1 ;;; html2text.el --- a simple html to plain text converter
3 ;; Copyright (C) 2002 Free Software Foundation, Inc.
5 ;; Author: Joakim Hove <hove@phys.ntnu.no>
7 ;; This file is part of GNU Emacs.
9 ;; GNU Emacs is free software; you can redistribute it and/or modify
10 ;; it under the terms of the GNU General Public License as published by
11 ;; the Free Software Foundation; either version 2, or (at your option)
14 ;; GNU Emacs is distributed in the hope that it will be useful,
15 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
16 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 ;; GNU General Public License for more details.
19 ;; You should have received a copy of the GNU General Public License
20 ;; along with GNU Emacs; see the file COPYING. If not, write to the
21 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
22 ;; Boston, MA 02111-1307, USA.
26 ;; These functions provide a simple way to wash/clean html infected
27 ;; mails. Definitely do not work in all cases, but some improvement
28 ;; in readability is generally obtained. Formatting is only done in
29 ;; the buffer, so the next time you enter the article it will be
32 ;; The main function is "html2text"
43 (defvar html2text-format-single-element-list '(("hr" . html2text-clean-hr)))
45 (defvar html2text-replace-list
46 '((" " . " ") (">" . ">") ("<" . "<") (""" . "\""))
47 "The map of entity to text.
49 This is an alist were each element is a dotted pair consisting of an
50 old string, and a replacement string. This replacement is done by the
51 function \"html2text-substitute\" which basically performs a
52 replace-string operation for every element in the list. This is
53 completely verbatim - without any use of REGEXP.")
55 (defvar html2text-remove-tag-list
56 '("html" "body" "p" "img" "dir" "head" "div" "br" "font" "title" "meta")
57 "A list of removable tags.
59 This is a list of tags which should be removed, without any
60 formatting. Observe that if you the tags in the list are presented
61 *without* any \"<\" or \">\". All occurences of a tag appearing in
62 this list are removed, irrespective of whether it is a closing or
63 opening tag, or if the tag has additional attributes. The actual
64 deletion is done by the function \"html2text-remove-tags\".
66 For instance the text:
68 \"Here comes something <font size\"+3\" face=\"Helvetica\"> big </font>.\"
72 \"Here comes something big.\"
74 If this list contains the element \"font\".")
76 (defvar html2text-format-tag-list
77 '(("b" . html2text-clean-bold)
78 ("u" . html2text-clean-underline)
79 ("i" . html2text-clean-italic)
80 ("blockquote" . html2text-clean-blockquote)
81 ("a" . html2text-clean-anchor)
82 ("ul" . html2text-clean-ul)
83 ("ol" . html2text-clean-ol)
84 ("dl" . html2text-clean-dl)
85 ("center" . html2text-clean-center))
86 "An alist of tags and processing functions.
88 This is an alist where each dotted pair consists of a tag, and then
89 the name of a function to be called when this tag is found. The
90 function is called with the arguments p1, p2, p3 and p4. These are
93 \"<b> This is bold text </b>\"
98 Then the called function will typically format the text somewhat and
101 (defvar html2text-remove-tag-list2 '("li" "dt" "dd" "meta")
102 "Another list of removable tags.
104 This is a list of tags which are removed similarly to the list
105 `html2text-remove-tag-list' - but these tags are retained for the
106 formatting, and then moved afterward.")
109 ;; </Global variables>
112 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
113 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
116 ;; <Utility functions>
119 (defun html2text-buffer-head ()
120 (if (string= mode-name "Article")
121 (beginning-of-buffer)
122 (beginning-of-buffer)
126 (defun html2text-replace-string (from-string to-string p1 p2)
128 (let ((delta (- (string-width to-string) (string-width from-string)))
130 (while (search-forward from-string p2 t)
131 (replace-match to-string)
132 (setq change (+ change delta))
139 ;; </Utility functions>
142 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
143 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
146 ;; <Functions related to attributes> i.e. <font size=+3>
149 (defun html2text-attr-value (attr-list attr)
150 (nth 1 (assoc attr attr-list))
153 (defun html2text-get-attr (p1 p2 tag)
155 (re-search-forward " +[^ ]" p2 t)
156 (let* ((attr-string (buffer-substring-no-properties (1- (point)) (1- p2)))
157 (tmp-list (split-string attr-string))
160 (prev (car tmp-list))
161 (this (nth 1 tmp-list))
162 (next (nth 2 tmp-list))
167 ((string-match "[^ ]=[^ ]" prev)
168 (let ((attr (nth 0 (split-string prev "=")))
169 (value (nth 1 (split-string prev "="))))
170 (setq attr-list (cons (list attr value) attr-list))
174 ((string-match "[^ ]=\\'" prev)
175 (setq attr-list (cons (list (substring prev 0 -1) this) attr-list))
179 (while (< index (length tmp-list))
182 ((string-match "[^ ]=[^ ]" this)
183 (let ((attr (nth 0 (split-string this "=")))
184 (value (nth 1 (split-string this "="))))
185 (setq attr-list (cons (list attr value) attr-list))
189 ((string-match "\\`=[^ ]" this)
190 (setq attr-list (cons (list prev (substring this 1)) attr-list)))
193 ((string-match "[^ ]=\\'" this)
194 (setq attr-list (cons (list (substring this 0 -1) next) attr-list))
199 (setq attr-list (cons (list prev next) attr-list))
202 (setq index (1+ index))
205 (setq next (nth (1+ index) tmp-list))
209 ;; Tags with no accompanying "=" i.e. value=nil
211 (setq prev (car tmp-list))
212 (setq this (nth 1 tmp-list))
213 (setq next (nth 2 tmp-list))
216 (if (not (string-match "=" prev))
218 (if (not (string= (substring this 0 1) "="))
219 (setq attr-list (cons (list prev nil) attr-list))
224 (while (< index (1- (length tmp-list)))
225 (if (not (string-match "=" this))
226 (if (not (or (string= (substring next 0 1) "=")
227 (string= (substring prev -1) "=")))
228 (setq attr-list (cons (list this nil) attr-list))
231 (setq index (1+ index))
234 (setq next (nth (1+ index) tmp-list))
239 (if (not (string-match "=" this))
241 (if (not (string= (substring prev -1) "="))
242 (setq attr-list (cons (list this nil) attr-list))
248 attr-list ;; return - value
253 ;; </Functions related to attributes>
256 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
257 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
260 ;; <Functions to be called to format a tag-pair>
262 (defun html2text-clean-list-items (p1 p2 list-type)
266 (while (re-search-forward "<li>" p2 t)
267 (setq items (1+ items)))
269 (while (< item-nr items)
270 (setq item-nr (1+ item-nr))
271 (re-search-forward "<li>" (point-max) t)
273 ((string= list-type "ul") (insert " o "))
274 ((string= list-type "ol") (insert (format " %s: " item-nr)))
280 (defun html2text-clean-dtdd (p1 p2)
284 (while (re-search-forward "<dt>" p2 t)
285 (setq items (1+ items)))
287 (while (< item-nr items)
288 (setq item-nr (1+ item-nr))
289 (re-search-forward "<dt>\\([ ]*\\)" (point-max) t)
291 (kill-region (point) (- (point) (string-width (match-string 1))))
293 (let ((def-p1 (point))
295 (re-search-forward "\\([ ]*\\)\\(</dt>\\|<dd>\\)" (point-max) t)
298 (let* ((mw1 (string-width (match-string 1)))
299 (mw2 (string-width (match-string 2)))
301 (goto-char (- (point) mw))
302 (kill-region (point) (+ (point) mw1))
303 (setq def-p2 (point))
306 (setq def-p2 (- (point) (string-width (match-string 2)))))
307 (put-text-property def-p1 def-p2 'face 'bold)
313 (defun html2text-delete-tags (p1 p2 p3 p4)
315 (kill-region (- p3 (- p2 p1)) (- p4 (- p2 p1)))
318 (defun html2text-delete-single-tag (p1 p2)
322 (defun html2text-clean-hr (p1 p2)
323 (html2text-delete-single-tag p1 p2)
326 (insert (make-string fill-column ?-))
329 (defun html2text-clean-ul (p1 p2 p3 p4)
330 (html2text-delete-tags p1 p2 p3 p4)
331 (html2text-clean-list-items p1 (- p3 (- p1 p2)) "ul")
334 (defun html2text-clean-ol (p1 p2 p3 p4)
335 (html2text-delete-tags p1 p2 p3 p4)
336 (html2text-clean-list-items p1 (- p3 (- p1 p2)) "ol")
339 (defun html2text-clean-dl (p1 p2 p3 p4)
340 (html2text-delete-tags p1 p2 p3 p4)
341 (html2text-clean-dtdd p1 (- p3 (- p1 p2)))
344 (defun html2text-clean-center (p1 p2 p3 p4)
345 (html2text-delete-tags p1 p2 p3 p4)
346 (center-region p1 (- p3 (- p2 p1)))
349 (defun html2text-clean-bold (p1 p2 p3 p4)
350 (put-text-property p2 p3 'face 'bold)
351 (html2text-delete-tags p1 p2 p3 p4)
354 (defun html2text-clean-title (p1 p2 p3 p4)
355 (put-text-property p2 p3 'face 'bold)
356 (html2text-delete-tags p1 p2 p3 p4)
359 (defun html2text-clean-underline (p1 p2 p3 p4)
360 (put-text-property p2 p3 'face 'underline)
361 (html2text-delete-tags p1 p2 p3 p4)
364 (defun html2text-clean-italic (p1 p2 p3 p4)
365 (put-text-property p2 p3 'face 'italic)
366 (html2text-delete-tags p1 p2 p3 p4)
369 (defun html2text-clean-font (p1 p2 p3 p4)
370 (html2text-delete-tags p1 p2 p3 p4)
373 (defun html2text-clean-blockquote (p1 p2 p3 p4)
374 (html2text-delete-tags p1 p2 p3 p4)
377 (defun html2text-clean-anchor (p1 p2 p3 p4)
378 ;; If someone can explain how to make the URL clickable I will
379 ;; surely improve upon this.
380 (let* ((attr-list (html2text-get-attr p1 p2 "a"))
381 (href (html2text-attr-value attr-list "href")))
385 (insert (substring href 1 -1 ))
386 (put-text-property p1 (point) 'face 'bold))))
389 ;; </Functions to be called to format a tag-pair>
392 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
393 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
396 ;; <Functions to be called to fix up paragraphs>
399 (defun html2text-fix-paragraph (p1 p2)
404 (if (re-search-forward "<br>$" p2 t)
410 (if (re-search-forward ".+[^<][^b][^r][^>]$" p2 t)
413 (setq refill-start (point))
415 (re-search-backward ".+[^<][^b][^r][^>]$" refill-start t)
418 ;; refill-stop should ideally be adjusted to
419 ;; accomodate the "<br>" strings which are removed
420 ;; between refill-start and refill-stop. Can simply
421 ;; be returned from my-replace-string
422 (setq refill-stop (+ (point)
423 (html2text-replace-string
425 refill-start (point))))
426 ;; (message "Point = %s refill-stop = %s" (point) refill-stop)
428 (fill-region refill-start refill-stop)
434 (html2text-replace-string "<br>" "" p1 p2)
438 ;; This one is interactive ...
440 (defun html2text-fix-paragraphs ()
441 "This _tries_ to fix up the paragraphs - this is done in quite a ad-hook
442 fashion, quite close to pure guess-work. It does work in some cases though."
444 (html2text-buffer-head)
445 (replace-regexp "^<br>$" "")
446 ;; Removing lonely <br> on a single line, if they are left intact we
447 ;; dont have any paragraphs at all.
448 (html2text-buffer-head)
449 (while (< (point) (point-max))
451 (forward-paragraph 1)
452 ;;(message "Kaller fix med p1=%s p2=%s " p1 (1- (point))) (sleep-for 5)
453 (html2text-fix-paragraph p1 (1- (point)))
455 (if (< (point) (point-max))
456 (forward-paragraph 1))
462 ;; </Functions to be called to fix up paragraphs>
465 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
466 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
469 ;; <Interactive functions>
472 (defun html2text-remove-tags (tag-list)
473 "Removes the tags listed in the list \"html2text-remove-tag-list\".
474 See the documentation for that variable."
476 (dolist (tag tag-list)
477 (html2text-buffer-head)
478 (while (re-search-forward (format "\\(</?%s[^>]*>\\)" tag) (point-max) t)
480 (search-backward "<")
481 (kill-region (point) p1)
487 (defun html2text-format-tags ()
488 "See the variable \"html2text-format-tag-list\" for documentation"
490 (dolist (tag-and-function html2text-format-tag-list)
491 (let ((tag (car tag-and-function))
492 (function (cdr tag-and-function)))
493 (html2text-buffer-head)
494 (while (re-search-forward (format "\\(<%s\\( [^>]*\\)?>\\)" tag)
499 (attr (match-string 1)))
500 (search-backward "<" (point-min) t)
502 (re-search-forward (format "</%s>" tag) (point-max) t)
504 (search-backward "</" (point-min) t)
506 (funcall function p1 p2 p3 p4)
514 (defun html2text-substitute ()
515 "See the variable \"html2text-replace-list\" for documentation"
517 (dolist (e html2text-replace-list)
518 (html2text-buffer-head)
519 (let ((old-string (car e))
520 (new-string (cdr e)))
521 (html2text-replace-string old-string new-string (point-min) (point-max))
526 (defun html2text-format-single-elements ()
529 (dolist (tag-and-function html2text-format-single-element-list)
530 (let ((tag (car tag-and-function))
531 (function (cdr tag-and-function)))
532 (html2text-buffer-head)
533 (while (re-search-forward (format "\\(<%s\\( [^>]*\\)?>\\)" tag)
537 (search-backward "<" (point-min) t)
539 (funcall function p1 p2)
552 "Convert HTML to plain text in the current buffer."
555 (let ((case-fold-search t)
557 (html2text-remove-tags html2text-remove-tag-list)
558 (html2text-format-tags)
559 (html2text-remove-tags html2text-remove-tag-list2)
560 (html2text-substitute)
561 (html2text-format-single-elements)
562 (html2text-fix-paragraphs))))
565 ;; </Interactive functions>
568 ;;; html2text.el ends here