1 ;;; codepage.el --- MS-DOS specific coding systems.
3 ;; Copyright (C) 1998 Free Software Foundation, Inc.
5 ;; Author: Eli Zaretskii
7 ;; Keywords: i18n ms-dos codepage
9 ;; This file is part of GNU Emacs.
11 ;; GNU Emacs is free software; you can redistribute it and/or modify
12 ;; it under the terms of the GNU General Public License as published by
13 ;; the Free Software Foundation; either version 2, or (at your option)
16 ;; GNU Emacs is distributed in the hope that it will be useful,
17 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 ;; GNU General Public License for more details.
21 ;; You should have received a copy of the GNU General Public License
22 ;; along with GNU Emacs; see the file COPYING. If not, write to the
23 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
24 ;; Boston, MA 02111-1307, USA.
28 ;; Special coding systems for DOS codepage support.
30 ;; These coding systems perform conversion from the DOS codepage encoding
31 ;; to one of the ISO-8859 character sets. Each codepage has its corresponding
32 ;; ISO-8859 charset, chosen so as to be able to convert all (or most) of the
33 ;; characters. The idea is that Emacs internally works with the usual MULE
34 ;; charsets, and the conversion to and from the DOS codepage is performed
36 ;; See term/internal.el for the complementary setup of the DOS terminal
37 ;; display and input methods.
39 ;; Thanks to Ken'ichi Handa <handa@etl.go.jp> for writing the CCL
40 ;; encoders/decoders, and for help in debugging this code.
44 (defun cp-coding-system-for-codepage-1 (coding mnemonic iso-name
46 "Make coding system CODING for a DOS codepage using translation tables.
47 MNEMONIC is a character to be displayed on mode line for the coding system.
48 ISO-NAME is the name of the ISO-8859 charset which corresponds to this
50 DECODER is a translation table for converting characters in the DOS codepage
51 encoding to Emacs multibyte characters.
52 ENCODER is a translation table for encoding Emacs multibyte characters into
53 external DOS codepage codes.
55 Note that the coding systems created by this function support automatic
56 detection of the EOL format. However, the decoders and encoders created
57 for these coding systems only support DOS and Unix style EOLs (the -mac
58 variety is actually just an alias for the -unix variety)."
60 (let* ((coding-name (symbol-name coding))
66 ((r0 = ,(charset-id 'ascii))
67 (translate-character ,decoder r0 r1)
68 (if (r0 == ,(charset-id 'ascii))
70 (write-multibyte-character r0 r1)))
77 ((r0 = ,(charset-id 'ascii))
78 (translate-character ,decoder r0 r1)
79 (if (r0 == ,(charset-id 'ascii))
81 (write-multibyte-character r0 r1)))
86 `(1 (loop (read-multibyte-character r0 r1)
89 (if (r0 != ,(charset-id 'ascii))
90 ((translate-character ,encoder r0 r1)
91 (if (r0 == ,(charset-id 'japanese-jisx0208))
97 `(1 (loop (read-multibyte-character r0 r1)
98 (if (r0 != ,(charset-id 'ascii))
99 ((translate-character ,encoder r0 r1)
100 (if (r0 == ,(charset-id 'japanese-jisx0208))
103 (write-repeat r1))))))
104 (if (memq coding coding-system-list)
105 (setq coding-system-list (delq coding coding-system-list)))
107 ;; Make coding system CODING.
110 (concat "8-bit encoding of " (symbol-name iso-name)
111 " characters using IBM codepage " coding-name)
112 (cons ccl-decoder-unix ccl-encoder-unix)
113 `((safe-charsets ascii ,iso-name)
114 (valid-codes (0 . 255))))
115 ;;; Make coding systems CODING-unix, CODING-dos, CODING-mac.
116 (make-subsidiary-coding-system coding)
117 (put coding 'eol-type (vector (intern (format "%s-unix" coding))
118 (intern (format "%s-dos" coding))
119 (intern (format "%s-mac" coding))))
120 ;; Change CCL code for CODING-dos.
121 (let ((coding-spec (copy-sequence (get coding 'coding-system))))
123 (cons (check-ccl-program
125 (intern (format "%s-dos-decoder" coding)))
128 (intern (format "%s-dos-encoder" coding)))))
129 (put (intern (concat coding-name "-dos")) 'coding-system
132 (defun cp-decoding-vector-for-codepage (table charset offset)
133 "Create a vector for decoding IBM PC characters using conversion table
134 TABLE into an ISO-8859 character set CHARSET whose first non-ASCII
135 character is generated by (make-char CHARSET OFFSET)."
136 (let* ((len (length table))
138 (if (eq system-type 'ms-dos)
139 (if dos-unsupported-char-glyph
140 (logand dos-unsupported-char-glyph 255)
143 (vec1 (make-vector 256 undefined-char))
151 (aset vec1 (aref table i) (make-char charset (+ i offset))))
155 ;;; You don't think I created all these tables below by hand, do you?
156 ;;; The following Awk script will create the table for cp850-to-Latin-1
157 ;;; conversion from the RFC 1345 file (the other tables are left as an
159 ;;; BEGIN { n_pages = 11;
160 ;;; pn["IBM437"] = 0; pn["IBM850"] = 1; pn["IBM851"] = 2;
161 ;;; pn["IBM852"] = 3; pn["IBM855"] = 4; pn["IBM860"] = 5;
162 ;;; pn["IBM861"] = 6; pn["IBM862"] = 7; pn["IBM863"] = 8;
163 ;;; pn["IBM864"] = 9; pn["IBM865"] = 10;
165 ;;; $1 == "&charset" { charset = $2; }
166 ;;; $1 == "&code" { code = $2; }
168 ;;; if ((charset ~ /^IBM(437|8(5[0125]|6[0-5]))$/) || (charset ~ /^ISO_8859-1/))
170 ;;; for (i = 1; i <= NF; i++)
171 ;;; chars[charset,code++] = $i;
176 ;;; for (i = 160; i < 256; i++)
178 ;;; c = chars["ISO_8859-1:1987",i];
179 ;;; if (c == "??") # skip unused positions
182 ;;; if ((i - 159)%16 == 0)
189 ;;; for (combined in chars)
191 ;;; candidate = chars[combined];
192 ;;; split (combined, separate, SUBSEP);
193 ;;; if (separate[1] == "IBM850" && candidate == c)
196 ;;; map[separate[1]] = separate[2];
199 ;;; printf " %s", map["IBM850"];
200 ;;; if ((i - 159)%16 == 0)
205 ;;; WARNING WARNING WARNING!!!
207 ;;; If you want to get fancy with these tables, remember that the inverse
208 ;;; tables, created by `cp-decoding-vector-for-codepage' above, are installed
209 ;;; on MS-DOS as nonascii-translation-table (see `dos-codepage-setup' on
210 ;;; internal.el). Therefore, you should NOT put any codes below 128 in
211 ;;; these tables! Otherwise, various Emacs commands and functions will
212 ;;; mysteriously fail! For example, a typical screwup is to map the Latin-N
213 ;;; acute accent character to the apostrophe, and have all regexps which
214 ;;; end with "\\'" begin to fail (e.g., the automatic setting of the major
215 ;;; mode by file name extension will stop working).
217 ;;; You HAVE BEEN warned!
219 ;; US/English/PC-8/IBM-2. This doesn't support Latin-1 characters very
220 ;; well, but why not use what we can salvage?
221 (defvar cp437-decode-table
222 ;; Nth element is the code of a cp437 glyph for the multibyte
223 ;; character created by (make-char 'latin-iso8859-1 (+ N 160)).
224 ;; The element nil means there's no corresponding cp437 glyph.
226 255 173 155 156 nil 157 179 nil nil nil 166 174 170 196 nil nil
227 248 241 253 nil nil nil nil 249 nil nil 167 175 172 171 nil 168
228 nil nil nil nil 142 143 146 128 nil 144 nil nil nil nil nil nil
229 nil 165 nil nil nil nil 153 nil nil nil nil nil 154 nil nil 225
230 133 160 131 nil 132 134 145 135 138 130 136 137 141 161 140 139
231 nil 164 149 162 147 nil 148 246 nil 151 163 150 129 nil nil 152]
232 "Table for converting ISO-8859-1 characters into codepage 437 glyphs.")
233 (setplist 'cp437-decode-table
234 '(charset latin-iso8859-1 language "Latin-1" offset 160))
236 ;; Multilingual (Latin-1)
237 (defvar cp850-decode-table
238 ;; Nth element is the code of a cp850 glyph for the multibyte
239 ;; character created by (make-char 'latin-iso8859-1 (+ N 160)).
240 ;; The element nil means there's no corresponding cp850 glyph.
242 255 173 189 156 207 190 221 245 249 184 166 174 170 240 169 nil
243 248 241 253 252 239 230 244 250 247 251 167 175 172 171 243 168
244 183 181 182 199 142 143 146 128 212 144 210 211 222 214 215 216
245 209 165 227 224 226 229 153 158 157 235 233 234 154 237 231 225
246 133 160 131 198 132 134 145 135 138 130 136 137 141 161 140 139
247 208 164 149 162 147 228 148 246 155 151 163 150 129 236 232 152]
248 "Table for converting ISO-8859-1 characters into codepage 850 glyphs.")
249 (setplist 'cp850-decode-table
250 '(charset latin-iso8859-1 language "Latin-1" offset 160))
253 (defvar cp851-decode-table
255 255 nil nil 156 nil nil nil 245 249 nil nil 174 nil 240 nil nil
256 248 241 nil nil 239 nil 134 nil 141 143 144 175 146 171 149 152
257 161 164 165 166 167 168 169 170 172 173 181 182 184 183 189 190
258 198 199 nil 207 208 209 210 211 212 213 nil nil 155 157 158 159
259 252 214 215 216 221 222 224 225 226 227 228 229 230 231 232 233
260 234 235 237 236 238 242 243 244 246 250 160 251 162 163 253 nil]
261 "Table for converting ISO-8859-7 characters into codepage 851 glyphs.")
262 (setplist 'cp851-decode-table
263 '(charset greek-iso8859-7 language "Greek" offset 160))
265 ;; Slavic/Eastern Europe (Latin-2)
266 (defvar cp852-decode-table
268 255 164 244 157 207 149 151 245 249 230 184 155 141 240 166 189
269 248 165 247 136 239 150 152 243 242 231 173 156 171 241 167 190
270 232 181 182 198 142 145 143 128 172 144 168 211 183 214 nil 210
271 209 227 213 224 226 138 153 158 252 222 233 235 154 237 221 225
272 234 160 131 199 132 146 134 135 159 130 169 137 216 161 140 212
273 208 228 229 162 147 139 148 246 253 133 163 251 129 236 238 250]
274 "Table for converting ISO-8859-2 characters into codepage 852 glyphs.")
275 (setplist 'cp852-decode-table
276 '(charset latin-iso8859-2 language "Latin-2" offset 160))
279 (defvar cp855-decode-table
281 255 133 129 131 135 137 139 141 143 145 147 149 151 240 153 155
282 161 163 236 173 167 169 234 244 184 190 199 209 211 213 215 221
283 226 228 230 232 171 182 165 252 246 250 159 242 238 248 157 224
284 160 162 235 172 166 168 233 243 183 189 198 208 210 212 214 216
285 225 227 229 231 170 181 164 251 245 249 158 241 237 247 156 222
286 239 132 128 130 134 136 138 140 142 144 146 148 150 253 152 154]
287 "Table for converting ISO-8859-5 characters into codepage 855 glyphs.")
288 (setplist 'cp855-decode-table
289 '(charset cyrillic-iso8859-5 language "Cyrillic-ISO" offset 160))
292 (defvar cp857-decode-table
294 255 nil nil 156 207 nil 245 249 152 158 166 nil 240 nil
295 248 nil 253 252 239 nil nil nil nil 141 159 167 nil 171 nil
296 183 181 182 142 nil nil 128 212 144 210 211 222 214 215 216
297 165 227 224 226 nil 153 232 nil 235 233 234 154 nil nil 225
298 133 160 131 132 nil nil 135 138 130 136 137 236 161 140 139
299 164 149 162 147 nil 148 246 nil 151 163 150 129 nil nil 250]
300 "Table for converting ISO-8859-3 characters into codepage 857 glyphs.")
301 (setplist 'cp857-decode-table
302 '(charset latin-iso8859-3 language "Latin-3" offset 160))
305 (defvar cp860-decode-table
307 255 173 155 156 nil nil 179 nil nil nil 166 174 170 nil nil nil
308 nil 241 253 nil nil nil nil 249 nil nil 167 175 172 171 nil 168
309 145 134 143 142 nil nil nil 128 146 144 137 nil 152 nil 139 nil
310 nil 165 159 169 140 153 nil nil nil 157 150 nil 154 nil nil nil
311 133 160 131 132 nil nil nil 135 138 130 136 nil 141 161 nil nil
312 nil 164 149 162 147 148 nil 246 nil 151 163 nil 129 nil nil nil]
313 "Table for converting ISO-8859-1 characters into codepage 860 glyphs.")
314 (setplist 'cp860-decode-table
315 '(charset latin-iso8859-1 language "Latin-1" offset 160))
318 (defvar cp861-decode-table
320 255 173 nil 156 nil nil nil nil nil nil nil 174 170 nil nil nil
321 nil 241 253 nil nil nil nil 249 nil nil nil 175 172 171 nil 168
322 nil 164 nil nil 142 143 146 128 nil 144 nil nil nil 165 nil nil
323 139 nil 159 166 nil nil 153 nil 157 nil 167 nil 154 151 141 nil
324 133 160 131 nil 132 134 145 135 138 130 136 137 nil 161 nil nil
325 140 nil nil 162 147 nil 148 246 155 nil 163 150 129 152 149 nil]
326 "Table for converting ISO-8859-1 characters into codepage 861 glyphs.")
327 (setplist 'cp861-decode-table
328 '(charset latin-iso8859-1 language "Latin-1" offset 160))
331 (defvar cp862-decode-table
332 ;; Nth element is the code of a cp862 glyph for the multibyte
333 ;; character created by (make-char 'hebrew-iso8859-8 (+ N 160)).
334 ;; The element nil means there's no corresponding cp850 glyph.
336 255 173 155 156 nil 157 179 nil nil nil nil 174 170 196 nil nil
337 248 241 253 nil nil 230 nil 249 nil nil 246 175 172 171 nil nil
338 nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil
339 nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil 205
340 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143
341 144 145 146 147 148 149 150 151 152 153 154 nil nil nil nil nil]
342 "Table for converting ISO-8859-8 characters into codepage 862 glyphs.")
343 (setplist 'cp862-decode-table
344 '(charset hebrew-iso8859-8 language "Hebrew" offset 160))
347 (defvar cp863-decode-table
349 255 nil 155 156 152 nil 160 143 164 nil nil 174 170 nil nil 167
350 nil 241 253 166 161 nil 134 249 165 nil nil 175 172 171 173 nil
351 142 nil 132 nil nil nil nil 128 145 144 146 148 nil nil 168 149
352 nil nil nil nil 153 nil nil nil nil 157 nil 158 154 nil nil nil
353 133 nil 131 nil nil nil nil 135 138 130 136 137 141 nil 140 139
354 nil nil nil 162 147 nil nil 246 nil 151 163 150 129 nil nil nil]
355 "Table for converting ISO-8859-1 characters into codepage 863 glyphs.")
356 (setplist 'cp863-decode-table
357 '(charset latin-iso8859-1 language "Latin-1" offset 160))
360 ;; FIXME: Emacs doesn't seem to support the "Arabic" language
361 ;; environment yet. So this is only partially usable, for now
362 (defvar cp864-decode-table
364 255 nil nil nil 164 nil nil nil nil nil nil nil 172 161 nil nil
365 nil nil nil nil nil nil nil nil nil nil nil 187 nil nil nil 191
366 nil 193 194 195 196 nil 198 199 169 201 170 171 173 174 175 207
367 208 209 210 188 189 190 235 215 216 223 238 nil nil nil nil nil
368 224 247 248 252 251 239 242 243 232 233 253 nil nil nil nil nil
369 nil 241 nil nil nil nil nil nil nil nil nil nil nil nil nil nil]
370 "Table for converting ISO-8859-1 characters into codepage 863 glyphs.")
371 (setplist 'cp864-decode-table
372 '(charset arabic-iso8859-6 language nil offset 160))
374 ;; Nordic (Norwegian/Danish)
375 (defvar cp865-decode-table
377 255 173 nil 156 nil nil nil nil nil nil 166 174 170 nil nil nil
378 nil 241 253 nil nil nil nil 249 nil nil 167 175 172 171 nil 168
379 nil nil nil nil 142 143 146 128 nil 144 nil nil nil nil nil nil
380 nil 165 nil nil nil nil 153 nil 157 nil nil nil 154 nil nil nil
381 133 160 131 nil 132 134 145 135 138 130 136 137 141 161 140 139
382 nil 164 149 162 147 nil 148 246 155 151 163 150 129 nil nil 152]
383 "Table for converting ISO-8859-1 characters into codepage 865 glyphs.")
384 (setplist 'cp865-decode-table
385 '(charset latin-iso8859-1 language "Latin-1" offset 160))
387 ;; Greek (yes, another one!)
388 (defvar cp869-decode-table
390 255 139 140 156 nil nil 138 245 249 151 nil 174 137 240 nil 142
391 248 241 153 154 239 247 134 136 141 143 144 175 146 171 149 152
392 161 164 165 166 167 168 169 170 172 173 181 182 183 184 189 190
393 198 199 nil 207 208 209 210 211 212 213 145 150 155 157 158 159
394 252 214 215 216 221 222 224 225 226 227 228 229 230 231 232 233
395 234 235 237 236 238 242 243 244 246 250 160 251 162 163 253 nil]
396 "Table for converting ISO-8859-7 characters into codepage 869 glyphs.")
397 (setplist 'cp869-decode-table
398 '(charset greek-iso8859-7 language "Greek" offset 160))
400 ;; Conversion from codepage 775 to Latin-4 for Baltic countries.
401 (defvar cp775-decode-table
403 255 181 nil 138 150 nil 234 245 166 190 237 149 173 240 207 nil
404 248 208 nil 139 239 nil 235 nil nil 213 137 133 nil nil 216 nil
405 160 nil nil nil 142 143 146 189 182 144 183 nil 184 nil nil 161
406 nil 238 226 232 nil 229 153 158 157 198 nil nil 154 nil 199 225
407 131 nil nil nil 132 134 145 212 209 130 210 nil 211 nil nil 140
408 nil 236 147 233 nil 228 148 nil 155 214 nil nil 129 nil 215 nil]
409 "Table for converting ISO-8859-4 characters into codepage 775 glyphs.")
410 (setplist 'cp775-decode-table
411 '(charset latin-iso8859-4 language "Latin-4" offset 160))
414 (defun cp-make-coding-systems-for-codepage (codepage iso-name offset)
415 "Create a coding system to convert IBM CODEPAGE into charset ISO-NAME
416 whose first character is at offset OFFSET from the beginning of 8-bit
419 The created coding system has the usual 3 subsidiary systems: for Unix-,
420 DOS- and Mac-style EOL conversion. However, unlike built-in coding
421 systems, the Mac-style EOL conversion is currently not supported by the
422 decoder and encoder created by this function."
423 (let* ((decode-table (intern (format "%s-decode-table" codepage)))
425 (intern (format "%s-nonascii-translation-table" codepage)))
427 (intern (format "%s-decode-translation-table" codepage)))
429 (intern (format "%s-encode-translation-table" codepage))))
431 (make-translation-table-from-vector
432 (cp-decoding-vector-for-codepage
433 (symbol-value decode-table) iso-name offset)))
434 (define-translation-table encode-translation
435 (char-table-extra-slot (symbol-value nonascii-table) 0))
436 ;; For charsets other than ascii and ISO-NAME, set `?' for
437 ;; one-column charsets, and some Japanese character for
438 ;; wide-column charsets. CCL encoder convert that Japanese
439 ;; character to "??".
440 (let ((tbl (char-table-extra-slot (symbol-value nonascii-table) 0))
441 (charsets (delq 'ascii (delq iso-name
442 (copy-sequence charset-list))))
443 (wide-column-char (make-char 'japanese-jisx0208 32 32)))
445 (aset tbl (make-char (car charsets))
446 (if (= (charset-width (car charsets)) 1) ?? wide-column-char))
447 (setq charsets (cdr charsets))))
448 (define-translation-table decode-translation
449 (symbol-value nonascii-table))
450 (cp-coding-system-for-codepage-1
451 (intern codepage) ?D iso-name decode-translation encode-translation)
454 (defun cp-codepage-decoder (codepage)
455 "If CODEPAGE is the name of a supported codepage, return its decode table;
456 otherwise return nil."
457 (let ((cp (if (symbolp codepage) (symbol-name codepage) codepage)))
460 (intern-soft (format "%s-decode-table" cp)))
464 (defun cp-charset-for-codepage (codepage)
465 "Return the charset for which there is a translation table to DOS CODEPAGE.
466 CODEPAGE must be the name of a DOS codepage, a string."
467 (let ((cp-decoder (cp-codepage-decoder codepage)))
468 (if (null cp-decoder)
469 (error "Unsupported codepage %s" codepage)
470 (get cp-decoder 'charset))))
473 (defun cp-language-for-codepage (codepage)
474 "Return the name of the MULE language environment for CODEPAGE.
475 CODEPAGE must be the name of a DOS codepage, a string."
476 (let ((cp-decoder (cp-codepage-decoder codepage)))
477 (if (null cp-decoder)
478 (error "Unsupported codepage %s" codepage)
479 (get cp-decoder 'language))))
482 (defun cp-offset-for-codepage (codepage)
483 "Return the offset to be used in setting up coding systems for CODEPAGE.
484 CODEPAGE must be the name of a DOS codepage, a string."
485 (let ((cp-decoder (cp-codepage-decoder codepage)))
486 (if (null cp-decoder)
487 (error "Unsupported codepage %s" codepage)
488 (get cp-decoder 'offset))))
491 (defun cp-supported-codepages ()
492 "Return an alist of supported codepages.
494 Each association in the alist has the form (NNN . CHARSET), where NNN is the
495 codepage number, and CHARSET is the MULE charset which is the closest match
496 for the character set supported by that codepage.
498 A codepage NNN is supported if a variable called `cpNNN-decode-table' exists,
499 is a vector, and has a charset property."
501 (let (alist chset sname)
505 (if (and (boundp sym)
506 (string-match "\\`cp\\([1-9][0-9][0-9]\\)-decode-table\\'"
507 (setq sname (symbol-name sym)))
508 (vectorp (symbol-value sym))
509 (setq chset (get sym 'charset)))
511 (cons (cons (match-string 1 sname) chset) alist))))))
515 (defun codepage-setup (codepage)
516 "Create a coding system cpCODEPAGE to support the IBM codepage CODEPAGE.
518 These coding systems are meant for encoding and decoding 8-bit non-ASCII
519 characters used by the IBM codepages, typically in conjunction with files
520 read/written by MS-DOS software, or for display on the MS-DOS terminal."
522 (let ((completion-ignore-case t)
523 (candidates (cp-supported-codepages)))
524 (list (completing-read "Setup DOS Codepage: (default 437) " candidates
525 nil t nil nil "437"))))
526 (let ((cp (format "cp%s" codepage)))
527 (cp-make-coding-systems-for-codepage
528 cp (cp-charset-for-codepage cp) (cp-offset-for-codepage cp))))
532 ;; codepage.el ends here