Store client's environment in terminal parameters, not server parameters.
[bpt/emacs.git] / lisp / international / mule-cmds.el
1 ;;; mule-cmds.el --- commands for mulitilingual environment -*-coding: iso-2022-7bit -*-
2
3 ;; Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005
4 ;; Free Software Foundation, Inc.
5 ;; Copyright (C) 1995, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005
6 ;; National Institute of Advanced Industrial Science and Technology (AIST)
7 ;; Registration Number H14PRO021
8
9 ;; Keywords: mule, multilingual
10
11 ;; This file is part of GNU Emacs.
12
13 ;; GNU Emacs is free software; you can redistribute it and/or modify
14 ;; it under the terms of the GNU General Public License as published by
15 ;; the Free Software Foundation; either version 2, or (at your option)
16 ;; any later version.
17
18 ;; GNU Emacs is distributed in the hope that it will be useful,
19 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
20 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
21 ;; GNU General Public License for more details.
22
23 ;; You should have received a copy of the GNU General Public License
24 ;; along with GNU Emacs; see the file COPYING. If not, write to the
25 ;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
26 ;; Boston, MA 02110-1301, USA.
27
28 ;;; Commentary:
29
30 ;;; Code:
31
32 (eval-when-compile
33 (defvar dos-codepage)
34 (autoload 'widget-value "wid-edit"))
35
36 (defvar mac-system-coding-system)
37 (defvar mac-system-locale)
38
39 ;;; MULE related key bindings and menus.
40
41 (defvar mule-keymap (make-sparse-keymap)
42 "Keymap for Mule (Multilingual environment) specific commands.")
43
44 ;; Keep "C-x C-m ..." for mule specific commands.
45 (define-key ctl-x-map "\C-m" mule-keymap)
46
47 (define-key mule-keymap "f" 'set-buffer-file-coding-system)
48 (define-key mule-keymap "r" 'revert-buffer-with-coding-system)
49 (define-key mule-keymap "F" 'set-file-name-coding-system)
50 (define-key mule-keymap "t" 'set-terminal-coding-system)
51 (define-key mule-keymap "k" 'set-keyboard-coding-system)
52 (define-key mule-keymap "p" 'set-buffer-process-coding-system)
53 (define-key mule-keymap "x" 'set-selection-coding-system)
54 (define-key mule-keymap "X" 'set-next-selection-coding-system)
55 (define-key mule-keymap "\C-\\" 'set-input-method)
56 (define-key mule-keymap "c" 'universal-coding-system-argument)
57 (define-key mule-keymap "l" 'set-language-environment)
58
59 (defvar mule-menu-keymap
60 (make-sparse-keymap "Mule (Multilingual Environment)")
61 "Keymap for Mule (Multilingual environment) menu specific commands.")
62
63 (defvar describe-language-environment-map
64 (make-sparse-keymap "Describe Language Environment"))
65
66 (defvar setup-language-environment-map
67 (make-sparse-keymap "Set Language Environment"))
68
69 (defvar set-coding-system-map
70 (make-sparse-keymap "Set Coding System"))
71
72 (define-key-after mule-menu-keymap [set-language-environment]
73 (list 'menu-item "Set Language Environment" setup-language-environment-map
74 :help "Multilingual environment suitable for a specific language"))
75 (define-key-after mule-menu-keymap [separator-mule]
76 '("--")
77 t)
78 (define-key-after mule-menu-keymap [toggle-input-method]
79 '(menu-item "Toggle Input Method" toggle-input-method)
80 t)
81 (define-key-after mule-menu-keymap [set-input-method]
82 '(menu-item "Select Input Method..." set-input-method)
83 t)
84 (define-key-after mule-menu-keymap [describe-input-method]
85 '(menu-item "Describe Input Method" describe-input-method))
86 (define-key-after mule-menu-keymap [separator-input-method]
87 '("--")
88 t)
89 (define-key-after mule-menu-keymap [set-various-coding-system]
90 (list 'menu-item "Set Coding Systems" set-coding-system-map
91 :enable 'default-enable-multibyte-characters))
92 (define-key-after mule-menu-keymap [view-hello-file]
93 '(menu-item "Show Multi-lingual Text" view-hello-file
94 :enable (file-readable-p
95 (expand-file-name "HELLO" data-directory))
96 :help "Display file which says HELLO in many languages")
97 t)
98 (define-key-after mule-menu-keymap [separator-coding-system]
99 '("--")
100 t)
101 (define-key-after mule-menu-keymap [describe-language-environment]
102 (list 'menu-item "Describe Language Environment"
103 describe-language-environment-map
104 :help "Show multilingual settings for a specific language")
105 t)
106 (define-key-after mule-menu-keymap [describe-input-method]
107 '(menu-item "Describe Input Method..." describe-input-method
108 :help "Keyboard layout for a specific input method")
109 t)
110 (define-key-after mule-menu-keymap [describe-coding-system]
111 '(menu-item "Describe Coding System..." describe-coding-system)
112 t)
113 (define-key-after mule-menu-keymap [list-character-sets]
114 '(menu-item "List Character Sets" list-character-sets
115 :help "Show table of available character sets"))
116 (define-key-after mule-menu-keymap [mule-diag]
117 '(menu-item "Show All of Mule Status" mule-diag
118 :help "Display multilingual environment settings")
119 t)
120
121 (define-key-after set-coding-system-map [universal-coding-system-argument]
122 '(menu-item "For Next Command" universal-coding-system-argument
123 :help "Coding system to be used by next command")
124 t)
125 (define-key-after set-coding-system-map [separator-1]
126 '("--")
127 t)
128 (define-key-after set-coding-system-map [set-buffer-file-coding-system]
129 '(menu-item "For Saving This Buffer" set-buffer-file-coding-system
130 :help "How to encode this buffer when saved")
131 t)
132 (define-key-after set-coding-system-map [revert-buffer-with-coding-system]
133 '(menu-item "For Reverting This File Now" revert-buffer-with-coding-system
134 :enable buffer-file-name
135 :help "Revisit this file immediately using specified coding system")
136 t)
137 (define-key-after set-coding-system-map [set-file-name-coding-system]
138 '(menu-item "For File Name" set-file-name-coding-system
139 :help "How to decode/encode file names")
140 t)
141 (define-key-after set-coding-system-map [separator-2]
142 '("--")
143 t)
144
145 (define-key-after set-coding-system-map [set-keyboard-coding-system]
146 '(menu-item "For Keyboard" set-keyboard-coding-system
147 :help "How to decode keyboard input")
148 t)
149 (define-key-after set-coding-system-map [set-terminal-coding-system]
150 '(menu-item "For Terminal" set-terminal-coding-system
151 :enable (null (memq initial-window-system '(x w32 mac)))
152 :help "How to encode terminal output")
153 t)
154 (define-key-after set-coding-system-map [separator-3]
155 '("--")
156 t)
157 (define-key-after set-coding-system-map [set-selection-coding-system]
158 '(menu-item "For X Selections/Clipboard" set-selection-coding-system
159 :visible (display-selections-p)
160 :help "How to en/decode data to/from selection/clipboard")
161 t)
162 (define-key-after set-coding-system-map [set-next-selection-coding-system]
163 '(menu-item "For Next X Selection" set-next-selection-coding-system
164 :visible (display-selections-p)
165 :help "How to en/decode next selection/clipboard operation")
166 t)
167 (define-key-after set-coding-system-map [set-buffer-process-coding-system]
168 '(menu-item "For I/O with Subprocess" set-buffer-process-coding-system
169 :visible (fboundp 'start-process)
170 :enable (get-buffer-process (current-buffer))
171 :help "How to en/decode I/O from/to subprocess connected to this buffer")
172 t)
173
174
175 (define-key setup-language-environment-map
176 [Default] '(menu-item "Default" setup-specified-language-environment))
177
178 (define-key describe-language-environment-map
179 [Default] '(menu-item "Default" describe-specified-language-support))
180
181 ;; This should be a single character key binding because users use it
182 ;; very frequently while editing multilingual text. Now we can use
183 ;; only two such keys: "\C-\\" and "\C-^", but the latter is not
184 ;; convenient because it requires shifting on most keyboards. An
185 ;; alternative is "\C-\]" which is now bound to `abort-recursive-edit'
186 ;; but it won't be used that frequently.
187 (define-key global-map "\C-\\" 'toggle-input-method)
188
189 ;; This is no good because people often type Shift-SPC
190 ;; meaning to type SPC. -- rms.
191 ;; ;; Here's an alternative key binding for X users (Shift-SPACE).
192 ;; (define-key global-map [?\S- ] 'toggle-input-method)
193
194 ;;; Mule related hyperlinks.
195 (defconst help-xref-mule-regexp-template
196 (purecopy (concat "\\(\\<\\("
197 "\\(coding system\\)\\|"
198 "\\(input method\\)\\|"
199 "\\(character set\\)\\|"
200 "\\(charset\\)"
201 "\\)\\s-+\\)?"
202 ;; Note starting with word-syntax character:
203 "`\\(\\sw\\(\\sw\\|\\s_\\)+\\)'")))
204
205 (defun coding-system-change-eol-conversion (coding-system eol-type)
206 "Return a coding system which differs from CODING-SYSTEM in eol conversion.
207 The returned coding system converts end-of-line by EOL-TYPE
208 but text as the same way as CODING-SYSTEM.
209 EOL-TYPE should be `unix', `dos', `mac', or nil.
210 If EOL-TYPE is nil, the returned coding system detects
211 how end-of-line is formatted automatically while decoding.
212
213 EOL-TYPE can be specified by an integer 0, 1, or 2.
214 They means `unix', `dos', and `mac' respectively."
215 (if (symbolp eol-type)
216 (setq eol-type (cond ((eq eol-type 'unix) 0)
217 ((eq eol-type 'dos) 1)
218 ((eq eol-type 'mac) 2)
219 (t eol-type))))
220 (let ((orig-eol-type (coding-system-eol-type coding-system)))
221 (if (vectorp orig-eol-type)
222 (if (not eol-type)
223 coding-system
224 (aref orig-eol-type eol-type))
225 (let ((base (coding-system-base coding-system)))
226 (if (not eol-type)
227 base
228 (if (= eol-type orig-eol-type)
229 coding-system
230 (setq orig-eol-type (coding-system-eol-type base))
231 (if (vectorp orig-eol-type)
232 (aref orig-eol-type eol-type))))))))
233
234 (defun coding-system-change-text-conversion (coding-system coding)
235 "Return a coding system which differs from CODING-SYSTEM in text conversion.
236 The returned coding system converts text by CODING
237 but end-of-line as the same way as CODING-SYSTEM.
238 If CODING is nil, the returned coding system detects
239 how text is formatted automatically while decoding."
240 (let ((eol-type (coding-system-eol-type coding-system)))
241 (coding-system-change-eol-conversion
242 (if coding coding 'undecided)
243 (if (numberp eol-type) (aref [unix dos mac] eol-type)))))
244
245 (defun toggle-enable-multibyte-characters (&optional arg)
246 "Change whether this buffer uses multibyte characters.
247 With arg, use multibyte characters if the arg is positive.
248
249 Note that this command does not convert the byte contents of
250 the buffer; it only changes the way those bytes are interpreted.
251 In general, therefore, this command *changes* the sequence of
252 characters that the current buffer contains.
253
254 We suggest you avoid using use this command unless you know what you
255 are doing. If you use it by mistake, and the buffer is now displayed
256 wrong, use this command again to toggle back to the right mode."
257 (interactive "P")
258 (let ((new-flag
259 (if (null arg) (null enable-multibyte-characters)
260 (> (prefix-numeric-value arg) 0))))
261 (set-buffer-multibyte new-flag))
262 (force-mode-line-update))
263
264 (defun view-hello-file ()
265 "Display the HELLO file which list up many languages and characters."
266 (interactive)
267 ;; We have to decode the file in any environment.
268 (let ((default-enable-multibyte-characters t)
269 (coding-system-for-read 'iso-2022-7bit))
270 (view-file (expand-file-name "HELLO" data-directory))))
271
272 (defun universal-coding-system-argument (coding-system)
273 "Execute an I/O command using the specified coding system."
274 (interactive
275 (let ((default (and buffer-file-coding-system
276 (not (eq (coding-system-type buffer-file-coding-system)
277 t))
278 buffer-file-coding-system)))
279 (list (read-coding-system
280 (if default
281 (format "Coding system for following command (default %s): " default)
282 "Coding system for following command: ")
283 default))))
284 (let* ((keyseq (read-key-sequence
285 (format "Command to execute with %s:" coding-system)))
286 (cmd (key-binding keyseq))
287 prefix)
288
289 (when (eq cmd 'universal-argument)
290 (call-interactively cmd)
291
292 ;; Process keys bound in `universal-argument-map'.
293 (while (progn
294 (setq keyseq (read-key-sequence nil t)
295 cmd (key-binding keyseq t))
296 (not (eq cmd 'universal-argument-other-key)))
297 (let ((current-prefix-arg prefix-arg)
298 ;; Have to bind `last-command-char' here so that
299 ;; `digit-argument', for instance, can compute the
300 ;; prefix arg.
301 (last-command-char (aref keyseq 0)))
302 (call-interactively cmd)))
303
304 ;; This is the final call to `universal-argument-other-key', which
305 ;; set's the final `prefix-arg.
306 (let ((current-prefix-arg prefix-arg))
307 (call-interactively cmd))
308
309 ;; Read the command to execute with the given prefix arg.
310 (setq prefix prefix-arg
311 keyseq (read-key-sequence nil t)
312 cmd (key-binding keyseq)))
313
314 (let ((coding-system-for-read coding-system)
315 (coding-system-for-write coding-system)
316 (coding-system-require-warning t)
317 (current-prefix-arg prefix))
318 (message "")
319 (call-interactively cmd))))
320
321 (defun set-default-coding-systems (coding-system)
322 "Set default value of various coding systems to CODING-SYSTEM.
323 This sets the following coding systems:
324 o coding system of a newly created buffer
325 o default coding system for subprocess I/O
326 This also sets the following values:
327 o default value used as `file-name-coding-system' for converting file names
328 if CODING-SYSTEM is ASCII-compatible.
329 o default value for the command `set-terminal-coding-system' (not on MSDOS)
330 o default value for the command `set-keyboard-coding-system'
331 if CODING-SYSTEM is ASCII-compatible.."
332 (check-coding-system coding-system)
333 (setq-default buffer-file-coding-system coding-system)
334 (if (fboundp 'ucs-set-table-for-input)
335 (dolist (buffer (buffer-list))
336 (or (local-variable-p 'buffer-file-coding-system buffer)
337 (ucs-set-table-for-input buffer))))
338
339 (if (and default-enable-multibyte-characters (not (eq system-type 'darwin))
340 (or (not coding-system)
341 (not (coding-system-get coding-system 'ascii-incompatible))))
342 ;; The file-name coding system on Darwin systems is always utf-8.
343 (setq default-file-name-coding-system coding-system))
344 ;; If coding-system is nil, honor that on MS-DOS as well, so
345 ;; that they could reset the terminal coding system.
346 (unless (and (eq window-system 'pc) coding-system)
347 (setq default-terminal-coding-system coding-system))
348 (if (or (not coding-system)
349 (not (coding-system-get coding-system 'ascii-incompatible)))
350 (setq default-keyboard-coding-system coding-system))
351 ;; Preserve eol-type from existing default-process-coding-systems.
352 ;; On non-unix-like systems in particular, these may have been set
353 ;; carefully by the user, or by the startup code, to deal with the
354 ;; users shell appropriately, so should not be altered by changing
355 ;; language environment.
356 (let ((output-coding
357 (coding-system-change-text-conversion
358 (car default-process-coding-system) coding-system))
359 (input-coding
360 (coding-system-change-text-conversion
361 (cdr default-process-coding-system) coding-system)))
362 (setq default-process-coding-system
363 (cons output-coding input-coding))))
364
365 (defun prefer-coding-system (coding-system)
366 "Add CODING-SYSTEM at the front of the priority list for automatic detection.
367 This also sets the following coding systems:
368 o coding system of a newly created buffer
369 o default coding system for subprocess I/O
370 This also sets the following values:
371 o default value used as `file-name-coding-system' for converting file names.
372 o default value for the command `set-terminal-coding-system' (not on MSDOS)
373 o default value for the command `set-keyboard-coding-system'
374
375 If CODING-SYSTEM specifies a certain type of EOL conversion, the coding
376 systems set by this function will use that type of EOL conversion.
377
378 This command does not change the default value of terminal coding system
379 for MS-DOS terminal, because DOS terminals only support a single coding
380 system, and Emacs automatically sets the default to that coding system at
381 startup.
382
383 A coding system that requires automatic detection of text
384 encoding (e.g. undecided, unix) can't be preferred.
385
386 See also `coding-category-list' and `coding-system-category'."
387 (interactive "zPrefer coding system: ")
388 (if (not (and coding-system (coding-system-p coding-system)))
389 (error "Invalid coding system `%s'" coding-system))
390 (let ((coding-category (coding-system-category coding-system))
391 (base (coding-system-base coding-system))
392 (eol-type (coding-system-eol-type coding-system)))
393 (if (not coding-category)
394 ;; CODING-SYSTEM is no-conversion or undecided.
395 (error "Can't prefer the coding system `%s'" coding-system))
396 (set coding-category (or base coding-system))
397 ;; Changing the binding of a coding category requires this call.
398 (update-coding-systems-internal)
399 (or (eq coding-category (car coding-category-list))
400 ;; We must change the order.
401 (set-coding-priority (list coding-category)))
402 (if (and base (interactive-p))
403 (message "Highest priority is set to %s (base of %s)"
404 base coding-system))
405 ;; If they asked for specific EOL conversion, honor that.
406 (if (memq eol-type '(0 1 2))
407 (setq coding-system
408 (coding-system-change-eol-conversion base eol-type))
409 (setq coding-system base))
410 (set-default-coding-systems coding-system)))
411
412 (defvar sort-coding-systems-predicate nil
413 "If non-nil, a predicate function to sort coding systems.
414
415 It is called with two coding systems, and should return t if the first
416 one is \"less\" than the second.
417
418 The function `sort-coding-systems' use it.")
419
420 (defun sort-coding-systems (codings)
421 "Sort coding system list CODINGS by a priority of each coding system.
422
423 If a coding system is most preferred, it has the highest priority.
424 Otherwise, a coding system corresponds to some MIME charset has higher
425 priorities. Among them, a coding system included in `coding-system'
426 key of the current language environment has higher priorities. See
427 also the documentation of `language-info-alist'.
428
429 If the variable `sort-coding-systems-predicate' (which see) is
430 non-nil, it is used to sort CODINGS in the different way than above."
431 (if sort-coding-systems-predicate
432 (sort codings sort-coding-systems-predicate)
433 (let* ((from-categories (mapcar #'(lambda (x) (symbol-value x))
434 coding-category-list))
435 (most-preferred (car from-categories))
436 (lang-preferred (get-language-info current-language-environment
437 'coding-system))
438 (func (function
439 (lambda (x)
440 (let ((base (coding-system-base x)))
441 ;; We calculate the priority number 0..255 by
442 ;; using the 8 bits PMMLCEII as this:
443 ;; P: 1 iff most preferred.
444 ;; MM: greater than 0 iff mime-charset.
445 ;; L: 1 iff one of the current lang. env.'s codings.
446 ;; C: 1 iff one of codings listed in the category list.
447 ;; E: 1 iff not XXX-with-esc
448 ;; II: if iso-2022 based, 0..3, else 1.
449 (logior
450 (lsh (if (eq base most-preferred) 1 0) 7)
451 (lsh
452 (let ((mime (coding-system-get base 'mime-charset)))
453 ;; Prefer coding systems corresponding to a
454 ;; MIME charset.
455 (if mime
456 ;; Lower utf-16 priority so that we
457 ;; normally prefer utf-8 to it, and put
458 ;; x-ctext below that.
459 (cond ((string-match "utf-16"
460 (symbol-name mime))
461 2)
462 ((string-match "^x-" (symbol-name mime))
463 1)
464 (t 3))
465 0))
466 5)
467 (lsh (if (memq base lang-preferred) 1 0) 4)
468 (lsh (if (memq base from-categories) 1 0) 3)
469 (lsh (if (string-match "-with-esc\\'"
470 (symbol-name base))
471 0 1) 2)
472 (if (eq (coding-system-type base) 2)
473 ;; For ISO based coding systems, prefer
474 ;; one that doesn't use escape sequences.
475 (let ((flags (coding-system-flags base)))
476 (if (or (consp (aref flags 0))
477 (consp (aref flags 1))
478 (consp (aref flags 2))
479 (consp (aref flags 3)))
480 (if (or (aref flags 8) (aref flags 9))
481 0
482 1)
483 2))
484 1)))))))
485 (sort codings (function (lambda (x y)
486 (> (funcall func x) (funcall func y))))))))
487
488 (defun find-coding-systems-region (from to)
489 "Return a list of proper coding systems to encode a text between FROM and TO.
490 If FROM is a string, find coding systems in that instead of the buffer.
491 All coding systems in the list can safely encode any multibyte characters
492 in the text.
493
494 If the text contains no multibyte characters, return a list of a single
495 element `undecided'."
496 (let ((codings (find-coding-systems-region-internal from to)))
497 (if (eq codings t)
498 ;; The text contains only ASCII characters. Any coding
499 ;; systems are safe.
500 '(undecided)
501 ;; We need copy-sequence because sorting will alter the argument.
502 (sort-coding-systems (copy-sequence codings)))))
503
504 (defun find-coding-systems-string (string)
505 "Return a list of proper coding systems to encode STRING.
506 All coding systems in the list can safely encode any multibyte characters
507 in STRING.
508
509 If STRING contains no multibyte characters, return a list of a single
510 element `undecided'."
511 (find-coding-systems-region string nil))
512
513 (defun find-coding-systems-for-charsets (charsets)
514 "Return a list of proper coding systems to encode characters of CHARSETS.
515 CHARSETS is a list of character sets.
516 It actually checks at most the first 96 characters of each charset.
517 So, if a charset of dimension two is included in CHARSETS, the value may
518 contain a coding system that can't encode all characters of the charset."
519 (cond ((or (null charsets)
520 (and (= (length charsets) 1)
521 (eq 'ascii (car charsets))))
522 '(undecided))
523 ((or (memq 'eight-bit-control charsets)
524 (memq 'eight-bit-graphic charsets))
525 '(raw-text emacs-mule))
526 (t
527 (let ((codings t)
528 charset l str)
529 (while (and codings charsets)
530 (setq charset (car charsets) charsets (cdr charsets))
531 (unless (eq charset 'ascii)
532 (setq str (make-string 96 32))
533 (if (= (charset-dimension charset) 1)
534 (if (= (charset-chars charset) 96)
535 (dotimes (i 96)
536 (aset str i (make-char charset (+ i 32))))
537 (dotimes (i 94)
538 (aset str i (make-char charset (+ i 33)))))
539 (if (= (charset-chars charset) 96)
540 (dotimes (i 96)
541 (aset str i (make-char charset 32 (+ i 32))))
542 (dotimes (i 94)
543 (aset str i (make-char charset 33 (+ i 33))))))
544 (setq l (find-coding-systems-string str))
545 (if (eq codings t)
546 (setq codings l)
547 (let ((ll nil))
548 (dolist (elt codings)
549 (if (memq elt l)
550 (setq ll (cons elt ll))))
551 (setq codings ll)))))
552 codings))))
553
554 (defun find-multibyte-characters (from to &optional maxcount excludes)
555 "Find multibyte characters in the region specified by FROM and TO.
556 If FROM is a string, find multibyte characters in the string.
557 The return value is an alist of the following format:
558 ((CHARSET COUNT CHAR ...) ...)
559 where
560 CHARSET is a character set,
561 COUNT is a number of characters,
562 CHARs are the characters found from the character set.
563 Optional 3rd arg MAXCOUNT limits how many CHARs are put in the above list.
564 Optional 4th arg EXCLUDE is a list of character sets to be ignored.
565
566 For invalid characters, CHARs are actually strings."
567 (let ((chars nil)
568 charset char)
569 (if (stringp from)
570 (let ((idx 0))
571 (while (setq idx (string-match "[^\000-\177]" from idx))
572 (setq char (aref from idx)
573 charset (char-charset char))
574 (if (eq charset 'unknown)
575 (setq char (match-string 0)))
576 (if (or (memq charset '(unknown
577 eight-bit-control eight-bit-graphic))
578 (not (or (eq excludes t) (memq charset excludes))))
579 (let ((slot (assq charset chars)))
580 (if slot
581 (if (not (memq char (nthcdr 2 slot)))
582 (let ((count (nth 1 slot)))
583 (setcar (cdr slot) (1+ count))
584 (if (or (not maxcount) (< count maxcount))
585 (nconc slot (list char)))))
586 (setq chars (cons (list charset 1 char) chars)))))
587 (setq idx (1+ idx))))
588 (save-excursion
589 (goto-char from)
590 (while (re-search-forward "[^\000-\177]" to t)
591 (setq char (preceding-char)
592 charset (char-charset char))
593 (if (eq charset 'unknown)
594 (setq char (match-string 0)))
595 (if (or (memq charset '(unknown eight-bit-control eight-bit-graphic))
596 (not (or (eq excludes t) (memq charset excludes))))
597 (let ((slot (assq charset chars)))
598 (if slot
599 (if (not (member char (nthcdr 2 slot)))
600 (let ((count (nth 1 slot)))
601 (setcar (cdr slot) (1+ count))
602 (if (or (not maxcount) (< count maxcount))
603 (nconc slot (list char)))))
604 (setq chars (cons (list charset 1 char) chars))))))))
605 (nreverse chars)))
606
607
608 (defun search-unencodable-char (coding-system)
609 "Search forward from point for a character that is not encodable.
610 It asks which coding system to check.
611 If such a character is found, set point after that character.
612 Otherwise, don't move point.
613
614 When called from a program, the value is a position of the found character,
615 or nil if all characters are encodable."
616 (interactive
617 (list (let ((default (or buffer-file-coding-system 'us-ascii)))
618 (read-coding-system
619 (format "Coding-system (default %s): " default)
620 default))))
621 (let ((pos (unencodable-char-position (point) (point-max) coding-system)))
622 (if pos
623 (goto-char (1+ pos))
624 (message "All following characters are encodable by %s" coding-system))
625 pos))
626
627
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.")
634
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.")
640
641 (defun select-safe-coding-system-interactively (from to codings unsafe
642 &optional rejected default)
643 "Select interactively a coding system for the region FROM ... TO.
644 FROM can be a string, as in `write-region'.
645 CODINGS is the list of base coding systems known to be safe for this region,
646 typically obtained with `find-coding-systems-region'.
647 UNSAFE is a list of coding systems known to be unsafe for this region.
648 REJECTED is a list of coding systems which were safe but for some reason
649 were not recommended in the particular context.
650 DEFAULT is the coding system to use by default in the query."
651 ;; At first, if some defaults are unsafe, record at most 11
652 ;; problematic characters and their positions for them by turning
653 ;; (CODING ...)
654 ;; into
655 ;; ((CODING (POS . CHAR) (POS . CHAR) ...) ...)
656 (if unsafe
657 (setq unsafe
658 (mapcar #'(lambda (coding)
659 (cons coding
660 (if (stringp from)
661 (mapcar #'(lambda (pos)
662 (cons pos (aref from pos)))
663 (unencodable-char-position
664 0 (length from) coding
665 11 from))
666 (mapcar #'(lambda (pos)
667 (cons pos (char-after pos)))
668 (unencodable-char-position
669 from to coding 11)))))
670 unsafe)))
671
672 ;; Change each safe coding system to the corresponding
673 ;; mime-charset name if it is also a coding system. Such a name
674 ;; is more friendly to users.
675 (let ((l codings)
676 mime-charset)
677 (while l
678 (setq mime-charset (coding-system-get (car l) 'mime-charset))
679 (if (and mime-charset (coding-system-p mime-charset))
680 (setcar l mime-charset))
681 (setq l (cdr l))))
682
683 ;; Don't offer variations with locking shift, which you
684 ;; basically never want.
685 (let (l)
686 (dolist (elt codings (setq codings (nreverse l)))
687 (unless (or (eq 'coding-category-iso-7-else
688 (coding-system-category elt))
689 (eq 'coding-category-iso-8-else
690 (coding-system-category elt)))
691 (push elt l))))
692
693 ;; Remove raw-text, emacs-mule and no-conversion unless nothing
694 ;; else is available.
695 (setq codings
696 (or (delq 'raw-text
697 (delq 'emacs-mule
698 (delq 'no-conversion codings)))
699 '(raw-text emacs-mule no-conversion)))
700
701 (let ((window-configuration (current-window-configuration))
702 (bufname (buffer-name))
703 coding-system)
704 (save-excursion
705 ;; If some defaults are unsafe, make sure the offending
706 ;; buffer is displayed.
707 (when (and unsafe (not (stringp from)))
708 (pop-to-buffer bufname)
709 (goto-char (apply 'min (mapcar #'(lambda (x) (car (cadr x)))
710 unsafe))))
711 ;; Then ask users to select one from CODINGS while showing
712 ;; the reason why none of the defaults are not used.
713 (with-output-to-temp-buffer "*Warning*"
714 (with-current-buffer standard-output
715 (if (and (null rejected) (null unsafe))
716 (insert "No default coding systems to try for "
717 (if (stringp from)
718 (format "string \"%s\"." from)
719 (format "buffer `%s'." bufname)))
720 (insert
721 "These default coding systems were tried to encode"
722 (if (stringp from)
723 (concat " \"" (if (> (length from) 10)
724 (concat (substring from 0 10) "...\"")
725 (concat from "\"")))
726 (format " text\nin the buffer `%s'" bufname))
727 ":\n")
728 (let ((pos (point))
729 (fill-prefix " "))
730 (dolist (x (append rejected unsafe))
731 (princ " ") (princ (car x)))
732 (insert "\n")
733 (fill-region-as-paragraph pos (point)))
734 (when rejected
735 (insert "These safely encodes the target text,
736 but it is not recommended for encoding text in this context,
737 e.g., for sending an email message.\n ")
738 (dolist (x rejected)
739 (princ " ") (princ x))
740 (insert "\n"))
741 (when unsafe
742 (insert (if rejected "And the others"
743 "However, each of them")
744 " encountered these problematic characters:\n")
745 (dolist (coding unsafe)
746 (insert (format " %s:" (car coding)))
747 (let ((i 0)
748 (func1
749 #'(lambda (bufname pos)
750 (when (buffer-live-p (get-buffer bufname))
751 (pop-to-buffer bufname)
752 (goto-char pos))))
753 (func2
754 #'(lambda (bufname pos coding)
755 (when (buffer-live-p (get-buffer bufname))
756 (pop-to-buffer bufname)
757 (if (< (point) pos)
758 (goto-char pos)
759 (forward-char 1)
760 (search-unencodable-char coding)
761 (forward-char -1))))))
762 (dolist (elt (cdr coding))
763 (insert " ")
764 (if (stringp from)
765 (insert (if (< i 10) (cdr elt) "..."))
766 (if (< i 10)
767 (insert-text-button
768 (cdr elt)
769 :type 'help-xref
770 'help-echo
771 "mouse-2, RET: jump to this character"
772 'help-function func1
773 'help-args (list bufname (car elt)))
774 (insert-text-button
775 "..."
776 :type 'help-xref
777 'help-echo
778 "mouse-2, RET: next unencodable character"
779 'help-function func2
780 'help-args (list bufname (car elt)
781 (car coding)))))
782 (setq i (1+ i))))
783 (insert "\n"))
784 (insert "\
785 The first problematic character is at point in the displayed buffer,\n"
786 (substitute-command-keys "\
787 and \\[universal-argument] \\[what-cursor-position] will give information about it.\n"))))
788 (insert "\nSelect \
789 one of the following safe coding systems, or edit the buffer:\n")
790 (let ((pos (point))
791 (fill-prefix " "))
792 (dolist (x codings)
793 (princ " ") (princ x))
794 (insert "\n")
795 (fill-region-as-paragraph pos (point)))
796 (insert "Or specify any other coding system
797 at the risk of losing the problematic characters.\n")))
798
799 ;; Read a coding system.
800 (setq coding-system
801 (read-coding-system
802 (format "Select coding system (default %s): " default)
803 default))
804 (setq last-coding-system-specified coding-system))
805
806 (kill-buffer "*Warning*")
807 (set-window-configuration window-configuration)
808 coding-system))
809
810 (defun select-safe-coding-system (from to &optional default-coding-system
811 accept-default-p file)
812 "Ask a user to select a safe coding system from candidates.
813 The candidates of coding systems which can safely encode a text
814 between FROM and TO are shown in a popup window. Among them, the most
815 proper one is suggested as the default.
816
817 The list of `buffer-file-coding-system' of the current buffer,
818 the `default-buffer-file-coding-system', and the
819 most preferred coding system (if it corresponds to a MIME charset) is
820 treated as the default coding system list. Among them, the first one
821 that safely encodes the text is normally selected silently and
822 returned without any user interaction. See also the command
823 `prefer-coding-system'.
824
825 However, the user is queried if the chosen coding system is
826 inconsistent with what would be selected by `find-auto-coding' from
827 coding cookies &c. if the contents of the region were read from a
828 file. (That could lead to data corruption in a file subsequently
829 re-visited and edited.)
830
831 Optional 3rd arg DEFAULT-CODING-SYSTEM specifies a coding system or a
832 list of coding systems to be prepended to the default coding system
833 list. However, if DEFAULT-CODING-SYSTEM is a list and the first
834 element is t, the cdr part is used as the defualt coding system list,
835 i.e. `buffer-file-coding-system', `default-buffer-file-coding-system',
836 and the most preferred coding system are not used.
837
838 Optional 4th arg ACCEPT-DEFAULT-P, if non-nil, is a function to
839 determine the acceptability of the silently selected coding system.
840 It is called with that coding system, and should return nil if it
841 should not be silently selected and thus user interaction is required.
842
843 Optional 5th arg FILE is the file name to use for this purpose.
844 That is different from `buffer-file-name' when handling `write-region'
845 \(for example).
846
847 The variable `select-safe-coding-system-accept-default-p', if
848 non-nil, overrides ACCEPT-DEFAULT-P.
849
850 Kludgy feature: if FROM is a string, the string is the target text,
851 and TO is ignored."
852 (if (and default-coding-system
853 (not (listp default-coding-system)))
854 (setq default-coding-system (list default-coding-system)))
855
856 (let ((no-other-defaults nil)
857 auto-cs)
858 (unless (or (stringp from) find-file-literally)
859 ;; Find an auto-coding that is specified for the the current
860 ;; buffer and file from the region FROM and TO.
861 (save-excursion
862 (save-restriction
863 (widen)
864 (goto-char from)
865 (setq auto-cs (find-auto-coding (or file buffer-file-name "")
866 (- to from)))
867 (if auto-cs
868 (if (coding-system-p (car auto-cs))
869 (setq auto-cs (car auto-cs))
870 (display-warning
871 :warning
872 (format "\
873 Invalid coding system `%s' is specified
874 for the current buffer/file by the %s.
875 It is highly recommended to fix it before writing to a file."
876 (car auto-cs)
877 (if (eq (cdr auto-cs) :coding) ":coding tag"
878 (format "variable `%s'" (cdr auto-cs)))))
879 (or (yes-or-no-p "Really proceed with writing? ")
880 (error "Save aborted"))
881 (setq auto-cs nil))))))
882
883 (if (eq (car default-coding-system) t)
884 (setq no-other-defaults t
885 default-coding-system (cdr default-coding-system)))
886
887 ;; Change elements of the list to (coding . base-coding).
888 (setq default-coding-system
889 (mapcar (function (lambda (x) (cons x (coding-system-base x))))
890 default-coding-system))
891
892 (if (and auto-cs (not no-other-defaults))
893 ;; If the file has a coding cookie, try to use it before anything
894 ;; else (i.e. before default-coding-system which will typically come
895 ;; from file-coding-system-alist).
896 (let ((base (coding-system-base auto-cs)))
897 (or (memq base '(nil undecided))
898 (rassq base default-coding-system)
899 (push (cons auto-cs base) default-coding-system))))
900
901 ;; From now on, the list of defaults is reversed.
902 (setq default-coding-system (nreverse default-coding-system))
903
904 (unless no-other-defaults
905 ;; If buffer-file-coding-system is not nil nor undecided, append it
906 ;; to the defaults.
907 (if buffer-file-coding-system
908 (let ((base (coding-system-base buffer-file-coding-system)))
909 (or (eq base 'undecided)
910 (rassq base default-coding-system)
911 (push (cons buffer-file-coding-system base)
912 default-coding-system))))
913
914 ;; If default-buffer-file-coding-system is not nil nor undecided,
915 ;; append it to the defaults.
916 (if default-buffer-file-coding-system
917 (let ((base (coding-system-base default-buffer-file-coding-system)))
918 (or (eq base 'undecided)
919 (rassq base default-coding-system)
920 (push (cons default-buffer-file-coding-system base)
921 default-coding-system))))
922
923 ;; If the most preferred coding system has the property mime-charset,
924 ;; append it to the defaults.
925 (let ((tail coding-category-list)
926 preferred base)
927 (while (and tail (not (setq preferred (symbol-value (car tail)))))
928 (setq tail (cdr tail)))
929 (and (coding-system-p preferred)
930 (setq base (coding-system-base preferred))
931 (coding-system-get preferred 'mime-charset)
932 (not (rassq base default-coding-system))
933 (push (cons preferred base)
934 default-coding-system))))
935
936 (if select-safe-coding-system-accept-default-p
937 (setq accept-default-p select-safe-coding-system-accept-default-p))
938
939 (let ((codings (find-coding-systems-region from to))
940 (coding-system nil)
941 safe rejected unsafe)
942 (if (eq (car codings) 'undecided)
943 ;; Any coding system is ok.
944 (setq coding-system t)
945 ;; Classify the defaults into safe, rejected, and unsafe.
946 (dolist (elt default-coding-system)
947 (if (memq (cdr elt) codings)
948 (if (and (functionp accept-default-p)
949 (not (funcall accept-default-p (cdr elt))))
950 (push (car elt) rejected)
951 (push (car elt) safe))
952 (push (car elt) unsafe)))
953 (if safe
954 (setq coding-system (car safe))))
955
956 ;; If all the defaults failed, ask a user.
957 (when (not coding-system)
958 (setq coding-system (select-safe-coding-system-interactively
959 from to codings unsafe rejected (car codings))))
960
961 (if (vectorp (coding-system-eol-type coding-system))
962 (let ((eol (coding-system-eol-type buffer-file-coding-system)))
963 (if (numberp eol)
964 (setq coding-system
965 (coding-system-change-eol-conversion coding-system eol)))))
966
967 (if (eq coding-system t)
968 (setq coding-system buffer-file-coding-system))
969 ;; Check we're not inconsistent with what `coding:' spec &c would
970 ;; give when file is re-read.
971 ;; But don't do this if we explicitly ignored the cookie
972 ;; by using `find-file-literally'.
973 (when (and auto-cs
974 (not (and
975 coding-system
976 (memq (coding-system-type coding-system) '(0 5)))))
977 ;; Merge coding-system and auto-cs as far as possible.
978 (if (not coding-system)
979 (setq coding-system auto-cs)
980 (if (not auto-cs)
981 (setq auto-cs coding-system)
982 (let ((eol-type-1 (coding-system-eol-type coding-system))
983 (eol-type-2 (coding-system-eol-type auto-cs)))
984 (if (eq (coding-system-base coding-system) 'undecided)
985 (setq coding-system (coding-system-change-text-conversion
986 coding-system auto-cs))
987 (if (eq (coding-system-base auto-cs) 'undecided)
988 (setq auto-cs (coding-system-change-text-conversion
989 auto-cs coding-system))))
990 (if (vectorp eol-type-1)
991 (or (vectorp eol-type-2)
992 (setq coding-system (coding-system-change-eol-conversion
993 coding-system eol-type-2)))
994 (if (vectorp eol-type-2)
995 (setq auto-cs (coding-system-change-eol-conversion
996 auto-cs eol-type-1)))))))
997
998 (if (and auto-cs
999 ;; Don't barf if writing a compressed file, say.
1000 ;; This check perhaps isn't ideal, but is probably
1001 ;; the best thing to do.
1002 (not (auto-coding-alist-lookup (or file buffer-file-name "")))
1003 (not (coding-system-equal coding-system auto-cs)))
1004 (unless (yes-or-no-p
1005 (format "Selected encoding %s disagrees with \
1006 %s specified by file contents. Really save (else edit coding cookies \
1007 and try again)? " coding-system auto-cs))
1008 (error "Save aborted"))))
1009 coding-system)))
1010
1011 (setq select-safe-coding-system-function 'select-safe-coding-system)
1012
1013 (defun select-message-coding-system ()
1014 "Return a coding system to encode the outgoing message of the current buffer.
1015 It at first tries the first coding system found in these variables
1016 in this order:
1017 (1) local value of `buffer-file-coding-system'
1018 (2) value of `sendmail-coding-system'
1019 (3) value of `default-sendmail-coding-system'
1020 (4) value of `default-buffer-file-coding-system'
1021 If the found coding system can't encode the current buffer,
1022 or none of them are bound to a coding system,
1023 it asks the user to select a proper coding system."
1024 (let ((coding (or (and (local-variable-p 'buffer-file-coding-system)
1025 buffer-file-coding-system)
1026 sendmail-coding-system
1027 default-sendmail-coding-system
1028 default-buffer-file-coding-system)))
1029 (if (eq coding 'no-conversion)
1030 ;; We should never use no-conversion for outgoing mail.
1031 (setq coding nil))
1032 (if (fboundp select-safe-coding-system-function)
1033 (setq coding
1034 (funcall select-safe-coding-system-function
1035 (point-min) (point-max) coding
1036 (function (lambda (x)
1037 (coding-system-get x 'mime-charset))))))
1038 (if coding
1039 ;; Be sure to use LF for end-of-line.
1040 (setq coding (coding-system-change-eol-conversion coding 'unix))
1041 ;; No coding system is decided. Usually this is the case that
1042 ;; the current buffer contains only ASCII. So, we hope
1043 ;; iso-8859-1 works.
1044 (setq coding 'iso-8859-1-unix))
1045 coding))
1046 \f
1047 ;;; Language support stuff.
1048
1049 (defvar language-info-alist nil
1050 "Alist of language environment definitions.
1051 Each element looks like:
1052 (LANGUAGE-NAME . ((KEY . INFO) ...))
1053 where LANGUAGE-NAME is a string, the name of the language environment,
1054 KEY is a symbol denoting the kind of information, and
1055 INFO is the data associated with KEY.
1056 Meaningful values for KEY include
1057
1058 documentation value is documentation of what this language environment
1059 is meant for, and how to use it.
1060 charset value is a list of the character sets used by this
1061 language environment.
1062 sample-text value is an expression which is evalled to generate
1063 a line of text written using characters appropriate
1064 for this language environment.
1065 setup-function value is a function to call to switch to this
1066 language environment.
1067 exit-function value is a function to call to leave this
1068 language environment.
1069 coding-system value is a list of coding systems that are good
1070 for saving text written in this language environment.
1071 This list serves as suggestions to the user;
1072 in effect, as a kind of documentation.
1073 coding-priority value is a list of coding systems for this language
1074 environment, in order of decreasing priority.
1075 This is used to set up the coding system priority
1076 list when you switch to this language environment.
1077 nonascii-translation
1078 value is a translation table to be set in the
1079 variable `nonascii-translation-table' in this
1080 language environment, or a character set from
1081 which `nonascii-insert-offset' is calculated.
1082 input-method value is a default input method for this language
1083 environment.
1084 features value is a list of features requested in this
1085 language environment.
1086 ctext-non-standard-encodings
1087 value is a list of non-standard encoding
1088 names used in extended segments of CTEXT.
1089 See the variable
1090 `ctext-non-standard-encodings' for more
1091 detail.
1092
1093 The following keys take effect only when multibyte characters are
1094 globally disabled, i.e. the value of `default-enable-multibyte-characters'
1095 is nil.
1096
1097 unibyte-syntax value is a library name to load to set
1098 unibyte 8-bit character syntaxes for this
1099 language environment.
1100
1101 unibyte-display value is a coding system to encode characters
1102 for the terminal. Characters in the range
1103 of 160 to 255 display not as octal escapes,
1104 but as non-ASCII characters in this language
1105 environment.")
1106
1107 (defun get-language-info (lang-env key)
1108 "Return information listed under KEY for language environment LANG-ENV.
1109 KEY is a symbol denoting the kind of information.
1110 For a list of useful values for KEY and their meanings,
1111 see `language-info-alist'."
1112 (if (symbolp lang-env)
1113 (setq lang-env (symbol-name lang-env)))
1114 (let ((lang-slot (assoc-string lang-env language-info-alist t)))
1115 (if lang-slot
1116 (cdr (assq key (cdr lang-slot))))))
1117
1118 (defun set-language-info (lang-env key info)
1119 "Modify part of the definition of language environment LANG-ENV.
1120 Specifically, this stores the information INFO under KEY
1121 in the definition of this language environment.
1122 KEY is a symbol denoting the kind of information.
1123 INFO is the value for that information.
1124
1125 For a list of useful values for KEY and their meanings,
1126 see `language-info-alist'."
1127 (if (symbolp lang-env)
1128 (setq lang-env (symbol-name lang-env)))
1129 (let (lang-slot key-slot)
1130 (setq lang-slot (assoc lang-env language-info-alist))
1131 (if (null lang-slot) ; If no slot for the language, add it.
1132 (setq lang-slot (list lang-env)
1133 language-info-alist (cons lang-slot language-info-alist)))
1134 (setq key-slot (assq key lang-slot))
1135 (if (null key-slot) ; If no slot for the key, add it.
1136 (progn
1137 (setq key-slot (list key))
1138 (setcdr lang-slot (cons key-slot (cdr lang-slot)))))
1139 (setcdr key-slot (purecopy info))
1140 ;; Update the custom-type of `current-language-environment'.
1141 (put 'current-language-environment 'custom-type
1142 (cons 'choice (mapcar
1143 (lambda (lang)
1144 (list 'const lang))
1145 (sort (mapcar 'car language-info-alist) 'string<))))))
1146
1147 (defun set-language-info-alist (lang-env alist &optional parents)
1148 "Store ALIST as the definition of language environment LANG-ENV.
1149 ALIST is an alist of KEY and INFO values. See the documentation of
1150 `language-info-alist' for the meanings of KEY and INFO.
1151
1152 Optional arg PARENTS is a list of parent menu names; it specifies
1153 where to put this language environment in the
1154 Describe Language Environment and Set Language Environment menus.
1155 For example, (\"European\") means to put this language environment
1156 in the European submenu in each of those two menus."
1157 (if (symbolp lang-env)
1158 (setq lang-env (symbol-name lang-env)))
1159 (let ((describe-map describe-language-environment-map)
1160 (setup-map setup-language-environment-map))
1161 (if parents
1162 (let ((l parents)
1163 map parent-symbol parent prompt)
1164 (while l
1165 (if (symbolp (setq parent-symbol (car l)))
1166 (setq parent (symbol-name parent))
1167 (setq parent parent-symbol parent-symbol (intern parent)))
1168 (setq map (lookup-key describe-map (vector parent-symbol)))
1169 ;; This prompt string is for define-prefix-command, so
1170 ;; that the map it creates will be suitable for a menu.
1171 (or map (setq prompt (format "%s Environment" parent)))
1172 (if (not map)
1173 (progn
1174 (setq map (intern (format "describe-%s-environment-map"
1175 (downcase parent))))
1176 (define-prefix-command map nil prompt)
1177 (define-key-after describe-map (vector parent-symbol)
1178 (cons parent map) t)))
1179 (setq describe-map (symbol-value map))
1180 (setq map (lookup-key setup-map (vector parent-symbol)))
1181 (if (not map)
1182 (progn
1183 (setq map (intern (format "setup-%s-environment-map"
1184 (downcase parent))))
1185 (define-prefix-command map nil prompt)
1186 (define-key-after setup-map (vector parent-symbol)
1187 (cons parent map) t)))
1188 (setq setup-map (symbol-value map))
1189 (setq l (cdr l)))))
1190
1191 ;; Set up menu items for this language env.
1192 (let ((doc (assq 'documentation alist)))
1193 (when doc
1194 (define-key-after describe-map (vector (intern lang-env))
1195 (cons lang-env 'describe-specified-language-support) t)))
1196 (define-key-after setup-map (vector (intern lang-env))
1197 (cons lang-env 'setup-specified-language-environment) t)
1198
1199 (while alist
1200 (set-language-info lang-env (car (car alist)) (cdr (car alist)))
1201 (setq alist (cdr alist)))))
1202
1203 (defun read-language-name (key prompt &optional default)
1204 "Read a language environment name which has information for KEY.
1205 If KEY is nil, read any language environment.
1206 Prompt with PROMPT. DEFAULT is the default choice of language environment.
1207 This returns a language environment name as a string."
1208 (let* ((completion-ignore-case t)
1209 (name (completing-read prompt
1210 language-info-alist
1211 (and key
1212 (function (lambda (elm) (and (listp elm) (assq key elm)))))
1213 t nil nil default)))
1214 (if (and (> (length name) 0)
1215 (or (not key)
1216 (get-language-info name key)))
1217 name)))
1218 \f
1219 ;;; Multilingual input methods.
1220 (defgroup leim nil
1221 "LEIM: Libraries of Emacs Input Methods."
1222 :group 'mule)
1223
1224 (defconst leim-list-file-name "leim-list.el"
1225 "Name of LEIM list file.
1226 This file contains a list of libraries of Emacs input methods (LEIM)
1227 in the format of Lisp expression for registering each input method.
1228 Emacs loads this file at startup time.")
1229
1230 (defvar leim-list-header (format
1231 ";;; %s -- list of LEIM (Library of Emacs Input Method) -*-coding: iso-2022-7bit;-*-
1232 ;;
1233 ;; This file contains a list of LEIM (Library of Emacs Input Method)
1234 ;; methods in the same directory as this file. Loading this file
1235 ;; registers all the input methods in Emacs.
1236 ;;
1237 ;; Each entry has the form:
1238 ;; (register-input-method
1239 ;; INPUT-METHOD LANGUAGE-NAME ACTIVATE-FUNC
1240 ;; TITLE DESCRIPTION
1241 ;; ARG ...)
1242 ;; See the function `register-input-method' for the meanings of the arguments.
1243 ;;
1244 ;; If this directory is included in load-path, Emacs automatically
1245 ;; loads this file at startup time.
1246
1247 "
1248 leim-list-file-name)
1249 "Header to be inserted in LEIM list file.")
1250
1251 (defvar leim-list-entry-regexp "^(register-input-method"
1252 "Regexp matching head of each entry in LEIM list file.
1253 See also the variable `leim-list-header'")
1254
1255 (defvar update-leim-list-functions
1256 '(quail-update-leim-list-file)
1257 "List of functions to call to update LEIM list file.
1258 Each function is called with one arg, LEIM directory name.")
1259
1260 (defun update-leim-list-file (&rest dirs)
1261 "Update LEIM list file in directories DIRS."
1262 (let ((functions update-leim-list-functions))
1263 (while functions
1264 (apply (car functions) dirs)
1265 (setq functions (cdr functions)))))
1266
1267 (defvar current-input-method nil
1268 "The current input method for multilingual text.
1269 If nil, that means no input method is activated now.")
1270 (make-variable-buffer-local 'current-input-method)
1271 (put 'current-input-method 'permanent-local t)
1272
1273 (defvar current-input-method-title nil
1274 "Title string of the current input method shown in mode line.")
1275 (make-variable-buffer-local 'current-input-method-title)
1276 (put 'current-input-method-title 'permanent-local t)
1277
1278 (defcustom default-input-method nil
1279 "*Default input method for multilingual text (a string).
1280 This is the input method activated automatically by the command
1281 `toggle-input-method' (\\[toggle-input-method])."
1282 :link '(custom-manual "(emacs)Input Methods")
1283 :group 'mule
1284 :type '(choice (const nil) (string
1285 :completion-ignore-case t
1286 :complete-function widget-string-complete
1287 :completion-alist input-method-alist
1288 :prompt-history input-method-history))
1289 :set-after '(current-language-environment))
1290
1291 (put 'input-method-function 'permanent-local t)
1292
1293 (defvar input-method-history nil
1294 "History list for some commands that read input methods.")
1295 (make-variable-buffer-local 'input-method-history)
1296 (put 'input-method-history 'permanent-local t)
1297
1298 (defvar inactivate-current-input-method-function nil
1299 "Function to call for inactivating the current input method.
1300 Every input method should set this to an appropriate value when activated.
1301 This function is called with no argument.
1302
1303 This function should never change the value of `current-input-method'.
1304 It is set to nil by the function `inactivate-input-method'.")
1305 (make-variable-buffer-local 'inactivate-current-input-method-function)
1306 (put 'inactivate-current-input-method-function 'permanent-local t)
1307
1308 (defvar describe-current-input-method-function nil
1309 "Function to call for describing the current input method.
1310 This function is called with no argument.")
1311 (make-variable-buffer-local 'describe-current-input-method-function)
1312 (put 'describe-current-input-method-function 'permanent-local t)
1313
1314 (defvar input-method-alist nil
1315 "Alist of input method names vs how to use them.
1316 Each element has the form:
1317 (INPUT-METHOD LANGUAGE-ENV ACTIVATE-FUNC TITLE DESCRIPTION ARGS...)
1318 See the function `register-input-method' for the meanings of the elements.")
1319
1320 (defun register-input-method (input-method lang-env &rest args)
1321 "Register INPUT-METHOD as an input method for language environment LANG-ENV.
1322 INPUT-METHOD and LANG-ENV are symbols or strings.
1323
1324 The remaining arguments are:
1325 ACTIVATE-FUNC, TITLE, DESCRIPTION, and ARGS...
1326 ACTIVATE-FUNC is a function to call to activate this method.
1327 TITLE is a string to show in the mode line when this method is active.
1328 DESCRIPTION is a string describing this method and what it is good for.
1329 The ARGS, if any, are passed as arguments to ACTIVATE-FUNC.
1330 All told, the arguments to ACTIVATE-FUNC are INPUT-METHOD and the ARGS.
1331
1332 This function is mainly used in the file \"leim-list.el\" which is
1333 created at Emacs build time, registering all Quail input methods
1334 contained in the Emacs distribution.
1335
1336 In case you want to register a new Quail input method by yourself, be
1337 careful to use the same input method title as given in the third
1338 parameter of `quail-define-package'. (If the values are different, the
1339 string specified in this function takes precedence.)
1340
1341 The commands `describe-input-method' and `list-input-methods' need
1342 these duplicated values to show some information about input methods
1343 without loading the relevant Quail packages."
1344 (if (symbolp lang-env)
1345 (setq lang-env (symbol-name lang-env)))
1346 (if (symbolp input-method)
1347 (setq input-method (symbol-name input-method)))
1348 (let ((info (cons lang-env args))
1349 (slot (assoc input-method input-method-alist)))
1350 (if slot
1351 (setcdr slot info)
1352 (setq slot (cons input-method info))
1353 (setq input-method-alist (cons slot input-method-alist)))))
1354
1355 (defun read-input-method-name (prompt &optional default inhibit-null)
1356 "Read a name of input method from a minibuffer prompting with PROMPT.
1357 If DEFAULT is non-nil, use that as the default,
1358 and substitute it into PROMPT at the first `%s'.
1359 If INHIBIT-NULL is non-nil, null input signals an error.
1360
1361 The return value is a string."
1362 (if default
1363 (setq prompt (format prompt default)))
1364 (let* ((completion-ignore-case t)
1365 ;; As it is quite normal to change input method in the
1366 ;; minibuffer, we must enable it even if
1367 ;; enable-recursive-minibuffers is currently nil.
1368 (enable-recursive-minibuffers t)
1369 ;; This binding is necessary because input-method-history is
1370 ;; buffer local.
1371 (input-method (completing-read prompt input-method-alist
1372 nil t nil 'input-method-history
1373 default)))
1374 (if (and input-method (symbolp input-method))
1375 (setq input-method (symbol-name input-method)))
1376 (if (> (length input-method) 0)
1377 input-method
1378 (if inhibit-null
1379 (error "No valid input method is specified")))))
1380
1381 (defun activate-input-method (input-method)
1382 "Switch to input method INPUT-METHOD for the current buffer.
1383 If some other input method is already active, turn it off first.
1384 If INPUT-METHOD is nil, deactivate any current input method."
1385 (if (and input-method (symbolp input-method))
1386 (setq input-method (symbol-name input-method)))
1387 (if (and current-input-method
1388 (not (string= current-input-method input-method)))
1389 (inactivate-input-method))
1390 (unless (or current-input-method (null input-method))
1391 (let ((slot (assoc input-method input-method-alist)))
1392 (if (null slot)
1393 (error "Can't activate input method `%s'" input-method))
1394 (setq current-input-method-title nil)
1395 (let ((func (nth 2 slot)))
1396 (if (functionp func)
1397 (apply (nth 2 slot) input-method (nthcdr 5 slot))
1398 (if (and (consp func) (symbolp (car func)) (symbolp (cdr func)))
1399 (progn
1400 (require (cdr func))
1401 (apply (car func) input-method (nthcdr 5 slot)))
1402 (error "Can't activate input method `%s'" input-method))))
1403 (setq current-input-method input-method)
1404 (or (stringp current-input-method-title)
1405 (setq current-input-method-title (nth 3 slot)))
1406 (unwind-protect
1407 (run-hooks 'input-method-activate-hook)
1408 (force-mode-line-update)))))
1409
1410 (defun inactivate-input-method ()
1411 "Turn off the current input method."
1412 (when current-input-method
1413 (if input-method-history
1414 (unless (string= current-input-method (car input-method-history))
1415 (setq input-method-history
1416 (cons current-input-method
1417 (delete current-input-method input-method-history))))
1418 (setq input-method-history (list current-input-method)))
1419 (unwind-protect
1420 (funcall inactivate-current-input-method-function)
1421 (unwind-protect
1422 (run-hooks 'input-method-inactivate-hook)
1423 (setq current-input-method nil
1424 input-method-function nil
1425 current-input-method-title nil)
1426 (force-mode-line-update)))))
1427
1428 (defun set-input-method (input-method &optional interactive)
1429 "Select and activate input method INPUT-METHOD for the current buffer.
1430 This also sets the default input method to the one you specify.
1431 If INPUT-METHOD is nil, this function turns off the input method, and
1432 also causes you to be prompted for a name of an input method the next
1433 time you invoke \\[toggle-input-method].
1434 When called interactively, the optional arg INTERACTIVE is non-nil,
1435 which marks the variable `default-input-method' as set for Custom buffers.
1436
1437 To deactivate the input method interactively, use \\[toggle-input-method].
1438 To deactivate it programmatically, use \\[inactivate-input-method]."
1439 (interactive
1440 (let* ((default (or (car input-method-history) default-input-method)))
1441 (list (read-input-method-name
1442 (if default "Select input method (default %s): " "Select input method: ")
1443 default t)
1444 t)))
1445 (activate-input-method input-method)
1446 (setq default-input-method input-method)
1447 (when interactive
1448 (customize-mark-as-set 'default-input-method))
1449 default-input-method)
1450
1451 (defun toggle-input-method (&optional arg interactive)
1452 "Enable or disable multilingual text input method for the current buffer.
1453 Only one input method can be enabled at any time in a given buffer.
1454
1455 The normal action is to enable an input method if none was
1456 enabled, and disable the current one otherwise. Which input method
1457 to enable can be determined in various ways--either the one most
1458 recently used, or the one specified by `default-input-method', or
1459 as a last resort by reading the name of an input method in the
1460 minibuffer.
1461
1462 With a prefix argument, read an input method name with the minibuffer
1463 and enable that one. The default is the most recent input method specified
1464 \(not including the currently active input method, if any).
1465
1466 When called interactively, the optional arg INTERACTIVE is non-nil,
1467 which marks the variable `default-input-method' as set for Custom buffers."
1468
1469 (interactive "P\np")
1470 (if (and current-input-method (not arg))
1471 (inactivate-input-method)
1472 (let ((default (or (car input-method-history) default-input-method)))
1473 (if (and arg default (equal current-input-method default)
1474 (> (length input-method-history) 1))
1475 (setq default (nth 1 input-method-history)))
1476 (activate-input-method
1477 (if (or arg (not default))
1478 (progn
1479 (read-input-method-name
1480 (if default "Input method (default %s): " "Input method: " )
1481 default t))
1482 default))
1483 (unless default-input-method
1484 (prog1
1485 (setq default-input-method current-input-method)
1486 (when interactive
1487 (customize-mark-as-set 'default-input-method)))))))
1488
1489 (defun describe-input-method (input-method)
1490 "Describe input method INPUT-METHOD."
1491 (interactive
1492 (list (read-input-method-name
1493 "Describe input method (default current choice): ")))
1494 (if (and input-method (symbolp input-method))
1495 (setq input-method (symbol-name input-method)))
1496 (help-setup-xref (list #'describe-input-method
1497 (or input-method current-input-method))
1498 (interactive-p))
1499
1500 (if (null input-method)
1501 (describe-current-input-method)
1502 (let ((current current-input-method))
1503 (condition-case nil
1504 (progn
1505 (save-excursion
1506 (activate-input-method input-method)
1507 (describe-current-input-method))
1508 (activate-input-method current))
1509 (error
1510 (activate-input-method current)
1511 (help-setup-xref (list #'describe-input-method input-method)
1512 (interactive-p))
1513 (with-output-to-temp-buffer (help-buffer)
1514 (let ((elt (assoc input-method input-method-alist)))
1515 (princ (format
1516 "Input method: %s (`%s' in mode line) for %s\n %s\n"
1517 input-method (nth 3 elt) (nth 1 elt) (nth 4 elt))))))))))
1518
1519 (defun describe-current-input-method ()
1520 "Describe the input method currently in use.
1521 This is a subroutine for `describe-input-method'."
1522 (if current-input-method
1523 (if (and (symbolp describe-current-input-method-function)
1524 (fboundp describe-current-input-method-function))
1525 (funcall describe-current-input-method-function)
1526 (message "No way to describe the current input method `%s'"
1527 current-input-method)
1528 (ding))
1529 (error "No input method is activated now")))
1530
1531 (defun read-multilingual-string (prompt &optional initial-input input-method)
1532 "Read a multilingual string from minibuffer, prompting with string PROMPT.
1533 The input method selected last time is activated in minibuffer.
1534 If optional second arg INITIAL-INPUT is non-nil, insert it in the minibuffer
1535 initially.
1536 Optional 3rd argument INPUT-METHOD specifies the input method
1537 to be activated instead of the one selected last time. It is a symbol
1538 or a string."
1539 (setq input-method
1540 (or input-method
1541 current-input-method
1542 default-input-method
1543 (read-input-method-name "Input method: " nil t)))
1544 (if (and input-method (symbolp input-method))
1545 (setq input-method (symbol-name input-method)))
1546 (let ((prev-input-method current-input-method))
1547 (unwind-protect
1548 (progn
1549 (activate-input-method input-method)
1550 (read-string prompt initial-input nil nil t))
1551 (activate-input-method prev-input-method))))
1552
1553 ;; Variables to control behavior of input methods. All input methods
1554 ;; should react to these variables.
1555
1556 (defcustom input-method-verbose-flag 'default
1557 "*A flag to control extra guidance given by input methods.
1558 The value should be nil, t, `complex-only', or `default'.
1559
1560 The extra guidance is done by showing list of available keys in echo
1561 area. When you use the input method in the minibuffer, the guidance
1562 is shown at the bottom short window (split from the existing window).
1563
1564 If the value is t, extra guidance is always given, if the value is
1565 nil, extra guidance is always suppressed.
1566
1567 If the value is `complex-only', only complex input methods such as
1568 `chinese-py' and `japanese' give extra guidance.
1569
1570 If the value is `default', complex input methods always give extra
1571 guidance, but simple input methods give it only when you are not in
1572 the minibuffer.
1573
1574 See also the variable `input-method-highlight-flag'."
1575 :type '(choice (const :tag "Always" t) (const :tag "Never" nil)
1576 (const complex-only) (const default))
1577 :group 'mule)
1578
1579 (defcustom input-method-highlight-flag t
1580 "*If this flag is non-nil, input methods highlight partially-entered text.
1581 For instance, while you are in the middle of a Quail input method sequence,
1582 the text inserted so far is temporarily underlined.
1583 The underlining goes away when you finish or abort the input method sequence.
1584 See also the variable `input-method-verbose-flag'."
1585 :type 'boolean
1586 :group 'mule)
1587
1588 (defvar input-method-activate-hook nil
1589 "Normal hook run just after an input method is activated.
1590
1591 The variable `current-input-method' keeps the input method name
1592 just activated.")
1593
1594 (defvar input-method-inactivate-hook nil
1595 "Normal hook run just after an input method is inactivated.
1596
1597 The variable `current-input-method' still keeps the input method name
1598 just inactivated.")
1599
1600 (defvar input-method-after-insert-chunk-hook nil
1601 "Normal hook run just after an input method insert some chunk of text.")
1602
1603 (defvar input-method-exit-on-first-char nil
1604 "This flag controls when an input method returns.
1605 Usually, the input method does not return while there's a possibility
1606 that it may find a different translation if a user types another key.
1607 But, it this flag is non-nil, the input method returns as soon as
1608 the current key sequence gets long enough to have some valid translation.")
1609
1610 (defvar input-method-use-echo-area nil
1611 "This flag controls how an input method shows an intermediate key sequence.
1612 Usually, the input method inserts the intermediate key sequence,
1613 or candidate translations corresponding to the sequence,
1614 at point in the current buffer.
1615 But, if this flag is non-nil, it displays them in echo area instead.")
1616
1617 (defvar input-method-exit-on-invalid-key nil
1618 "This flag controls the behavior of an input method on invalid key input.
1619 Usually, when a user types a key which doesn't start any character
1620 handled by the input method, the key is handled by turning off the
1621 input method temporarily. After that key, the input method is re-enabled.
1622 But, if this flag is non-nil, the input method is never back on.")
1623
1624 \f
1625 (defvar set-language-environment-hook nil
1626 "Normal hook run after some language environment is set.
1627
1628 When you set some hook function here, that effect usually should not
1629 be inherited to another language environment. So, you had better set
1630 another function in `exit-language-environment-hook' (which see) to
1631 cancel the effect.")
1632
1633 (defvar exit-language-environment-hook nil
1634 "Normal hook run after exiting from some language environment.
1635 When this hook is run, the variable `current-language-environment'
1636 is still bound to the language environment being exited.
1637
1638 This hook is mainly used for canceling the effect of
1639 `set-language-environment-hook' (which-see).")
1640
1641 (put 'setup-specified-language-environment 'apropos-inhibit t)
1642
1643 (defun setup-specified-language-environment ()
1644 "Switch to a specified language environment."
1645 (interactive)
1646 (let (language-name)
1647 (if (and (symbolp last-command-event)
1648 (or (not (eq last-command-event 'Default))
1649 (setq last-command-event 'English))
1650 (setq language-name (symbol-name last-command-event)))
1651 (prog1
1652 (set-language-environment language-name)
1653 (customize-mark-as-set 'current-language-environment))
1654 (error "Bogus calling sequence"))))
1655
1656 (defcustom current-language-environment "English"
1657 "The last language environment specified with `set-language-environment'.
1658 This variable should be set only with \\[customize], which is equivalent
1659 to using the function `set-language-environment'."
1660 :link '(custom-manual "(emacs)Language Environments")
1661 :set (lambda (symbol value) (set-language-environment value))
1662 :get (lambda (x)
1663 (or (car-safe (assoc-string
1664 (if (symbolp current-language-environment)
1665 (symbol-name current-language-environment)
1666 current-language-environment)
1667 language-info-alist t))
1668 "English"))
1669 ;; custom type will be updated with `set-language-info'.
1670 :type (if language-info-alist
1671 (cons 'choice (mapcar
1672 (lambda (lang)
1673 (list 'const lang))
1674 (sort (mapcar 'car language-info-alist) 'string<)))
1675 'string)
1676 :initialize 'custom-initialize-default
1677 :group 'mule)
1678
1679 (defun reset-language-environment ()
1680 "Reset multilingual environment of Emacs to the default status.
1681
1682 The default status is as follows:
1683
1684 The default value of `buffer-file-coding-system' is nil.
1685 The default coding system for process I/O is nil.
1686 The default value for the command `set-terminal-coding-system' is nil.
1687 The default value for the command `set-keyboard-coding-system' is nil.
1688
1689 The order of priorities of coding categories and the coding system
1690 bound to each category are as follows
1691 coding category coding system
1692 --------------------------------------------------
1693 coding-category-iso-8-1 iso-latin-1
1694 coding-category-iso-8-2 iso-latin-1
1695 coding-category-utf-8 mule-utf-8
1696 coding-category-utf-16-be mule-utf-16be-with-signature
1697 coding-category-utf-16-le mule-utf-16le-with-signature
1698 coding-category-iso-7-tight iso-2022-jp
1699 coding-category-iso-7 iso-2022-7bit
1700 coding-category-iso-7-else iso-2022-7bit-lock
1701 coding-category-iso-8-else iso-2022-8bit-ss2
1702 coding-category-emacs-mule emacs-mule
1703 coding-category-raw-text raw-text
1704 coding-category-sjis japanese-shift-jis
1705 coding-category-big5 chinese-big5
1706 coding-category-ccl nil
1707 coding-category-binary no-conversion"
1708 (interactive)
1709 ;; This function formerly set default-enable-multibyte-characters to t,
1710 ;; but that is incorrect. It should not alter the unibyte/multibyte choice.
1711
1712 (setq coding-category-iso-7-tight 'iso-2022-jp
1713 coding-category-iso-7 'iso-2022-7bit
1714 coding-category-iso-8-1 'iso-latin-1
1715 coding-category-iso-8-2 'iso-latin-1
1716 coding-category-iso-7-else 'iso-2022-7bit-lock
1717 coding-category-iso-8-else 'iso-2022-8bit-ss2
1718 coding-category-emacs-mule 'emacs-mule
1719 coding-category-raw-text 'raw-text
1720 coding-category-sjis 'japanese-shift-jis
1721 coding-category-big5 'chinese-big5
1722 coding-category-utf-16-be 'mule-utf-16be-with-signature
1723 coding-category-utf-16-le 'mule-utf-16le-with-signature
1724 coding-category-utf-8 'mule-utf-8
1725 coding-category-ccl nil
1726 coding-category-binary 'no-conversion)
1727
1728 (set-coding-priority
1729 '(coding-category-iso-8-1
1730 coding-category-iso-8-2
1731 coding-category-utf-8
1732 coding-category-utf-16-be
1733 coding-category-utf-16-le
1734 coding-category-iso-7-tight
1735 coding-category-iso-7
1736 coding-category-iso-7-else
1737 coding-category-iso-8-else
1738 coding-category-emacs-mule
1739 coding-category-raw-text
1740 coding-category-sjis
1741 coding-category-big5
1742 coding-category-ccl
1743 coding-category-binary))
1744
1745 ;; Changing the binding of a coding category requires this call.
1746 (update-coding-systems-internal)
1747
1748 (set-default-coding-systems nil)
1749 (setq default-sendmail-coding-system 'iso-latin-1)
1750 ;; On Darwin systems, this should be utf-8, but when this file is loaded
1751 ;; utf-8 is not yet defined, so we set it in set-locale-environment instead.
1752 (setq default-file-name-coding-system 'iso-latin-1)
1753 ;; Preserve eol-type from existing default-process-coding-systems.
1754 ;; On non-unix-like systems in particular, these may have been set
1755 ;; carefully by the user, or by the startup code, to deal with the
1756 ;; users shell appropriately, so should not be altered by changing
1757 ;; language environment.
1758 (let ((output-coding
1759 ;; When bootstrapping, coding-systems are not defined yet, so
1760 ;; we need to catch the error from check-coding-system.
1761 (condition-case nil
1762 (coding-system-change-text-conversion
1763 (car default-process-coding-system) 'undecided)
1764 (coding-system-error 'undecided)))
1765 (input-coding
1766 (condition-case nil
1767 (coding-system-change-text-conversion
1768 (cdr default-process-coding-system) 'iso-latin-1)
1769 (coding-system-error 'iso-latin-1))))
1770 (setq default-process-coding-system
1771 (cons output-coding input-coding)))
1772
1773 ;; Don't alter the terminal and keyboard coding systems here.
1774 ;; The terminal still supports the same coding system
1775 ;; that it supported a minute ago.
1776 ;; (set-terminal-coding-system-internal nil)
1777 ;; (set-keyboard-coding-system-internal nil)
1778
1779 (setq nonascii-translation-table nil
1780 nonascii-insert-offset 0)
1781
1782 ;; Don't invoke fontset-related functions if fontsets aren't
1783 ;; supported in this build of Emacs.
1784 (and (fboundp 'fontset-list)
1785 (set-overriding-fontspec-internal nil)))
1786
1787 (reset-language-environment)
1788
1789 (defun set-display-table-and-terminal-coding-system (language-name &optional coding-system display)
1790 "Set up the display table and terminal coding system for LANGUAGE-NAME."
1791 (let ((coding (get-language-info language-name 'unibyte-display)))
1792 (if coding
1793 (standard-display-european-internal)
1794 ;; The following 2 lines undo the 8-bit display that we set up
1795 ;; in standard-display-european-internal, which see. This is in
1796 ;; case the user has used standard-display-european earlier in
1797 ;; this session. (The MS-DOS port doesn't use that setup, so it
1798 ;; doesn't need to undo it.)
1799 (when standard-display-table
1800 (dotimes (i 128)
1801 (aset standard-display-table (+ i 128) nil))))
1802 (or (eq window-system 'pc)
1803 (set-terminal-coding-system (or coding-system coding) display))))
1804
1805 (defun set-language-environment (language-name)
1806 "Set up multi-lingual environment for using LANGUAGE-NAME.
1807 This sets the coding system priority and the default input method
1808 and sometimes other things. LANGUAGE-NAME should be a string
1809 which is the name of a language environment. For example, \"Latin-1\"
1810 specifies the character set for the major languages of Western Europe."
1811 (interactive (list (read-language-name
1812 nil
1813 "Set language environment (default English): ")))
1814 (if language-name
1815 (if (symbolp language-name)
1816 (setq language-name (symbol-name language-name)))
1817 (setq language-name "English"))
1818 (let ((slot (assoc-string language-name language-info-alist t)))
1819 (unless slot
1820 (error "Language environment not defined: %S" language-name))
1821 (setq language-name (car slot)))
1822 (if current-language-environment
1823 (let ((func (get-language-info current-language-environment
1824 'exit-function)))
1825 (run-hooks 'exit-language-environment-hook)
1826 (if (functionp func) (funcall func))))
1827 (let ((default-eol-type (coding-system-eol-type
1828 default-buffer-file-coding-system)))
1829 (reset-language-environment)
1830
1831 ;; The features might set up coding systems.
1832 (let ((required-features (get-language-info language-name 'features)))
1833 (while required-features
1834 (require (car required-features))
1835 (setq required-features (cdr required-features))))
1836
1837 (setq current-language-environment language-name)
1838 (set-language-environment-coding-systems language-name default-eol-type))
1839 (let ((input-method (get-language-info language-name 'input-method)))
1840 (when input-method
1841 (setq default-input-method input-method)
1842 (if input-method-history
1843 (setq input-method-history
1844 (cons input-method
1845 (delete input-method input-method-history))))))
1846 (let ((nonascii (get-language-info language-name 'nonascii-translation))
1847 (dos-table
1848 (if (eq window-system 'pc)
1849 (intern
1850 (format "cp%d-nonascii-translation-table" dos-codepage)))))
1851 (cond
1852 ((char-table-p nonascii)
1853 (setq nonascii-translation-table nonascii))
1854 ((and (eq window-system 'pc) (boundp dos-table))
1855 ;; DOS terminals' default is to use a special non-ASCII translation
1856 ;; table as appropriate for the installed codepage.
1857 (setq nonascii-translation-table (symbol-value dos-table)))
1858 ((charsetp nonascii)
1859 (setq nonascii-insert-offset (- (make-char nonascii) 128)))))
1860
1861 ;; Unibyte setups if necessary.
1862 (unless default-enable-multibyte-characters
1863 ;; Syntax and case table.
1864 (let ((syntax (get-language-info language-name 'unibyte-syntax)))
1865 (if syntax
1866 (let ((set-case-syntax-set-multibyte nil))
1867 (load syntax nil t))
1868 ;; No information for syntax and case. Reset to the defaults.
1869 (let ((syntax-table (standard-syntax-table))
1870 (standard-table (standard-case-table))
1871 (case-table (make-char-table 'case-table))
1872 (ch (if (eq window-system 'pc) 128 160)))
1873 (while (< ch 256)
1874 (modify-syntax-entry ch " " syntax-table)
1875 (setq ch (1+ ch)))
1876 (dotimes (i 128)
1877 (aset case-table i (aref standard-table i)))
1878 (set-char-table-extra-slot case-table 0 nil)
1879 (set-char-table-extra-slot case-table 1 nil)
1880 (set-char-table-extra-slot case-table 2 nil)
1881 (set-standard-case-table case-table))
1882 (let ((list (buffer-list)))
1883 (while list
1884 (with-current-buffer (car list)
1885 (set-case-table (standard-case-table)))
1886 (setq list (cdr list))))))
1887 (set-display-table-and-terminal-coding-system language-name))
1888
1889 (let ((required-features (get-language-info language-name 'features)))
1890 (while required-features
1891 (require (car required-features))
1892 (setq required-features (cdr required-features))))
1893
1894 ;; Don't invoke fontset-related functions if fontsets aren't
1895 ;; supported in this build of Emacs.
1896 (when (fboundp 'fontset-list)
1897 (let ((overriding-fontspec (get-language-info language-name
1898 'overriding-fontspec)))
1899 (if overriding-fontspec
1900 (set-overriding-fontspec-internal overriding-fontspec))))
1901
1902 (let ((func (get-language-info language-name 'setup-function)))
1903 (if (functionp func)
1904 (funcall func)))
1905 (if (and utf-translate-cjk-mode
1906 (not (eq utf-translate-cjk-lang-env language-name))
1907 (catch 'tag
1908 (dolist (charset (get-language-info language-name 'charset))
1909 (if (memq charset utf-translate-cjk-charsets)
1910 (throw 'tag t)))
1911 nil))
1912 (utf-translate-cjk-load-tables))
1913 (run-hooks 'set-language-environment-hook)
1914 (force-mode-line-update t))
1915
1916 (defun standard-display-european-internal ()
1917 ;; Actually set up direct output of non-ASCII characters.
1918 (standard-display-8bit (if (eq window-system 'pc) 128 160) 255)
1919 ;; Unibyte Emacs on MS-DOS wants to display all 8-bit characters with
1920 ;; the native font, and codes 160 and 146 stand for something very
1921 ;; different there.
1922 (or (and (eq window-system 'pc) (not default-enable-multibyte-characters))
1923 (progn
1924 ;; Most X fonts used to do the wrong thing for latin-1 code 160.
1925 (unless (and (eq window-system 'x)
1926 ;; XFree86 4 has fixed the fonts.
1927 (string= "The XFree86 Project, Inc" (x-server-vendor))
1928 (> (aref (number-to-string (nth 2 (x-server-version))) 0)
1929 ?3))
1930 ;; Make non-line-break space display as a plain space.
1931 (aset standard-display-table 160 [32]))
1932 ;; Most Windows programs send out apostrophes as \222. Most X fonts
1933 ;; don't contain a character at that position. Map it to the ASCII
1934 ;; apostrophe. [This is actually RIGHT SINGLE QUOTATION MARK,
1935 ;; U+2019, normally from the windows-1252 character set. XFree 4
1936 ;; fonts probably have the appropriate glyph at this position,
1937 ;; so they could use standard-display-8bit. It's better to use a
1938 ;; proper windows-1252 coding system. --fx]
1939 (aset standard-display-table 146 [39]))))
1940
1941 (defun set-language-environment-coding-systems (language-name
1942 &optional eol-type)
1943 "Do various coding system setups for language environment LANGUAGE-NAME.
1944
1945 The optional arg EOL-TYPE specifies the eol-type of the default value
1946 of `buffer-file-coding-system' set by this function."
1947 (let* ((priority (get-language-info language-name 'coding-priority))
1948 (default-coding (car priority)))
1949 (if priority
1950 (let ((categories (mapcar 'coding-system-category priority)))
1951 (set-default-coding-systems
1952 (if (memq eol-type '(0 1 2 unix dos mac))
1953 (coding-system-change-eol-conversion default-coding eol-type)
1954 default-coding))
1955 (setq default-sendmail-coding-system default-coding)
1956 (set-coding-priority categories)
1957 (while priority
1958 (set (car categories) (car priority))
1959 (setq priority (cdr priority) categories (cdr categories)))
1960 ;; Changing the binding of a coding category requires this call.
1961 (update-coding-systems-internal)))))
1962
1963 (defsubst princ-list (&rest args)
1964 "Print all arguments with `princ', then print \"\n\"."
1965 (while args (princ (car args)) (setq args (cdr args)))
1966 (princ "\n"))
1967
1968 (put 'describe-specified-language-support 'apropos-inhibit t)
1969
1970 ;; Print language-specific information such as input methods,
1971 ;; charsets, and coding systems. This function is intended to be
1972 ;; called from the menu:
1973 ;; [menu-bar mule describe-language-environment LANGUAGE]
1974 ;; and should not run it by `M-x describe-current-input-method-function'.
1975 (defun describe-specified-language-support ()
1976 "Describe how Emacs supports the specified language environment."
1977 (interactive)
1978 (let (language-name)
1979 (if (not (and (symbolp last-command-event)
1980 (or (not (eq last-command-event 'Default))
1981 (setq last-command-event 'English))
1982 (setq language-name (symbol-name last-command-event))))
1983 (error "Bogus calling sequence"))
1984 (describe-language-environment language-name)))
1985
1986 (defun describe-language-environment (language-name)
1987 "Describe how Emacs supports language environment LANGUAGE-NAME."
1988 (interactive
1989 (list (read-language-name
1990 'documentation
1991 "Describe language environment (default current choice): ")))
1992 (if (null language-name)
1993 (setq language-name current-language-environment))
1994 (if (or (null language-name)
1995 (null (get-language-info language-name 'documentation)))
1996 (error "No documentation for the specified language"))
1997 (if (symbolp language-name)
1998 (setq language-name (symbol-name language-name)))
1999 (dolist (feature (get-language-info language-name 'features))
2000 (require feature))
2001 (let ((doc (get-language-info language-name 'documentation)))
2002 (help-setup-xref (list #'describe-language-environment language-name)
2003 (interactive-p))
2004 (with-output-to-temp-buffer (help-buffer)
2005 (save-excursion
2006 (set-buffer standard-output)
2007 (insert language-name " language environment\n\n")
2008 (if (stringp doc)
2009 (insert doc "\n\n"))
2010 (condition-case nil
2011 (let ((str (eval (get-language-info language-name 'sample-text))))
2012 (if (stringp str)
2013 (insert "Sample text:\n " str "\n\n")))
2014 (error nil))
2015 (let ((input-method (get-language-info language-name 'input-method))
2016 (l (copy-sequence input-method-alist)))
2017 (insert "Input methods")
2018 (when input-method
2019 (insert " (default " input-method ")")
2020 (setq input-method (assoc input-method input-method-alist))
2021 (setq l (cons input-method (delete input-method l))))
2022 (insert ":\n")
2023 (while l
2024 (when (string= language-name (nth 1 (car l)))
2025 (insert " " (car (car l)))
2026 (search-backward (car (car l)))
2027 (help-xref-button 0 'help-input-method (car (car l)))
2028 (goto-char (point-max))
2029 (insert " (\""
2030 (if (stringp (nth 3 (car l)))
2031 (nth 3 (car l))
2032 (car (nth 3 (car l))))
2033 "\" in mode line)\n"))
2034 (setq l (cdr l)))
2035 (insert "\n"))
2036 (insert "Character sets:\n")
2037 (let ((l (get-language-info language-name 'charset)))
2038 (if (null l)
2039 (insert " nothing specific to " language-name "\n")
2040 (while l
2041 (insert " " (symbol-name (car l)))
2042 (search-backward (symbol-name (car l)))
2043 (help-xref-button 0 'help-character-set (car l))
2044 (goto-char (point-max))
2045 (insert ": " (charset-description (car l)) "\n")
2046 (setq l (cdr l)))))
2047 (insert "\n")
2048 (insert "Coding systems:\n")
2049 (let ((l (get-language-info language-name 'coding-system)))
2050 (if (null l)
2051 (insert " nothing specific to " language-name "\n")
2052 (while l
2053 (insert " " (symbol-name (car l)))
2054 (search-backward (symbol-name (car l)))
2055 (help-xref-button 0 'help-coding-system (car l))
2056 (goto-char (point-max))
2057 (insert " (`"
2058 (coding-system-mnemonic (car l))
2059 "' in mode line):\n\t"
2060 (coding-system-doc-string (car l))
2061 "\n")
2062 (let ((aliases (coding-system-get (car l)
2063 'alias-coding-systems)))
2064 (when aliases
2065 (insert "\t(alias:")
2066 (while aliases
2067 (insert " " (symbol-name (car aliases)))
2068 (setq aliases (cdr aliases)))
2069 (insert ")\n")))
2070 (setq l (cdr l)))))))))
2071 \f
2072 ;;; Locales.
2073
2074 (defvar locale-translation-file-name nil
2075 "File name for the system's file of locale-name aliases, or nil if none.")
2076
2077 ;; The following definitions might as well be marked as constants and
2078 ;; purecopied, since they're normally used on startup, and probably
2079 ;; should reflect the facilities of the base Emacs.
2080 (defconst locale-language-names
2081 (purecopy
2082 '(
2083 ;; Locale names of the form LANGUAGE[_TERRITORY][.CODESET][@MODIFIER]
2084 ;; as specified in the Single Unix Spec, Version 2.
2085 ;; LANGUAGE is a language code taken from ISO 639:1988 (E/F)
2086 ;; with additions from ISO 639/RA Newsletter No.1/1989;
2087 ;; see Internet RFC 2165 (1997-06) and
2088 ;; http://www.evertype.com/standards/iso639/iso639-en.html
2089 ;; TERRITORY is a country code taken from ISO 3166
2090 ;; http://www.din.de/gremien/nas/nabd/iso3166ma/codlstp1/en_listp1.html.
2091 ;; CODESET and MODIFIER are implementation-dependent.
2092
2093 ;; jasonr comments: MS Windows uses three letter codes for
2094 ;; languages instead of the two letter ISO codes that POSIX
2095 ;; uses. In most cases the first two letters are the same, so
2096 ;; most of the regexps in locale-language-names work. Japanese
2097 ;; and Chinese are exceptions, which are listed in the
2098 ;; non-standard section at the bottom of locale-language-names.
2099
2100 ("aa_DJ" . "Latin-1") ; Afar
2101 ("aa" . "UTF-8")
2102 ;; ab Abkhazian
2103 ("af" . "Latin-1") ; Afrikaans
2104 ("am" "Ethiopic" utf-8) ; Amharic
2105 ("an" . "Latin-9") ; Aragonese
2106 ; ar Arabic glibc uses 8859-6
2107 ; as Assamese
2108 ; ay Aymara
2109 ("az" . "UTF-8") ; Azerbaijani
2110 ; ba Bashkir
2111 ("be" "Belarusian" cp1251) ; Belarusian [Byelorussian until early 1990s]
2112 ("bg" "Bulgarian" cp1251) ; Bulgarian
2113 ; bh Bihari
2114 ; bi Bislama
2115 ("bn" . "UTF-8") ; Bengali, Bangla
2116 ("bo" . "Tibetan")
2117 ("br" . "Latin-1") ; Breton
2118 ("bs" . "Latin-2") ; Bosnian
2119 ("byn" . "UTF-8") ; Bilin; Blin
2120 ("ca" . "Latin-1") ; Catalan
2121 ; co Corsican
2122 ("cs" "Czech" iso-8859-2)
2123 ("cy" "Welsh" iso-8859-14)
2124 ("da" . "Latin-1") ; Danish
2125 ("de" "German" iso-8859-1)
2126 ; dz Bhutani
2127 ("el" "Greek" iso-8859-7)
2128 ;; Users who specify "en" explicitly typically want Latin-1, not ASCII.
2129 ;; That's actually what the GNU locales define, modulo things like
2130 ;; en_IN -- fx.
2131 ("en_IN" "English" utf-8) ; glibc uses utf-8 for English in India
2132 ("en" . "Latin-1") ; English
2133 ("eo" . "Latin-3") ; Esperanto
2134 ("es" "Spanish" iso-8859-1)
2135 ("et" . "Latin-1") ; Estonian
2136 ("eu" . "Latin-1") ; Basque
2137 ("fa" . "UTF-8") ; Persian
2138 ("fi" . "Latin-1") ; Finnish
2139 ("fj" . "Latin-1") ; Fiji
2140 ("fo" . "Latin-1") ; Faroese
2141 ("fr" "French" iso-8859-1) ; French
2142 ("fy" . "Latin-1") ; Frisian
2143 ("ga" . "Latin-1") ; Irish Gaelic (new orthography)
2144 ("gd" . "Latin-9") ; Scots Gaelic
2145 ("gez" "Ethiopic" utf-8) ; Geez
2146 ("gl" . "Latin-1") ; Gallegan; Galician
2147 ; gn Guarani
2148 ("gu" . "UTF-8") ; Gujarati
2149 ("gv" . "Latin-1") ; Manx Gaelic
2150 ; ha Hausa
2151 ("he" "Hebrew" iso-8859-8)
2152 ("hi" "Devanagari" utf-8) ; Hindi
2153 ("hr" "Croatian" iso-8859-2) ; Croatian
2154 ("hu" . "Latin-2") ; Hungarian
2155 ; hy Armenian
2156 ; ia Interlingua
2157 ("id" . "Latin-1") ; Indonesian
2158 ; ie Interlingue
2159 ; ik Inupiak
2160 ("is" . "Latin-1") ; Icelandic
2161 ("it" "Italian" iso-8859-1) ; Italian
2162 ; iu Inuktitut
2163 ("iw" "Hebrew" iso-8859-8)
2164 ("ja" "Japanese" euc-jp)
2165 ; jw Javanese
2166 ("ka" "Georgian" georgian-ps) ; Georgian
2167 ; kk Kazakh
2168 ("kl" . "Latin-1") ; Greenlandic
2169 ; km Cambodian
2170 ("kn" "Kannada" utf-8)
2171 ("ko" "Korean" euc-kr)
2172 ; ks Kashmiri
2173 ; ku Kurdish
2174 ("kw" . "Latin-1") ; Cornish
2175 ; ky Kirghiz
2176 ("la" . "Latin-1") ; Latin
2177 ("lb" . "Latin-1") ; Luxemburgish
2178 ("lg" . "Laint-6") ; Ganda
2179 ; ln Lingala
2180 ("lo" "Lao" utf-8) ; Laothian
2181 ("lt" "Lithuanian" iso-8859-13)
2182 ("lv" . "Latvian") ; Latvian, Lettish
2183 ; mg Malagasy
2184 ("mi" . "Latin-7") ; Maori
2185 ("mk" "Cyrillic-ISO" iso-8859-5) ; Macedonian
2186 ("ml" "Malayalam" utf-8)
2187 ("mn" . "UTF-8") ; Mongolian
2188 ; mo Moldavian
2189 ("mr" "Devanagari" utf-8) ; Marathi
2190 ("ms" . "Latin-1") ; Malay
2191 ("mt" . "Latin-3") ; Maltese
2192 ; my Burmese
2193 ; na Nauru
2194 ("nb" . "Latin-1") ; Norwegian
2195 ("ne" "Devanagari" utf-8) ; Nepali
2196 ("nl" "Dutch" iso-8859-1)
2197 ("no" . "Latin-1") ; Norwegian
2198 ("oc" . "Latin-1") ; Occitan
2199 ("om_ET" . "UTF-8") ; (Afan) Oromo
2200 ("om" . "Latin-1") ; (Afan) Oromo
2201 ; or Oriya
2202 ("pa" . "UTF-8") ; Punjabi
2203 ("pl" . "Latin-2") ; Polish
2204 ; ps Pashto, Pushto
2205 ("pt" . "Latin-1") ; Portuguese
2206 ; qu Quechua
2207 ("rm" . "Latin-1") ; Rhaeto-Romanic
2208 ; rn Kirundi
2209 ("ro" "Romanian" iso-8859-2)
2210 ("ru_RU" "Russian" iso-8859-5)
2211 ("ru_UA" "Russian" koi8-u)
2212 ; rw Kinyarwanda
2213 ("sa" . "Devanagari") ; Sanskrit
2214 ; sd Sindhi
2215 ("se" . "UTF-8") ; Northern Sami
2216 ; sg Sangho
2217 ("sh" . "Latin-2") ; Serbo-Croatian
2218 ; si Sinhalese
2219 ("sid" . "UTF-8") ; Sidamo
2220 ("sk" "Slovak" iso-8859-2)
2221 ("sl" "Slovenian" iso-8859-2)
2222 ; sm Samoan
2223 ; sn Shona
2224 ("so_ET" "UTF-8") ; Somali
2225 ("so" "Latin-1") ; Somali
2226 ("sq" . "Latin-1") ; Albanian
2227 ("sr_YU@cyrillic" . "Cyrillic-ISO") ; Serbian (Cyrillic alphabet)
2228 ("sr" . "Latin-2") ; Serbian (Latin alphabet)
2229 ; ss Siswati
2230 ("st" . "Latin-1") ; Sesotho
2231 ; su Sundanese
2232 ("sv" "Swedish" iso-8859-1) ; Swedish
2233 ("sw" . "Latin-1") ; Swahili
2234 ("ta" "Tamil" utf-8)
2235 ("te" . "UTF-8") ; Telugu
2236 ("tg" "Tajik" koi8-t)
2237 ("th" "Thai" tis-620)
2238 ("ti" "Ethiopic" utf-8) ; Tigrinya
2239 ("tig_ER" . "UTF-8") ; Tigre
2240 ; tk Turkmen
2241 ("tl" . "Latin-1") ; Tagalog
2242 ; tn Setswana
2243 ; to Tonga
2244 ("tr" "Turkish" iso-8859-9)
2245 ; ts Tsonga
2246 ("tt" . "UTF-8") ; Tatar
2247 ; tw Twi
2248 ; ug Uighur
2249 ("uk" "Ukrainian" koi8-u)
2250 ("ur" . "UTF-8") ; Urdu
2251 ("uz_UZ@cyrillic" . "UTF-8"); Uzbek
2252 ("uz" . "Latin-1") ; Uzbek
2253 ("vi" "Vietnamese" utf-8)
2254 ; vo Volapuk
2255 ("wa" . "Latin-1") ; Walloon
2256 ; wo Wolof
2257 ("xh" . "Latin-1") ; Xhosa
2258 ("yi" . "Windows-1255") ; Yiddish
2259 ; yo Yoruba
2260 ; za Zhuang
2261 ("zh_HK" . "Chinese-Big5")
2262 ("zh_TW" . "Chinese-Big5")
2263 ("zh_CN" . "Chinese-GB")
2264 ("zh" . "Chinese-GB")
2265 ; zh_CN.GB18030/GB18030 \
2266 ; zh_CN.GBK/GBK \
2267 ; zh_HK/BIG5-HKSCS \
2268 ("zu" . "Latin-1") ; Zulu
2269
2270 ;; ISO standard locales
2271 ("c$" . "ASCII")
2272 ("posix$" . "ASCII")
2273
2274 ;; The "IPA" Emacs language environment does not correspond
2275 ;; to any ISO 639 code, so let it stand for itself.
2276 ("ipa$" . "IPA")
2277
2278 ;; Nonstandard or obsolete language codes
2279 ("cz" . "Czech") ; e.g. Solaris 2.6
2280 ("ee" . "Latin-4") ; Estonian, e.g. X11R6.4
2281 ("iw" . "Hebrew") ; e.g. X11R6.4
2282 ("sp" . "Cyrillic-ISO") ; Serbian (Cyrillic alphabet), e.g. X11R6.4
2283 ("su" . "Latin-1") ; Finnish, e.g. Solaris 2.6
2284 ("jp" . "Japanese") ; e.g. MS Windows
2285 ("chs" . "Chinese-GB") ; MS Windows Chinese Simplified
2286 ("cht" . "Chinese-BIG5") ; MS Windows Chinese Traditional
2287 ))
2288 "Alist of locale regexps vs the corresponding languages and coding systems.
2289 Each element has these form:
2290 \(LOCALE-REGEXP LANG-ENV CODING-SYSTEM)
2291 The first element whose LOCALE-REGEXP matches the start of a
2292 downcased locale specifies the LANG-ENV \(language environtment)
2293 and CODING-SYSTEM corresponding to that locale. If there is no
2294 appropriate language environment, the element may have this form:
2295 \(LOCALE-REGEXP . LANG-ENV)
2296 In this case, LANG-ENV is one of generic language environments for an
2297 specific encoding such as \"Latin-1\" and \"UTF-8\".")
2298
2299 (defconst locale-charset-language-names
2300 (purecopy
2301 '((".*8859[-_]?1\\>" . "Latin-1")
2302 (".*8859[-_]?2\\>" . "Latin-2")
2303 (".*8859[-_]?3\\>" . "Latin-3")
2304 (".*8859[-_]?4\\>" . "Latin-4")
2305 (".*8859[-_]?9\\>" . "Latin-5")
2306 (".*8859[-_]?14\\>" . "Latin-8")
2307 (".*8859[-_]?15\\>" . "Latin-9")
2308 (".*utf\\(?:-?8\\)?\\>" . "UTF-8")
2309 ;; utf-8@euro exists, so put this last. (@euro really specifies
2310 ;; the currency, rather than the charset.)
2311 (".*@euro\\>" . "Latin-9")))
2312 "List of pairs of locale regexps and charset language names.
2313 The first element whose locale regexp matches the start of a downcased locale
2314 specifies the language name whose charset corresponds to that locale.
2315 This language name is used if the locale is not listed in
2316 `locale-language-names'")
2317
2318 (defconst locale-preferred-coding-systems
2319 (purecopy
2320 '((".*8859[-_]?1\\>" . iso-8859-1)
2321 (".*8859[-_]?2\\>" . iso-8859-2)
2322 (".*8859[-_]?3\\>" . iso-8859-3)
2323 (".*8859[-_]?4\\>" . iso-8859-4)
2324 (".*8859[-_]?9\\>" . iso-8859-9)
2325 (".*8859[-_]?14\\>" . iso-8859-14)
2326 (".*8859[-_]?15\\>" . iso-8859-15)
2327 (".*utf\\(?:-?8\\)?" . utf-8)
2328 ;; utf-8@euro exists, so put this after utf-8. (@euro really
2329 ;; specifies the currency, rather than the charset.)
2330 (".*@euro" . iso-8859-15)
2331 ("koi8-?r" . koi8-r)
2332 ("koi8-?u" . koi8-u)
2333 ("tcvn" . tcvn)
2334 ("big5" . big5)
2335 ("euc-?tw" . euc-tw)
2336 ;; We don't support GBK, but as it is upper compatible with
2337 ;; GB-2312, we setup the default coding system to gb2312.
2338 ("gbk" . gb2312)
2339 ;; We don't support BIG5-HKSCS, but as it is upper compatible with
2340 ;; BIG5, we setup the default coding system to big5.
2341 ("big5hkscs" . big5)
2342 ("ja.*[._]euc" . japanese-iso-8bit)
2343 ("ja.*[._]jis7" . iso-2022-jp)
2344 ("ja.*[._]pck" . japanese-shift-jis)
2345 ("ja.*[._]sjis" . japanese-shift-jis)
2346 ("jpn" . japanese-shift-jis) ; MS-Windows uses this.
2347 ))
2348 "List of pairs of locale regexps and preferred coding systems.
2349 The first element whose locale regexp matches the start of a downcased locale
2350 specifies the coding system to prefer when using that locale.
2351 This coding system is used if the locale specifies a specific charset.")
2352
2353 (defun locale-name-match (key alist)
2354 "Search for KEY in ALIST, which should be a list of regexp-value pairs.
2355 Return the value corresponding to the first regexp that matches the
2356 start of KEY, or nil if there is no match."
2357 (let (element)
2358 (while (and alist (not element))
2359 (if (string-match (concat "\\`\\(?:" (car (car alist)) "\\)") key)
2360 (setq element (car alist)))
2361 (setq alist (cdr alist)))
2362 (cdr element)))
2363
2364 (defun locale-charset-match-p (charset1 charset2)
2365 "Whether charset names (strings) CHARSET1 and CHARSET2 are equivalent.
2366 Matching is done ignoring case and any hyphens and underscores in the
2367 names. E.g. `ISO_8859-1' and `iso88591' both match `iso-8859-1'."
2368 (setq charset1 (replace-regexp-in-string "[-_]" "" charset1))
2369 (setq charset2 (replace-regexp-in-string "[-_]" "" charset2))
2370 (eq t (compare-strings charset1 nil nil charset2 nil nil t)))
2371
2372 (defvar locale-charset-alist nil
2373 "Coding system alist keyed on locale-style charset name.
2374 Used by `locale-charset-to-coding-system'.")
2375
2376 (defun locale-charset-to-coding-system (charset)
2377 "Find coding system corresponding to CHARSET.
2378 CHARSET is any sort of non-Emacs charset name, such as might be used
2379 in a locale codeset, or elsewhere. It is matched to a coding system
2380 first by case-insensitive lookup in `locale-charset-alist'. Then
2381 matches are looked for in the coding system list, treating case and
2382 the characters `-' and `_' as insignificant. The coding system base
2383 is returned. Thus, for instance, if charset \"ISO8859-2\",
2384 `iso-latin-2' is returned."
2385 (or (car (assoc-string charset locale-charset-alist t))
2386 (let ((cs coding-system-alist)
2387 c)
2388 (while (and (not c) cs)
2389 (if (locale-charset-match-p charset (caar cs))
2390 (setq c (intern (caar cs)))
2391 (pop cs)))
2392 (if c (coding-system-base c)))))
2393
2394 ;; Fixme: This ought to deal with the territory part of the locale
2395 ;; too, for setting things such as calendar holidays, ps-print paper
2396 ;; size, spelling dictionary.
2397
2398 (defun set-locale-environment (&optional locale-name display)
2399 "Set up multi-lingual environment for using LOCALE-NAME.
2400 This sets the language environment, the coding system priority,
2401 the default input method and sometimes other things.
2402
2403 LOCALE-NAME should be a string which is the name of a locale supported
2404 by the system. Often it is of the form xx_XX.CODE, where xx is a
2405 language, XX is a country, and CODE specifies a character set and
2406 coding system. For example, the locale name \"ja_JP.EUC\" might name
2407 a locale for Japanese in Japan using the `japanese-iso-8bit'
2408 coding-system. The name may also have a modifier suffix, e.g. `@euro'
2409 or `@cyrillic'.
2410
2411 If LOCALE-NAME is nil, its value is taken from the environment
2412 variables LC_ALL, LC_CTYPE and LANG (the first one that is set).
2413
2414 The locale names supported by your system can typically be found in a
2415 directory named `/usr/share/locale' or `/usr/lib/locale'. LOCALE-NAME
2416 will be translated according to the table specified by
2417 `locale-translation-file-name'.
2418
2419 If DISPLAY is non-nil, only set the keyboard coding system and
2420 the terminal coding system for the given display, and don't touch
2421 session-global parameters like the language environment. DISPLAY
2422 may be a display id or a frame.
2423
2424 See also `locale-charset-language-names', `locale-language-names',
2425 `locale-preferred-coding-systems' and `locale-coding-system'."
2426 (interactive "sSet environment for locale: ")
2427
2428 ;; Do this at runtime for the sake of binaries possibly transported
2429 ;; to a system without X.
2430 (setq locale-translation-file-name
2431 (let ((files
2432 '("/usr/lib/X11/locale/locale.alias" ; e.g. X11R6.4
2433 "/usr/X11R6/lib/X11/locale/locale.alias" ; XFree86, e.g. RedHat 4.2
2434 "/usr/openwin/lib/locale/locale.alias" ; e.g. Solaris 2.6
2435 ;;
2436 ;; The following name appears after the X-related names above,
2437 ;; since the X-related names are what X actually uses.
2438 "/usr/share/locale/locale.alias" ; GNU/Linux sans X
2439 )))
2440 (while (and files (not (file-exists-p (car files))))
2441 (setq files (cdr files)))
2442 (car files)))
2443
2444 (let ((locale locale-name))
2445
2446 (unless locale
2447 ;; Use the first of these three environment variables
2448 ;; that has a nonempty value.
2449 (let ((vars '("LC_ALL" "LC_CTYPE" "LANG")))
2450 (while (and vars
2451 (= 0 (length locale))) ; nil or empty string
2452 (setq locale (terminal-getenv (pop vars))))))
2453
2454 (unless locale
2455 ;; The two tests are kept separate so the byte-compiler sees
2456 ;; that mac-get-preference is only called after checking its existence.
2457 (when (fboundp 'mac-get-preference)
2458 (setq locale (mac-get-preference "AppleLocale"))
2459 (unless locale
2460 (let ((languages (mac-get-preference "AppleLanguages")))
2461 (unless (= (length languages) 0) ; nil or empty vector
2462 (setq locale (aref languages 0)))))))
2463 (unless (or locale (not (boundp 'mac-system-locale)))
2464 (setq locale mac-system-locale))
2465
2466 (when locale
2467
2468 ;; Translate "swedish" into "sv_SE.ISO8859-1", and so on,
2469 ;; using the translation file that many systems have.
2470 (when locale-translation-file-name
2471 (with-temp-buffer
2472 (insert-file-contents locale-translation-file-name)
2473 (when (re-search-forward
2474 (concat "^" (regexp-quote locale) ":?[ \t]+") nil t)
2475 (setq locale (buffer-substring (point) (line-end-position))))))
2476
2477 ;; Leave the system locales alone if the caller did not specify
2478 ;; an explicit locale name, as their defaults are set from
2479 ;; LC_MESSAGES and LC_TIME, not LC_CTYPE, and the user might not
2480 ;; want to set them to the same value as LC_CTYPE.
2481 (when locale-name
2482 (setq system-messages-locale locale)
2483 (setq system-time-locale locale))
2484
2485 (setq locale (downcase locale))
2486
2487 (let ((language-name
2488 (locale-name-match locale locale-language-names))
2489 (charset-language-name
2490 (locale-name-match locale locale-charset-language-names))
2491 (coding-system
2492 (or (locale-name-match locale locale-preferred-coding-systems)
2493 (when locale
2494 (if (string-match "\\.\\([^@]+\\)" locale)
2495 (locale-charset-to-coding-system
2496 (match-string 1 locale))))
2497 (and (eq system-type 'macos) mac-system-coding-system))))
2498
2499 (if (consp language-name)
2500 ;; locale-language-names specify both lang-env and coding.
2501 ;; But, what specified in locale-preferred-coding-systems
2502 ;; has higher priority.
2503 (setq coding-system (or coding-system
2504 (nth 1 language-name))
2505 language-name (car language-name))
2506 ;; Otherwise, if locale is not listed in locale-language-names,
2507 ;; use what listed in locale-charset-language-names.
2508 (if (not language-name)
2509 (setq language-name charset-language-name)))
2510
2511 (when language-name
2512
2513 ;; Set up for this character set. This is now the right way
2514 ;; to do it for both unibyte and multibyte modes.
2515 (unless display
2516 (set-language-environment language-name))
2517
2518 ;; If default-enable-multibyte-characters is nil,
2519 ;; we are using single-byte characters,
2520 ;; so the display table and terminal coding system are irrelevant.
2521 (when default-enable-multibyte-characters
2522 (set-display-table-and-terminal-coding-system
2523 language-name coding-system display))
2524
2525 ;; Set the `keyboard-coding-system' if appropriate (tty
2526 ;; only). At least X and MS Windows can generate
2527 ;; multilingual input.
2528 ;; XXX This was disabled unless `window-system', but that
2529 ;; leads to buggy behaviour when a tty frame is opened
2530 ;; later. Setting the keyboard coding system has no adverse
2531 ;; effect on X, so let's do it anyway. -- Lorentey
2532 (let ((kcs (or coding-system
2533 (car (get-language-info language-name
2534 'coding-system)))))
2535 (if kcs (set-keyboard-coding-system kcs display)))
2536
2537 (unless display
2538 (setq locale-coding-system
2539 (car (get-language-info language-name 'coding-priority)))))
2540
2541 (when (and (not display)
2542 coding-system
2543 (not (coding-system-equal coding-system
2544 locale-coding-system)))
2545 (prefer-coding-system coding-system)
2546 (setq locale-coding-system coding-system))))
2547
2548 ;; On Windows, override locale-coding-system,
2549 ;; keyboard-coding-system with system codepage. Note:
2550 ;; selection-coding-system is already set in w32select.c.
2551 (when (boundp 'w32-ansi-code-page)
2552 (let ((code-page-coding (intern (format "cp%d" w32-ansi-code-page))))
2553 (when (coding-system-p code-page-coding)
2554 (unless display (setq locale-coding-system code-page-coding))
2555 (set-keyboard-coding-system code-page-coding display)
2556 (set-terminal-coding-system code-page-coding display))))
2557
2558 (when (eq system-type 'darwin)
2559 ;; On Darwin, file names are always encoded in utf-8, no matter
2560 ;; the locale.
2561 (setq default-file-name-coding-system 'utf-8)
2562 ;; Mac OS X's Terminal.app by default uses utf-8 regardless of
2563 ;; the locale.
2564 (when (and (null window-system)
2565 (equal (terminal-getenv "TERM_PROGRAM") "Apple_Terminal"))
2566 (set-terminal-coding-system 'utf-8)
2567 (set-keyboard-coding-system 'utf-8)))
2568
2569 ;; Default to A4 paper if we're not in a C, POSIX or US locale.
2570 ;; (See comments in Flocale_info.)
2571 (unless display
2572 (let ((locale locale)
2573 (paper (locale-info 'paper)))
2574 (if paper
2575 ;; This will always be null at the time of writing.
2576 (cond
2577 ((equal paper '(216 279))
2578 (setq ps-paper-type 'letter))
2579 ((equal paper '(210 297))
2580 (setq ps-paper-type 'a4)))
2581 (let ((vars '("LC_ALL" "LC_PAPER" "LANG")))
2582 (while (and vars (= 0 (length locale)))
2583 (setq locale (terminal-getenv (pop vars)))))
2584 (when locale
2585 ;; As of glibc 2.2.5, these are the only US Letter locales,
2586 ;; and the rest are A4.
2587 (setq ps-paper-type
2588 (or (locale-name-match locale '(("c$" . letter)
2589 ("posix$" . letter)
2590 (".._us" . letter)
2591 (".._pr" . letter)
2592 (".._ca" . letter)
2593 ("enu$" . letter) ; Windows
2594 ("esu$" . letter)
2595 ("enc$" . letter)
2596 ("frc$" . letter)))
2597 'a4)))))))
2598 nil)
2599 \f
2600 ;;; Charset property
2601
2602 (defun get-charset-property (charset propname)
2603 "Return the value of CHARSET's PROPNAME property.
2604 This is the last value stored with
2605 (put-charset-property CHARSET PROPNAME VALUE)."
2606 (and (not (eq charset 'composition))
2607 (plist-get (charset-plist charset) propname)))
2608
2609 (defun put-charset-property (charset propname value)
2610 "Store CHARSETS's PROPNAME property with value VALUE.
2611 It can be retrieved with `(get-charset-property CHARSET PROPNAME)'."
2612 (or (eq charset 'composition)
2613 (set-charset-plist charset
2614 (plist-put (charset-plist charset) propname value))))
2615
2616 ;;; Character code property
2617 (put 'char-code-property-table 'char-table-extra-slots 0)
2618
2619 (defvar char-code-property-table
2620 (make-char-table 'char-code-property-table)
2621 "Char-table containing a property list of each character code.
2622
2623 See also the documentation of `get-char-code-property' and
2624 `put-char-code-property'.")
2625
2626 (defun get-char-code-property (char propname)
2627 "Return the value of CHAR's PROPNAME property in `char-code-property-table'."
2628 (let ((plist (aref char-code-property-table char)))
2629 (if (listp plist)
2630 (car (cdr (memq propname plist))))))
2631
2632 (defun put-char-code-property (char propname value)
2633 "Store CHAR's PROPNAME property with VALUE in `char-code-property-table'.
2634 It can be retrieved with `(get-char-code-property CHAR PROPNAME)'."
2635 (let ((plist (aref char-code-property-table char)))
2636 (if plist
2637 (let ((slot (memq propname plist)))
2638 (if slot
2639 (setcar (cdr slot) value)
2640 (nconc plist (list propname value))))
2641 (aset char-code-property-table char (list propname value)))))
2642
2643 \f
2644 ;; Pretty description of encoded string
2645
2646 ;; Alist of ISO 2022 control code vs the corresponding mnemonic string.
2647 (defvar iso-2022-control-alist
2648 '((?\x1b . "ESC")
2649 (?\x0e . "SO")
2650 (?\x0f . "SI")
2651 (?\x8e . "SS2")
2652 (?\x8f . "SS3")
2653 (?\x9b . "CSI")))
2654
2655 (defun encoded-string-description (str coding-system)
2656 "Return a pretty description of STR that is encoded by CODING-SYSTEM."
2657 (setq str (string-as-unibyte str))
2658 (mapconcat
2659 (if (and coding-system (eq (coding-system-type coding-system) 2))
2660 ;; Try to get a pretty description for ISO 2022 escape sequences.
2661 (function (lambda (x) (or (cdr (assq x iso-2022-control-alist))
2662 (format "0x%02X" x))))
2663 (function (lambda (x) (format "0x%02X" x))))
2664 str " "))
2665
2666 (defun encode-coding-char (char coding-system)
2667 "Encode CHAR by CODING-SYSTEM and return the resulting string.
2668 If CODING-SYSTEM can't safely encode CHAR, return nil."
2669 (let ((str1 (string-as-multibyte (char-to-string char)))
2670 (str2 (string-as-multibyte (make-string 2 char)))
2671 (safe-chars (and coding-system
2672 (coding-system-get coding-system 'safe-chars)))
2673 (charset (char-charset char))
2674 enc1 enc2 i1 i2)
2675 (when (or (eq safe-chars t)
2676 (eq charset 'ascii)
2677 (and safe-chars (aref safe-chars char)))
2678 ;; We must find the encoded string of CHAR. But, just encoding
2679 ;; CHAR will put extra control sequences (usually to designate
2680 ;; ASCII charset) at the tail if type of CODING is ISO 2022.
2681 ;; To exclude such tailing bytes, we at first encode one-char
2682 ;; string and two-char string, then check how many bytes at the
2683 ;; tail of both encoded strings are the same.
2684
2685 (setq enc1 (encode-coding-string str1 coding-system)
2686 i1 (length enc1)
2687 enc2 (encode-coding-string str2 coding-system)
2688 i2 (length enc2))
2689 (while (and (> i1 0) (= (aref enc1 (1- i1)) (aref enc2 (1- i2))))
2690 (setq i1 (1- i1) i2 (1- i2)))
2691
2692 ;; Now (substring enc1 i1) and (substring enc2 i2) are the same,
2693 ;; and they are the extra control sequences at the tail to
2694 ;; exclude.
2695 (substring enc2 0 i2))))
2696
2697
2698 ;; arch-tag: b382c432-4b36-460e-bf4c-05efd0bb18dc
2699 ;;; mule-cmds.el ends here