]> code.delx.au - gnu-emacs/blob - lisp/international/mule-cmds.el
2002-02-12 Per Abrahamsen <abraham@dina.kvl.dk>
[gnu-emacs] / lisp / international / mule-cmds.el
1 ;;; mule-cmds.el --- commands for mulitilingual environment
2
3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
4 ;; Licensed to the Free Software Foundation.
5 ;; Copyright (C) 2000, 2001 Free Software Foundation, Inc.
6
7 ;; Keywords: mule, multilingual
8
9 ;; This file is part of GNU Emacs.
10
11 ;; GNU Emacs is free software; you can redistribute it and/or modify
12 ;; it under the terms of the GNU General Public License as published by
13 ;; the Free Software Foundation; either version 2, or (at your option)
14 ;; any later version.
15
16 ;; GNU Emacs is distributed in the hope that it will be useful,
17 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 ;; GNU General Public License for more details.
20
21 ;; You should have received a copy of the GNU General Public License
22 ;; along with GNU Emacs; see the file COPYING. If not, write to the
23 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
24 ;; Boston, MA 02111-1307, USA.
25
26 ;;; Commentary:
27
28 ;;; Code:
29
30 (eval-when-compile (defvar dos-codepage))
31
32 ;;; MULE related key bindings and menus.
33
34 (defvar mule-keymap (make-sparse-keymap)
35 "Keymap for Mule (Multilingual environment) specific commands.")
36
37 ;; Keep "C-x C-m ..." for mule specific commands.
38 (define-key ctl-x-map "\C-m" mule-keymap)
39
40 (define-key mule-keymap "f" 'set-buffer-file-coding-system)
41 (define-key mule-keymap "t" 'set-terminal-coding-system)
42 (define-key mule-keymap "k" 'set-keyboard-coding-system)
43 (define-key mule-keymap "p" 'set-buffer-process-coding-system)
44 (define-key mule-keymap "x" 'set-selection-coding-system)
45 (define-key mule-keymap "X" 'set-next-selection-coding-system)
46 (define-key mule-keymap "\C-\\" 'set-input-method)
47 (define-key mule-keymap "c" 'universal-coding-system-argument)
48 (define-key mule-keymap "l" 'set-language-environment)
49
50 (defvar mule-menu-keymap
51 (make-sparse-keymap "Mule (Multilingual Environment)")
52 "Keymap for Mule (Multilingual environment) menu specific commands.")
53
54 (defvar describe-language-environment-map
55 (make-sparse-keymap "Describe Language Environment"))
56
57 (defvar setup-language-environment-map
58 (make-sparse-keymap "Set Language Environment"))
59
60 (defvar set-coding-system-map
61 (make-sparse-keymap "Set Coding System"))
62
63 (define-key-after mule-menu-keymap [set-language-environment]
64 (list 'menu-item "Set Language Environment" setup-language-environment-map
65 :help "Multilingual environment suitable for a specific language"))
66 (define-key-after mule-menu-keymap [mouse-set-font]
67 '(menu-item "Set Font/Fontset" mouse-set-font
68 :visible (fboundp 'generate-fontset-menu)
69 :help "Select a font from list of known fonts/fontsets"))
70 (define-key-after mule-menu-keymap [separator-mule]
71 '("--")
72 t)
73 (define-key-after mule-menu-keymap [toggle-input-method]
74 '(menu-item "Toggle Input Method" toggle-input-method)
75 t)
76 (define-key-after mule-menu-keymap [set-input-method]
77 '(menu-item "Select Input Method..." set-input-method)
78 t)
79 (define-key-after mule-menu-keymap [describe-input-method]
80 '(menu-item "Describe Input Method" describe-input-method))
81 (define-key-after mule-menu-keymap [separator-input-method]
82 '("--")
83 t)
84 (define-key-after mule-menu-keymap [set-various-coding-system]
85 (list 'menu-item "Set Coding Systems" set-coding-system-map
86 :enable 'enable-multibyte-characters))
87 (define-key-after mule-menu-keymap [view-hello-file]
88 '(menu-item "Show Multi-lingual Text" view-hello-file
89 :enable (file-readable-p
90 (expand-file-name "HELLO" data-directory))
91 :help "Display file which says HELLO in many languages")
92 t)
93 (define-key-after mule-menu-keymap [separator-coding-system]
94 '("--")
95 t)
96 (define-key-after mule-menu-keymap [describe-language-environment]
97 (list 'menu-item "Describe Language Environment"
98 describe-language-environment-map
99 :help "Show multilingual settings for a specific language")
100 t)
101 (define-key-after mule-menu-keymap [describe-input-method]
102 '(menu-item "Describe Input Method..." describe-input-method
103 :help "Keyboard layout for a specific input method")
104 t)
105 (define-key-after mule-menu-keymap [describe-coding-system]
106 '(menu-item "Describe Coding System..." describe-coding-system)
107 t)
108 (define-key-after mule-menu-keymap [list-character-sets]
109 '(menu-item "List Character Sets" list-character-sets
110 :help "Show table of available character sets"))
111 (define-key-after mule-menu-keymap [mule-diag]
112 '(menu-item "Show All of Mule Status" mule-diag
113 :help "Display multilingual environment settings")
114 t)
115
116 (define-key-after set-coding-system-map [set-buffer-file-coding-system]
117 '(menu-item "For Saving this Buffer" set-buffer-file-coding-system
118 :help "How to encode this buffer on disk")
119 t)
120 (define-key-after set-coding-system-map [universal-coding-system-argument]
121 '(menu-item "For Next Command" universal-coding-system-argument
122 :help "Coding system to be used by next command")
123 t)
124 (define-key-after set-coding-system-map [set-terminal-coding-system]
125 '(menu-item "For Terminal" set-terminal-coding-system
126 :enable (null (memq window-system '(x w32 mac)))
127 :help "How to encode terminal output")
128 t)
129 (define-key-after set-coding-system-map [set-keyboard-coding-system]
130 '(menu-item "For Keyboard" set-keyboard-coding-system
131 :help "How to decode keyboard input")
132 t)
133 (define-key-after set-coding-system-map [set-buffer-process-coding-system]
134 '(menu-item "For I/O with Subprocess" set-buffer-process-coding-system
135 :visible (fboundp 'start-process)
136 :enable (get-buffer-process (current-buffer))
137 :help "How to en/decode I/O from/to subprocess connected to this buffer")
138 t)
139 (define-key-after set-coding-system-map [set-selection-coding-system]
140 '(menu-item "For X Selections/Clipboard" set-selection-coding-system
141 :visible (display-selections-p)
142 :help "How to en/decode data to/from selection/clipboard")
143 t)
144 (define-key-after set-coding-system-map [set-next-selection-coding-system]
145 '(menu-item "For Next X Selection" set-next-selection-coding-system
146 :visible (display-selections-p)
147 :help "How to en/decode next selection/clipboard operation")
148 t)
149 (define-key setup-language-environment-map
150 [Default] '(menu-item "Default" setup-specified-language-environment))
151
152 (define-key describe-language-environment-map
153 [Default] '(menu-item "Default" describe-specified-language-support))
154
155 ;; This should be a single character key binding because users use it
156 ;; very frequently while editing multilingual text. Now we can use
157 ;; only two such keys: "\C-\\" and "\C-^", but the latter is not
158 ;; convenient because it requires shifting on most keyboards. An
159 ;; alternative is "\C-\]" which is now bound to `abort-recursive-edit'
160 ;; but it won't be used that frequently.
161 (define-key global-map "\C-\\" 'toggle-input-method)
162
163 ;; This is no good because people often type Shift-SPC
164 ;; meaning to type SPC. -- rms.
165 ;; ;; Here's an alternative key binding for X users (Shift-SPACE).
166 ;; (define-key global-map [?\S- ] 'toggle-input-method)
167
168 ;;; Mule related hyperlinks.
169 (defconst help-xref-mule-regexp-template
170 (purecopy (concat "\\(\\<\\("
171 "\\(coding system\\)\\|"
172 "\\(input method\\)\\|"
173 "\\(character set\\)\\|"
174 "\\(charset\\)"
175 "\\)\\s-+\\)?"
176 ;; Note starting with word-syntax character:
177 "`\\(\\sw\\(\\sw\\|\\s_\\)+\\)'")))
178
179 (defun coding-system-change-eol-conversion (coding-system eol-type)
180 "Return a coding system which differs from CODING-SYSTEM in eol conversion.
181 The returned coding system converts end-of-line by EOL-TYPE
182 but text as the same way as CODING-SYSTEM.
183 EOL-TYPE should be `unix', `dos', `mac', or nil.
184 If EOL-TYPE is nil, the returned coding system detects
185 how end-of-line is formatted automatically while decoding.
186
187 EOL-TYPE can be specified by an integer 0, 1, or 2.
188 They means `unix', `dos', and `mac' respectively."
189 (if (symbolp eol-type)
190 (setq eol-type (cond ((eq eol-type 'unix) 0)
191 ((eq eol-type 'dos) 1)
192 ((eq eol-type 'mac) 2)
193 (t eol-type))))
194 (let ((orig-eol-type (coding-system-eol-type coding-system)))
195 (if (vectorp orig-eol-type)
196 (if (not eol-type)
197 coding-system
198 (aref orig-eol-type eol-type))
199 (let ((base (coding-system-base coding-system)))
200 (if (not eol-type)
201 base
202 (if (= eol-type orig-eol-type)
203 coding-system
204 (setq orig-eol-type (coding-system-eol-type base))
205 (if (vectorp orig-eol-type)
206 (aref orig-eol-type eol-type))))))))
207
208 (defun coding-system-change-text-conversion (coding-system coding)
209 "Return a coding system which differs from CODING-SYSTEM in text conversion.
210 The returned coding system converts text by CODING
211 but end-of-line as the same way as CODING-SYSTEM.
212 If CODING is nil, the returned coding system detects
213 how text is formatted automatically while decoding."
214 (if (not coding)
215 (coding-system-base coding-system)
216 (let ((eol-type (coding-system-eol-type coding-system)))
217 (coding-system-change-eol-conversion
218 coding
219 (if (numberp eol-type) (aref [unix dos mac] eol-type))))))
220
221 (defun toggle-enable-multibyte-characters (&optional arg)
222 "Change whether this buffer uses multibyte characters.
223 With arg, use multibyte characters if the arg is positive.
224
225 Note that this command does not convert the byte contents of
226 the buffer; it only changes the way those bytes are interpreted.
227 In general, therefore, this command *changes* the sequence of
228 characters that the current buffer contains.
229
230 We suggest you avoid using use this command unless you know what you
231 are doing. If you use it by mistake, and the buffer is now displayed
232 wrong, use this command again to toggle back to the right mode."
233 (interactive "P")
234 (let ((new-flag
235 (if (null arg) (null enable-multibyte-characters)
236 (> (prefix-numeric-value arg) 0))))
237 (set-buffer-multibyte new-flag))
238 (force-mode-line-update))
239
240 (defun view-hello-file ()
241 "Display the HELLO file which list up many languages and characters."
242 (interactive)
243 ;; We have to decode the file in any environment.
244 (let ((default-enable-multibyte-characters t)
245 (coding-system-for-read 'iso-2022-7bit))
246 (find-file-read-only (expand-file-name "HELLO" data-directory))))
247
248 (defun universal-coding-system-argument ()
249 "Execute an I/O command using the specified coding system."
250 (interactive)
251 (let* ((default (and buffer-file-coding-system
252 (not (eq (coding-system-type buffer-file-coding-system)
253 t))
254 buffer-file-coding-system))
255 (coding-system (read-coding-system
256 (if default
257 (format "Coding system for following command (default, %s): " default)
258 "Coding system for following command: ")
259 default))
260 (keyseq (read-key-sequence
261 (format "Command to execute with %s:" coding-system)))
262 (cmd (key-binding keyseq))
263 prefix)
264
265 (when (eq cmd 'universal-argument)
266 (call-interactively cmd)
267
268 ;; Process keys bound in `universal-argument-map'.
269 (while (progn
270 (setq keyseq (read-key-sequence nil t)
271 cmd (key-binding keyseq t))
272 (not (eq cmd 'universal-argument-other-key)))
273 (let ((current-prefix-arg prefix-arg)
274 ;; Have to bind `last-command-char' here so that
275 ;; `digit-argument', for isntance, can compute the
276 ;; prefix arg.
277 (last-command-char (aref keyseq 0)))
278 (call-interactively cmd)))
279
280 ;; This is the final call to `univeral-argument-other-key', which
281 ;; set's the final `prefix-arg.
282 (let ((current-prefix-arg prefix-arg))
283 (call-interactively cmd))
284
285 ;; Read the command to execute with the given prefix arg.
286 (setq prefix prefix-arg
287 keyseq (read-key-sequence nil t)
288 cmd (key-binding keyseq)))
289
290 (let ((coding-system-for-read coding-system)
291 (coding-system-for-write coding-system)
292 (current-prefix-arg prefix))
293 (message "")
294 (call-interactively cmd))))
295
296 (defun set-default-coding-systems (coding-system)
297 "Set default value of various coding systems to CODING-SYSTEM.
298 This sets the following coding systems:
299 o coding system of a newly created buffer
300 o default coding system for subprocess I/O
301 This also sets the following values:
302 o default value used as file-name-coding-system for converting file names.
303 o default value for the command `set-terminal-coding-system' (not on MSDOS)
304 o default value for the command `set-keyboard-coding-system'."
305 (check-coding-system coding-system)
306 (setq-default buffer-file-coding-system coding-system)
307 (if default-enable-multibyte-characters
308 (setq default-file-name-coding-system coding-system))
309 ;; If coding-system is nil, honor that on MS-DOS as well, so
310 ;; that they could reset the terminal coding system.
311 (unless (and (eq window-system 'pc) coding-system)
312 (setq default-terminal-coding-system coding-system))
313 (setq default-keyboard-coding-system coding-system)
314 (setq default-process-coding-system (cons coding-system coding-system)))
315
316 (defalias 'update-iso-coding-systems 'update-coding-systems-internal)
317 (make-obsolete 'update-iso-coding-systems 'update-coding-systems-internal "20.3")
318
319 (defun prefer-coding-system (coding-system)
320 "Add CODING-SYSTEM at the front of the priority list for automatic detection.
321 This also sets the following coding systems:
322 o coding system of a newly created buffer
323 o default coding system for subprocess I/O
324 This also sets the following values:
325 o default value used as `file-name-coding-system' for converting file names.
326 o default value for the command `set-terminal-coding-system' (not on MSDOS)
327 o default value for the command `set-keyboard-coding-system'
328
329 If CODING-SYSTEM specifies a certain type of EOL conversion, the coding
330 systems set by this function will use that type of EOL conversion.
331
332 This command does not change the default value of terminal coding system
333 for MS-DOS terminal, because DOS terminals only support a single coding
334 system, and Emacs automatically sets the default to that coding system at
335 startup.
336
337 A coding system that requires automatic detection of text
338 encoding (e.g. undecided, unix) can't be preferred.
339
340 See also `coding-category-list' and `coding-system-category'."
341 (interactive "zPrefer coding system: ")
342 (if (not (and coding-system (coding-system-p coding-system)))
343 (error "Invalid coding system `%s'" coding-system))
344 (let ((coding-category (coding-system-category coding-system))
345 (base (coding-system-base coding-system))
346 (eol-type (coding-system-eol-type coding-system)))
347 (if (not coding-category)
348 ;; CODING-SYSTEM is no-conversion or undecided.
349 (error "Can't prefer the coding system `%s'" coding-system))
350 (set coding-category (or base coding-system))
351 (update-coding-systems-internal)
352 (or (eq coding-category (car coding-category-list))
353 ;; We must change the order.
354 (set-coding-priority (list coding-category)))
355 (if (and base (interactive-p))
356 (message "Highest priority is set to %s (base of %s)"
357 base coding-system))
358 ;; If they asked for specific EOL conversion, honor that.
359 (if (memq eol-type '(0 1 2))
360 (setq coding-system
361 (coding-system-change-eol-conversion base eol-type))
362 (setq coding-system base))
363 (set-default-coding-systems coding-system)))
364
365 (defvar sort-coding-systems-predicate nil
366 "If non-nil, a predicate function to sort coding systems.
367
368 It is called with two coding systems, and should return t if the first
369 one is \"less\" than the second.
370
371 The function `sort-coding-systems' use it.")
372
373 (defun sort-coding-systems (codings)
374 "Sort coding system list CODINGS by a priority of each coding system.
375
376 If a coding system is most preferred, it has the highest priority.
377 Otherwise, a coding system corresponds to some MIME charset has higher
378 priorities. Among them, a coding system included in `coding-system'
379 key of the current language environment has higher priorities. See
380 also the documentation of `language-info-alist'.
381
382 If the variable `sort-coding-systems-predicate' (which see) is
383 non-nil, it is used to sort CODINGS in the different way than above."
384 (if sort-coding-systems-predicate
385 (sort codings sort-coding-systems-predicate)
386 (let* ((most-preferred (symbol-value (car coding-category-list)))
387 (lang-preferred (get-language-info current-language-environment
388 'coding-system))
389 (func (function
390 (lambda (x)
391 (let ((base (coding-system-base x)))
392 (+ (if (eq base most-preferred) 64 0)
393 (let ((mime (coding-system-get base 'mime-charset)))
394 (if mime
395 (if (string-match "^x-" (symbol-name mime))
396 16 32)
397 0))
398 (if (memq base lang-preferred) 8 0)
399 (if (string-match "-with-esc$" (symbol-name base))
400 0 4)
401 (if (eq (coding-system-type base) 2)
402 ;; For ISO based coding systems, prefer
403 ;; one that doesn't use escape sequences.
404 (let ((flags (coding-system-flags base)))
405 (if (or (consp (aref flags 0))
406 (consp (aref flags 1))
407 (consp (aref flags 2))
408 (consp (aref flags 3)))
409 (if (or (aref flags 8) (aref flags 9))
410 0
411 1)
412 2))
413 1)))))))
414 (sort codings (function (lambda (x y)
415 (> (funcall func x) (funcall func y))))))))
416
417 (defun find-coding-systems-region (from to)
418 "Return a list of proper coding systems to encode a text between FROM and TO.
419 All coding systems in the list can safely encode any multibyte characters
420 in the text.
421
422 If the text contains no multibyte characters, return a list of a single
423 element `undecided'."
424 (let ((codings (find-coding-systems-region-internal from to)))
425 (if (eq codings t)
426 ;; The text contains only ASCII characters. Any coding
427 ;; systems are safe.
428 '(undecided)
429 ;; We need copy-sequence because sorting will alter the argument.
430 (sort-coding-systems (copy-sequence codings)))))
431
432 (defun find-coding-systems-string (string)
433 "Return a list of proper coding systems to encode STRING.
434 All coding systems in the list can safely encode any multibyte characters
435 in STRING.
436
437 If STRING contains no multibyte characters, return a list of a single
438 element `undecided'."
439 (find-coding-systems-region string nil))
440
441 (defun find-coding-systems-for-charsets (charsets)
442 "Return a list of proper coding systems to encode characters of CHARSETS.
443 CHARSETS is a list of character sets."
444 (cond ((or (null charsets)
445 (and (= (length charsets) 1)
446 (eq 'ascii (car charsets))))
447 '(undecided))
448 ((or (memq 'eight-bit-control charsets)
449 (memq 'eight-bit-graphic charsets))
450 '(raw-text emacs-mule))
451 (t
452 (let ((codings t)
453 charset l ll)
454 (while (and codings charsets)
455 (setq charset (car charsets) charsets (cdr charsets))
456 (unless (eq charset 'ascii)
457 (setq l (aref char-coding-system-table (make-char charset)))
458 (if (eq codings t)
459 (setq codings l)
460 (let ((ll nil))
461 (while codings
462 (if (memq (car codings) l)
463 (setq ll (cons (car codings) ll)))
464 (setq codings (cdr codings)))
465 (setq codings ll)))))
466 (append codings
467 (char-table-extra-slot char-coding-system-table 0))))))
468
469 (defun find-multibyte-characters (from to &optional maxcount excludes)
470 "Find multibyte characters in the region specified by FROM and TO.
471 If FROM is a string, find multibyte characters in the string.
472 The return value is an alist of the following format:
473 ((CHARSET COUNT CHAR ...) ...)
474 where
475 CHARSET is a character set,
476 COUNT is a number of characters,
477 CHARs are found characters of the character set.
478 Optional 3rd arg MAXCOUNT limits how many CHARs are put in the above list.
479 Optional 4th arg EXCLUDE is a list of character sets to be ignored.
480
481 For invalid characters, CHARs are actually strings."
482 (let ((chars nil)
483 charset char)
484 (if (stringp from)
485 (let ((idx 0))
486 (while (setq idx (string-match "[^\000-\177]" from idx))
487 (setq char (aref from idx)
488 charset (char-charset char))
489 (if (eq charset 'unknown)
490 (setq char (match-string 0)))
491 (if (or (memq charset '(unknown
492 eight-bit-control eight-bit-graphic))
493 (not (or (eq excludes t) (memq charset excludes))))
494 (let ((slot (assq charset chars)))
495 (if slot
496 (if (not (memq char (nthcdr 2 slot)))
497 (let ((count (nth 1 slot)))
498 (setcar (cdr slot) (1+ count))
499 (if (or (not maxcount) (< count maxcount))
500 (nconc slot (list char)))))
501 (setq chars (cons (list charset 1 char) chars)))))
502 (setq idx (1+ idx))))
503 (save-excursion
504 (goto-char from)
505 (while (re-search-forward "[^\000-\177]" to t)
506 (setq char (preceding-char)
507 charset (char-charset char))
508 (if (eq charset 'unknown)
509 (setq char (match-string 0)))
510 (if (or (memq charset '(unknown eight-bit-control eight-bit-graphic))
511 (not (or (eq excludes t) (memq charset excludes))))
512 (let ((slot (assq charset chars)))
513 (if slot
514 (if (not (member char (nthcdr 2 slot)))
515 (let ((count (nth 1 slot)))
516 (setcar (cdr slot) (1+ count))
517 (if (or (not maxcount) (< count maxcount))
518 (nconc slot (list char)))))
519 (setq chars (cons (list charset 1 char) chars))))))))
520 (nreverse chars)))
521
522 (defvar last-coding-system-specified nil
523 "Most recent coding system explicitly specified by the user when asked.
524 This variable is set whenever Emacs asks the user which coding system
525 to use in order to write a file. If you set it to nil explicitly,
526 then call `write-region', then afterward this variable will be non-nil
527 only if the user was explicitly asked and specified a coding system.")
528
529 (defvar select-safe-coding-system-accept-default-p nil
530 "If non-nil, a function to control the behaviour of coding system selection.
531 The meaning is the same as the argument ACCEPT-DEFAULT-P of the
532 function `select-safe-coding-system' (which see). This variable
533 overrides that argument.")
534
535 (defun select-safe-coding-system (from to &optional default-coding-system
536 accept-default-p)
537 "Ask a user to select a safe coding system from candidates.
538 The candidates of coding systems which can safely encode a text
539 between FROM and TO are shown in a popup window. Among them, the most
540 proper one is suggested as the default.
541
542 The list of `buffer-file-coding-system' of the current buffer and the
543 most preferred coding system (if it corresponds to a MIME charset) is
544 treated as the default coding system list. Among them, the first one
545 that safely encodes the text is silently selected and returned without
546 any user interaction. See also the command `prefer-coding-system'.
547
548 Optional 3rd arg DEFAULT-CODING-SYSTEM specifies a coding system or a
549 list of coding systems to be prepended to the default coding system
550 list.
551
552 Optional 4th arg ACCEPT-DEFAULT-P, if non-nil, is a function to
553 determine the acceptability of the silently selected coding system.
554 It is called with that coding system, and should return nil if it
555 should not be silently selected and thus user interaction is required.
556
557 The variable `select-safe-coding-system-accept-default-p', if
558 non-nil, overrides ACCEPT-DEFAULT-P.
559
560 Kludgy feature: if FROM is a string, the string is the target text,
561 and TO is ignored."
562 (if (and default-coding-system
563 (not (listp default-coding-system)))
564 (setq default-coding-system (list default-coding-system)))
565
566 ;; Change elements of the list to (coding . base-coding).
567 (setq default-coding-system
568 (mapcar (function (lambda (x) (cons x (coding-system-base x))))
569 default-coding-system))
570
571 ;; If buffer-file-coding-system is not nil nor undecided, append it
572 ;; to the defaults.
573 (if buffer-file-coding-system
574 (let ((base (coding-system-base buffer-file-coding-system)))
575 (or (eq base 'undecided)
576 (assq buffer-file-coding-system default-coding-system)
577 (rassq base default-coding-system)
578 (setq default-coding-system
579 (append default-coding-system
580 (list (cons buffer-file-coding-system base)))))))
581
582 ;; If the most preferred coding system has the property mime-charset,
583 ;; append it to the defaults.
584 (let ((tail coding-category-list)
585 preferred base)
586 (while (and tail
587 (not (setq preferred (symbol-value (car tail)))))
588 (setq tail (cdr tail)))
589 (and (coding-system-p preferred)
590 (setq base (coding-system-base preferred))
591 (coding-system-get preferred 'mime-charset)
592 (not (assq preferred default-coding-system))
593 (not (rassq base default-coding-system))
594 (setq default-coding-system
595 (append default-coding-system (list (cons preferred base))))))
596
597 (if select-safe-coding-system-accept-default-p
598 (setq accept-default-p select-safe-coding-system-accept-default-p))
599
600 (let ((codings (find-coding-systems-region from to))
601 (coding-system nil)
602 (bufname (buffer-name))
603 (l default-coding-system))
604 (if (eq (car codings) 'undecided)
605 ;; Any coding system is ok.
606 (setq coding-system t)
607 ;; Try the defaults.
608 (while (and l (not coding-system))
609 (if (memq (cdr (car l)) codings)
610 (setq coding-system (car (car l)))
611 (setq l (cdr l))))
612 (if (and coding-system accept-default-p)
613 (or (funcall accept-default-p coding-system)
614 (setq coding-system (list coding-system)))))
615
616 ;; If all the defaults failed, ask a user.
617 (when (or (not coding-system) (consp coding-system))
618 ;; At first, change each coding system to the corresponding
619 ;; mime-charset name if it is also a coding system. Such a name
620 ;; is more friendly to users.
621 (let ((l codings)
622 mime-charset)
623 (while l
624 (setq mime-charset (coding-system-get (car l) 'mime-charset))
625 (if (and mime-charset (coding-system-p mime-charset))
626 (setcar l mime-charset))
627 (setq l (cdr l))))
628
629 ;; Make sure the offending buffer is displayed.
630 (or (stringp from)
631 (pop-to-buffer bufname))
632 ;; Then ask users to select one form CODINGS.
633 (unwind-protect
634 (save-window-excursion
635 (with-output-to-temp-buffer "*Warning*"
636 (save-excursion
637 (set-buffer standard-output)
638 (if (not default-coding-system)
639 (insert "No default coding systems to try for "
640 (if (stringp from)
641 (format "string \"%s\"." from)
642 (format "buffer `%s'." bufname)))
643 (insert
644 "These default coding systems were tried to encode"
645 (if (stringp from)
646 (concat " \"" (if (> (length from) 10)
647 (concat (substring from 0 10) "...\"")
648 (concat from "\"")))
649 (format " text\nin the buffer `%s'" bufname))
650 ":\n")
651 (let ((pos (point))
652 (fill-prefix " "))
653 (mapcar (function (lambda (x)
654 (princ " ") (princ (car x))))
655 default-coding-system)
656 (insert "\n")
657 (fill-region-as-paragraph pos (point)))
658 (insert
659 (if (consp coding-system)
660 (concat (format "%s safely encodes the target text,\n"
661 (car coding-system))
662 "but it is not recommended for encoding text in this context,\n"
663 "e.g., for sending an email message.\n")
664 "However, none of them safely encodes the target text.\n")))
665 (insert (if (consp coding-system)
666 "\nSelect the above, or "
667 "\nSelect ")
668 "one of the following safe coding systems:\n")
669 (let ((pos (point))
670 (fill-prefix " "))
671 (mapcar (function (lambda (x) (princ " ") (princ x)))
672 codings)
673 (insert "\n")
674 (fill-region-as-paragraph pos (point)))))
675
676 ;; Read a coding system.
677 (if (consp coding-system)
678 (setq codings (cons (car coding-system) codings)))
679 (let* ((safe-names (mapcar (lambda (x) (list (symbol-name x)))
680 codings))
681 (name (completing-read
682 (format "Select coding system (default %s): "
683 (car codings))
684 safe-names nil t nil nil
685 (car (car safe-names)))))
686 (setq last-coding-system-specified (intern name)
687 coding-system last-coding-system-specified)))
688 (kill-buffer "*Warning*")))
689
690 (if (vectorp (coding-system-eol-type coding-system))
691 (let ((eol (coding-system-eol-type buffer-file-coding-system)))
692 (if (numberp eol)
693 (setq coding-system
694 (coding-system-change-eol-conversion coding-system eol)))))
695
696 (if (eq coding-system t)
697 (setq coding-system buffer-file-coding-system))
698 coding-system))
699
700 (setq select-safe-coding-system-function 'select-safe-coding-system)
701
702 (defun select-message-coding-system ()
703 "Return a coding system to encode the outgoing message of the current buffer.
704 It at first tries the first coding system found in these variables
705 in this order:
706 (1) local value of `buffer-file-coding-system'
707 (2) value of `sendmail-coding-system'
708 (3) value of `default-sendmail-coding-system'
709 (4) value of `default-buffer-file-coding-system'
710 If the found coding system can't encode the current buffer,
711 or none of them are bound to a coding system,
712 it asks the user to select a proper coding system."
713 (let ((coding (or (and (local-variable-p 'buffer-file-coding-system)
714 buffer-file-coding-system)
715 sendmail-coding-system
716 default-sendmail-coding-system
717 default-buffer-file-coding-system)))
718 (if (eq coding 'no-conversion)
719 ;; We should never use no-conversion for outgoing mails.
720 (setq coding nil))
721 (if (fboundp select-safe-coding-system-function)
722 (funcall select-safe-coding-system-function
723 (point-min) (point-max) coding
724 (function (lambda (x) (coding-system-get x 'mime-charset))))
725 coding)))
726 \f
727 ;;; Language support stuff.
728
729 (defvar language-info-alist nil
730 "Alist of language environment definitions.
731 Each element looks like:
732 (LANGUAGE-NAME . ((KEY . INFO) ...))
733 where LANGUAGE-NAME is a string, the name of the language environment,
734 KEY is a symbol denoting the kind of information, and
735 INFO is the data associated with KEY.
736 Meaningful values for KEY include
737
738 documentation value is documentation of what this language environment
739 is meant for, and how to use it.
740 charset value is a list of the character sets used by this
741 language environment.
742 sample-text value is one line of text,
743 written using those character sets,
744 appropriate for this language environment.
745 setup-function value is a function to call to switch to this
746 language environment.
747 exit-function value is a function to call to leave this
748 language environment.
749 coding-system value is a list of coding systems that are good
750 for saving text written in this language environment.
751 This list serves as suggestions to the user;
752 in effect, as a kind of documentation.
753 coding-priority value is a list of coding systems for this language
754 environment, in order of decreasing priority.
755 This is used to set up the coding system priority
756 list when you switch to this language environment.
757 nonascii-translation
758 value is a translation table to be set in the
759 variable `nonascii-translation-table' in this
760 language environment, or a character set from
761 which `nonascii-insert-offset' is calculated.
762 input-method value is a default input method for this language
763 environment.
764 features value is a list of features requested in this
765 language environment.
766
767 The following keys take effect only when multibyte characters are
768 globally disabled, i.e. the value of `default-enable-multibyte-characters'
769 is nil.
770
771 unibyte-syntax value is a library name to load to set
772 unibyte 8-bit character syntaxes for this
773 language environment.
774
775 unibyte-display value is a coding system to encode characters
776 for the terminal. Characters in the range
777 of 160 to 255 display not as octal escapes,
778 but as non-ASCII characters in this language
779 environment.")
780
781 (defun get-language-info (lang-env key)
782 "Return information listed under KEY for language environment LANG-ENV.
783 KEY is a symbol denoting the kind of information.
784 For a list of useful values for KEY and their meanings,
785 see `language-info-alist'."
786 (if (symbolp lang-env)
787 (setq lang-env (symbol-name lang-env)))
788 (let ((lang-slot (assoc-ignore-case lang-env language-info-alist)))
789 (if lang-slot
790 (cdr (assq key (cdr lang-slot))))))
791
792 (defun set-language-info (lang-env key info)
793 "Modify part of the definition of language environment LANG-ENV.
794 Specifically, this stores the information INFO under KEY
795 in the definition of this language environment.
796 KEY is a symbol denoting the kind of information.
797 INFO is the value for that information.
798
799 For a list of useful values for KEY and their meanings,
800 see `language-info-alist'."
801 (if (symbolp lang-env)
802 (setq lang-env (symbol-name lang-env)))
803 (let (lang-slot key-slot)
804 (setq lang-slot (assoc lang-env language-info-alist))
805 (if (null lang-slot) ; If no slot for the language, add it.
806 (setq lang-slot (list lang-env)
807 language-info-alist (cons lang-slot language-info-alist)))
808 (setq key-slot (assq key lang-slot))
809 (if (null key-slot) ; If no slot for the key, add it.
810 (progn
811 (setq key-slot (list key))
812 (setcdr lang-slot (cons key-slot (cdr lang-slot)))))
813 (setcdr key-slot (purecopy info))))
814
815 (defun set-language-info-alist (lang-env alist &optional parents)
816 "Store ALIST as the definition of language environment LANG-ENV.
817 ALIST is an alist of KEY and INFO values. See the documentation of
818 `language-info-alist' for the meanings of KEY and INFO.
819
820 Optional arg PARENTS is a list of parent menu names; it specifies
821 where to put this language environment in the
822 Describe Language Environment and Set Language Environment menus.
823 For example, (\"European\") means to put this language environment
824 in the European submenu in each of those two menus."
825 (if (symbolp lang-env)
826 (setq lang-env (symbol-name lang-env)))
827 (let ((describe-map describe-language-environment-map)
828 (setup-map setup-language-environment-map))
829 (if parents
830 (let ((l parents)
831 map parent-symbol parent prompt)
832 (while l
833 (if (symbolp (setq parent-symbol (car l)))
834 (setq parent (symbol-name parent))
835 (setq parent parent-symbol parent-symbol (intern parent)))
836 (setq map (lookup-key describe-map (vector parent-symbol)))
837 ;; This prompt string is for define-prefix-command, so
838 ;; that the map it creates will be suitable for a menu.
839 (or map (setq prompt (format "%s Environment" parent)))
840 (if (not map)
841 (progn
842 (setq map (intern (format "describe-%s-environment-map"
843 (downcase parent))))
844 (define-prefix-command map nil prompt)
845 (define-key-after describe-map (vector parent-symbol)
846 (cons parent map) t)))
847 (setq describe-map (symbol-value map))
848 (setq map (lookup-key setup-map (vector parent-symbol)))
849 (if (not map)
850 (progn
851 (setq map (intern (format "setup-%s-environment-map"
852 (downcase parent))))
853 (define-prefix-command map nil prompt)
854 (define-key-after setup-map (vector parent-symbol)
855 (cons parent map) t)))
856 (setq setup-map (symbol-value map))
857 (setq l (cdr l)))))
858
859 ;; Set up menu items for this language env.
860 (let ((doc (assq 'documentation alist)))
861 (when doc
862 (define-key-after describe-map (vector (intern lang-env))
863 (cons lang-env 'describe-specified-language-support) t)))
864 (define-key-after setup-map (vector (intern lang-env))
865 (cons lang-env 'setup-specified-language-environment) t)
866
867 (while alist
868 (set-language-info lang-env (car (car alist)) (cdr (car alist)))
869 (setq alist (cdr alist)))))
870
871 (defun read-language-name (key prompt &optional default)
872 "Read a language environment name which has information for KEY.
873 If KEY is nil, read any language environment.
874 Prompt with PROMPT. DEFAULT is the default choice of language environment.
875 This returns a language environment name as a string."
876 (let* ((completion-ignore-case t)
877 (name (completing-read prompt
878 language-info-alist
879 (and key
880 (function (lambda (elm) (assq key elm))))
881 t nil nil default)))
882 (if (and (> (length name) 0)
883 (or (not key)
884 (get-language-info name key)))
885 name)))
886 \f
887 ;;; Multilingual input methods.
888 (defgroup leim nil
889 "LEIM: Libraries of Emacs Input Methods."
890 :group 'mule)
891
892 (defconst leim-list-file-name "leim-list.el"
893 "Name of LEIM list file.
894 This file contains a list of libraries of Emacs input methods (LEIM)
895 in the format of Lisp expression for registering each input method.
896 Emacs loads this file at startup time.")
897
898 (defvar leim-list-header (format
899 ";;; %s -- list of LEIM (Library of Emacs Input Method)
900 ;;
901 ;; This file contains a list of LEIM (Library of Emacs Input Method)
902 ;; methods in the same directory as this file. Loading this file
903 ;; registers all the input methods in Emacs.
904 ;;
905 ;; Each entry has the form:
906 ;; (register-input-method
907 ;; INPUT-METHOD LANGUAGE-NAME ACTIVATE-FUNC
908 ;; TITLE DESCRIPTION
909 ;; ARG ...)
910 ;; See the function `register-input-method' for the meanings of the arguments.
911 ;;
912 ;; If this directory is included in load-path, Emacs automatically
913 ;; loads this file at startup time.
914
915 "
916 leim-list-file-name)
917 "Header to be inserted in LEIM list file.")
918
919 (defvar leim-list-entry-regexp "^(register-input-method"
920 "Regexp matching head of each entry in LEIM list file.
921 See also the variable `leim-list-header'")
922
923 (defvar update-leim-list-functions
924 '(quail-update-leim-list-file)
925 "List of functions to call to update LEIM list file.
926 Each function is called with one arg, LEIM directory name.")
927
928 (defun update-leim-list-file (&rest dirs)
929 "Update LEIM list file in directories DIRS."
930 (let ((functions update-leim-list-functions))
931 (while functions
932 (apply (car functions) dirs)
933 (setq functions (cdr functions)))))
934
935 (defvar current-input-method nil
936 "The current input method for multilingual text.
937 If nil, that means no input method is activated now.")
938 (make-variable-buffer-local 'current-input-method)
939 (put 'current-input-method 'permanent-local t)
940
941 (defvar current-input-method-title nil
942 "Title string of the current input method shown in mode line.")
943 (make-variable-buffer-local 'current-input-method-title)
944 (put 'current-input-method-title 'permanent-local t)
945
946 (defcustom default-input-method nil
947 "*Default input method for multilingual text (a string).
948 This is the input method activated automatically by the command
949 `toggle-input-method' (\\[toggle-input-method])."
950 :group 'mule
951 :type '(choice (const nil) string)
952 :set-after '(current-language-environment))
953
954 (put 'input-method-function 'permanent-local t)
955
956 (defvar input-method-history nil
957 "History list for some commands that read input methods.")
958 (make-variable-buffer-local 'input-method-history)
959 (put 'input-method-history 'permanent-local t)
960
961 (defvar inactivate-current-input-method-function nil
962 "Function to call for inactivating the current input method.
963 Every input method should set this to an appropriate value when activated.
964 This function is called with no argument.
965
966 This function should never change the value of `current-input-method'.
967 It is set to nil by the function `inactivate-input-method'.")
968 (make-variable-buffer-local 'inactivate-current-input-method-function)
969 (put 'inactivate-current-input-method-function 'permanent-local t)
970
971 (defvar describe-current-input-method-function nil
972 "Function to call for describing the current input method.
973 This function is called with no argument.")
974 (make-variable-buffer-local 'describe-current-input-method-function)
975 (put 'describe-current-input-method-function 'permanent-local t)
976
977 (defvar input-method-alist nil
978 "Alist of input method names vs how to use them.
979 Each element has the form:
980 (INPUT-METHOD LANGUAGE-ENV ACTIVATE-FUNC TITLE DESCRIPTION ARGS...)
981 See the function `register-input-method' for the meanings of the elements.")
982
983 (defun register-input-method (input-method lang-env &rest args)
984 "Register INPUT-METHOD as an input method for language environment LANG-ENV.
985 INPUT-METHOD and LANG-ENV are symbols or strings.
986
987 The remaining arguments are:
988 ACTIVATE-FUNC, TITLE, DESCRIPTION, and ARGS...
989 ACTIVATE-FUNC is a function to call to activate this method.
990 TITLE is a string to show in the mode line when this method is active.
991 DESCRIPTION is a string describing this method and what it is good for.
992 The ARGS, if any, are passed as arguments to ACTIVATE-FUNC.
993 All told, the arguments to ACTIVATE-FUNC are INPUT-METHOD and the ARGS.
994
995 This function is mainly used in the file \"leim-list.el\" which is
996 created at Emacs build time, registering all Quail input methods
997 contained in the Emacs distribution.
998
999 In case you want to register a new Quail input method by yourself, be
1000 careful to use the same input method title as given in the third
1001 parameter of `quail-define-package'. (If the values are different, the
1002 string specified in this function takes precedence.)
1003
1004 The commands `describe-input-method' and `list-input-methods' need
1005 these duplicated values to show some information about input methods
1006 without loading the relevant Quail packages."
1007 (if (symbolp lang-env)
1008 (setq lang-env (symbol-name lang-env)))
1009 (if (symbolp input-method)
1010 (setq input-method (symbol-name input-method)))
1011 (let ((info (cons lang-env args))
1012 (slot (assoc input-method input-method-alist)))
1013 (if slot
1014 (setcdr slot info)
1015 (setq slot (cons input-method info))
1016 (setq input-method-alist (cons slot input-method-alist)))))
1017
1018 (defun read-input-method-name (prompt &optional default inhibit-null)
1019 "Read a name of input method from a minibuffer prompting with PROMPT.
1020 If DEFAULT is non-nil, use that as the default,
1021 and substitute it into PROMPT at the first `%s'.
1022 If INHIBIT-NULL is non-nil, null input signals an error.
1023
1024 The return value is a string."
1025 (if default
1026 (setq prompt (format prompt default)))
1027 (let* ((completion-ignore-case t)
1028 ;; As it is quite normal to change input method in the
1029 ;; minibuffer, we must enable it even if
1030 ;; enable-recursive-minibuffers is currently nil.
1031 (enable-recursive-minibuffers t)
1032 ;; This binding is necessary because input-method-history is
1033 ;; buffer local.
1034 (input-method (completing-read prompt input-method-alist
1035 nil t nil 'input-method-history
1036 default)))
1037 (if (and input-method (symbolp input-method))
1038 (setq input-method (symbol-name input-method)))
1039 (if (> (length input-method) 0)
1040 input-method
1041 (if inhibit-null
1042 (error "No valid input method is specified")))))
1043
1044 (defun activate-input-method (input-method)
1045 "Switch to input method INPUT-METHOD for the current buffer.
1046 If some other input method is already active, turn it off first.
1047 If INPUT-METHOD is nil, deactivate any current input method."
1048 (if (and input-method (symbolp input-method))
1049 (setq input-method (symbol-name input-method)))
1050 (if (and current-input-method
1051 (not (string= current-input-method input-method)))
1052 (inactivate-input-method))
1053 (unless (or current-input-method (null input-method))
1054 (let ((slot (assoc input-method input-method-alist)))
1055 (if (null slot)
1056 (error "Can't activate input method `%s'" input-method))
1057 (setq current-input-method-title nil)
1058 (let ((func (nth 2 slot)))
1059 (if (functionp func)
1060 (apply (nth 2 slot) input-method (nthcdr 5 slot))
1061 (if (and (consp func) (symbolp (car func)) (symbolp (cdr func)))
1062 (progn
1063 (require (cdr func))
1064 (apply (car func) input-method (nthcdr 5 slot)))
1065 (error "Can't activate input method `%s'" input-method))))
1066 (setq current-input-method input-method)
1067 (or (stringp current-input-method-title)
1068 (setq current-input-method-title (nth 3 slot)))
1069 (unwind-protect
1070 (run-hooks 'input-method-activate-hook)
1071 (force-mode-line-update)))))
1072
1073 (defun inactivate-input-method ()
1074 "Turn off the current input method."
1075 (when current-input-method
1076 (if input-method-history
1077 (unless (string= current-input-method (car input-method-history))
1078 (setq input-method-history
1079 (cons current-input-method
1080 (delete current-input-method input-method-history))))
1081 (setq input-method-history (list current-input-method)))
1082 (unwind-protect
1083 (funcall inactivate-current-input-method-function)
1084 (unwind-protect
1085 (run-hooks 'input-method-inactivate-hook)
1086 (setq current-input-method nil
1087 input-method-function nil
1088 current-input-method-title nil)
1089 (force-mode-line-update)))))
1090
1091 (defun set-input-method (input-method)
1092 "Select and activate input method INPUT-METHOD for the current buffer.
1093 This also sets the default input method to the one you specify.
1094 If INPUT-METHOD is nil, this function turns off the input method, and
1095 also causes you to be prompted for a name of an input method the next
1096 time you invoke \\[toggle-input-method].
1097
1098 To deactivate the input method interactively, use \\[toggle-input-method].
1099 To deactivate it programmatically, use \\[inactivate-input-method]."
1100 (interactive
1101 (let* ((default (or (car input-method-history) default-input-method)))
1102 (list (read-input-method-name
1103 (if default "Select input method (default %s): " "Select input method: ")
1104 default t))))
1105 (activate-input-method input-method)
1106 (setq default-input-method input-method)
1107 (when (interactive-p)
1108 (customize-mark-as-set 'default-input-method))
1109 default-input-method)
1110
1111 (defun toggle-input-method (&optional arg)
1112 "Enable or disable multilingual text input method for the current buffer.
1113 Only one input method can be enabled at any time in a given buffer.
1114
1115 The normal action is to enable an input method if none was
1116 enabled, and disable the current one otherwise. Which input method
1117 to enable can be determined in various ways--either the one most
1118 recently used, or the one specified by `default-input-method', or
1119 as a last resort by reading the name of an input method in the
1120 minibuffer.
1121
1122 With a prefix argument, read an input method name with the minibuffer
1123 and enable that one. The default is the most recent input method specified
1124 \(not including the currently active input method, if any)."
1125
1126 (interactive "P")
1127 (if (and current-input-method (not arg))
1128 (inactivate-input-method)
1129 (let ((default (or (car input-method-history) default-input-method)))
1130 (if (and arg default (equal current-input-method default)
1131 (> (length input-method-history) 1))
1132 (setq default (nth 1 input-method-history)))
1133 (activate-input-method
1134 (if (or arg (not default))
1135 (progn
1136 (read-input-method-name
1137 (if default "Input method (default %s): " "Input method: " )
1138 default t))
1139 default))
1140 (unless default-input-method
1141 (prog1
1142 (setq default-input-method current-input-method)
1143 (when (interactive-p)
1144 (customize-mark-as-set 'default-input-method)))))))
1145
1146 (defun describe-input-method (input-method)
1147 "Describe input method INPUT-METHOD."
1148 (interactive
1149 (list (read-input-method-name
1150 "Describe input method (default, current choice): ")))
1151 (if (and input-method (symbolp input-method))
1152 (setq input-method (symbol-name input-method)))
1153 (help-setup-xref (list #'describe-input-method (or input-method current-input-method))
1154 (interactive-p))
1155
1156 (if (null input-method)
1157 (describe-current-input-method)
1158 (let ((current current-input-method))
1159 (condition-case nil
1160 (progn
1161 (save-excursion
1162 (activate-input-method input-method)
1163 (describe-current-input-method))
1164 (activate-input-method current))
1165 (error
1166 (activate-input-method current)
1167 (help-setup-xref (list #'describe-input-method input-method)
1168 (interactive-p))
1169 (with-output-to-temp-buffer (help-buffer)
1170 (let ((elt (assoc input-method input-method-alist)))
1171 (princ (format
1172 "Input method: %s (`%s' in mode line) for %s\n %s\n"
1173 input-method (nth 3 elt) (nth 1 elt) (nth 4 elt))))))))))
1174
1175 (defun describe-current-input-method ()
1176 "Describe the input method currently in use.
1177 This is a subroutine for `describe-input-method'."
1178 (if current-input-method
1179 (if (and (symbolp describe-current-input-method-function)
1180 (fboundp describe-current-input-method-function))
1181 (funcall describe-current-input-method-function)
1182 (message "No way to describe the current input method `%s'"
1183 current-input-method)
1184 (ding))
1185 (error "No input method is activated now")))
1186
1187 (defun read-multilingual-string (prompt &optional initial-input input-method)
1188 "Read a multilingual string from minibuffer, prompting with string PROMPT.
1189 The input method selected last time is activated in minibuffer.
1190 If optional second arg INITIAL-INPUT is non-nil, insert it in the minibuffer
1191 initially.
1192 Optional 3rd argument INPUT-METHOD specifies the input method
1193 to be activated instead of the one selected last time. It is a symbol
1194 or a string."
1195 (setq input-method
1196 (or input-method
1197 current-input-method
1198 default-input-method
1199 (read-input-method-name "Input method: " nil t)))
1200 (if (and input-method (symbolp input-method))
1201 (setq input-method (symbol-name input-method)))
1202 (let ((prev-input-method current-input-method))
1203 (unwind-protect
1204 (progn
1205 (activate-input-method input-method)
1206 (read-string prompt initial-input nil nil t))
1207 (activate-input-method prev-input-method))))
1208
1209 ;; Variables to control behavior of input methods. All input methods
1210 ;; should react to these variables.
1211
1212 (defcustom input-method-verbose-flag 'default
1213 "*A flag to control extra guidance given by input methods.
1214 The value should be nil, t, `complex-only', or `default'.
1215
1216 The extra guidance is done by showing list of available keys in echo
1217 area. When you use the input method in the minibuffer, the guidance
1218 is shown at the bottom short window (split from the existing window).
1219
1220 If the value is t, extra guidance is always given, if the value is
1221 nil, extra guidance is always suppressed.
1222
1223 If the value is `complex-only', only complex input methods such as
1224 `chinese-py' and `japanese' give extra guidance.
1225
1226 If the value is `default', complex input methods always give extra
1227 guidance, but simple input methods give it only when you are not in
1228 the minibuffer.
1229
1230 See also the variable `input-method-highlight-flag'."
1231 :type '(choice (const t) (const nil) (const complex-only) (const default))
1232 :group 'mule)
1233
1234 (defcustom input-method-highlight-flag t
1235 "*If this flag is non-nil, input methods highlight partially-entered text.
1236 For instance, while you are in the middle of a Quail input method sequence,
1237 the text inserted so far is temporarily underlined.
1238 The underlining goes away when you finish or abort the input method sequence.
1239 See also the variable `input-method-verbose-flag'."
1240 :type 'boolean
1241 :group 'mule)
1242
1243 (defvar input-method-activate-hook nil
1244 "Normal hook run just after an input method is activated.
1245
1246 The variable `current-input-method' keeps the input method name
1247 just activated.")
1248
1249 (defvar input-method-inactivate-hook nil
1250 "Normal hook run just after an input method is inactivated.
1251
1252 The variable `current-input-method' still keeps the input method name
1253 just inactivated.")
1254
1255 (defvar input-method-after-insert-chunk-hook nil
1256 "Normal hook run just after an input method insert some chunk of text.")
1257
1258 (defvar input-method-exit-on-first-char nil
1259 "This flag controls when an input method returns.
1260 Usually, the input method does not return while there's a possibility
1261 that it may find a different translation if a user types another key.
1262 But, it this flag is non-nil, the input method returns as soon as
1263 the current key sequence gets long enough to have some valid translation.")
1264
1265 (defvar input-method-use-echo-area nil
1266 "This flag controls how an input method shows an intermediate key sequence.
1267 Usually, the input method inserts the intermediate key sequence,
1268 or candidate translations corresponding to the sequence,
1269 at point in the current buffer.
1270 But, if this flag is non-nil, it displays them in echo area instead.")
1271
1272 (defvar input-method-exit-on-invalid-key nil
1273 "This flag controls the behaviour of an input method on invalid key input.
1274 Usually, when a user types a key which doesn't start any character
1275 handled by the input method, the key is handled by turning off the
1276 input method temporarily. After that key, the input method is re-enabled.
1277 But, if this flag is non-nil, the input method is never back on.")
1278
1279 \f
1280 (defvar set-language-environment-hook nil
1281 "Normal hook run after some language environment is set.
1282
1283 When you set some hook function here, that effect usually should not
1284 be inherited to another language environment. So, you had better set
1285 another function in `exit-language-environment-hook' (which see) to
1286 cancel the effect.")
1287
1288 (defvar exit-language-environment-hook nil
1289 "Normal hook run after exiting from some language environment.
1290 When this hook is run, the variable `current-language-environment'
1291 is still bound to the language environment being exited.
1292
1293 This hook is mainly used for canceling the effect of
1294 `set-language-environment-hook' (which-see).")
1295
1296 (put 'setup-specified-language-environment 'apropos-inhibit t)
1297
1298 (defun setup-specified-language-environment ()
1299 "Switch to a specified language environment."
1300 (interactive)
1301 (let (language-name)
1302 (if (and (symbolp last-command-event)
1303 (or (not (eq last-command-event 'Default))
1304 (setq last-command-event 'English))
1305 (setq language-name (symbol-name last-command-event)))
1306 (prog1
1307 (set-language-environment language-name)
1308 (customize-mark-as-set 'current-language-environment))
1309 (error "Bogus calling sequence"))))
1310
1311 (defcustom current-language-environment "English"
1312 "The last language environment specified with `set-language-environment'.
1313 This variable should be set only with \\[customize], which is equivalent
1314 to using the function `set-language-environment'."
1315 :link '(custom-manual "(emacs)Language Environments")
1316 :set (lambda (symbol value) (set-language-environment value))
1317 :get (lambda (x)
1318 (or (car-safe (assoc-ignore-case
1319 (if (symbolp current-language-environment)
1320 (symbol-name current-language-environment)
1321 current-language-environment)
1322 language-info-alist))
1323 "English"))
1324 :type (cons 'choice (mapcar (lambda (lang)
1325 (list 'const (car lang)))
1326 language-info-alist))
1327 :initialize 'custom-initialize-default
1328 :group 'mule
1329 :type 'string)
1330
1331 (defun reset-language-environment ()
1332 "Reset multilingual environment of Emacs to the default status.
1333
1334 The default status is as follows:
1335
1336 The default value of buffer-file-coding-system is nil.
1337 The default coding system for process I/O is nil.
1338 The default value for the command `set-terminal-coding-system' is nil.
1339 The default value for the command `set-keyboard-coding-system' is nil.
1340
1341 The order of priorities of coding categories and the coding system
1342 bound to each category are as follows
1343 coding category coding system
1344 --------------------------------------------------
1345 coding-category-iso-8-2 iso-latin-1
1346 coding-category-iso-8-1 iso-latin-1
1347 coding-category-iso-7-tight iso-2022-jp
1348 coding-category-iso-7 iso-2022-7bit
1349 coding-category-iso-7-else iso-2022-7bit-lock
1350 coding-category-iso-8-else iso-2022-8bit-ss2
1351 coding-category-emacs-mule emacs-mule
1352 coding-category-raw-text raw-text
1353 coding-category-sjis japanese-shift-jis
1354 coding-category-big5 chinese-big5
1355 coding-category-ccl nil
1356 coding-category-binary no-conversion
1357 coding-category-utf-16-be nil
1358 coding-category-utf-16-le nil
1359 coding-category-utf-8 mule-utf-8"
1360 (interactive)
1361 ;; This function formerly set default-enable-multibyte-characters to t,
1362 ;; but that is incorrect. It should not alter the unibyte/multibyte choice.
1363
1364 (setq coding-category-iso-7-tight 'iso-2022-jp
1365 coding-category-iso-7 'iso-2022-7bit
1366 coding-category-iso-8-1 'iso-latin-1
1367 coding-category-iso-8-2 'iso-latin-1
1368 coding-category-iso-7-else 'iso-2022-7bit-lock
1369 coding-category-iso-8-else 'iso-2022-8bit-ss2
1370 coding-category-emacs-mule 'emacs-mule
1371 coding-category-raw-text 'raw-text
1372 coding-category-sjis 'japanese-shift-jis
1373 coding-category-big5 'chinese-big5
1374 coding-category-utf-16-be nil
1375 coding-category-utf-16-le nil
1376 coding-category-utf-8 'mule-utf-8
1377 coding-category-ccl nil
1378 coding-category-binary 'no-conversion)
1379
1380 (set-coding-priority
1381 '(coding-category-iso-8-1
1382 coding-category-iso-8-2
1383 coding-category-iso-7-tight
1384 coding-category-iso-7
1385 coding-category-iso-7-else
1386 coding-category-iso-8-else
1387 coding-category-emacs-mule
1388 coding-category-raw-text
1389 coding-category-sjis
1390 coding-category-big5
1391 coding-category-ccl
1392 coding-category-binary
1393 coding-category-utf-16-be
1394 coding-category-utf-16-le
1395 coding-category-utf-8))
1396
1397 (update-coding-systems-internal)
1398
1399 (set-default-coding-systems nil)
1400 (setq default-sendmail-coding-system 'iso-latin-1)
1401 (setq default-process-coding-system '(undecided . iso-latin-1))
1402
1403 ;; Don't alter the terminal and keyboard coding systems here.
1404 ;; The terminal still supports the same coding system
1405 ;; that it supported a minute ago.
1406 ;;; (set-terminal-coding-system-internal nil)
1407 ;;; (set-keyboard-coding-system-internal nil)
1408
1409 (setq nonascii-translation-table nil
1410 nonascii-insert-offset 0))
1411
1412 (reset-language-environment)
1413
1414 (defun set-display-table-and-terminal-coding-system (language-name)
1415 "Set up the display table and terminal coding system for LANGUAGE-NAME."
1416 (let ((coding (get-language-info language-name 'unibyte-display)))
1417 (if coding
1418 (standard-display-european-internal)
1419 (standard-display-default (if (eq window-system 'pc) 128 160) 255)
1420 (aset standard-display-table 146 nil))
1421 (or (eq window-system 'pc)
1422 (set-terminal-coding-system coding))))
1423
1424 (defun set-language-environment (language-name)
1425 "Set up multi-lingual environment for using LANGUAGE-NAME.
1426 This sets the coding system priority and the default input method
1427 and sometimes other things. LANGUAGE-NAME should be a string
1428 which is the name of a language environment. For example, \"Latin-1\"
1429 specifies the character set for the major languages of Western Europe."
1430 (interactive (list (read-language-name
1431 nil
1432 "Set language environment (default, English): ")))
1433 (if language-name
1434 (if (symbolp language-name)
1435 (setq language-name (symbol-name language-name)))
1436 (setq language-name "English"))
1437 (or (assoc-ignore-case language-name language-info-alist)
1438 (error "Language environment not defined: %S" language-name))
1439 (if current-language-environment
1440 (let ((func (get-language-info current-language-environment
1441 'exit-function)))
1442 (run-hooks 'exit-language-environment-hook)
1443 (if (functionp func) (funcall func))))
1444 (let ((default-eol-type (coding-system-eol-type
1445 default-buffer-file-coding-system)))
1446 (reset-language-environment)
1447
1448 ;; The fetaures might set up coding systems.
1449 (let ((required-features (get-language-info language-name 'features)))
1450 (while required-features
1451 (require (car required-features))
1452 (setq required-features (cdr required-features))))
1453
1454 (setq current-language-environment language-name)
1455 (set-language-environment-coding-systems language-name default-eol-type))
1456 (let ((input-method (get-language-info language-name 'input-method)))
1457 (when input-method
1458 (setq default-input-method input-method)
1459 (if input-method-history
1460 (setq input-method-history
1461 (cons input-method
1462 (delete input-method input-method-history))))))
1463 (let ((nonascii (get-language-info language-name 'nonascii-translation))
1464 (dos-table
1465 (if (eq window-system 'pc)
1466 (intern
1467 (format "cp%d-nonascii-translation-table" dos-codepage)))))
1468 (cond
1469 ((char-table-p nonascii)
1470 (setq nonascii-translation-table nonascii))
1471 ((and (eq window-system 'pc) (boundp dos-table))
1472 ;; DOS terminals' default is to use a special non-ASCII translation
1473 ;; table as appropriate for the installed codepage.
1474 (setq nonascii-translation-table (symbol-value dos-table)))
1475 ((charsetp nonascii)
1476 (setq nonascii-insert-offset (- (make-char nonascii) 128)))))
1477
1478 ;; Unibyte setups if necessary.
1479 (unless default-enable-multibyte-characters
1480 ;; Syntax and case table.
1481 (let ((syntax (get-language-info language-name 'unibyte-syntax)))
1482 (if syntax
1483 (let ((set-case-syntax-set-multibyte nil))
1484 (load syntax nil t))
1485 ;; No information for syntax and case. Reset to the defaults.
1486 (let ((syntax-table (standard-syntax-table))
1487 (case-table (standard-case-table))
1488 (ch (if (eq window-system 'pc) 128 160)))
1489 (while (< ch 256)
1490 (modify-syntax-entry ch " " syntax-table)
1491 (aset case-table ch ch)
1492 (setq ch (1+ ch)))
1493 (set-char-table-extra-slot case-table 0 nil)
1494 (set-char-table-extra-slot case-table 1 nil)
1495 (set-char-table-extra-slot case-table 2 nil))
1496 (set-standard-case-table (standard-case-table))
1497 (let ((list (buffer-list)))
1498 (while list
1499 (with-current-buffer (car list)
1500 (set-case-table (standard-case-table)))
1501 (setq list (cdr list))))))
1502 (set-display-table-and-terminal-coding-system language-name))
1503
1504 (let ((required-features (get-language-info language-name 'features)))
1505 (while required-features
1506 (require (car required-features))
1507 (setq required-features (cdr required-features))))
1508 (let ((func (get-language-info language-name 'setup-function)))
1509 (if (functionp func)
1510 (funcall func)))
1511 (run-hooks 'set-language-environment-hook)
1512 (force-mode-line-update t))
1513
1514 (defun standard-display-european-internal ()
1515 ;; Actually set up direct output of non-ASCII characters.
1516 (standard-display-8bit (if (eq window-system 'pc) 128 160) 255)
1517 ;; Unibyte Emacs on MS-DOS wants to display all 8-bit characters with
1518 ;; the native font, and codes 160 and 146 stand for something very
1519 ;; different there.
1520 (or (and (eq window-system 'pc) (not default-enable-multibyte-characters))
1521 (progn
1522 ;; Make non-line-break space display as a plain space.
1523 ;; Most X fonts do the wrong thing for code 160.
1524 (aset standard-display-table 160 [32])
1525 ;; With luck, non-Latin-1 fonts are more recent and so don't
1526 ;; have this bug.
1527 (aset standard-display-table 2208 [32]) ; Latin-1 NBSP
1528 ;; Most Windows programs send out apostrophes as \222. Most X fonts
1529 ;; don't contain a character at that position. Map it to the ASCII
1530 ;; apostrophe. [This is actually RIGHT SINGLE QUOTATION MARK,
1531 ;; U+2019, normally from the windows-1252 character set. XFree 4
1532 ;; fonts probably have the appropriate glyph at this position,
1533 ;; so they could use standard-display-8bit. It's better to use a
1534 ;; proper windows-1252 coding system. --fx]
1535 (aset standard-display-table 146 [39]))))
1536
1537 (defun set-language-environment-coding-systems (language-name
1538 &optional eol-type)
1539 "Do various coding system setups for language environment LANGUAGE-NAME.
1540
1541 The optional arg EOL-TYPE specifies the eol-type of the default value
1542 of buffer-file-coding-system set by this function."
1543 (let* ((priority (get-language-info language-name 'coding-priority))
1544 (default-coding (car priority)))
1545 (if priority
1546 (let ((categories (mapcar 'coding-system-category priority)))
1547 (set-default-coding-systems
1548 (if (memq eol-type '(0 1 2 unix dos mac))
1549 (coding-system-change-eol-conversion default-coding eol-type)
1550 default-coding))
1551 (setq default-sendmail-coding-system default-coding)
1552 (set-coding-priority categories)
1553 (while priority
1554 (set (car categories) (car priority))
1555 (setq priority (cdr priority) categories (cdr categories)))
1556 (update-coding-systems-internal)))))
1557
1558 ;; Print all arguments with `princ', then print "\n".
1559 (defsubst princ-list (&rest args)
1560 (while args (princ (car args)) (setq args (cdr args)))
1561 (princ "\n"))
1562
1563 (put 'describe-specified-language-support 'apropos-inhibit t)
1564
1565 ;; Print a language specific information such as input methods,
1566 ;; charsets, and coding systems. This function is intended to be
1567 ;; called from the menu:
1568 ;; [menu-bar mule describe-language-environment LANGUAGE]
1569 ;; and should not run it by `M-x describe-current-input-method-function'.
1570 (defun describe-specified-language-support ()
1571 "Describe how Emacs supports the specified language environment."
1572 (interactive)
1573 (let (language-name)
1574 (if (not (and (symbolp last-command-event)
1575 (or (not (eq last-command-event 'Default))
1576 (setq last-command-event 'English))
1577 (setq language-name (symbol-name last-command-event))))
1578 (error "Bogus calling sequence"))
1579 (describe-language-environment language-name)))
1580
1581 (defun describe-language-environment (language-name)
1582 "Describe how Emacs supports language environment LANGUAGE-NAME."
1583 (interactive
1584 (list (read-language-name
1585 'documentation
1586 "Describe language environment (default, current choice): ")))
1587 (if (null language-name)
1588 (setq language-name current-language-environment))
1589 (if (or (null language-name)
1590 (null (get-language-info language-name 'documentation)))
1591 (error "No documentation for the specified language"))
1592 (if (symbolp language-name)
1593 (setq language-name (symbol-name language-name)))
1594 (let ((doc (get-language-info language-name 'documentation))
1595 pos)
1596 (help-setup-xref (list #'describe-language-environment language-name)
1597 (interactive-p))
1598 (with-output-to-temp-buffer (help-buffer)
1599 (save-excursion
1600 (set-buffer standard-output)
1601 (insert language-name " language environment\n\n")
1602 (if (stringp doc)
1603 (insert doc "\n\n"))
1604 (condition-case nil
1605 (let ((str (eval (get-language-info language-name 'sample-text))))
1606 (if (stringp str)
1607 (insert "Sample text:\n " str "\n\n")))
1608 (error nil))
1609 (let ((input-method (get-language-info language-name 'input-method))
1610 (l (copy-sequence input-method-alist)))
1611 (insert "Input methods")
1612 (when input-method
1613 (insert " (default, " input-method ")")
1614 (setq input-method (assoc input-method input-method-alist))
1615 (setq l (cons input-method (delete input-method l))))
1616 (insert ":\n")
1617 (while l
1618 (when (string= language-name (nth 1 (car l)))
1619 (insert " " (car (car l)))
1620 (search-backward (car (car l)))
1621 (help-xref-button 0 'help-input-method (car (car l)))
1622 (goto-char (point-max))
1623 (insert " (\""
1624 (if (stringp (nth 3 (car l)))
1625 (nth 3 (car l))
1626 (car (nth 3 (car l))))
1627 "\" in mode line)\n"))
1628 (setq l (cdr l)))
1629 (insert "\n"))
1630 (insert "Character sets:\n")
1631 (let ((l (get-language-info language-name 'charset)))
1632 (if (null l)
1633 (insert " nothing specific to " language-name "\n")
1634 (while l
1635 (insert " " (symbol-name (car l)))
1636 (search-backward (symbol-name (car l)))
1637 (help-xref-button 0 'help-character-set (car l))
1638 (goto-char (point-max))
1639 (insert ": " (charset-description (car l)) "\n")
1640 (setq l (cdr l)))))
1641 (insert "\n")
1642 (insert "Coding systems:\n")
1643 (let ((l (get-language-info language-name 'coding-system)))
1644 (if (null l)
1645 (insert " nothing specific to " language-name "\n")
1646 (while l
1647 (insert " " (symbol-name (car l)))
1648 (search-backward (symbol-name (car l)))
1649 (help-xref-button 0 'help-coding-system (car l))
1650 (goto-char (point-max))
1651 (insert " (`"
1652 (coding-system-mnemonic (car l))
1653 "' in mode line):\n\t"
1654 (coding-system-doc-string (car l))
1655 "\n")
1656 (let ((aliases (coding-system-get (car l)
1657 'alias-coding-systems)))
1658 (when aliases
1659 (insert "\t(alias:")
1660 (while aliases
1661 (insert " " (symbol-name (car aliases)))
1662 (setq aliases (cdr aliases)))
1663 (insert ")\n")))
1664 (setq l (cdr l)))))))))
1665 \f
1666 ;;; Locales.
1667
1668 (defvar locale-translation-file-name nil
1669 "File name for the system's file of locale-name aliases, or nil if none.")
1670
1671 ;; The following definitions might as well be marked as constants and
1672 ;; purecopied, since they're normally used on startup, and probably
1673 ;; should reflect the facilities of the base Emacs.
1674 (defconst locale-language-names
1675 (purecopy
1676 '(
1677 ;; Locale names of the form LANGUAGE[_TERRITORY][.CODESET][@MODIFIER]
1678 ;; as specified in the Single Unix Spec, Version 2.
1679 ;; LANGUAGE is a language code taken from ISO 639:1988 (E/F)
1680 ;; with additions from ISO 639/RA Newsletter No.1/1989;
1681 ;; see Internet RFC 2165 (1997-06) and
1682 ;; http://www.evertype.com/standards/iso639/iso639-en.html
1683 ;; TERRITORY is a country code taken from ISO 3166
1684 ;; http://www.din.de/gremien/nas/nabd/iso3166ma/codlstp1/en_listp1.html.
1685 ;; CODESET and MODIFIER are implementation-dependent.
1686
1687 ; aa Afar
1688 ; ab Abkhazian
1689 ("af" . "Latin-1") ; Afrikaans
1690 ("am" . "Ethiopic") ; Amharic
1691 ; ar Arabic glibc uses 8859-6
1692 ; as Assamese
1693 ; ay Aymara
1694 ; az Azerbaijani
1695 ; ba Bashkir
1696 ("be" . "Belarussian") ; Belarussian [Byelorussian]
1697 ("bg" . "Bulgarian") ; Bulgarian
1698 ; bh Bihari
1699 ; bi Bislama
1700 ; bn Bengali, Bangla
1701 ("bo" . "Tibetan")
1702 ("br" . "Latin-1") ; Breton
1703 ("ca" . "Latin-1") ; Catalan
1704 ; co Corsican
1705 ("cs" . "Czech")
1706 ("cy" . "Latin-8") ; Welsh
1707 ("da" . "Latin-1") ; Danish
1708 ("de" . "German")
1709 ; dz Bhutani
1710 ("el" . "Greek")
1711 ;; Users who specify "en" explicitly typically want Latin-1, not ASCII.
1712 ("en" . "Latin-1") ; English
1713 ("eo" . "Latin-3") ; Esperanto
1714 ("es" . "Spanish")
1715 ("et" . "Latin-4") ; Estonian
1716 ("eu" . "Latin-1") ; Basque
1717 ; fa Persian
1718 ("fi" . "Latin-1") ; Finnish
1719 ; fj Fiji
1720 ("fo" . "Latin-1") ; Faroese
1721 ("fr" . "French") ; French
1722 ("fy" . "Latin-1") ; Frisian
1723 ("ga" . "Latin-1") ; Irish Gaelic (new orthography)
1724 ("gd" . "Latin-1") ; Scots Gaelic
1725 ("gl" . "Latin-1") ; Galician
1726 ; gn Guarani
1727 ; gu Gujarati
1728 ("gv" . "Latin-8") ; Manx Gaelic
1729 ; ha Hausa
1730 ("he" . "Hebrew")
1731 ("hi" . "Devanagari") ; Hindi glibc uses utf-8
1732 ("hr" . "Latin-2") ; Croatian
1733 ("hu" . "Latin-2") ; Hungarian
1734 ; hy Armenian
1735 ; ia Interlingua
1736 ("id" . "Latin-1") ; Indonesian
1737 ; ie Interlingue
1738 ; ik Inupiak
1739 ("is" . "Latin-1") ; Icelandic
1740 ("it" . "Latin-1") ; Italian
1741 ; iu Inuktitut
1742 ("ja" . "Japanese")
1743 ; jw Javanese
1744 ("ka" . "Georgian") ; Georgian
1745 ; kk Kazakh
1746 ("kl" . "Latin-1") ; Greenlandic
1747 ; km Cambodian
1748 ; kn Kannada
1749 ("ko" . "Korean")
1750 ; ks Kashmiri
1751 ; ku Kurdish
1752 ("kw" . "Latin-1") ; Cornish
1753 ; ky Kirghiz
1754 ("la" . "Latin-1") ; Latin
1755 ("lb" . "Latin-1") ; Luxemburgish
1756 ; ln Lingala
1757 ("lo" . "Lao") ; Laothian
1758 ("lt" . "Lithuanian")
1759 ("lv" . "Latvian") ; Latvian, Lettish
1760 ; mg Malagasy
1761 ("mi" . "Latin-7") ; Maori
1762 ("mk" . "Latin-5") ; Macedonian
1763 ; ml Malayalam
1764 ; mn Mongolian
1765 ; mo Moldavian
1766 ("mr" . "Devanagari") ; Marathi glibc uses utf-8
1767 ("ms" . "Latin-1") ; Malay
1768 ("mt" . "Latin-3") ; Maltese
1769 ; my Burmese
1770 ; na Nauru
1771 ("ne" . "Devanagari") ; Nepali
1772 ("nl" . "Dutch")
1773 ("no" . "Latin-1") ; Norwegian
1774 ("oc" . "Latin-1") ; Occitan
1775 ; om (Afan) Oromo
1776 ; or Oriya
1777 ; pa Punjabi
1778 ("pl" . "Latin-2") ; Polish
1779 ; ps Pashto, Pushto
1780 ("pt" . "Latin-1") ; Portuguese
1781 ; qu Quechua
1782 ("rm" . "Latin-1") ; Rhaeto-Romanic
1783 ; rn Kirundi
1784 ("ro" . "Romanian")
1785 ("ru.*[_.]koi8" . "Cyrillic-KOI8") ; Russian
1786 ("ru" . "Latin-5") ; Russian
1787 ; rw Kinyarwanda
1788 ("sa" . "Devanagari") ; Sanskrit
1789 ; sd Sindhi
1790 ; se Northern Sami
1791 ; sg Sangho
1792 ("sh" . "Latin-2") ; Serbo-Croatian
1793 ; si Sinhalese
1794 ("sk" . "Slovak")
1795 ("sl" . "Slovenian")
1796 ; sm Samoan
1797 ; sn Shona
1798 ; so Somali
1799 ("sq" . "Latin-1") ; Albanian
1800 ("sr" . "Latin-2") ; Serbian (Latin alphabet)
1801 ; ss Siswati
1802 ; st Sesotho
1803 ; su Sundanese
1804 ("sv" . "Latin-1") ; Swedish
1805 ("sw" . "Latin-1") ; Swahili
1806 ; ta Tamil glibc uses utf-8
1807 ; te Telugu glibc uses utf-8
1808 ("tg" . "Cyrillic-KOI8-T") ; Tajik
1809 ("th" . "Thai")
1810 ; ti Tigrinya
1811 ; tk Turkmen
1812 ("tl" . "Latin-1") ; Tagalog
1813 ; tn Setswana
1814 ; to Tonga
1815 ("tr" . "Latin-5") ; Turkish
1816 ; ts Tsonga
1817 ; tt Tatar
1818 ; tw Twi
1819 ; ug Uighur
1820 ("uk" . "Ukrainian") ; Ukrainian
1821 ; ur Urdu glibc uses utf-8
1822 ("uz" . "Latin-1") ; Uzbek
1823 ("vi" . "Vietnamese") ; glibc uses utf-8
1824 ; vo Volapuk
1825 ; wo Wolof
1826 ; xh Xhosa
1827 ("yi" . "Windows-1255") ; Yiddish
1828 ; yo Yoruba
1829 ; za Zhuang
1830
1831 ; glibc:
1832 ; zh_CN.GB18030/GB18030 \
1833 ; zh_CN.GBK/GBK \
1834 ; zh_HK/BIG5-HKSCS \
1835 ; zh_TW/BIG5 \
1836 ; zh_TW.EUC-TW/EUC-TW \
1837
1838 ("zh.*[._]big5" . "Chinese-BIG5")
1839 ("zh.*[._]gbk" . nil) ; Solaris 2.7; has gbk-0 as well as GB 2312.1980-0
1840 ("zh_tw" . "Chinese-CNS")
1841 ("zh" . "Chinese-GB")
1842 ; zu Zulu
1843
1844 ;; ISO standard locales
1845 ("c$" . "ASCII")
1846 ("posix$" . "ASCII")
1847
1848 ;; The "IPA" Emacs language environment does not correspond
1849 ;; to any ISO 639 code, so let it stand for itself.
1850 ("ipa$" . "IPA")
1851
1852 ;; Nonstandard or obsolete language codes
1853 ("cz" . "Czech") ; e.g. Solaris 2.6
1854 ("ee" . "Latin-4") ; Estonian, e.g. X11R6.4
1855 ("iw" . "Hebrew") ; e.g. X11R6.4
1856 ("sp" . "Latin-5") ; Serbian (Cyrillic alphabet), e.g. X11R6.4
1857 ("su" . "Latin-1") ; Finnish, e.g. Solaris 2.6
1858 ("jp" . "Japanese") ; e.g. MS Windows
1859 ("chs" . "Chinese-GB") ; MS Windows Chinese Simplified
1860 ("cht" . "Chinese-BIG5") ; MS Windows Chinese Traditional
1861 ))
1862 "List of pairs of locale regexps and language names.
1863 The first element whose locale regexp matches the start of a downcased locale
1864 specifies the language name corresponding to that locale.
1865 If the language name is nil, there is no corresponding language environment.")
1866
1867 (defconst locale-charset-language-names
1868 (purecopy
1869 '((".*8859[-_]?1\\>" . "Latin-1")
1870 (".*8859[-_]?2\\>" . "Latin-2")
1871 (".*8859[-_]?3\\>" . "Latin-3")
1872 (".*8859[-_]?4\\>" . "Latin-4")
1873 (".*8859[-_]?9\\>" . "Latin-5")
1874 (".*8859[-_]?14\\>" . "Latin-8")
1875 (".*8859[-_]?15\\>" . "Latin-9")
1876 (".*@euro\\>" . "Latin-9")
1877 (".*utf\\(-?8\\)\\>" . "UTF-8")))
1878 "List of pairs of locale regexps and charset language names.
1879 The first element whose locale regexp matches the start of a downcased locale
1880 specifies the language name whose charsets corresponds to that locale.
1881 This language name is used if its charsets disagree with the charsets of
1882 the language name that would otherwise be used for this locale.")
1883
1884 (defconst locale-preferred-coding-systems
1885 (purecopy
1886 '(("ja.*[._]euc" . japanese-iso-8bit)
1887 ("ja.*[._]jis7" . iso-2022-jp)
1888 ("ja.*[._]pck" . japanese-shift-jis)
1889 ("ja.*[._]sjis" . japanese-shift-jis)
1890 (".*[._]utf" . utf-8)))
1891 "List of pairs of locale regexps and preferred coding systems.
1892 The first element whose locale regexp matches the start of a downcased locale
1893 specifies the coding system to prefer when using that locale.")
1894
1895 (defconst standard-keyboard-coding-systems
1896 (purecopy
1897 '(iso-latin-1 iso-latin-2 iso-latin-3 iso-latin-4 iso-latin-5
1898 iso-latin-6 iso-latin-7 iso-latin-8 iso-latin-9))
1899 "Coding systems that are commonly used for keyboards.
1900 `set-locale-environment' will set the `keyboard-coding-system' if the
1901 coding-system specified by the locale setting is a member of this list.")
1902
1903 (defun locale-name-match (key alist)
1904 "Search for KEY in ALIST, which should be a list of regexp-value pairs.
1905 Return the value corresponding to the first regexp that matches the
1906 start of KEY, or nil if there is no match."
1907 (let (element)
1908 (while (and alist (not element))
1909 (if (string-match (concat "\\`\\(?:" (car (car alist)) "\\)") key)
1910 (setq element (car alist)))
1911 (setq alist (cdr alist)))
1912 (cdr element)))
1913
1914 (defun set-locale-environment (&optional locale-name)
1915 "Set up multi-lingual environment for using LOCALE-NAME.
1916 This sets the language environment, the coding system priority,
1917 the default input method and sometimes other things.
1918
1919 LOCALE-NAME should be a string
1920 which is the name of a locale supported by the system;
1921 often it is of the form xx_XX.CODE, where xx is a language,
1922 XX is a country, and CODE specifies a character set and coding system.
1923 For example, the locale name \"ja_JP.EUC\" might name a locale
1924 for Japanese in Japan using the `japanese-iso-8bit' coding-system.
1925
1926 If LOCALE-NAME is nil, its value is taken from the environment
1927 variables LC_ALL, LC_CTYLE and LANG (the first one that is set).
1928
1929 The locale names supported by your system can typically be found in a
1930 directory named `/usr/share/locale' or `/usr/lib/locale'. LOCALE-NAME
1931 will be translated according to the table specified by
1932 `locale-translation-file-name'.
1933
1934 See also `locale-charset-language-names', `locale-language-names',
1935 `locale-preferred-coding-systems' and `locale-coding-system'."
1936 (interactive "sSet environment for locale: ")
1937
1938 ;; Do this at runtime for the sake of binaries possibly transported
1939 ;; to a system without X.
1940 (setq locale-translation-file-name
1941 (let ((files
1942 '("/usr/lib/X11/locale/locale.alias" ; e.g. X11R6.4
1943 "/usr/X11R6/lib/X11/locale/locale.alias" ; e.g. RedHat 4.2
1944 "/usr/openwin/lib/locale/locale.alias" ; e.g. Solaris 2.6
1945 ;;
1946 ;; The following name appears after the X-related names above,
1947 ;; since the X-related names are what X actually uses.
1948 "/usr/share/locale/locale.alias" ; GNU/Linux sans X
1949 )))
1950 (while (and files (not (file-exists-p (car files))))
1951 (setq files (cdr files)))
1952 (car files)))
1953
1954 (let ((locale locale-name))
1955
1956 (unless locale
1957 ;; Use the first of these three environment variables
1958 ;; that has a nonempty value.
1959 (let ((vars '("LC_ALL" "LC_CTYPE" "LANG")))
1960 (while (and vars (not (setq locale (getenv (car vars)))))
1961 (setq vars (cdr vars)))))
1962
1963 (when locale
1964
1965 ;; Translate "swedish" into "sv_SE.ISO8859-1", and so on,
1966 ;; using the translation file that many systems have.
1967 (when locale-translation-file-name
1968 (with-temp-buffer
1969 (insert-file-contents locale-translation-file-name)
1970 (when (re-search-forward
1971 (concat "^" (regexp-quote locale) ":?[ \t]+") nil t)
1972 (setq locale (buffer-substring (point) (line-end-position))))))
1973
1974 ;; Leave the system locales alone if the caller did not specify
1975 ;; an explicit locale name, as their defaults are set from
1976 ;; LC_MESSAGES and LC_TIME, not LC_CTYPE, and the user might not
1977 ;; want to set them to the same value as LC_CTYPE.
1978 (when locale-name
1979 (setq system-messages-locale locale)
1980 (setq system-time-locale locale))
1981
1982 (setq locale (downcase locale))
1983
1984 (let ((language-name
1985 (locale-name-match locale locale-language-names))
1986 (charset-language-name
1987 (locale-name-match locale locale-charset-language-names))
1988 (coding-system
1989 (locale-name-match locale locale-preferred-coding-systems)))
1990
1991 ;; Give preference to charset-language-name over language-name.
1992 (if (and charset-language-name
1993 (not
1994 (equal (get-language-info language-name 'charset)
1995 (get-language-info charset-language-name 'charset))))
1996 (setq language-name charset-language-name))
1997
1998 (when language-name
1999
2000 ;; Set up for this character set. This is now the right way
2001 ;; to do it for both unibyte and multibyte modes.
2002 (set-language-environment language-name)
2003
2004 ;; If default-enable-multibyte-characters is nil,
2005 ;; we are using single-byte characters,
2006 ;; so the display table and terminal coding system are irrelevant.
2007 (when default-enable-multibyte-characters
2008 (set-display-table-and-terminal-coding-system language-name))
2009
2010 ;; Set the `keyboard-coding-system' if appropriate.
2011 (let ((kcs (or coding-system
2012 (car (get-language-info language-name
2013 'coding-system)))))
2014 (if (memq kcs standard-keyboard-coding-systems)
2015 (set-keyboard-coding-system kcs)))
2016
2017 (setq locale-coding-system
2018 (car (get-language-info language-name 'coding-priority))))
2019
2020 (when coding-system
2021 (prefer-coding-system coding-system)
2022 (setq locale-coding-system coding-system))))))
2023 \f
2024 ;;; Charset property
2025
2026 (defun get-charset-property (charset propname)
2027 "Return the value of CHARSET's PROPNAME property.
2028 This is the last value stored with
2029 (put-charset-property CHARSET PROPNAME VALUE)."
2030 (and (not (eq charset 'composition))
2031 (plist-get (charset-plist charset) propname)))
2032
2033 (defun put-charset-property (charset propname value)
2034 "Store CHARSETS's PROPNAME property with value VALUE.
2035 It can be retrieved with `(get-charset-property CHARSET PROPNAME)'."
2036 (or (eq charset 'composition)
2037 (set-charset-plist charset
2038 (plist-put (charset-plist charset) propname value))))
2039
2040 ;;; Character code property
2041 (put 'char-code-property-table 'char-table-extra-slots 0)
2042
2043 (defvar char-code-property-table
2044 (make-char-table 'char-code-property-table)
2045 "Char-table containing a property list of each character code.
2046
2047 See also the documentation of `get-char-code-property' and
2048 `put-char-code-property'.")
2049
2050 (defun get-char-code-property (char propname)
2051 "Return the value of CHAR's PROPNAME property in `char-code-property-table'."
2052 (let ((plist (aref char-code-property-table char)))
2053 (if (listp plist)
2054 (car (cdr (memq propname plist))))))
2055
2056 (defun put-char-code-property (char propname value)
2057 "Store CHAR's PROPNAME property with VALUE in `char-code-property-table'.
2058 It can be retrieved with `(get-char-code-property CHAR PROPNAME)'."
2059 (let ((plist (aref char-code-property-table char)))
2060 (if plist
2061 (let ((slot (memq propname plist)))
2062 (if slot
2063 (setcar (cdr slot) value)
2064 (nconc plist (list propname value))))
2065 (aset char-code-property-table char (list propname value)))))
2066
2067 \f
2068 ;; Pretty description of encoded string
2069
2070 ;; Alist of ISO 2022 control code vs the corresponding mnemonic string.
2071 (defvar iso-2022-control-alist
2072 '((?\x1b . "ESC")
2073 (?\x0e . "SO")
2074 (?\x0f . "SI")
2075 (?\x8e . "SS2")
2076 (?\x8f . "SS3")
2077 (?\x9b . "CSI")))
2078
2079 (defun encoded-string-description (str coding-system)
2080 "Return a pretty description of STR that is encoded by CODING-SYSTEM."
2081 (setq str (string-as-unibyte str))
2082 (mapconcat
2083 (if (and coding-system (eq (coding-system-type coding-system) 2))
2084 ;; Try to get a pretty description for ISO 2022 escape sequences.
2085 (function (lambda (x) (or (cdr (assq x iso-2022-control-alist))
2086 (format "%02X" x))))
2087 (function (lambda (x) (format "0x%02X" x))))
2088 str " "))
2089
2090 (defun encode-coding-char (char coding-system)
2091 "Encode CHAR by CODING-SYSTEM and return the resulting string.
2092 If CODING-SYSTEM can't safely encode CHAR, return nil."
2093 (let ((str1 (string-as-multibyte (char-to-string char)))
2094 (str2 (string-as-multibyte (make-string 2 char)))
2095 (safe-chars (and coding-system
2096 (coding-system-get coding-system 'safe-chars)))
2097 (charset (char-charset char))
2098 enc1 enc2 i1 i2)
2099 (when (or (eq safe-chars t)
2100 (eq charset 'ascii)
2101 (and safe-chars (aref safe-chars char)))
2102 ;; We must find the encoded string of CHAR. But, just encoding
2103 ;; CHAR will put extra control sequences (usually to designate
2104 ;; ASCII charaset) at the tail if type of CODING is ISO 2022.
2105 ;; To exclude such tailing bytes, we at first encode one-char
2106 ;; string and two-char string, then check how many bytes at the
2107 ;; tail of both encoded strings are the same.
2108
2109 (setq enc1 (encode-coding-string str1 coding-system)
2110 i1 (length enc1)
2111 enc2 (encode-coding-string str2 coding-system)
2112 i2 (length enc2))
2113 (while (and (> i1 0) (= (aref enc1 (1- i1)) (aref enc2 (1- i2))))
2114 (setq i1 (1- i1) i2 (1- i2)))
2115
2116 ;; Now (substring enc1 i1) and (substring enc2 i2) are the same,
2117 ;; and they are the extra control sequences at the tail to
2118 ;; exclude.
2119 (substring enc2 0 i2))))
2120
2121
2122 ;;; mule-cmds.el ends here