;; doesn't have a new name and whose old-name is shadowed by a newer
2965 ;; char with that name.
2966 (setq ucs-names `(("BELL (BEL)" . 7) ,@names)))))
2968 (defun mule--ucs-names-annotation (name)
2969 ;; FIXME: It would be much better to add this annotation before rather than
2970 ;; after the char name, so the annotations are aligned.
2971 ;; FIXME: The default behavior of displaying annotations in italics
2972 ;; doesn't work well here.
2973 (let ((char (assoc name ucs-names)))
2974 (when char (format " (%c)" (cdr char)))))
2976 (defun read-char-by-name (prompt)
2977 "Read a character by its Unicode name or hex number string.
2978 Display PROMPT and read a string that represents a character by its
2979 Unicode property `name' or `old-name'.
2981 This function returns the character as a number.
2983 You can type a few of the first letters of the Unicode name and
2984 use completion. If you type a substring of the Unicode name
2985 preceded by an asterisk `*' and use completion, it will show all
2986 the characters whose names include that substring, not necessarily
2987 at the beginning of the name.
2989 This function also accepts a hexadecimal number of Unicode code
2990 point or a number in hash notation, e.g. #o21430 for octal,
2991 #x2318 for hex, or #10r8984 for decimal."
2992 (let* ((enable-recursive-minibuffers t)
2993 (completion-ignore-case t)
2997 (lambda (string pred action)
2998 (if (eq action 'metadata)
3000 (annotation-function . mule--ucs-names-annotation)
3001 (category . unicode-name))
3002 (complete-with-action action (ucs-names) string pred)))))
3005 ((string-match-p "\\`[0-9a-fA-F]+\\'" input)
3006 (string-to-number input 16))
3007 ((string-match-p "\\`#" input)
3010 (cdr (assoc-string input (ucs-names) t))))))
3011 (unless (characterp char)
3012 (error "Invalid character"))
3015 (define-obsolete-function-alias 'ucs-insert 'insert-char "24.3")
3016 (define-key ctl-x-map "8\r" 'insert-char)
3018 ;;; mule-cmds.el ends here