Merge from emacs-24; up to 2013-01-03T02:31:36Z!rgm@gnu.org
[emacs.git] / lisp / international / mule.el
blob9af6bc369a4d297341d8bbe0e165542bcc0b1d8d
1 ;;; mule.el --- basic commands for multilingual environment
3 ;; Copyright (C) 1997-2013 Free Software Foundation, Inc.
4 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
5 ;; 2005, 2006, 2007, 2008, 2009, 2010, 2011
6 ;; National Institute of Advanced Industrial Science and Technology (AIST)
7 ;; Registration Number H14PRO021
8 ;; Copyright (C) 2003
9 ;; National Institute of Advanced Industrial Science and Technology (AIST)
10 ;; Registration Number H13PRO009
12 ;; Keywords: mule, multilingual, character set, coding system
14 ;; This file is part of GNU Emacs.
16 ;; GNU Emacs is free software: you can redistribute it and/or modify
17 ;; it under the terms of the GNU General Public License as published by
18 ;; the Free Software Foundation, either version 3 of the License, or
19 ;; (at your option) any later version.
21 ;; GNU Emacs is distributed in the hope that it will be useful,
22 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
23 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
24 ;; GNU General Public License for more details.
26 ;; You should have received a copy of the GNU General Public License
27 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
29 ;;; Commentary:
31 ;;; Code:
33 ;; FIXME? Are these still relevant? Nothing uses them AFAICS.
34 (defconst mule-version "6.0 (HANACHIRUSATO)" "\
35 Version number and name of this version of MULE (multilingual environment).")
37 (defconst mule-version-date "2003.9.1" "\
38 Distribution date of this version of MULE (multilingual environment).")
41 ;;; CHARSET
43 ;; Backward compatibility code for handling emacs-mule charsets.
44 (defvar private-char-area-1-min #xF0000)
45 (defvar private-char-area-1-max #xFFFFE)
46 (defvar private-char-area-2-min #x100000)
47 (defvar private-char-area-2-max #x10FFFE)
49 ;; Table of emacs-mule charsets indexed by their emacs-mule ID.
50 (defvar emacs-mule-charset-table (make-vector 256 nil))
51 (aset emacs-mule-charset-table 0 'ascii)
53 ;; Convert the argument of old-style call of define-charset to a
54 ;; property list used by the new-style.
55 ;; INFO-VECTOR is a vector of the format:
56 ;; [DIMENSION CHARS WIDTH DIRECTION ISO-FINAL-CHAR ISO-GRAPHIC-PLANE
57 ;; SHORT-NAME LONG-NAME DESCRIPTION]
59 (defun convert-define-charset-argument (emacs-mule-id info-vector)
60 (let* ((dim (aref info-vector 0))
61 (chars (aref info-vector 1))
62 (total (if (= dim 1) chars (* chars chars)))
63 (code-space (if (= dim 1) (if (= chars 96) [32 127] [33 126])
64 (if (= chars 96) [32 127 32 127] [33 126 33 126])))
65 code-offset)
66 (if (integerp emacs-mule-id)
67 (or (= emacs-mule-id 0)
68 (and (>= emacs-mule-id 129) (< emacs-mule-id 256))
69 (error "Invalid CHARSET-ID: %d" emacs-mule-id))
70 (let (from-id to-id)
71 (if (= dim 1) (setq from-id 160 to-id 224)
72 (setq from-id 224 to-id 255))
73 (while (and (< from-id to-id)
74 (not (aref emacs-mule-charset-table from-id)))
75 (setq from-id (1+ from-id)))
76 (if (= from-id to-id)
77 (error "No more room for the new Emacs-mule charset"))
78 (setq emacs-mule-id from-id)))
79 (if (> (- private-char-area-1-max private-char-area-1-min) total)
80 (setq code-offset private-char-area-1-min
81 private-char-area-1-min (+ private-char-area-1-min total))
82 (if (> (- private-char-area-2-max private-char-area-2-min) total)
83 (setq code-offset private-char-area-2-min
84 private-char-area-2-min (+ private-char-area-2-min total))
85 (error "No more space for a new charset")))
86 (list :dimension dim
87 :code-space code-space
88 :iso-final-char (aref info-vector 4)
89 :code-offset code-offset
90 :emacs-mule-id emacs-mule-id)))
92 (defun define-charset (name docstring &rest props)
93 "Define NAME (symbol) as a charset with DOCSTRING.
94 The remaining arguments must come in pairs ATTRIBUTE VALUE. ATTRIBUTE
95 may be any symbol. The following have special meanings, and one of
96 `:code-offset', `:map', `:subset', `:superset' must be specified.
98 `:short-name'
100 VALUE must be a short string to identify the charset. If omitted,
101 NAME is used.
103 `:long-name'
105 VALUE must be a string longer than `:short-name' to identify the
106 charset. If omitted, the value of the `:short-name' attribute is used.
108 `:dimension'
110 VALUE must be an integer 0, 1, 2, or 3, specifying the dimension of
111 code-points of the charsets. If omitted, it is calculated from the
112 value of the `:code-space' attribute.
114 `:code-space'
116 VALUE must be a vector of length at most 8 specifying the byte code
117 range of each dimension in this format:
118 [ MIN-1 MAX-1 MIN-2 MAX-2 ... ]
119 where MIN-N is the minimum byte value of Nth dimension of code-point,
120 MAX-N is the maximum byte value of that.
122 `:min-code'
124 VALUE must be an integer specifying the minimum code point of the
125 charset. If omitted, it is calculated from `:code-space'. VALUE may
126 be a cons (HIGH . LOW), where HIGH is the most significant 16 bits of
127 the code point and LOW is the least significant 16 bits.
129 `:max-code'
131 VALUE must be an integer specifying the maximum code point of the
132 charset. If omitted, it is calculated from `:code-space'. VALUE may
133 be a cons (HIGH . LOW), where HIGH is the most significant 16 bits of
134 the code point and LOW is the least significant 16 bits.
136 `:iso-final-char'
138 VALUE must be a character in the range 32 to 127 (inclusive)
139 specifying the final char of the charset for ISO-2022 encoding. If
140 omitted, the charset can't be encoded by ISO-2022 based
141 coding-systems.
143 `:iso-revision-number'
145 VALUE must be an integer in the range 0..63, specifying the revision
146 number of the charset for ISO-2022 encoding.
148 `:emacs-mule-id'
150 VALUE must be an integer of 0, 129..255. If omitted, the charset
151 can't be encoded by coding-systems of type `emacs-mule'.
153 `:ascii-compatible-p'
155 VALUE must be nil or t (default nil). If VALUE is t, the charset is
156 compatible with ASCII, i.e. the first 128 code points map to ASCII.
158 `:supplementary-p'
160 VALUE must be nil or t. If the VALUE is t, the charset is
161 supplementary, which means it is used only as a parent or a
162 subset of some other charset, or it is provided just for backward
163 compatibility.
165 `:invalid-code'
167 VALUE must be a nonnegative integer that can be used as an invalid
168 code point of the charset. If the minimum code is 0 and the maximum
169 code is greater than Emacs's maximum integer value, `:invalid-code'
170 should not be omitted.
172 `:code-offset'
174 VALUE must be an integer added to the index number of a character to
175 get the corresponding character code.
177 `:map'
179 VALUE must be vector or string.
181 If it is a vector, the format is [ CODE-1 CHAR-1 CODE-2 CHAR-2 ... ],
182 where CODE-n is a code-point of the charset, and CHAR-n is the
183 corresponding character code.
185 If it is a string, it is a name of file that contains the above
186 information. Each line of the file must be this format:
187 0xXXX 0xYYY
188 where XXX is a hexadecimal representation of CODE-n and YYY is a
189 hexadecimal representation of CHAR-n. A line starting with `#' is a
190 comment line.
192 `:subset'
194 VALUE must be a list:
195 ( PARENT MIN-CODE MAX-CODE OFFSET )
196 PARENT is a parent charset. MIN-CODE and MAX-CODE specify the range
197 of characters inherited from the parent. OFFSET is an integer value
198 to add to a code point of the parent charset to get the corresponding
199 code point of this charset.
201 `:superset'
203 VALUE must be a list of parent charsets. The charset inherits
204 characters from them. Each element of the list may be a cons (PARENT
205 . OFFSET), where PARENT is a parent charset, and OFFSET is an offset
206 value to add to a code point of PARENT to get the corresponding code
207 point of this charset.
209 `:unify-map'
211 VALUE must be vector or string.
213 If it is a vector, the format is [ CODE-1 CHAR-1 CODE-2 CHAR-2 ... ],
214 where CODE-n is a code-point of the charset, and CHAR-n is the
215 corresponding Unicode character code.
217 If it is a string, it is a name of file that contains the above
218 information. The file format is the same as what described for `:map'
219 attribute."
220 (when (vectorp (car props))
221 ;; Old style code:
222 ;; (define-charset CHARSET-ID CHARSET-SYMBOL INFO-VECTOR)
223 ;; Convert the argument to make it fit with the current style.
224 (let ((vec (car props)))
225 (setq props (convert-define-charset-argument name vec)
226 name docstring
227 docstring (aref vec 8))))
228 (let ((attrs (mapcar 'list '(:dimension
229 :code-space
230 :min-code
231 :max-code
232 :iso-final-char
233 :iso-revision-number
234 :emacs-mule-id
235 :ascii-compatible-p
236 :supplementary-p
237 :invalid-code
238 :code-offset
239 :map
240 :subset
241 :superset
242 :unify-map
243 :plist))))
245 ;; If :dimension is omitted, get the dimension from :code-space.
246 (let ((dimension (plist-get props :dimension)))
247 (or dimension
248 (let ((code-space (plist-get props :code-space)))
249 (setq dimension (if code-space (/ (length code-space) 2) 4))
250 (setq props (plist-put props :dimension dimension)))))
252 (let ((code-space (plist-get props :code-space)))
253 (or code-space
254 (let ((dimension (plist-get props :dimension)))
255 (setq code-space (make-vector 8 0))
256 (dotimes (i dimension)
257 (aset code-space (1+ (* i 2)) #xFF))
258 (setq props (plist-put props :code-space code-space)))))
260 ;; If :emacs-mule-id is specified, update emacs-mule-charset-table.
261 (let ((emacs-mule-id (plist-get props :emacs-mule-id)))
262 (if (integerp emacs-mule-id)
263 (aset emacs-mule-charset-table emacs-mule-id name)))
265 (dolist (slot attrs)
266 (setcdr slot (purecopy (plist-get props (car slot)))))
268 ;; Make sure that the value of :code-space is a vector of 8
269 ;; elements.
270 (let* ((slot (assq :code-space attrs))
271 (val (cdr slot))
272 (len (length val)))
273 (if (< len 8)
274 (setcdr slot
275 (vconcat val (make-vector (- 8 len) 0)))))
277 ;; Add :name and :docstring properties to PROPS.
278 (setq props
279 (cons :name (cons name (cons :docstring (cons (purecopy docstring) props)))))
280 (or (plist-get props :short-name)
281 (plist-put props :short-name (symbol-name name)))
282 (or (plist-get props :long-name)
283 (plist-put props :long-name (plist-get props :short-name)))
284 (plist-put props :base name)
285 ;; We can probably get a worthwhile amount in purespace.
286 (setq props
287 (mapcar (lambda (elt)
288 (if (stringp elt)
289 (purecopy elt)
290 elt))
291 props))
292 (setcdr (assq :plist attrs) props)
294 (apply 'define-charset-internal name (mapcar 'cdr attrs))))
297 (defun load-with-code-conversion (fullname file &optional noerror nomessage)
298 "Execute a file of Lisp code named FILE whose absolute name is FULLNAME.
299 The file contents are decoded before evaluation if necessary.
300 If optional third arg NOERROR is non-nil,
301 report no error if FILE doesn't exist.
302 Print messages at start and end of loading unless
303 optional fourth arg NOMESSAGE is non-nil.
304 Return t if file exists."
305 (if (null (file-readable-p fullname))
306 (and (null noerror)
307 (signal 'file-error (list "Cannot open load file" file)))
308 ;; Read file with code conversion, and then eval.
309 (let* ((buffer
310 ;; We can't use `generate-new-buffer' because files.el
311 ;; is not yet loaded.
312 (get-buffer-create (generate-new-buffer-name " *load*")))
313 (load-in-progress t)
314 (source (save-match-data (string-match "\\.el\\'" fullname))))
315 (unless nomessage
316 (if source
317 (message "Loading %s (source)..." file)
318 (message "Loading %s..." file)))
319 (when purify-flag
320 (push (purecopy file) preloaded-file-list))
321 (unwind-protect
322 (let ((load-file-name fullname)
323 (set-auto-coding-for-load t)
324 (inhibit-file-name-operation nil))
325 (with-current-buffer buffer
326 ;; So that we don't get completely screwed if the
327 ;; file is encoded in some complicated character set,
328 ;; read it with real decoding, as a multibyte buffer.
329 (set-buffer-multibyte t)
330 ;; Don't let deactivate-mark remain set.
331 (let (deactivate-mark)
332 (insert-file-contents fullname))
333 ;; If the loaded file was inserted with no-conversion or
334 ;; raw-text coding system, make the buffer unibyte.
335 ;; Otherwise, eval-buffer might try to interpret random
336 ;; binary junk as multibyte characters.
337 (if (and enable-multibyte-characters
338 (or (eq (coding-system-type last-coding-system-used)
339 'raw-text)))
340 (set-buffer-multibyte nil))
341 ;; Make `kill-buffer' quiet.
342 (set-buffer-modified-p nil))
343 ;; Have the original buffer current while we eval.
344 (eval-buffer buffer nil
345 ;; This is compatible with what `load' does.
346 (if purify-flag file fullname)
347 nil t))
348 (let (kill-buffer-hook kill-buffer-query-functions)
349 (kill-buffer buffer)))
350 (do-after-load-evaluation fullname)
352 (unless (or nomessage noninteractive)
353 (if source
354 (message "Loading %s (source)...done" file)
355 (message "Loading %s...done" file)))
356 t)))
358 (defun charset-info (charset)
359 "Return a vector of information of CHARSET.
360 This function is provided for backward compatibility.
362 The elements of the vector are:
363 CHARSET-ID, BYTES, DIMENSION, CHARS, WIDTH, DIRECTION,
364 LEADING-CODE-BASE, LEADING-CODE-EXT,
365 ISO-FINAL-CHAR, ISO-GRAPHIC-PLANE,
366 REVERSE-CHARSET, SHORT-NAME, LONG-NAME, DESCRIPTION,
367 PLIST.
368 where
369 CHARSET-ID is always 0.
370 BYTES is always 0.
371 DIMENSION is the number of bytes of a code-point of the charset:
372 1, 2, 3, or 4.
373 CHARS is the number of characters in a dimension:
374 94, 96, 128, or 256.
375 WIDTH is always 0.
376 DIRECTION is always 0.
377 LEADING-CODE-BASE is always 0.
378 LEADING-CODE-EXT is always 0.
379 ISO-FINAL-CHAR (character) is the final character of the
380 corresponding ISO 2022 charset. If the charset is not assigned
381 any final character, the value is -1.
382 ISO-GRAPHIC-PLANE is always 0.
383 REVERSE-CHARSET is always -1.
384 SHORT-NAME (string) is the short name to refer to the charset.
385 LONG-NAME (string) is the long name to refer to the charset
386 DESCRIPTION (string) is the description string of the charset.
387 PLIST (property list) may contain any type of information a user
388 want to put and get by functions `put-charset-property' and
389 `get-charset-property' respectively."
390 (vector 0
392 (charset-dimension charset)
393 (charset-chars charset)
398 (charset-iso-final-char charset)
401 (get-charset-property charset :short-name)
402 (get-charset-property charset :short-name)
403 (charset-description charset)
404 (charset-plist charset)))
406 ;; It is better not to use backquote in this file,
407 ;; because that makes a bootstrapping problem
408 ;; if you need to recompile all the Lisp files using interpreted code.
410 (defun charset-id (_charset)
411 "Always return 0. This is provided for backward compatibility."
412 (declare (obsolete nil "23.1"))
415 (defmacro charset-bytes (_charset)
416 "Always return 0. This is provided for backward compatibility."
417 (declare (obsolete nil "23.1"))
420 (defun get-charset-property (charset propname)
421 "Return the value of CHARSET's PROPNAME property.
422 This is the last value stored with
423 (put-charset-property CHARSET PROPNAME VALUE)."
424 (plist-get (charset-plist charset) propname))
426 (defun put-charset-property (charset propname value)
427 "Set CHARSETS's PROPNAME property to value VALUE.
428 It can be retrieved with `(get-charset-property CHARSET PROPNAME)'."
429 (set-charset-plist charset
430 (plist-put (charset-plist charset) propname
431 (if (stringp value)
432 (purecopy value)
433 value))))
435 (defun charset-description (charset)
436 "Return description string of CHARSET."
437 (plist-get (charset-plist charset) :docstring))
439 (defun charset-dimension (charset)
440 "Return dimension of CHARSET."
441 (plist-get (charset-plist charset) :dimension))
443 (defun charset-chars (charset &optional dimension)
444 "Return number of characters contained in DIMENSION of CHARSET.
445 DIMENSION defaults to the first dimension."
446 (unless dimension (setq dimension 1))
447 (let ((code-space (plist-get (charset-plist charset) :code-space)))
448 (1+ (- (aref code-space (1- (* 2 dimension)))
449 (aref code-space (- (* 2 dimension) 2))))))
451 (defun charset-iso-final-char (charset)
452 "Return ISO-2022 final character of CHARSET.
453 Return -1 if charset isn't an ISO 2022 one."
454 (or (plist-get (charset-plist charset) :iso-final-char)
455 -1))
457 (defmacro charset-short-name (charset)
458 "Return short name of CHARSET."
459 (plist-get (charset-plist charset) :short-name))
461 (defmacro charset-long-name (charset)
462 "Return long name of CHARSET."
463 (plist-get (charset-plist charset) :long-name))
465 (defun charset-list ()
466 "Return list of all charsets ever defined."
467 (declare (obsolete charset-list "23.1"))
468 charset-list)
471 ;;; CHARACTER
472 (define-obsolete-function-alias 'char-valid-p 'characterp "23.1")
474 (defun generic-char-p (_char)
475 "Always return nil. This is provided for backward compatibility."
476 (declare (obsolete nil "23.1"))
477 nil)
479 (defun make-char-internal (charset-id &optional code1 code2)
480 (let ((charset (aref emacs-mule-charset-table charset-id)))
481 (or charset
482 (error "Invalid Emacs-mule charset ID: %d" charset-id))
483 (make-char charset code1 code2)))
485 ;; Save the ASCII case table in case we need it later. Some locales
486 ;; (such as Turkish) modify the case behavior of ASCII characters,
487 ;; which can interfere with networking code that uses ASCII strings.
489 (defvar ascii-case-table
490 ;; Code copied from copy-case-table to avoid requiring case-table.el
491 (let ((tbl (copy-sequence (standard-case-table)))
492 (up (char-table-extra-slot (standard-case-table) 0)))
493 (if up (set-char-table-extra-slot tbl 0 (copy-sequence up)))
494 (set-char-table-extra-slot tbl 1 nil)
495 (set-char-table-extra-slot tbl 2 nil)
496 tbl)
497 "Case table for the ASCII character set.")
499 ;; Coding system stuff
501 ;; Coding system is a symbol that has been defined by the function
502 ;; `define-coding-system'.
504 (defconst coding-system-iso-2022-flags
505 '(long-form
506 ascii-at-eol
507 ascii-at-cntl
508 7-bit
509 locking-shift
510 single-shift
511 designation
512 revision
513 direction
514 init-at-bol
515 designate-at-bol
516 safe
517 latin-extra
518 composition
519 euc-tw-shift
520 use-roman
521 use-oldjis
522 8-bit-level-4)
523 "List of symbols that control ISO-2022 encoder/decoder.
525 The value of the `:flags' attribute in the argument of the function
526 `define-coding-system' must be one of them.
528 If `long-form' is specified, use a long designation sequence on
529 encoding for the charsets `japanese-jisx0208-1978', `chinese-gb2312',
530 and `japanese-jisx0208'. The long designation sequence doesn't
531 conform to ISO 2022, but is used by such coding systems as
532 `compound-text'.
534 If `ascii-at-eol' is specified, designate ASCII to g0 at end of line
535 on encoding.
537 If `ascii-at-cntl' is specified, designate ASCII to g0 before control
538 codes and SPC on encoding.
540 If `7-bit' is specified, use 7-bit code only on encoding.
542 If `locking-shift' is specified, decode locking-shift code correctly
543 on decoding, and use locking-shift to invoke a graphic element on
544 encoding.
546 If `single-shift' is specified, decode single-shift code
547 correctly on decoding, and use single-shift to invoke a graphic
548 element on encoding. See also `8-bit-level-4' specification.
550 If `designation' is specified, decode designation code correctly on
551 decoding, and use designation to designate a charset to a graphic
552 element on encoding.
554 If `revision' is specified, produce an escape sequence to specify
555 revision number of a charset on encoding. Such an escape sequence is
556 always correctly decoded on decoding.
558 If `direction' is specified, decode ISO6429's code for specifying
559 direction correctly, and produce the code on encoding.
561 If `init-at-bol' is specified, on encoding, it is assumed that
562 invocation and designation statuses are reset at each beginning of
563 line even if `ascii-at-eol' is not specified; thus no codes for
564 resetting them are produced.
566 If `safe' is specified, on encoding, characters not supported by a
567 coding are replaced with `?'.
569 If `latin-extra' is specified, the code-detection routine assumes that a
570 code specified in `latin-extra-code-table' (which see) is valid.
572 If `composition' is specified, an escape sequence to specify
573 composition sequence is correctly decoded on decoding, and is produced
574 on encoding.
576 If `euc-tw-shift' is specified, the EUC-TW specific shifting code is
577 correctly decoded on decoding, and is produced on encoding.
579 If `use-roman' is specified, JIS0201-1976-Roman is designated instead
580 of ASCII.
582 If `use-oldjis' is specified, JIS0208-1976 is designated instead of
583 JIS0208-1983.
585 If `8-bit-level-4' is specified, the decoder assumes the
586 implementation level \"4\" for 8-bit codes which means that GL is
587 identified as the single-shift area. The default implementation
588 level for 8-bit code is \"4A\" which means that GR is identified
589 as the single-shift area.")
591 (defun define-coding-system (name docstring &rest props)
592 "Define NAME (a symbol) as a coding system with DOCSTRING and attributes.
593 The remaining arguments must come in pairs ATTRIBUTE VALUE. ATTRIBUTE
594 may be any symbol.
596 The following attributes have special meanings. Those labeled as
597 \"(required)\" should not be omitted.
599 `:mnemonic' (required)
601 VALUE is a character to display on mode line for the coding system.
603 `:coding-type' (required)
605 VALUE must be one of `charset', `utf-8', `utf-16', `iso-2022',
606 `emacs-mule', `shift-jis', `ccl', `raw-text', `undecided'.
608 `:eol-type'
610 VALUE is the EOL (end-of-line) format of the coding system. It must be
611 one of `unix', `dos', `mac'. The symbol `unix' means Unix-like EOL
612 \(i.e. single LF), `dos' means DOS-like EOL \(i.e. sequence of CR LF),
613 and `mac' means Mac-like EOL \(i.e. single CR). If omitted, Emacs
614 detects the EOL format automatically when decoding.
616 `:charset-list'
618 VALUE must be a list of charsets supported by the coding system. On
619 encoding by the coding system, if a character belongs to multiple
620 charsets in the list, a charset that comes earlier in the list is
621 selected. If `:coding-type' is `iso-2022', VALUE may be `iso-2022',
622 which indicates that the coding system supports all ISO-2022 based
623 charsets. If `:coding-type' is `emacs-mule', VALUE may be
624 `emacs-mule', which indicates that the coding system supports all
625 charsets that have the `:emacs-mule-id' property.
627 `:ascii-compatible-p'
629 If VALUE is non-nil, the coding system decodes all 7-bit bytes into
630 the corresponding ASCII characters, and encodes all ASCII characters
631 back to the corresponding 7-bit bytes. VALUE defaults to nil.
633 `:decode-translation-table'
635 VALUE must be a translation table to use on decoding.
637 `:encode-translation-table'
639 VALUE must be a translation table to use on encoding.
641 `:post-read-conversion'
643 VALUE must be a function to call after some text is inserted and
644 decoded by the coding system itself and before any functions in
645 `after-insert-functions' are called. This function is passed one
646 argument; the number of characters in the text to convert, with
647 point at the start of the text. The function should leave point
648 the same, and return the new character count.
650 `:pre-write-conversion'
652 VALUE must be a function to call after all functions in
653 `write-region-annotate-functions' and `buffer-file-format' are
654 called, and before the text is encoded by the coding system
655 itself. This function should convert the whole text in the
656 current buffer. For backward compatibility, this function is
657 passed two arguments which can be ignored.
659 `:default-char'
661 VALUE must be a character. On encoding, a character not supported by
662 the coding system is replaced with VALUE.
664 `:for-unibyte'
666 VALUE non-nil means that visiting a file with the coding system
667 results in a unibyte buffer.
669 `:mime-charset'
671 VALUE must be a symbol whose name is that of a MIME charset converted
672 to lower case.
674 `:mime-text-unsuitable'
676 VALUE non-nil means the `:mime-charset' property names a charset which
677 is unsuitable for the top-level media type \"text\".
679 `:flags'
681 VALUE must be a list of symbols that control the ISO-2022 converter.
682 Each must be a member of the list `coding-system-iso-2022-flags'
683 \(which see). This attribute is meaningful only when `:coding-type'
684 is `iso-2022'.
686 `:designation'
688 VALUE must be a vector [G0-USAGE G1-USAGE G2-USAGE G3-USAGE].
689 GN-USAGE specifies the usage of graphic register GN as follows.
691 If it is nil, no charset can be designated to GN.
693 If it is a charset, the charset is initially designated to GN, and
694 never used by the other charsets.
696 If it is a list, the elements must be charsets, nil, 94, or 96. GN
697 can be used by all the listed charsets. If the list contains 94, any
698 iso-2022 charset whose code-space ranges are 94 long can be designated
699 to GN. If the list contains 96, any charsets whose whose ranges are
700 96 long can be designated to GN. If the first element is a charset,
701 that charset is initially designated to GN.
703 This attribute is meaningful only when `:coding-type' is `iso-2022'.
705 `:bom'
707 This attributes specifies whether the coding system uses a `byte order
708 mark'. VALUE must be nil, t, or cons of coding systems whose
709 `:coding-type' is `utf-16' or `utf-8'.
711 If the value is nil, on decoding, don't treat the first two-byte as
712 BOM, and on encoding, don't produce BOM bytes.
714 If the value is t, on decoding, skip the first two-byte as BOM, and on
715 encoding, produce BOM bytes according to the value of `:endian'.
717 If the value is cons, on decoding, check the first two-byte. If they
718 are 0xFE 0xFF, use the car part coding system of the value. If they
719 are 0xFF 0xFE, use the cdr part coding system of the value.
720 Otherwise, treat them as bytes for a normal character. On encoding,
721 produce BOM bytes according to the value of `:endian'.
723 This attribute is meaningful only when `:coding-type' is `utf-16' or
724 `utf-8'.
726 `:endian'
728 VALUE must be `big' or `little' specifying big-endian and
729 little-endian respectively. The default value is `big'.
731 This attribute is meaningful only when `:coding-type' is `utf-16'.
733 `:ccl-decoder'
735 VALUE is a symbol representing the registered CCL program used for
736 decoding. This attribute is meaningful only when `:coding-type' is
737 `ccl'.
739 `:ccl-encoder'
741 VALUE is a symbol representing the registered CCL program used for
742 encoding. This attribute is meaningful only when `:coding-type' is
743 `ccl'.
745 `:inhibit-null-byte-detection'
747 VALUE non-nil means Emacs ignore null bytes on code detection.
748 See the variable `inhibit-null-byte-detection'. This attribute
749 is meaningful only when `:coding-type' is `undecided'.
751 `:inhibit-iso-escape-detection'
753 VALUE non-nil means Emacs ignores ISO-2022 escape sequences on
754 code detection. See the variable `inhibit-iso-escape-detection'.
755 This attribute is meaningful only when `:coding-type' is
756 `undecided'.
758 `:prefer-utf-8'
760 VALUE non-nil means Emacs prefers UTF-8 on code detection for
761 non-ASCII files. This attribute is meaningful only when
762 `:coding-type' is `undecided'."
763 (let* ((common-attrs (mapcar 'list
764 '(:mnemonic
765 :coding-type
766 :charset-list
767 :ascii-compatible-p
768 :decode-translation-table
769 :encode-translation-table
770 :post-read-conversion
771 :pre-write-conversion
772 :default-char
773 :for-unibyte
774 :plist
775 :eol-type)))
776 (coding-type (plist-get props :coding-type))
777 (spec-attrs (mapcar 'list
778 (cond ((eq coding-type 'iso-2022)
779 '(:initial
780 :reg-usage
781 :request
782 :flags))
783 ((eq coding-type 'utf-8)
784 '(:bom))
785 ((eq coding-type 'utf-16)
786 '(:bom
787 :endian))
788 ((eq coding-type 'ccl)
789 '(:ccl-decoder
790 :ccl-encoder
791 :valids))
792 ((eq coding-type 'undecided)
793 '(:inhibit-null-byte-detection
794 :inhibit-iso-escape-detection
795 :prefer-utf-8))))))
797 (dolist (slot common-attrs)
798 (setcdr slot (plist-get props (car slot))))
800 (dolist (slot spec-attrs)
801 (setcdr slot (plist-get props (car slot))))
803 (if (eq coding-type 'iso-2022)
804 (let ((designation (plist-get props :designation))
805 (flags (plist-get props :flags))
806 (initial (make-vector 4 nil))
807 (reg-usage (cons 4 4))
808 request elt)
809 (dotimes (i 4)
810 (setq elt (aref designation i))
811 (cond ((charsetp elt)
812 (aset initial i elt)
813 (setq request (cons (cons elt i) request)))
814 ((consp elt)
815 (aset initial i (car elt))
816 (if (charsetp (car elt))
817 (setq request (cons (cons (car elt) i) request)))
818 (dolist (e (cdr elt))
819 (cond ((charsetp e)
820 (setq request (cons (cons e i) request)))
821 ((eq e 94)
822 (setcar reg-usage i))
823 ((eq e 96)
824 (setcdr reg-usage i))
825 ((eq e t)
826 (setcar reg-usage i)
827 (setcdr reg-usage i)))))))
828 (setcdr (assq :initial spec-attrs) initial)
829 (setcdr (assq :reg-usage spec-attrs) reg-usage)
830 (setcdr (assq :request spec-attrs) request)
832 ;; Change :flags value from a list to a bit-mask.
833 (let ((bits 0)
834 (i 0))
835 (dolist (elt coding-system-iso-2022-flags)
836 (if (memq elt flags)
837 (setq bits (logior bits (lsh 1 i))))
838 (setq i (1+ i)))
839 (setcdr (assq :flags spec-attrs) bits))))
841 ;; Add :name and :docstring properties to PROPS.
842 (setq props
843 (cons :name (cons name (cons :docstring (cons (purecopy docstring)
844 props)))))
845 (setcdr (assq :plist common-attrs) props)
846 (apply 'define-coding-system-internal
847 name (mapcar 'cdr (append common-attrs spec-attrs)))))
849 (defun coding-system-doc-string (coding-system)
850 "Return the documentation string for CODING-SYSTEM."
851 (plist-get (coding-system-plist coding-system) :docstring))
853 (defun coding-system-mnemonic (coding-system)
854 "Return the mnemonic character of CODING-SYSTEM.
855 The mnemonic character of a coding system is used in mode line to
856 indicate the coding system. If CODING-SYSTEM is nil, return ?=."
857 (plist-get (coding-system-plist coding-system) :mnemonic))
859 (defun coding-system-type (coding-system)
860 "Return the coding type of CODING-SYSTEM.
861 A coding type is a symbol indicating the encoding method of CODING-SYSTEM.
862 See the function `define-coding-system' for more detail."
863 (plist-get (coding-system-plist coding-system) :coding-type))
865 (defun coding-system-charset-list (coding-system)
866 "Return list of charsets supported by CODING-SYSTEM.
867 If CODING-SYSTEM supports all ISO-2022 charsets, return `iso-2022'.
868 If CODING-SYSTEM supports all emacs-mule charsets, return `emacs-mule'."
869 (plist-get (coding-system-plist coding-system) :charset-list))
871 (defun coding-system-category (coding-system)
872 "Return a category symbol of CODING-SYSTEM."
873 (plist-get (coding-system-plist coding-system) :category))
875 (defun coding-system-get (coding-system prop)
876 "Extract a value from CODING-SYSTEM's property list for property PROP.
877 For compatibility with Emacs 20/21, this accepts old-style symbols
878 like `mime-charset' as well as the current style like `:mime-charset'."
879 (or (plist-get (coding-system-plist coding-system) prop)
880 (if (not (keywordp prop))
881 ;; For backward compatibility.
882 (if (eq prop 'ascii-incompatible)
883 (not (plist-get (coding-system-plist coding-system)
884 :ascii-compatible-p))
885 (plist-get (coding-system-plist coding-system)
886 (intern (concat ":" (symbol-name prop))))))))
888 (defun coding-system-eol-type-mnemonic (coding-system)
889 "Return the string indicating end-of-line format of CODING-SYSTEM."
890 (let* ((eol-type (coding-system-eol-type coding-system))
891 (val (cond ((eq eol-type 0) eol-mnemonic-unix)
892 ((eq eol-type 1) eol-mnemonic-dos)
893 ((eq eol-type 2) eol-mnemonic-mac)
894 (t eol-mnemonic-undecided))))
895 (if (stringp val)
897 (char-to-string val))))
899 (defun coding-system-lessp (x y)
900 (cond ((eq x 'no-conversion) t)
901 ((eq y 'no-conversion) nil)
902 ((eq x 'emacs-mule) t)
903 ((eq y 'emacs-mule) nil)
904 ((eq x 'undecided) t)
905 ((eq y 'undecided) nil)
906 (t (let ((c1 (coding-system-mnemonic x))
907 (c2 (coding-system-mnemonic y)))
908 (or (< (downcase c1) (downcase c2))
909 (and (not (> (downcase c1) (downcase c2)))
910 (< c1 c2)))))))
912 (defun coding-system-equal (coding-system-1 coding-system-2)
913 "Return t if and only if CODING-SYSTEM-1 and CODING-SYSTEM-2 are identical.
914 Two coding systems are identical if both symbols are equal
915 or one is an alias of the other."
916 (or (eq coding-system-1 coding-system-2)
917 (and (equal (coding-system-plist coding-system-1)
918 (coding-system-plist coding-system-2))
919 (let ((eol-type-1 (coding-system-eol-type coding-system-1))
920 (eol-type-2 (coding-system-eol-type coding-system-2)))
921 (or (eq eol-type-1 eol-type-2)
922 (and (vectorp eol-type-1) (vectorp eol-type-2)))))))
924 (defun add-to-coding-system-list (coding-system)
925 "Add CODING-SYSTEM to variable `coding-system-list' while keeping it sorted."
926 (if (or (null coding-system-list)
927 (coding-system-lessp coding-system (car coding-system-list)))
928 (setq coding-system-list (cons coding-system coding-system-list))
929 (let ((len (length coding-system-list))
930 mid (tem coding-system-list))
931 (while (> len 1)
932 (setq mid (nthcdr (/ len 2) tem))
933 (if (coding-system-lessp (car mid) coding-system)
934 (setq tem mid
935 len (- len (/ len 2)))
936 (setq len (/ len 2))))
937 (setcdr tem (cons coding-system (cdr tem))))))
939 (defun coding-system-list (&optional base-only)
940 "Return a list of all existing non-subsidiary coding systems.
941 If optional arg BASE-ONLY is non-nil, only base coding systems are
942 listed. The value doesn't include subsidiary coding systems which are
943 made from bases and aliases automatically for various end-of-line
944 formats (e.g. iso-latin-1-unix, koi8-r-dos)."
945 (let ((codings nil))
946 (dolist (coding coding-system-list)
947 (if (eq (coding-system-base coding) coding)
948 (if base-only
949 (setq codings (cons coding codings))
950 (dolist (alias (coding-system-aliases coding))
951 (setq codings (cons alias codings))))))
952 codings))
954 (defconst char-coding-system-table nil
955 "It exists just for backward compatibility, and the value is always nil.")
956 (make-obsolete-variable 'char-coding-system-table nil "23.1")
958 (defun transform-make-coding-system-args (name type &optional doc-string props)
959 "For internal use only.
960 Transform XEmacs style args for `make-coding-system' to Emacs style.
961 Value is a list of transformed arguments."
962 (let ((mnemonic (string-to-char (or (plist-get props 'mnemonic) "?")))
963 (eol-type (plist-get props 'eol-type))
964 properties tmp)
965 (cond
966 ((eq eol-type 'lf) (setq eol-type 'unix))
967 ((eq eol-type 'crlf) (setq eol-type 'dos))
968 ((eq eol-type 'cr) (setq eol-type 'mac)))
969 (if (setq tmp (plist-get props 'post-read-conversion))
970 (setq properties (plist-put properties 'post-read-conversion tmp)))
971 (if (setq tmp (plist-get props 'pre-write-conversion))
972 (setq properties (plist-put properties 'pre-write-conversion tmp)))
973 (cond
974 ((eq type 'shift-jis)
975 `(,name 1 ,mnemonic ,doc-string () ,properties ,eol-type))
976 ((eq type 'iso2022) ; This is not perfect.
977 (if (plist-get props 'escape-quoted)
978 (error "escape-quoted is not supported: %S"
979 `(,name ,type ,doc-string ,props)))
980 (let ((g0 (plist-get props 'charset-g0))
981 (g1 (plist-get props 'charset-g1))
982 (g2 (plist-get props 'charset-g2))
983 (g3 (plist-get props 'charset-g3))
984 (use-roman
985 (and
986 (eq (cadr (assoc 'latin-jisx0201
987 (plist-get props 'input-charset-conversion)))
988 'ascii)
989 (eq (cadr (assoc 'ascii
990 (plist-get props 'output-charset-conversion)))
991 'latin-jisx0201)))
992 (use-oldjis
993 (and
994 (eq (cadr (assoc 'japanese-jisx0208-1978
995 (plist-get props 'input-charset-conversion)))
996 'japanese-jisx0208)
997 (eq (cadr (assoc 'japanese-jisx0208
998 (plist-get props 'output-charset-conversion)))
999 'japanese-jisx0208-1978))))
1000 (if (charsetp g0)
1001 (if (plist-get props 'force-g0-on-output)
1002 (setq g0 `(nil ,g0))
1003 (setq g0 `(,g0 t))))
1004 (if (charsetp g1)
1005 (if (plist-get props 'force-g1-on-output)
1006 (setq g1 `(nil ,g1))
1007 (setq g1 `(,g1 t))))
1008 (if (charsetp g2)
1009 (if (plist-get props 'force-g2-on-output)
1010 (setq g2 `(nil ,g2))
1011 (setq g2 `(,g2 t))))
1012 (if (charsetp g3)
1013 (if (plist-get props 'force-g3-on-output)
1014 (setq g3 `(nil ,g3))
1015 (setq g3 `(,g3 t))))
1016 `(,name 2 ,mnemonic ,doc-string
1017 (,g0 ,g1 ,g2 ,g3
1018 ,(plist-get props 'short)
1019 ,(not (plist-get props 'no-ascii-eol))
1020 ,(not (plist-get props 'no-ascii-cntl))
1021 ,(plist-get props 'seven)
1023 ,(not (plist-get props 'lock-shift))
1024 ,use-roman
1025 ,use-oldjis
1026 ,(plist-get props 'no-iso6429)
1027 nil nil nil nil)
1028 ,properties ,eol-type)))
1029 ((eq type 'big5)
1030 `(,name 3 ,mnemonic ,doc-string () ,properties ,eol-type))
1031 ((eq type 'ccl)
1032 `(,name 4 ,mnemonic ,doc-string
1033 (,(plist-get props 'decode) . ,(plist-get props 'encode))
1034 ,properties ,eol-type))
1036 (error "unsupported XEmacs style make-coding-style arguments: %S"
1037 `(,name ,type ,doc-string ,props))))))
1039 (defun make-coding-system (coding-system type mnemonic doc-string
1040 &optional
1041 flags
1042 properties
1043 eol-type)
1044 "Define a new coding system CODING-SYSTEM (symbol).
1045 This function is provided for backward compatibility."
1046 (declare (obsolete define-coding-system "23.1"))
1047 ;; For compatibility with XEmacs, we check the type of TYPE. If it
1048 ;; is a symbol, perhaps, this function is called with XEmacs-style
1049 ;; arguments. Here, try to transform that kind of arguments to
1050 ;; Emacs style.
1051 (if (symbolp type)
1052 (let ((args (transform-make-coding-system-args coding-system type
1053 mnemonic doc-string)))
1054 (setq coding-system (car args)
1055 type (nth 1 args)
1056 mnemonic (nth 2 args)
1057 doc-string (nth 3 args)
1058 flags (nth 4 args)
1059 properties (nth 5 args)
1060 eol-type (nth 6 args))))
1062 (setq type
1063 (cond ((eq type 0) 'emacs-mule)
1064 ((eq type 1) 'shift-jis)
1065 ((eq type 2) 'iso2022)
1066 ((eq type 3) 'big5)
1067 ((eq type 4) 'ccl)
1068 ((eq type 5) 'raw-text)
1070 (error "Invalid coding system type: %s" type))))
1072 (setq properties
1073 (let ((plist nil) key)
1074 (dolist (elt properties)
1075 (setq key (car elt))
1076 (cond ((eq key 'post-read-conversion)
1077 (setq key :post-read-conversion))
1078 ((eq key 'pre-write-conversion)
1079 (setq key :pre-write-conversion))
1080 ((eq key 'translation-table-for-decode)
1081 (setq key :decode-translation-table))
1082 ((eq key 'translation-table-for-encode)
1083 (setq key :encode-translation-table))
1084 ((eq key 'safe-charsets)
1085 (setq key :charset-list))
1086 ((eq key 'mime-charset)
1087 (setq key :mime-charset))
1088 ((eq key 'valid-codes)
1089 (setq key :valids)))
1090 (setq plist (plist-put plist key (cdr elt))))
1091 plist))
1092 (setq properties (plist-put properties :mnemonic mnemonic))
1093 (plist-put properties :coding-type type)
1094 (cond ((eq eol-type 0) (setq eol-type 'unix))
1095 ((eq eol-type 1) (setq eol-type 'dos))
1096 ((eq eol-type 2) (setq eol-type 'mac))
1097 ((vectorp eol-type) (setq eol-type nil)))
1098 (plist-put properties :eol-type eol-type)
1100 (cond
1101 ((eq type 'iso2022)
1102 (plist-put properties :flags
1103 (list (and (or (consp (nth 0 flags))
1104 (consp (nth 1 flags))
1105 (consp (nth 2 flags))
1106 (consp (nth 3 flags))) 'designation)
1107 (or (nth 4 flags) 'long-form)
1108 (and (nth 5 flags) 'ascii-at-eol)
1109 (and (nth 6 flags) 'ascii-at-cntl)
1110 (and (nth 7 flags) '7-bit)
1111 (and (nth 8 flags) 'locking-shift)
1112 (and (nth 9 flags) 'single-shift)
1113 (and (nth 10 flags) 'use-roman)
1114 (and (nth 11 flags) 'use-oldjis)
1115 (or (nth 12 flags) 'direction)
1116 (and (nth 13 flags) 'init-at-bol)
1117 (and (nth 14 flags) 'designate-at-bol)
1118 (and (nth 15 flags) 'safe)
1119 (and (nth 16 flags) 'latin-extra)))
1120 (plist-put properties :designation
1121 (let ((vec (make-vector 4 nil)))
1122 (dotimes (i 4)
1123 (let ((spec (nth i flags)))
1124 (if (eq spec t)
1125 (aset vec i '(94 96))
1126 (if (consp spec)
1127 (progn
1128 (if (memq t spec)
1129 (setq spec (append (delq t spec) '(94 96))))
1130 (aset vec i spec))))))
1131 vec)))
1133 ((eq type 'ccl)
1134 (plist-put properties :ccl-decoder (car flags))
1135 (plist-put properties :ccl-encoder (cdr flags))))
1137 (apply 'define-coding-system coding-system doc-string properties))
1139 (defun merge-coding-systems (first second)
1140 "Fill in any unspecified aspects of coding system FIRST from SECOND.
1141 Return the resulting coding system."
1142 (let ((base (coding-system-base second))
1143 (eol (coding-system-eol-type second)))
1144 ;; If FIRST doesn't specify text conversion, merge with that of SECOND.
1145 (if (eq (coding-system-base first) 'undecided)
1146 (setq first (coding-system-change-text-conversion first base)))
1147 ;; If FIRST doesn't specify eol conversion, merge with that of SECOND.
1148 (if (and (vectorp (coding-system-eol-type first))
1149 (numberp eol) (>= eol 0) (<= eol 2))
1150 (setq first (coding-system-change-eol-conversion
1151 first eol)))
1152 first))
1154 (defun autoload-coding-system (symbol form)
1155 "Define SYMBOL as a coding-system that is defined on demand.
1157 FORM is a form to evaluate to define the coding-system."
1158 (put symbol 'coding-system-define-form form)
1159 (setq coding-system-alist (cons (list (symbol-name symbol))
1160 coding-system-alist))
1161 (dolist (elt '("-unix" "-dos" "-mac"))
1162 (let ((name (concat (symbol-name symbol) elt)))
1163 (put (intern name) 'coding-system-define-form form)
1164 (setq coding-system-alist (cons (list name) coding-system-alist)))))
1166 ;; This variable is set in these two cases:
1167 ;; (1) A file is read by a coding system specified explicitly.
1168 ;; `after-insert-file-set-coding' sets the car of this value to
1169 ;; `coding-system-for-read', and sets the cdr to nil.
1170 ;; (2) `set-buffer-file-coding-system' is called.
1171 ;; The cdr of this value is set to the specified coding system.
1172 ;; This variable is used for decoding in `revert-buffer' and encoding
1173 ;; in `select-safe-coding-system'.
1175 ;; When saving a buffer, if `buffer-file-coding-system-explicit' is
1176 ;; already non-nil, `basic-save-buffer-1' sets its CAR to the value of
1177 ;; `last-coding-system-used'. (It used to set it unconditionally, but
1178 ;; that seems unnecessary; see Bug#4533.)
1180 (defvar buffer-file-coding-system-explicit nil
1181 "The file coding system explicitly specified for the current buffer.
1182 The value is a cons of coding systems for reading (decoding) and
1183 writing (encoding).
1184 Internal use only.")
1185 (make-variable-buffer-local 'buffer-file-coding-system-explicit)
1186 (put 'buffer-file-coding-system-explicit 'permanent-local t)
1188 (defun read-buffer-file-coding-system ()
1189 (let* ((bcss (find-coding-systems-region (point-min) (point-max)))
1190 (css-table
1191 (unless (equal bcss '(undecided))
1192 (append '("dos" "unix" "mac")
1193 (delq nil (mapcar (lambda (cs)
1194 (if (memq (coding-system-base cs) bcss)
1195 (symbol-name cs)))
1196 coding-system-list)))))
1197 (combined-table
1198 (if css-table
1199 (completion-table-in-turn css-table coding-system-alist)
1200 coding-system-alist))
1201 (auto-cs
1202 (unless find-file-literally
1203 (save-excursion
1204 (save-restriction
1205 (widen)
1206 (goto-char (point-min))
1207 (funcall set-auto-coding-function
1208 (or buffer-file-name "") (buffer-size))))))
1209 (preferred
1210 (let ((bfcs (default-value 'buffer-file-coding-system)))
1211 (cons (and (or (equal bcss '(undecided))
1212 (memq (coding-system-base bfcs) bcss))
1213 bfcs)
1214 (mapcar (lambda (cs)
1215 (and (coding-system-p cs)
1216 (coding-system-get cs :mime-charset)
1217 (or (equal bcss '(undecided))
1218 (memq (coding-system-base cs) bcss))
1219 cs))
1220 (coding-system-priority-list)))))
1221 (default
1222 (let ((current (coding-system-base buffer-file-coding-system)))
1223 ;; Generally use as a default the first preferred coding-system
1224 ;; different from the current coding-system, except for
1225 ;; the case of auto-cs since choosing anything else is asking
1226 ;; for trouble (would lead to using a different coding
1227 ;; system than specified in the coding tag).
1228 (or auto-cs
1229 (car (delq nil
1230 (mapcar (lambda (cs)
1231 (if (eq current (coding-system-base cs))
1233 cs))
1234 preferred))))))
1235 (completion-ignore-case t)
1236 (completion-pcm--delim-wild-regex ; Let "u8" complete to "utf-8".
1237 (concat completion-pcm--delim-wild-regex
1238 "\\|\\([[:alpha:]]\\)[[:digit:]]"))
1239 (cs (completing-read
1240 (format "Coding system for saving file (default %s): " default)
1241 combined-table
1242 nil t nil 'coding-system-history
1243 (if default (symbol-name default)))))
1244 (unless (zerop (length cs)) (intern cs))))
1246 (defun set-buffer-file-coding-system (coding-system &optional force nomodify)
1247 "Set the file coding-system of the current buffer to CODING-SYSTEM.
1248 This means that when you save the buffer, it will be converted
1249 according to CODING-SYSTEM. For a list of possible values of
1250 CODING-SYSTEM, use \\[list-coding-systems].
1252 If CODING-SYSTEM leaves the text conversion unspecified, or if it leaves
1253 the end-of-line conversion unspecified, FORCE controls what to do.
1254 If FORCE is nil, get the unspecified aspect (or aspects) from the buffer's
1255 previous `buffer-file-coding-system' value (if it is specified there).
1256 Otherwise, leave it unspecified.
1258 This marks the buffer modified so that the succeeding \\[save-buffer]
1259 surely saves the buffer with CODING-SYSTEM. From a program, if you
1260 don't want to mark the buffer modified, specify t for NOMODIFY.
1261 If you know exactly what coding system you want to use,
1262 just set the variable `buffer-file-coding-system' directly."
1263 (interactive
1264 (list (read-buffer-file-coding-system)
1265 current-prefix-arg))
1266 (check-coding-system coding-system)
1267 (if (and coding-system buffer-file-coding-system (null force))
1268 (setq coding-system
1269 (merge-coding-systems coding-system buffer-file-coding-system)))
1270 (when (and (called-interactively-p 'interactive)
1271 (not (memq 'emacs (coding-system-get coding-system
1272 :charset-list))))
1273 ;; Check whether save would succeed, and jump to the offending char(s)
1274 ;; if not.
1275 (let ((css (find-coding-systems-region (point-min) (point-max))))
1276 (unless (or (eq (car css) 'undecided)
1277 (memq (coding-system-base coding-system) css))
1278 (setq coding-system (select-safe-coding-system-interactively
1279 (point-min) (point-max) css
1280 (list coding-system))))))
1281 (setq buffer-file-coding-system coding-system)
1282 (if buffer-file-coding-system-explicit
1283 (setcdr buffer-file-coding-system-explicit coding-system)
1284 (setq buffer-file-coding-system-explicit (cons nil coding-system)))
1285 (unless nomodify
1286 (set-buffer-modified-p t))
1287 (force-mode-line-update))
1289 (defun revert-buffer-with-coding-system (coding-system &optional force)
1290 "Visit the current buffer's file again using coding system CODING-SYSTEM.
1291 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems].
1293 If CODING-SYSTEM leaves the text conversion unspecified, or if it leaves
1294 the end-of-line conversion unspecified, FORCE controls what to do.
1295 If FORCE is nil, get the unspecified aspect (or aspects) from the buffer's
1296 previous `buffer-file-coding-system' value (if it is specified there).
1297 Otherwise, determine it from the file contents as usual for visiting a file."
1298 (interactive "zCoding system for visited file (default nil): \nP")
1299 (check-coding-system coding-system)
1300 (if (and coding-system buffer-file-coding-system (null force))
1301 (setq coding-system
1302 (merge-coding-systems coding-system buffer-file-coding-system)))
1303 (let ((coding-system-for-read coding-system))
1304 (revert-buffer)))
1306 (defun set-file-name-coding-system (coding-system)
1307 "Set coding system for decoding and encoding file names to CODING-SYSTEM.
1308 It actually just set the variable `file-name-coding-system' (which see)
1309 to CODING-SYSTEM."
1310 (interactive "zCoding system for file names (default nil): ")
1311 (check-coding-system coding-system)
1312 (if (and coding-system
1313 (not (coding-system-get coding-system :ascii-compatible-p))
1314 (not (coding-system-get coding-system :suitable-for-file-name)))
1315 (error "%s is not suitable for file names" coding-system))
1316 (setq file-name-coding-system coding-system))
1318 (defvar default-terminal-coding-system nil
1319 "Default value for the terminal coding system.
1320 This is normally set according to the selected language environment.
1321 See also the command `set-terminal-coding-system'.")
1323 (defun set-terminal-coding-system (coding-system &optional terminal)
1324 "Set coding system of terminal output to CODING-SYSTEM.
1325 All text output to TERMINAL will be encoded
1326 with the specified coding system.
1328 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems].
1329 The default is determined by the selected language environment
1330 or by the previous use of this command.
1332 TERMINAL may be a terminal object, a frame, or nil for the
1333 selected frame's terminal. The setting has no effect on
1334 graphical terminals."
1335 (interactive
1336 (list (let ((default (if (and (not (terminal-coding-system))
1337 default-terminal-coding-system)
1338 default-terminal-coding-system)))
1339 (read-coding-system
1340 (format "Coding system for terminal display (default %s): "
1341 default)
1342 default))))
1343 (if (and (not coding-system)
1344 (not (terminal-coding-system)))
1345 (setq coding-system default-terminal-coding-system))
1346 (if coding-system
1347 (setq default-terminal-coding-system coding-system))
1348 (set-terminal-coding-system-internal coding-system terminal)
1349 (redraw-frame))
1351 (defvar default-keyboard-coding-system nil
1352 "Default value of the keyboard coding system.
1353 This is normally set according to the selected language environment.
1354 See also the command `set-keyboard-coding-system'.")
1356 (defun set-keyboard-coding-system (coding-system &optional terminal)
1357 "Set coding system for keyboard input on TERMINAL to CODING-SYSTEM.
1359 For a list of possible values of CODING-SYSTEM, use \\[list-coding-systems].
1360 The default is determined by the selected language environment
1361 or by the previous use of this command.
1363 If CODING-SYSTEM is nil or the coding-type of CODING-SYSTEM is
1364 `raw-text', the decoding of keyboard input is disabled.
1366 TERMINAL may be a terminal object, a frame, or nil for the
1367 selected frame's terminal. The setting has no effect on
1368 graphical terminals."
1369 (interactive
1370 (list (let* ((coding (keyboard-coding-system nil))
1371 (default (if (eq (coding-system-type coding) 'raw-text)
1372 default-keyboard-coding-system)))
1373 (read-coding-system
1374 (format "Coding system for keyboard input (default %s): "
1375 default)
1376 default))))
1377 (let ((coding-type (coding-system-type coding-system))
1378 (saved-meta-mode
1379 (terminal-parameter terminal 'keyboard-coding-saved-meta-mode)))
1380 (if (not (eq coding-type 'raw-text))
1381 (let (accept-8-bit)
1382 (if (not (or (coding-system-get coding-system :suitable-for-keyboard)
1383 (coding-system-get coding-system :ascii-compatible-p)))
1384 (error "Unsuitable coding system for keyboard: %s" coding-system))
1385 (cond ((memq coding-type '(charset utf-8 shift-jis big5 ccl))
1386 (setq accept-8-bit t))
1387 ((eq coding-type 'iso-2022)
1388 (let ((flags (coding-system-get coding-system :flags)))
1389 (or (memq '7-bit flags)
1390 (setq accept-8-bit t))))
1392 (error "Unsupported coding system for keyboard: %s"
1393 coding-system)))
1394 (if accept-8-bit
1395 (progn
1396 (or saved-meta-mode
1397 (set-terminal-parameter terminal
1398 'keyboard-coding-saved-meta-mode
1399 (cons (nth 2 (current-input-mode))
1400 nil)))
1401 (set-input-meta-mode 8 terminal))
1402 (when saved-meta-mode
1403 (set-input-meta-mode (car saved-meta-mode) terminal)
1404 (set-terminal-parameter terminal
1405 'keyboard-coding-saved-meta-mode
1406 nil)))
1407 ;; Avoid end-of-line conversion.
1408 (setq coding-system
1409 (coding-system-change-eol-conversion coding-system 'unix)))
1411 (when saved-meta-mode
1412 (set-input-meta-mode (car saved-meta-mode) terminal)
1413 (set-terminal-parameter terminal
1414 'keyboard-coding-saved-meta-mode
1415 nil))))
1416 (set-keyboard-coding-system-internal coding-system terminal)
1417 (setq keyboard-coding-system coding-system))
1419 (defcustom keyboard-coding-system nil
1420 "Specify coding system for keyboard input.
1421 If you set this on a terminal which can't distinguish Meta keys from
1422 8-bit characters, you will have to use ESC to type Meta characters.
1423 See Info node `Terminal Coding' and Info node `Unibyte Mode'.
1425 On non-windowing terminals, this is set from the locale by default.
1427 Setting this variable directly does not take effect;
1428 use either \\[customize] or \\[set-keyboard-coding-system]."
1429 :type '(coding-system :tag "Coding system")
1430 :link '(info-link "(emacs)Terminal Coding")
1431 :link '(info-link "(emacs)Unibyte Mode")
1432 :set (lambda (_symbol value)
1433 ;; Don't load encoded-kb unnecessarily.
1434 (if (or value (boundp 'encoded-kbd-setup-display))
1435 (set-keyboard-coding-system value)
1436 (set-default 'keyboard-coding-system nil))) ; must initialize
1437 :version "22.1"
1438 :group 'keyboard
1439 :group 'mule)
1441 (defun set-buffer-process-coding-system (decoding encoding)
1442 "Set coding systems for the process associated with the current buffer.
1443 DECODING is the coding system to be used to decode input from the process,
1444 ENCODING is the coding system to be used to encode output to the process.
1446 For a list of possible coding systems, use \\[list-coding-systems]."
1447 (interactive
1448 "zCoding-system for output from the process: \nzCoding-system for input to the process: ")
1449 (let ((proc (get-buffer-process (current-buffer))))
1450 (if (null proc)
1451 (error "No process")
1452 (check-coding-system decoding)
1453 (check-coding-system encoding)
1454 (set-process-coding-system proc decoding encoding)))
1455 (force-mode-line-update))
1457 (defalias 'set-clipboard-coding-system 'set-selection-coding-system)
1459 (defun set-selection-coding-system (coding-system)
1460 "Make CODING-SYSTEM used for communicating with other X clients.
1461 When sending or receiving text via cut_buffer, selection, and clipboard,
1462 the text is encoded or decoded by CODING-SYSTEM."
1463 (interactive "zCoding system for X selection: ")
1464 (check-coding-system coding-system)
1465 (setq selection-coding-system coding-system))
1467 ;; Coding system lastly specified by the command
1468 ;; set-next-selection-coding-system.
1469 (defvar last-next-selection-coding-system nil)
1471 (defun set-next-selection-coding-system (coding-system)
1472 "Use CODING-SYSTEM for next communication with other window system clients.
1473 This setting is effective for the next communication only."
1474 (interactive
1475 (list (read-coding-system
1476 (if last-next-selection-coding-system
1477 (format "Coding system for the next selection (default %S): "
1478 last-next-selection-coding-system)
1479 "Coding system for the next selection: ")
1480 last-next-selection-coding-system)))
1481 (if coding-system
1482 (setq last-next-selection-coding-system coding-system)
1483 (setq coding-system last-next-selection-coding-system))
1484 (check-coding-system coding-system)
1486 (setq next-selection-coding-system coding-system))
1488 (defun set-coding-priority (arg)
1489 "Set priority of coding categories according to ARG.
1490 ARG is a list of coding categories ordered by priority.
1492 This function is provided for backward compatibility."
1493 (declare (obsolete set-coding-system-priority "23.1"))
1494 (apply 'set-coding-system-priority
1495 (mapcar #'(lambda (x) (symbol-value x)) arg)))
1497 ;;; X selections
1499 (defvar ctext-non-standard-encodings-alist
1500 (mapcar 'purecopy
1501 '(("big5-0" big5 2 big5)
1502 ("ISO8859-14" iso-8859-14 1 latin-iso8859-14)
1503 ("ISO8859-15" iso-8859-15 1 latin-iso8859-15)
1504 ("gbk-0" gbk 2 chinese-gbk)
1505 ("koi8-r" koi8-r 1 koi8-r)
1506 ("microsoft-cp1251" windows-1251 1 windows-1251)))
1507 "Alist of non-standard encoding names vs the corresponding usages in CTEXT.
1509 It controls how extended segments of a compound text are handled
1510 by the coding system `compound-text-with-extensions'.
1512 Each element has the form (ENCODING-NAME CODING-SYSTEM N-OCTET CHARSET).
1514 ENCODING-NAME is an encoding name of an \"extended segment\".
1516 CODING-SYSTEM is the coding-system to encode (or decode) the
1517 characters into (or from) the extended segment.
1519 N-OCTET is the number of octets (bytes) that encodes a character
1520 in the segment. It can be 0 (meaning the number of octets per
1521 character is variable), 1, 2, 3, or 4.
1523 CHARSET is a character set containing characters that are encoded
1524 in the segment. It can be a list of character sets.
1526 On decoding CTEXT, all encoding names listed here are recognized.
1528 On encoding CTEXT, encoding names in the variable
1529 `ctext-non-standard-encodings' (which see) and in the information
1530 listed for the current language environment under the key
1531 `ctext-non-standard-encodings' are used.")
1533 (defvar ctext-non-standard-encodings nil
1534 "List of non-standard encoding names used in extended segments of CTEXT.
1535 Each element must be one of the names listed in the variable
1536 `ctext-non-standard-encodings-alist' (which see).")
1538 (defvar ctext-non-standard-encodings-regexp
1539 (purecopy
1540 (string-to-multibyte
1541 (concat
1542 ;; For non-standard encodings.
1543 "\\(\e%/[0-4][\200-\377][\200-\377]\\([^\002]+\\)\002\\)"
1544 "\\|"
1545 ;; For UTF-8 encoding.
1546 "\\(\e%G[^\e]*\e%@\\)"))))
1548 ;; Functions to support "Non-Standard Character Set Encodings" defined
1549 ;; by the COMPOUND-TEXT spec. They also support "The UTF-8 encoding"
1550 ;; described in the section 7 of the documentation of COMPOUND-TEXT
1551 ;; distributed with XFree86.
1553 (defun ctext-post-read-conversion (len)
1554 "Decode LEN characters encoded as Compound Text with Extended Segments."
1555 ;; We don't need the following because it is expected that this
1556 ;; function is mainly used for decoding X selection which is not
1557 ;; that big data.
1558 ;;(buffer-disable-undo) ; minimize consing due to insertions and deletions
1559 (save-match-data
1560 (save-restriction
1561 (narrow-to-region (point) (+ (point) len))
1562 (let ((case-fold-search nil)
1563 last-coding-system-used
1564 pos bytes)
1565 (decode-coding-region (point-min) (point-max) 'ctext)
1566 (while (re-search-forward ctext-non-standard-encodings-regexp
1567 nil 'move)
1568 (setq pos (match-beginning 0))
1569 (if (match-beginning 1)
1570 ;; ESC % / [0-4] M L --ENCODING-NAME-- \002 --BYTES--
1571 (let* ((M (multibyte-char-to-unibyte (char-after (+ pos 4))))
1572 (L (multibyte-char-to-unibyte (char-after (+ pos 5))))
1573 (encoding (match-string 2))
1574 (encoding-info (assoc-string
1575 encoding
1576 ctext-non-standard-encodings-alist t))
1577 (coding (if encoding-info
1578 (nth 1 encoding-info)
1579 (setq encoding (intern (downcase encoding)))
1580 (and (coding-system-p encoding)
1581 encoding))))
1582 (setq bytes (- (+ (* (- M 128) 128) (- L 128))
1583 (- (point) (+ pos 6))))
1584 (when coding
1585 (delete-region pos (point))
1586 (forward-char bytes)
1587 (decode-coding-region (- (point) bytes) (point) coding)))
1588 ;; ESC % G --UTF-8-BYTES-- ESC % @
1589 (delete-char -3)
1590 (delete-region pos (+ pos 3))
1591 (decode-coding-region pos (point) 'utf-8))))
1592 (goto-char (point-min))
1593 (- (point-max) (point)))))
1595 (defvar ctext-standard-encodings
1596 '(ascii latin-jisx0201 katakana-jisx0201
1597 latin-iso8859-1 latin-iso8859-2 latin-iso8859-3 latin-iso8859-4
1598 greek-iso8859-7 arabic-iso8859-6 hebrew-iso8859-8 cyrillic-iso8859-5
1599 latin-iso8859-9
1600 chinese-gb2312 japanese-jisx0208 korean-ksc5601)
1601 "List of approved standard encodings (i.e. charsets) of X's Compound Text.
1602 Coding-system `compound-text-with-extensions' encodes a character
1603 belonging to any of those charsets using the normal ISO2022
1604 designation sequence unless the current language environment or
1605 the variable `ctext-non-standard-encodings' decide to use an extended
1606 segment of CTEXT for that character. See also the documentation
1607 of `ctext-non-standard-encodings-alist'.")
1609 ;; Return an alist of CHARSET vs CTEXT-USAGE-INFO generated from
1610 ;; `ctext-non-standard-encodings' and a list specified by the key
1611 ;; `ctext-non-standard-encodings' for the current language
1612 ;; environment. CTEXT-USAGE-INFO is one of the element of
1613 ;; `ctext-non-standard-encodings-alist' or nil. In the former case, a
1614 ;; character in CHARSET is encoded using extended segment. In the
1615 ;; latter case, a character in CHARSET is encoded using normal ISO2022
1616 ;; designation sequence. If a character is not in any of CHARSETs, it
1617 ;; is encoded using UTF-8 encoding extension.
1619 (defun ctext-non-standard-encodings-table ()
1620 (let* ((table (append ctext-non-standard-encodings
1621 (copy-sequence
1622 (get-language-info current-language-environment
1623 'ctext-non-standard-encodings))))
1624 (tail table)
1625 elt)
1626 (while tail
1627 (setq elt (car tail))
1628 (let* ((slot (assoc elt ctext-non-standard-encodings-alist))
1629 (charset (nth 3 slot)))
1630 (if (charsetp charset)
1631 (setcar tail
1632 (cons (plist-get (charset-plist charset) :base) slot))
1633 (setcar tail (cons (car charset) slot))
1634 (dolist (cs (cdr charset))
1635 (setcdr tail
1636 (cons (cons (plist-get (charset-plist (car cs)) :base) slot)
1637 (cdr tail)))
1638 (setq tail (cdr tail))))
1639 (setq tail (cdr tail))))
1640 table))
1642 (defun ctext-pre-write-conversion (from to)
1643 "Encode characters between FROM and TO as Compound Text w/Extended Segments.
1645 If FROM is a string, generate a new temp buffer, insert the text,
1646 and convert it in the temporary buffer. Otherwise, convert
1647 in-place."
1648 (save-match-data
1649 ;; Setup a working buffer if necessary.
1650 (when (stringp from)
1651 (set-buffer (generate-new-buffer " *temp"))
1652 (set-buffer-multibyte (multibyte-string-p from))
1653 (insert from)
1654 (setq from (point-min) to (point-max)))
1655 (save-restriction
1656 (narrow-to-region from to)
1657 (goto-char from)
1658 (let ((encoding-table (ctext-non-standard-encodings-table))
1659 (charset-list (sort-charsets
1660 (copy-sequence ctext-standard-encodings)))
1661 (end-pos (make-marker))
1662 last-coding-system-used
1663 last-pos charset encoding-info)
1664 (dolist (elt encoding-table)
1665 (push (car elt) charset-list))
1666 (setq end-pos (point-marker))
1667 (while (re-search-forward "[^\0-\177]+" nil t)
1668 ;; Found a sequence of non-ASCII characters.
1669 (set-marker end-pos (match-end 0))
1670 (goto-char (match-beginning 0))
1671 (setq last-pos (point)
1672 charset (char-charset (following-char) charset-list))
1673 (forward-char 1)
1674 (while (and (< (point) end-pos)
1675 (eq charset (char-charset (following-char) charset-list)))
1676 (forward-char 1))
1677 (if charset
1678 (if (setq encoding-info (cdr (assq charset encoding-table)))
1679 ;; Encode this range using an extended segment.
1680 (let ((encoding-name (car encoding-info))
1681 (coding-system (nth 1 encoding-info))
1682 (noctets (nth 2 encoding-info))
1683 len)
1684 (encode-coding-region last-pos (point) coding-system)
1685 (setq len (+ (length encoding-name) 1
1686 (- (point) last-pos)))
1687 ;; According to the spec of CTEXT, it is not
1688 ;; necessary to produce this extra designation
1689 ;; sequence, but some buggy application
1690 ;; (e.g. crxvt-gb) requires it.
1691 (insert "\e(B")
1692 (save-excursion
1693 (goto-char last-pos)
1694 (insert (format "\e%%/%d" noctets))
1695 (insert-byte (+ (/ len 128) 128) 1)
1696 (insert-byte (+ (% len 128) 128) 1)
1697 (insert encoding-name)
1698 (insert 2)))
1699 ;; Encode this range as characters in CHARSET.
1700 (put-text-property last-pos (point) 'charset charset))
1701 ;; Encode this range using UTF-8 encoding extension.
1702 (encode-coding-region last-pos (point) 'mule-utf-8)
1703 (save-excursion
1704 (goto-char last-pos)
1705 (insert "\e%G"))
1706 (insert "\e%@")))
1707 (goto-char (point-min)))))
1708 ;; Must return nil, as build_annotations_2 expects that.
1709 nil)
1711 ;;; FILE I/O
1713 ;; TODO many elements of this list are also in inhibit-local-variables-regexps.
1714 (defcustom auto-coding-alist
1715 ;; .exe and .EXE are added to support archive-mode looking at DOS
1716 ;; self-extracting exe archives.
1717 (mapcar (lambda (arg) (cons (purecopy (car arg)) (cdr arg)))
1718 '(("\\.\\(\
1719 arc\\|zip\\|lzh\\|lha\\|zoo\\|[jew]ar\\|xpi\\|rar\\|7z\\|\
1720 ARC\\|ZIP\\|LZH\\|LHA\\|ZOO\\|[JEW]AR\\|XPI\\|RAR\\|7Z\\)\\'"
1721 . no-conversion-multibyte)
1722 ("\\.\\(exe\\|EXE\\)\\'" . no-conversion)
1723 ("\\.\\(sx[dmicw]\\|odt\\|tar\\|t[bg]z\\)\\'" . no-conversion)
1724 ("\\.\\(gz\\|Z\\|bz\\|bz2\\|xz\\|gpg\\)\\'" . no-conversion)
1725 ("\\.\\(jpe?g\\|png\\|gif\\|tiff?\\|p[bpgn]m\\)\\'" . no-conversion)
1726 ("\\.pdf\\'" . no-conversion)
1727 ("/#[^/]+#\\'" . utf-8-emacs-unix)))
1728 "Alist of filename patterns vs corresponding coding systems.
1729 Each element looks like (REGEXP . CODING-SYSTEM).
1730 A file whose name matches REGEXP is decoded by CODING-SYSTEM on reading.
1732 The settings in this alist take priority over `coding:' tags
1733 in the file (see the function `set-auto-coding')
1734 and the contents of `file-coding-system-alist'."
1735 :version "24.1" ; added xz
1736 :group 'files
1737 :group 'mule
1738 :type '(repeat (cons (regexp :tag "File name regexp")
1739 (symbol :tag "Coding system"))))
1741 (defcustom auto-coding-regexp-alist
1742 (mapcar (lambda (arg) (cons (purecopy (car arg)) (cdr arg)))
1743 '(("\\`BABYL OPTIONS:[ \t]*-\\*-[ \t]*rmail[ \t]*-\\*-" . no-conversion)
1744 ("\\`\xFE\xFF" . utf-16be-with-signature)
1745 ("\\`\xFF\xFE" . utf-16le-with-signature)
1746 ("\\`\xEF\xBB\xBF" . utf-8-with-signature)
1747 ("\\`;ELC\024\0\0\0" . emacs-mule))) ; Emacs 20-compiled
1748 "Alist of patterns vs corresponding coding systems.
1749 Each element looks like (REGEXP . CODING-SYSTEM).
1750 A file whose first bytes match REGEXP is decoded by CODING-SYSTEM on reading.
1752 The settings in this alist take priority over `coding:' tags
1753 in the file (see the function `set-auto-coding')
1754 and the contents of `file-coding-system-alist'."
1755 :group 'files
1756 :group 'mule
1757 :type '(repeat (cons (regexp :tag "Regexp")
1758 (symbol :tag "Coding system"))))
1760 (defun auto-coding-regexp-alist-lookup (from to)
1761 "Lookup `auto-coding-regexp-alist' for the contents of the current buffer.
1762 The value is a coding system is specified for the region FROM and TO,
1763 or nil."
1764 (save-excursion
1765 (goto-char from)
1766 (let ((alist auto-coding-regexp-alist)
1767 coding-system)
1768 (while (and alist (not coding-system))
1769 (let ((regexp (car (car alist))))
1770 (if enable-multibyte-characters
1771 (setq regexp (string-to-multibyte regexp)))
1772 (if (re-search-forward regexp to t)
1773 (setq coding-system (cdr (car alist)))
1774 (setq alist (cdr alist)))))
1775 coding-system)))
1777 ;; See the bottom of this file for built-in auto coding functions.
1778 (defcustom auto-coding-functions '(sgml-xml-auto-coding-function
1779 sgml-html-meta-auto-coding-function)
1780 "A list of functions which attempt to determine a coding system.
1782 Each function in this list should be written to operate on the
1783 current buffer, but should not modify it in any way. The buffer
1784 will contain undecoded text of parts of the file. Each function
1785 should take one argument, SIZE, which says how many characters
1786 \(starting from point) it should look at.
1788 If one of these functions succeeds in determining a coding
1789 system, it should return that coding system. Otherwise, it
1790 should return nil.
1792 If a file has a `coding:' tag, that takes precedence over these
1793 functions, so they won't be called at all."
1794 :group 'files
1795 :group 'mule
1796 :type '(repeat function))
1798 (defvar set-auto-coding-for-load nil
1799 "Non-nil means respect a \"unibyte: t\" entry in file local variables.
1800 Emacs binds this variable to t when loading or byte-compiling Emacs Lisp
1801 files.")
1803 (defun auto-coding-alist-lookup (filename)
1804 "Return the coding system specified by `auto-coding-alist' for FILENAME."
1805 (let ((alist auto-coding-alist)
1806 (case-fold-search (memq system-type '(windows-nt ms-dos cygwin)))
1807 coding-system)
1808 (while (and alist (not coding-system))
1809 (if (string-match (car (car alist)) filename)
1810 (setq coding-system (cdr (car alist)))
1811 (setq alist (cdr alist))))
1812 coding-system))
1814 (put 'enable-character-translation 'permanent-local t)
1815 (put 'enable-character-translation 'safe-local-variable 'booleanp)
1817 (defun find-auto-coding (filename size)
1818 "Find a coding system for a file FILENAME of which SIZE bytes follow point.
1819 These bytes should include at least the first 1k of the file
1820 and the last 3k of the file, but the middle may be omitted.
1822 The function checks FILENAME against the variable `auto-coding-alist'.
1823 If FILENAME doesn't match any entries in the variable, it checks the
1824 contents of the current buffer following point against
1825 `auto-coding-regexp-alist'. If no match is found, it checks for a
1826 `coding:' tag in the first one or two lines following point. If no
1827 `coding:' tag is found, it checks any local variables list in the last
1828 3K bytes out of the SIZE bytes. Finally, if none of these methods
1829 succeed, it checks to see if any function in `auto-coding-functions'
1830 gives a match.
1832 If a coding system is specified, the return value is a cons
1833 \(CODING . SOURCE), where CODING is the specified coding system and
1834 SOURCE is a symbol `auto-coding-alist', `auto-coding-regexp-alist',
1835 `:coding', or `auto-coding-functions' indicating by what CODING is
1836 specified. Note that the validity of CODING is not checked;
1837 it's the caller's responsibility to check it.
1839 If nothing is specified, the return value is nil."
1840 (or (let ((coding-system (auto-coding-alist-lookup filename)))
1841 (if coding-system
1842 (cons coding-system 'auto-coding-alist)))
1843 ;; Try using `auto-coding-regexp-alist'.
1844 (let ((coding-system (auto-coding-regexp-alist-lookup (point)
1845 (+ (point) size))))
1846 (if coding-system
1847 (cons coding-system 'auto-coding-regexp-alist)))
1848 (let* ((case-fold-search t)
1849 (head-start (point))
1850 (head-end (+ head-start (min size 1024)))
1851 (tail-start (+ head-start (max (- size 3072) 0)))
1852 (tail-end (+ head-start size))
1853 coding-system head-found tail-found char-trans)
1854 ;; Try a short cut by searching for the string "coding:"
1855 ;; and for "unibyte:" at the head and tail of SIZE bytes.
1856 (setq head-found (or (search-forward "coding:" head-end t)
1857 (search-forward "unibyte:" head-end t)
1858 (search-forward "enable-character-translation:"
1859 head-end t)))
1860 (if (and head-found (> head-found tail-start))
1861 ;; Head and tail are overlapped.
1862 (setq tail-found head-found)
1863 (goto-char tail-start)
1864 (setq tail-found (or (search-forward "coding:" tail-end t)
1865 (search-forward "unibyte:" tail-end t)
1866 (search-forward "enable-character-translation:"
1867 tail-end t))))
1869 ;; At first check the head.
1870 (when head-found
1871 (goto-char head-start)
1872 (setq head-end (set-auto-mode-1))
1873 (setq head-start (point))
1874 (when (and head-end (< head-found head-end))
1875 (goto-char head-start)
1876 (when (and set-auto-coding-for-load
1877 (re-search-forward
1878 "\\(.*;\\)?[ \t]*unibyte:[ \t]*\\([^ ;]+\\)"
1879 head-end t))
1880 (display-warning 'mule
1881 (format "\"unibyte: t\" (in %s) is obsolete; \
1882 use \"coding: 'raw-text\" instead."
1883 (file-relative-name filename))
1884 :warning)
1885 (setq coding-system 'raw-text))
1886 (when (and (not coding-system)
1887 (re-search-forward
1888 "\\(.*;\\)?[ \t]*coding:[ \t]*\\([^ ;]+\\)"
1889 head-end t))
1890 (setq coding-system (intern (match-string 2))))
1891 (when (re-search-forward
1892 "\\(.*;\\)?[ \t]*enable-character-translation:[ \t]*\\([^ ;]+\\)"
1893 head-end t)
1894 (setq char-trans (match-string 2)))))
1896 ;; If no coding: tag in the head, check the tail.
1897 ;; Here we must pay attention to the case that the end-of-line
1898 ;; is just "\r" and we can't use "^" nor "$" in regexp.
1899 (when (and tail-found (or (not coding-system) (not char-trans)))
1900 (goto-char tail-start)
1901 (re-search-forward "[\r\n]\^L" tail-end t)
1902 (if (re-search-forward
1903 "[\r\n]\\([^[\r\n]*\\)[ \t]*Local Variables:[ \t]*\\([^\r\n]*\\)[\r\n]"
1904 tail-end t)
1905 ;; The prefix is what comes before "local variables:" in its
1906 ;; line. The suffix is what comes after "local variables:"
1907 ;; in its line.
1908 (let* ((prefix (regexp-quote (match-string 1)))
1909 (suffix (regexp-quote (match-string 2)))
1910 (re-coding
1911 (concat
1912 "[\r\n]" prefix
1913 ;; N.B. without the \n below, the regexp can
1914 ;; eat newlines.
1915 "[ \t]*coding[ \t]*:[ \t]*\\([^ \t\r\n]+\\)[ \t]*"
1916 suffix "[\r\n]"))
1917 (re-unibyte
1918 (concat
1919 "[\r\n]" prefix
1920 "[ \t]*unibyte[ \t]*:[ \t]*\\([^ \t\r\n]+\\)[ \t]*"
1921 suffix "[\r\n]"))
1922 (re-char-trans
1923 (concat
1924 "[\r\n]" prefix
1925 "[ \t]*enable-character-translation[ \t]*:[ \t]*\\([^ \t\r\n]+\\)[ \t]*"
1926 suffix "[\r\n]"))
1927 (re-end
1928 (concat "[\r\n]" prefix "[ \t]*End *:[ \t]*" suffix
1929 "[\r\n]?"))
1930 (pos (1- (point))))
1931 (forward-char -1) ; skip back \r or \n.
1932 (re-search-forward re-end tail-end 'move)
1933 (setq tail-end (point))
1934 (goto-char pos)
1935 (when (and set-auto-coding-for-load
1936 (re-search-forward re-unibyte tail-end t))
1937 (display-warning 'mule "`unibyte: t' is obsolete; \
1938 use \"coding: 'raw-text\" instead." :warning)
1939 (setq coding-system 'raw-text))
1940 (when (and (not coding-system)
1941 (re-search-forward re-coding tail-end t))
1942 (setq coding-system (intern (match-string 1))))
1943 (when (and (not char-trans)
1944 (re-search-forward re-char-trans tail-end t))
1945 (setq char-trans (match-string 1))))))
1946 (if coding-system
1947 ;; If the coding-system name ends with "!", remove it and
1948 ;; set char-trans to "nil".
1949 (let ((name (symbol-name coding-system)))
1950 (if (= (aref name (1- (length name))) ?!)
1951 (setq coding-system (intern (substring name 0 -1))
1952 char-trans "nil"))))
1953 (when (and char-trans
1954 (not (setq char-trans (intern char-trans))))
1955 (make-local-variable 'enable-character-translation)
1956 (setq enable-character-translation nil))
1957 (if coding-system
1958 (cons coding-system :coding)))
1959 ;; Finally, try all the `auto-coding-functions'.
1960 (let ((funcs auto-coding-functions)
1961 (coding-system nil))
1962 (while (and funcs (not coding-system))
1963 (setq coding-system (ignore-errors
1964 (save-excursion
1965 (goto-char (point-min))
1966 (funcall (pop funcs) size)))))
1967 (if coding-system
1968 (cons coding-system 'auto-coding-functions)))))
1970 (defun set-auto-coding (filename size)
1971 "Return coding system for a file FILENAME of which SIZE bytes follow point.
1972 See `find-auto-coding' for how the coding system is found.
1973 Return nil if an invalid coding system is found.
1975 The variable `set-auto-coding-function' (which see) is set to this
1976 function by default."
1977 (let ((found (find-auto-coding filename size)))
1978 (if (and found (coding-system-p (car found)))
1979 (car found))))
1981 (setq set-auto-coding-function 'set-auto-coding)
1983 (defun after-insert-file-set-coding (inserted &optional visit)
1984 "Set `buffer-file-coding-system' of current buffer after text is inserted.
1985 INSERTED is the number of characters that were inserted, as figured
1986 in the situation before this function. Return the number of characters
1987 inserted, as figured in the situation after. The two numbers can be
1988 different if the buffer has become unibyte.
1989 The optional second arg VISIT non-nil means that we are visiting a file."
1990 (if (and visit
1991 coding-system-for-read
1992 (not (eq coding-system-for-read 'auto-save-coding)))
1993 (setq buffer-file-coding-system-explicit
1994 (cons coding-system-for-read nil)))
1995 (if last-coding-system-used
1996 (let ((coding-system
1997 (find-new-buffer-file-coding-system last-coding-system-used)))
1998 (if coding-system
1999 (setq buffer-file-coding-system coding-system))))
2000 inserted)
2002 ;; The coding-spec and eol-type of coding-system returned is decided
2003 ;; independently in the following order.
2004 ;; 1. That of buffer-file-coding-system locally bound.
2005 ;; 2. That of CODING.
2007 (defun find-new-buffer-file-coding-system (coding)
2008 "Return a coding system for a buffer when a file of CODING is inserted.
2009 The local variable `buffer-file-coding-system' of the current buffer
2010 is set to the returned value.
2011 Return nil if there's no need to set `buffer-file-coding-system'."
2012 (let (local-coding local-eol
2013 found-coding found-eol
2014 new-coding new-eol)
2015 (if (null coding)
2016 ;; Nothing found about coding.
2019 ;; Get information of `buffer-file-coding-system' in LOCAL-EOL
2020 ;; and LOCAL-CODING.
2021 (setq local-eol (coding-system-eol-type buffer-file-coding-system))
2022 (if (null (numberp local-eol))
2023 ;; But eol-type is not yet set.
2024 (setq local-eol nil))
2025 (if (and buffer-file-coding-system
2026 (not (eq (coding-system-type buffer-file-coding-system)
2027 'undecided)))
2028 (setq local-coding (coding-system-base buffer-file-coding-system)))
2030 (if (and (local-variable-p 'buffer-file-coding-system)
2031 local-eol local-coding)
2032 ;; The current buffer has already set full coding-system, we
2033 ;; had better not change it.
2036 (setq found-eol (coding-system-eol-type coding))
2037 (if (null (numberp found-eol))
2038 ;; But eol-type is not found.
2039 ;; If EOL conversions are inhibited, force unix eol-type.
2040 (setq found-eol (if inhibit-eol-conversion 0)))
2041 (setq found-coding (coding-system-base coding))
2043 (if (and (not found-eol) (eq found-coding 'undecided))
2044 ;; No valid coding information found.
2047 ;; Some coding information (eol or text) found.
2049 ;; The local setting takes precedence over the found one.
2050 (setq new-coding (if (local-variable-p 'buffer-file-coding-system)
2051 (or local-coding found-coding)
2052 (or found-coding local-coding)))
2053 (setq new-eol (if (local-variable-p 'buffer-file-coding-system)
2054 (or local-eol found-eol)
2055 (or found-eol local-eol)))
2057 (let ((eol-type (coding-system-eol-type new-coding)))
2058 (if (and (numberp new-eol) (vectorp eol-type))
2059 (aref eol-type new-eol)
2060 new-coding)))))))
2062 (defun modify-coding-system-alist (target-type regexp coding-system)
2063 "Modify one of look up tables for finding a coding system on I/O operation.
2064 There are three of such tables, `file-coding-system-alist',
2065 `process-coding-system-alist', and `network-coding-system-alist'.
2067 TARGET-TYPE specifies which of them to modify.
2068 If it is `file', it affects `file-coding-system-alist' (which see).
2069 If it is `process', it affects `process-coding-system-alist' (which see).
2070 If it is `network', it affects `network-coding-system-alist' (which see).
2072 REGEXP is a regular expression matching a target of I/O operation.
2073 The target is a file name if TARGET-TYPE is `file', a program name if
2074 TARGET-TYPE is `process', or a network service name or a port number
2075 to connect to if TARGET-TYPE is `network'.
2077 CODING-SYSTEM is a coding system to perform code conversion on the I/O
2078 operation, or a cons cell (DECODING . ENCODING) specifying the coding
2079 systems for decoding and encoding respectively, or a function symbol
2080 which, when called, returns such a cons cell."
2081 (or (memq target-type '(file process network))
2082 (error "Invalid target type: %s" target-type))
2083 (or (stringp regexp)
2084 (and (eq target-type 'network) (integerp regexp))
2085 (error "Invalid regular expression: %s" regexp))
2086 (if (symbolp coding-system)
2087 (if (not (fboundp coding-system))
2088 (progn
2089 (check-coding-system coding-system)
2090 (setq coding-system (cons coding-system coding-system))))
2091 (check-coding-system (car coding-system))
2092 (check-coding-system (cdr coding-system)))
2093 (cond ((eq target-type 'file)
2094 (let ((slot (assoc regexp file-coding-system-alist)))
2095 (if slot
2096 (setcdr slot coding-system)
2097 (setq file-coding-system-alist
2098 (cons (cons regexp coding-system)
2099 file-coding-system-alist)))))
2100 ((eq target-type 'process)
2101 (let ((slot (assoc regexp process-coding-system-alist)))
2102 (if slot
2103 (setcdr slot coding-system)
2104 (setq process-coding-system-alist
2105 (cons (cons regexp coding-system)
2106 process-coding-system-alist)))))
2108 (let ((slot (assoc regexp network-coding-system-alist)))
2109 (if slot
2110 (setcdr slot coding-system)
2111 (setq network-coding-system-alist
2112 (cons (cons regexp coding-system)
2113 network-coding-system-alist)))))))
2115 (defun decode-coding-inserted-region (from to filename
2116 &optional visit beg end replace)
2117 "Decode the region between FROM and TO as if it is read from file FILENAME.
2118 The idea is that the text between FROM and TO was just inserted somehow.
2119 Optional arguments VISIT, BEG, END, and REPLACE are the same as those
2120 of the function `insert-file-contents'.
2121 Part of the job of this function is setting `buffer-undo-list' appropriately."
2122 (save-excursion
2123 (save-restriction
2124 (let ((coding coding-system-for-read)
2125 undo-list-saved)
2126 (if visit
2127 ;; Temporarily turn off undo recording, if we're decoding the
2128 ;; text of a visited file.
2129 (setq buffer-undo-list t)
2130 ;; Otherwise, if we can recognize the undo elt for the insertion,
2131 ;; remove it and get ready to replace it later.
2132 ;; In the mean time, turn off undo recording.
2133 (let ((last (car-safe buffer-undo-list)))
2134 (if (and (consp last) (eql (car last) from) (eql (cdr last) to))
2135 (setq undo-list-saved (cdr buffer-undo-list)
2136 buffer-undo-list t))))
2137 (narrow-to-region from to)
2138 (goto-char (point-min))
2139 (or coding
2140 (setq coding (funcall set-auto-coding-function
2141 filename (- (point-max) (point-min)))))
2142 (or coding
2143 (setq coding (car (find-operation-coding-system
2144 'insert-file-contents
2145 (cons filename (current-buffer))
2146 visit beg end replace))))
2147 (if (coding-system-p coding)
2148 (or enable-multibyte-characters
2149 (setq coding
2150 (coding-system-change-text-conversion coding 'raw-text)))
2151 (setq coding nil))
2152 (if coding
2153 (decode-coding-region (point-min) (point-max) coding)
2154 (setq last-coding-system-used coding))
2155 ;; If we're decoding the text of a visited file,
2156 ;; the undo list should start out empty.
2157 (if visit
2158 (setq buffer-undo-list nil)
2159 ;; If we decided to replace the undo entry for the insertion,
2160 ;; do so now.
2161 (if undo-list-saved
2162 (setq buffer-undo-list
2163 (cons (cons from (point-max)) undo-list-saved))))))))
2165 (defun recode-region (start end new-coding coding)
2166 "Re-decode the region (previously decoded by CODING) by NEW-CODING."
2167 (interactive
2168 (list (region-beginning) (region-end)
2169 (read-coding-system "Text was really in: ")
2170 (let ((coding (or buffer-file-coding-system last-coding-system-used)))
2171 (read-coding-system
2172 (concat "But was interpreted as"
2173 (if coding (format " (default %S): " coding) ": "))
2174 coding))))
2175 (or (and new-coding coding)
2176 (error "Coding system not specified"))
2177 ;; Check it before we encode the region.
2178 (check-coding-system new-coding)
2179 (save-restriction
2180 (narrow-to-region start end)
2181 (encode-coding-region (point-min) (point-max) coding)
2182 (decode-coding-region (point-min) (point-max) new-coding))
2183 (if (region-active-p)
2184 (deactivate-mark)))
2186 (defun make-translation-table (&rest args)
2187 "Make a translation table from arguments.
2188 A translation table is a char table intended for character
2189 translation in CCL programs.
2191 Each argument is a list of elements of the form (FROM . TO), where FROM
2192 is a character to be translated to TO.
2194 The arguments and forms in each argument are processed in the given
2195 order, and if a previous form already translates TO to some other
2196 character, say TO-ALT, FROM is also translated to TO-ALT."
2197 (let ((table (make-char-table 'translation-table))
2198 revlist)
2199 (dolist (elts args)
2200 (dolist (elt elts)
2201 (let ((from (car elt))
2202 (to (cdr elt))
2203 to-alt rev-from rev-to)
2204 ;; If we have already translated TO to TO-ALT, FROM should
2205 ;; also be translated to TO-ALT.
2206 (if (setq to-alt (aref table to))
2207 (setq to to-alt))
2208 (aset table from to)
2209 ;; If we have already translated some chars to FROM, they
2210 ;; should also be translated to TO.
2211 (when (setq rev-from (assq from revlist))
2212 (dolist (elt (cdr rev-from))
2213 (aset table elt to))
2214 (setq revlist (delq rev-from revlist)
2215 rev-from (cdr rev-from)))
2216 ;; Now update REVLIST.
2217 (setq rev-to (assq to revlist))
2218 (if rev-to
2219 (setcdr rev-to (cons from (cdr rev-to)))
2220 (setq rev-to (list to from)
2221 revlist (cons rev-to revlist)))
2222 (if rev-from
2223 (setcdr rev-to (append rev-from (cdr rev-to)))))))
2224 ;; Return TABLE just created.
2225 (set-char-table-extra-slot table 1 1)
2226 table))
2228 (defun make-translation-table-from-vector (vec)
2229 "Make translation table from decoding vector VEC.
2230 VEC is an array of 256 elements to map unibyte codes to multibyte
2231 characters. Elements may be nil for undefined code points."
2232 (let ((table (make-char-table 'translation-table))
2233 (rev-table (make-char-table 'translation-table))
2235 (dotimes (i 256)
2236 (setq ch (aref vec i))
2237 (when ch
2238 (aset table i ch)
2239 (if (>= ch 256)
2240 (aset rev-table ch i))))
2241 (set-char-table-extra-slot table 0 rev-table)
2242 (set-char-table-extra-slot table 1 1)
2243 (set-char-table-extra-slot rev-table 1 1)
2244 table))
2246 (defun make-translation-table-from-alist (alist)
2247 "Make translation table from N<->M mapping in ALIST.
2248 ALIST is an alist, each element has the form (FROM . TO).
2249 FROM and TO are a character or a vector of characters.
2250 If FROM is a character, that character is translated to TO.
2251 If FROM is a vector of characters, that sequence is translated to TO.
2252 The first extra-slot of the value is a translation table for reverse mapping."
2253 (let ((tables (vector (make-char-table 'translation-table)
2254 (make-char-table 'translation-table)))
2255 table max-lookup from to idx val)
2256 (dotimes (i 2)
2257 (setq table (aref tables i))
2258 (setq max-lookup 1)
2259 (dolist (elt alist)
2260 (if (= i 0)
2261 (setq from (car elt) to (cdr elt))
2262 (setq from (cdr elt) to (car elt)))
2263 (if (characterp from)
2264 (setq idx from)
2265 (setq idx (aref from 0)
2266 max-lookup (max max-lookup (length from))))
2267 (setq val (aref table idx))
2268 (if val
2269 (progn
2270 (or (consp val)
2271 (setq val (list (cons (vector idx) val))))
2272 (if (characterp from)
2273 (setq from (vector from)))
2274 (setq val (nconc val (list (cons from to)))))
2275 (if (characterp from)
2276 (setq val to)
2277 (setq val (list (cons from to)))))
2278 (aset table idx val))
2279 (set-char-table-extra-slot table 1 max-lookup))
2280 (set-char-table-extra-slot (aref tables 0) 0 (aref tables 1))
2281 (aref tables 0)))
2283 (defun define-translation-table (symbol &rest args)
2284 "Define SYMBOL as the name of translation table made by ARGS.
2285 This sets up information so that the table can be used for
2286 translations in a CCL program.
2288 If the first element of ARGS is a char-table whose purpose is
2289 `translation-table', just define SYMBOL to name it. (Note that this
2290 function does not bind SYMBOL.)
2292 Any other ARGS should be suitable as arguments of the function
2293 `make-translation-table' (which see).
2295 This function sets properties `translation-table' and
2296 `translation-table-id' of SYMBOL to the created table itself and the
2297 identification number of the table respectively. It also registers
2298 the table in `translation-table-vector'."
2299 (let ((table (if (and (char-table-p (car args))
2300 (eq (char-table-subtype (car args))
2301 'translation-table))
2302 (car args)
2303 (apply 'make-translation-table args)))
2304 (len (length translation-table-vector))
2305 (id 0)
2306 (done nil))
2307 (put symbol 'translation-table table)
2308 (while (not done)
2309 (if (>= id len)
2310 (setq translation-table-vector
2311 (vconcat translation-table-vector (make-vector len nil))))
2312 (let ((slot (aref translation-table-vector id)))
2313 (if (or (not slot)
2314 (eq (car slot) symbol))
2315 (progn
2316 (aset translation-table-vector id (cons symbol table))
2317 (setq done t))
2318 (setq id (1+ id)))))
2319 (put symbol 'translation-table-id id)
2320 id))
2322 (defun translate-region (start end table)
2323 "From START to END, translate characters according to TABLE.
2324 TABLE is a string or a char-table.
2325 If TABLE is a string, the Nth character in it is the mapping
2326 for the character with code N.
2327 If TABLE is a char-table, the element for character N is the mapping
2328 for the character with code N.
2329 It returns the number of characters changed."
2330 (interactive
2331 (list (region-beginning)
2332 (region-end)
2333 (let (table l)
2334 (dotimes (i (length translation-table-vector))
2335 (if (consp (aref translation-table-vector i))
2336 (push (list (symbol-name
2337 (car (aref translation-table-vector i)))) l)))
2338 (if (not l)
2339 (error "No translation table defined"))
2340 (while (not table)
2341 (setq table (completing-read "Translation table: " l nil t)))
2342 (intern table))))
2343 (if (symbolp table)
2344 (let ((val (get table 'translation-table)))
2345 (or (char-table-p val)
2346 (error "Invalid translation table name: %s" table))
2347 (setq table val)))
2348 (translate-region-internal start end table))
2350 (defmacro with-category-table (table &rest body)
2351 "Execute BODY like `progn' with TABLE the current category table.
2352 The category table of the current buffer is saved, BODY is evaluated,
2353 then the saved table is restored, even in case of an abnormal exit.
2354 Value is what BODY returns."
2355 (declare (indent 1) (debug t))
2356 (let ((old-table (make-symbol "old-table"))
2357 (old-buffer (make-symbol "old-buffer")))
2358 `(let ((,old-table (category-table))
2359 (,old-buffer (current-buffer)))
2360 (unwind-protect
2361 (progn
2362 (set-category-table ,table)
2363 ,@body)
2364 (with-current-buffer ,old-buffer
2365 (set-category-table ,old-table))))))
2367 (defun define-translation-hash-table (symbol table)
2368 "Define SYMBOL as the name of the hash translation TABLE for use in CCL.
2370 Analogous to `define-translation-table', but updates
2371 `translation-hash-table-vector' and the table is for use in the CCL
2372 `lookup-integer' and `lookup-character' functions."
2373 (unless (and (symbolp symbol)
2374 (hash-table-p table))
2375 (error "Bad args to define-translation-hash-table"))
2376 (let ((len (length translation-hash-table-vector))
2377 (id 0)
2378 done)
2379 (put symbol 'translation-hash-table table)
2380 (while (not done)
2381 (if (>= id len)
2382 (setq translation-hash-table-vector
2383 (vconcat translation-hash-table-vector [nil])))
2384 (let ((slot (aref translation-hash-table-vector id)))
2385 (if (or (not slot)
2386 (eq (car slot) symbol))
2387 (progn
2388 (aset translation-hash-table-vector id (cons symbol table))
2389 (setq done t))
2390 (setq id (1+ id)))))
2391 (put symbol 'translation-hash-table-id id)
2392 id))
2394 ;;; Initialize some variables.
2396 (put 'use-default-ascent 'char-table-extra-slots 0)
2397 (setq use-default-ascent (make-char-table 'use-default-ascent))
2398 (put 'ignore-relative-composition 'char-table-extra-slots 0)
2399 (setq ignore-relative-composition
2400 (make-char-table 'ignore-relative-composition))
2402 ;;; Built-in auto-coding-functions:
2404 (defun sgml-xml-auto-coding-function (size)
2405 "Determine whether the buffer is XML, and if so, its encoding.
2406 This function is intended to be added to `auto-coding-functions'."
2407 (setq size (+ (point) size))
2408 (when (re-search-forward "\\`[[:space:]\n]*<\\?xml" size t)
2409 (let ((end (save-excursion
2410 ;; This is a hack.
2411 (re-search-forward "[\"']\\s-*\\?>" size t))))
2412 (when end
2413 (if (re-search-forward "encoding=[\"']\\(.+?\\)[\"']" end t)
2414 (let* ((match (match-string 1))
2415 (sym (intern (downcase match))))
2416 (if (coding-system-p sym)
2418 (message "Warning: unknown coding system \"%s\"" match)
2419 nil))
2420 ;; Files without an encoding tag should be UTF-8. But users
2421 ;; may be naive about encodings, and have saved the file from
2422 ;; another editor that does not help them get the encoding right.
2423 ;; Detect the encoding and warn the user if it is detected as
2424 ;; something other than UTF-8.
2425 (let ((detected
2426 (with-coding-priority '(utf-8)
2427 (coding-system-base
2428 (detect-coding-region (point-min) size t)))))
2429 ;; Pure ASCII always comes back as undecided.
2430 (if (memq detected '(utf-8 undecided))
2431 'utf-8
2432 (warn "File contents detected as %s.
2433 Consider adding an encoding attribute to the xml declaration,
2434 or saving as utf-8, as mandated by the xml specification." detected)
2435 detected)))))))
2437 (defun sgml-html-meta-auto-coding-function (size)
2438 "If the buffer has an HTML meta tag, use it to determine encoding.
2439 This function is intended to be added to `auto-coding-functions'."
2440 (let ((case-fold-search t))
2441 (setq size (min (+ (point) size)
2442 (save-excursion
2443 ;; Limit the search by the end of the HTML header.
2444 (or (search-forward "</head>" (+ (point) size) t)
2445 ;; In case of no header, search only 10 lines.
2446 (forward-line 10))
2447 (point))))
2448 ;; Make sure that the buffer really contains an HTML document, by
2449 ;; checking that it starts with a doctype or a <HTML> start tag
2450 ;; (allowing for whitespace at bob). Note: 'DOCTYPE NETSCAPE' is
2451 ;; useful for Mozilla bookmark files.
2452 (when (and (re-search-forward "\\`[[:space:]\n]*\\(<!doctype[[:space:]\n]+\\(html\\|netscape\\)\\|<html\\)" size t)
2453 (re-search-forward "<meta\\s-+\\(http-equiv=[\"']?content-type[\"']?\\s-+content=[\"']text/\\sw+;\\s-*\\)?charset=[\"']?\\(.+?\\)[\"'\\s-/>]" size t))
2454 (let* ((match (match-string 2))
2455 (sym (intern (downcase match))))
2456 (if (coding-system-p sym)
2458 (message "Warning: unknown coding system \"%s\"" match)
2459 nil)))))
2461 (defun xml-find-file-coding-system (args)
2462 "Determine the coding system of an XML file without a declaration.
2463 Strictly speaking, the file should be utf-8, but mistakes are
2464 made, and there are genuine cases where XML fragments are saved,
2465 with the encoding properly specified in a master document, or
2466 added by processing software."
2467 (if (eq (car args) 'insert-file-contents)
2468 (let ((detected
2469 (with-coding-priority '(utf-8)
2470 (coding-system-base
2471 (detect-coding-region (point-min) (point-max) t)))))
2472 ;; Pure ASCII always comes back as undecided.
2473 (cond
2474 ((memq detected '(utf-8 undecided))
2475 'utf-8)
2476 ((eq detected 'utf-16le-with-signature) 'utf-16le-with-signature)
2477 ((eq detected 'utf-16be-with-signature) 'utf-16be-with-signature)
2479 (warn "File contents detected as %s.
2480 Consider adding an xml declaration with the encoding specified,
2481 or saving as utf-8, as mandated by the xml specification." detected)
2482 detected)))
2483 ;; Don't interfere with the user's wishes for saving the buffer.
2484 ;; We did what we could when the buffer was created to ensure the
2485 ;; correct encoding was used, or the user was warned, so any
2486 ;; non-conformity here is deliberate on the part of the user.
2487 'undecided))
2490 (provide 'mule)
2492 ;;; mule.el ends here