1 ;;; deuglify.el --- deuglify broken Outlook (Express) articles
3 ;; Copyright (C) 2002, 2003, 2004, 2005 Free Software Foundation, Inc.
4 ;; Copyright (C) 2001, 2002 Raymond Scholz
6 ;; Author: Raymond Scholz <rscholz@zonix.de>
7 ;; Thomas Steffen (unwrapping algorithm,
8 ;; based on an idea of Stefan Monnier)
9 ;; Keywords: mail, news
11 ;; This file is part of GNU Emacs.
13 ;; GNU Emacs is free software; you can redistribute it and/or modify
14 ;; it under the terms of the GNU General Public License as published by
15 ;; the Free Software Foundation; either version 2, or (at your option)
18 ;; GNU Emacs is distributed in the hope that it will be useful,
19 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
20 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
21 ;; GNU General Public License for more details.
23 ;; You should have received a copy of the GNU General Public License
24 ;; along with GNU Emacs; see the file COPYING. If not, write to the
25 ;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
26 ;; Boston, MA 02110-1301, USA.
30 ;; This file enables Gnus to repair broken citations produced by
31 ;; common user agents like MS Outlook (Express). It may repair
32 ;; articles of other user agents too.
34 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
37 ;; Outlook sometimes wraps cited lines before sending a message as
38 ;; seen in this example:
45 ;; > This sentence no verb. This sentence no verb. This sentence
47 ;; > verb. This sentence no verb. This sentence no verb. This
48 ;; > sentence no verb.
50 ;; The function `gnus-article-outlook-unwrap-lines' tries to recognize those
51 ;; erroneously wrapped lines and will unwrap them. I.e. putting the
52 ;; wrapped parts ("no" in this example) back where they belong (at the
53 ;; end of the cited line above).
55 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
57 ;; Note that some people not only use broken user agents but also
58 ;; practice a bad citation style by omitting blank lines between the
59 ;; cited text and their own text.
66 ;; > This sentence no verb. This sentence no verb. This sentence no
67 ;; You forgot in all your sentences.
68 ;; > verb. This sentence no verb. This sentence no verb. This
69 ;; > sentence no verb.
71 ;; Unwrapping "You forgot in all your sentences." would be illegal as
72 ;; this part wasn't intended to be cited text.
73 ;; `gnus-article-outlook-unwrap-lines' will only unwrap lines if the resulting
74 ;; citation line will be of a certain maximum length. You can control
75 ;; this by adjusting `gnus-outlook-deuglify-unwrap-max'. Also
76 ;; unwrapping will only be done if the line above the (possibly)
77 ;; wrapped line has a minimum length of `gnus-outlook-deuglify-unwrap-min'.
79 ;; Furthermore no unwrapping will be undertaken if the last character
80 ;; is one of the chars specified in
81 ;; `gnus-outlook-deuglify-unwrap-stop-chars'. Setting this to ".?!"
82 ;; inhibits unwrapping if the cited line ends with a full stop,
83 ;; question mark or exclamation mark. Note that this variable
84 ;; defaults to `nil', triggering a few false positives but generally
85 ;; giving you better results.
87 ;; Unwrapping works on every level of citation. Thus you will be able
88 ;; repair broken citations of broken user agents citing broken
89 ;; citations of broken user agents citing broken citations...
91 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
93 ;; Citations are commonly introduced with an attribution line
94 ;; indicating who wrote the cited text. Outlook adds superfluous
95 ;; information that can be found in the header of the message to this
96 ;; line and often wraps it.
98 ;; If that weren't enough, lots of people write their own text above
99 ;; the cited text and cite the complete original article below.
104 ;; Hey, John. There's no in all your sentences!
106 ;; John Doe <john.doe@some.domain> wrote in message
107 ;; news:a87usw8$dklsssa$2@some.news.server...
108 ;; > This sentence no verb. This sentence no verb. This sentence
110 ;; > verb. This sentence no verb. This sentence no verb. This
111 ;; > sentence no verb.
115 ;; Repairing the attribution line will be done by function
116 ;; `gnus-article-outlook-repair-attribution which calls other function that
117 ;; try to recognize and repair broken attribution lines. See variable
118 ;; `gnus-outlook-deuglify-attrib-cut-regexp' for stuff that should be
119 ;; cut off from the beginning of an attribution line and variable
120 ;; `gnus-outlook-deuglify-attrib-verb-regexp' for the verbs that are
121 ;; required to be found in an attribution line. These function return
122 ;; the point where the repaired attribution line starts.
124 ;; Rearranging the article so that the cited text appears above the
125 ;; new text will be done by function
126 ;; `gnus-article-outlook-rearrange-citation'. This function calls
127 ;; `gnus-article-outlook-repair-attribution to find and repair an attribution
130 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
132 ;; Well, and that's what the message will look like after applying
135 ;; Example #3 (deuglified)
136 ;; -----------------------
138 ;; John Doe <john.doe@some.domain> wrote:
140 ;; > This sentence no verb. This sentence no verb. This sentence no
141 ;; > verb. This sentence no verb. This sentence no verb. This
142 ;; > sentence no verb.
146 ;; Hey, John. There's no in all your sentences!
148 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
153 ;; Press `W k' in the Summary Buffer.
155 ;; Non recommended usage :-)
156 ;; ---------------------
158 ;; To automatically invoke deuglification on every article you read,
159 ;; put something like that in your .gnus:
161 ;; (add-hook 'gnus-article-decode-hook 'gnus-article-outlook-unwrap-lines)
163 ;; or _one_ of the following lines:
165 ;; ;; repair broken attribution lines
166 ;; (add-hook 'gnus-article-decode-hook 'gnus-article-outlook-repair-attribution)
168 ;; ;; repair broken attribution lines and citations
169 ;; (add-hook 'gnus-article-decode-hook 'gnus-article-outlook-rearrange-citation)
171 ;; Note that there always may be some false positives, so I suggest
172 ;; using the manual invocation. After deuglification you may want to
173 ;; refill the whole article using `W w'.
175 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
180 ;; As I said before there may (or will) be a few false positives on
181 ;; unwrapping cited lines with `gnus-article-outlook-unwrap-lines'.
183 ;; `gnus-article-outlook-repair-attribution will only fix the first
184 ;; attribution line found in the article. Furthermore it fixed to
185 ;; certain kinds of attributions. And there may be horribly many
186 ;; false positives, vanishing lines and so on -- so don't trust your
187 ;; eyes. Again I recommend manual invocation.
189 ;; `gnus-article-outlook-rearrange-citation' carries all the limitations of
190 ;; `gnus-article-outlook-repair-attribution.
192 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
194 ;; See ChangeLog for other changes.
196 ;; Revision 1.5 2002/01/27 14:39:17 rscholz
197 ;; * New variable `gnus-outlook-deuglify-no-wrap-chars' to inhibit
198 ;; unwrapping if one these chars is first in the possibly wrapped line.
199 ;; * Improved rearranging of the article.
200 ;; * New function `gnus-outlook-repair-attribution-block' for repairing
201 ;; those big "Original Message (following some headers)" attributions.
203 ;; Revision 1.4 2002/01/03 14:05:00 rscholz
204 ;; Renamed `gnus-outlook-deuglify-article' to
205 ;; `gnus-article-outlook-deuglify-article'.
206 ;; Made it easier to deuglify the article while being in Gnus' Article
207 ;; Edit Mode. (suggested by Phil Nitschke)
210 ;; Revision 1.3 2002/01/02 23:35:54 rscholz
211 ;; Fix a bug that caused succeeding long attribution lines to be
212 ;; unwrapped. Minor doc fixes and regular expression tuning.
214 ;; Revision 1.2 2001/12/30 20:14:34 rscholz
217 ;; Revision 1.1 2001/12/30 20:13:32 rscholz
220 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
227 (defconst gnus-outlook-deuglify-version "1.5 Gnus version"
228 "Version of gnus-outlook-deuglify.")
230 ;;; User Customizable Variables:
232 (defgroup gnus-outlook-deuglify nil
233 "Deuglify articles generated by broken user agents like MS Outlook (Express)."
238 (defcustom gnus-outlook-deuglify-unwrap-min 45
239 "Minimum length of the cited line above the (possibly) wrapped line."
242 :group 'gnus-outlook-deuglify)
245 (defcustom gnus-outlook-deuglify-unwrap-max 95
246 "Maximum length of the cited line after unwrapping."
249 :group 'gnus-outlook-deuglify)
251 (defcustom gnus-outlook-deuglify-cite-marks ">|#%"
252 "Characters that indicate cited lines."
255 :group 'gnus-outlook-deuglify)
257 (defcustom gnus-outlook-deuglify-unwrap-stop-chars nil ;; ".?!" or nil
258 "Characters that inhibit unwrapping if they are the last one on the cited line above the possible wrapped line."
260 :type '(radio (const :format "None " nil)
261 (string :value ".?!"))
262 :group 'gnus-outlook-deuglify)
264 (defcustom gnus-outlook-deuglify-no-wrap-chars "`"
265 "Characters that inhibit unwrapping if they are the first one in the possibly wrapped line."
268 :group 'gnus-outlook-deuglify)
270 (defcustom gnus-outlook-deuglify-attrib-cut-regexp
271 "\\(On \\|Am \\)?\\(Mon\\|Tue\\|Wed\\|Thu\\|Fri\\|Sat\\|Sun\\),[^,]+, "
272 "Regular expression matching the beginning of an attribution line that should be cut off."
275 :group 'gnus-outlook-deuglify)
277 (defcustom gnus-outlook-deuglify-attrib-verb-regexp
278 "wrote\\|writes\\|says\\|schrieb\\|schreibt\\|meinte\\|skrev\\|a écrit\\|schreef\\|escribió"
279 "Regular expression matching the verb used in an attribution line."
282 :group 'gnus-outlook-deuglify)
284 (defcustom gnus-outlook-deuglify-attrib-end-regexp
286 "Regular expression matching the end of an attribution line."
289 :group 'gnus-outlook-deuglify)
292 (defcustom gnus-outlook-display-hook nil
293 "A hook called after an deuglified article has been prepared.
294 It is run after `gnus-article-prepare-hook'."
297 :group 'gnus-outlook-deuglify)
301 (defun gnus-outlook-display-article-buffer ()
302 "Redisplay current buffer or article buffer."
303 (with-current-buffer (or gnus-article-buffer (current-buffer))
304 ;; "Emulate" `gnus-article-prepare-display' without calling
305 ;; it. Calling `gnus-article-prepare-display' on an already
306 ;; prepared article removes all MIME parts. I'm unsure whether
307 ;; this is a bug or not.
308 (gnus-article-highlight t)
309 (gnus-treat-article nil)
310 (gnus-run-hooks 'gnus-article-prepare-hook
311 'gnus-outlook-display-hook)))
314 (defun gnus-article-outlook-unwrap-lines (&optional nodisplay)
315 "Unwrap lines that appear to be wrapped citation lines.
316 You can control what lines will be unwrapped by frobbing
317 `gnus-outlook-deuglify-unwrap-min' and `gnus-outlook-deuglify-unwrap-max',
318 indicating the minimum and maximum length of an unwrapped citation line. If
319 NODISPLAY is non-nil, don't redisplay the article buffer."
321 (let ((case-fold-search nil)
322 (inhibit-read-only t)
323 (cite-marks gnus-outlook-deuglify-cite-marks)
324 (no-wrap gnus-outlook-deuglify-no-wrap-chars)
325 (stop-chars gnus-outlook-deuglify-unwrap-stop-chars))
326 (gnus-with-article-buffer
328 (while (re-search-forward
330 "^\\([ \t" cite-marks "]*\\)"
331 "\\([" cite-marks "].*[^\n " stop-chars "]\\)[ \t]?\n"
332 "\\1\\([^\n " cite-marks no-wrap "]+.*\\)$")
334 (let ((len12 (- (match-end 2) (match-beginning 1)))
335 (len3 (- (match-end 3) (match-beginning 3))))
336 (when (and (> len12 gnus-outlook-deuglify-unwrap-min)
337 (< (+ len12 len3) gnus-outlook-deuglify-unwrap-max))
338 (replace-match "\\1\\2 \\3")
339 (goto-char (match-beginning 0)))))))
340 (unless nodisplay (gnus-outlook-display-article-buffer)))
342 (defun gnus-outlook-rearrange-article (attr-start)
343 "Put the text from ATTR-START to the end of buffer at the top of the article buffer."
344 (let ((inhibit-read-only t)
345 (cite-marks gnus-outlook-deuglify-cite-marks))
346 (gnus-with-article-buffer
348 ;; article does not start with attribution
349 (unless (= (point) attr-start)
350 (gnus-kill-all-overlays)
352 ;; before signature or end of buffer
353 (to (if (gnus-article-search-signature)
356 ;; handle the case where the full quote is below the
358 (if (< to attr-start)
359 (setq to (point-max)))
360 (transpose-regions cur attr-start attr-start to))))))
362 ;; John Doe <john.doe@some.domain> wrote in message
363 ;; news:a87usw8$dklsssa$2@some.news.server...
365 (defun gnus-outlook-repair-attribution-outlook ()
366 "Repair a broken attribution line (Outlook)."
367 (let ((case-fold-search nil)
368 (inhibit-read-only t)
369 (cite-marks gnus-outlook-deuglify-cite-marks))
370 (gnus-with-article-buffer
372 (when (re-search-forward
373 (concat "^\\([^" cite-marks "].+\\)"
374 "\\(" gnus-outlook-deuglify-attrib-verb-regexp "\\)"
375 "\\(.*\n?[^\n" cite-marks "].*\\)?"
376 "\\(" gnus-outlook-deuglify-attrib-end-regexp "\\)$")
378 (gnus-kill-all-overlays)
379 (replace-match "\\1\\2\\4")
380 (match-beginning 0)))))
383 ;; ----- Original Message -----
384 ;; From: "John Doe" <john.doe@some.domain>
385 ;; To: "Doe Foundation" <info@doefnd.org>
386 ;; Sent: Monday, November 19, 2001 12:13 PM
387 ;; Subject: More Doenuts
389 (defun gnus-outlook-repair-attribution-block ()
390 "Repair a big broken attribution block."
391 (let ((case-fold-search nil)
392 (inhibit-read-only t)
393 (cite-marks gnus-outlook-deuglify-cite-marks))
394 (gnus-with-article-buffer
396 (when (re-search-forward
397 (concat "^[" cite-marks " \t]*--* ?[^-]+ [^-]+ ?--*\\s *\n"
398 "[^\n:]+:[ \t]*\\([^\n]+\\)\n"
399 "\\([^\n:]+:[ \t]*[^\n]+\n\\)+")
401 (gnus-kill-all-overlays)
402 (replace-match "\\1 wrote:\n")
403 (match-beginning 0)))))
405 ;; On Wed, 16 Jan 2002 23:23:30 +0100, John Doe <john.doe@some.domain> wrote:
407 (defun gnus-outlook-repair-attribution-other ()
408 "Repair a broken attribution line (other user agents than Outlook)."
409 (let ((case-fold-search nil)
410 (inhibit-read-only t)
411 (cite-marks gnus-outlook-deuglify-cite-marks))
412 (gnus-with-article-buffer
414 (when (re-search-forward
415 (concat "^\\("gnus-outlook-deuglify-attrib-cut-regexp"\\)?"
416 "\\([^" cite-marks "].+\\)\n\\([^\n" cite-marks "].*\\)?"
417 "\\(" gnus-outlook-deuglify-attrib-verb-regexp "\\).*"
418 "\\(" gnus-outlook-deuglify-attrib-end-regexp "\\)$")
420 (gnus-kill-all-overlays)
421 (replace-match "\\4 \\5\\6\\7")
422 (match-beginning 0)))))
425 (defun gnus-article-outlook-repair-attribution (&optional nodisplay)
426 "Repair a broken attribution line.
427 If NODISPLAY is non-nil, don't redisplay the article buffer."
431 (gnus-outlook-repair-attribution-other)
432 (gnus-outlook-repair-attribution-block)
433 (gnus-outlook-repair-attribution-outlook))))
434 (unless nodisplay (gnus-outlook-display-article-buffer))
437 (defun gnus-article-outlook-rearrange-citation (&optional nodisplay)
438 "Repair broken citations.
439 If NODISPLAY is non-nil, don't redisplay the article buffer."
441 (let ((attrib-start (gnus-article-outlook-repair-attribution 'nodisplay)))
442 ;; rearrange citations if an attribution line has been recognized
444 (gnus-outlook-rearrange-article attrib-start)))
445 (unless nodisplay (gnus-outlook-display-article-buffer)))
448 (defun gnus-outlook-deuglify-article (&optional nodisplay)
449 "Full deuglify of broken Outlook (Express) articles.
450 Treat dumbquotes, unwrap lines, repair attribution and rearrange citation. If
451 NODISPLAY is non-nil, don't redisplay the article buffer."
453 ;; apply treatment of dumb quotes
454 (gnus-article-treat-dumbquotes)
455 ;; repair wrapped cited lines
456 (gnus-article-outlook-unwrap-lines 'nodisplay)
457 ;; repair attribution line and rearrange citation.
458 (gnus-article-outlook-rearrange-citation 'nodisplay)
459 (unless nodisplay (gnus-outlook-display-article-buffer)))
462 (defun gnus-article-outlook-deuglify-article ()
463 "Deuglify broken Outlook (Express) articles and redisplay."
465 (gnus-outlook-deuglify-article nil))
470 ;; coding: iso-8859-1
473 ;;; deuglify.el ends here