annotate lisp/gnus/html2text.el @ 96215:ad9760e68890

(Todo): Update.
author Dan Nicolaescu <dann@ics.uci.edu>
date Tue, 24 Jun 2008 03:24:38 +0000
parents 2763e6e1fbe8
children a9dc0e7c3f2b
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
1 ;;; html2text.el --- a simple html to plain text converter
64754
fafd692d1e40 Update years in copyright notice; nfc.
Thien-Thi Nguyen <ttn@gnuvola.org>
parents: 64085
diff changeset
2
79708
1cb31606209f Add 2008 to copyright years.
Glenn Morris <rgm@gnu.org>
parents: 79236
diff changeset
3 ;; Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
4
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
5 ;; Author: Joakim Hove <hove@phys.ntnu.no>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
6
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
7 ;; This file is part of GNU Emacs.
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
8
94662
f42ef85caf91 Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents: 93975
diff changeset
9 ;; GNU Emacs is free software: you can redistribute it and/or modify
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
10 ;; it under the terms of the GNU General Public License as published by
94662
f42ef85caf91 Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents: 93975
diff changeset
11 ;; the Free Software Foundation, either version 3 of the License, or
f42ef85caf91 Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents: 93975
diff changeset
12 ;; (at your option) any later version.
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
13
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
14 ;; GNU Emacs is distributed in the hope that it will be useful,
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
15 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
94662
f42ef85caf91 Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents: 93975
diff changeset
16 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
17 ;; GNU General Public License for more details.
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
18
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
19 ;; You should have received a copy of the GNU General Public License
94662
f42ef85caf91 Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents: 93975
diff changeset
20 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
21
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
22 ;;; Commentary:
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
23
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
24 ;; These functions provide a simple way to wash/clean html infected
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
25 ;; mails. Definitely do not work in all cases, but some improvement
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
26 ;; in readability is generally obtained. Formatting is only done in
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
27 ;; the buffer, so the next time you enter the article it will be
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
28 ;; "re-htmlized".
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
29 ;;
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
30 ;; The main function is `html2text'.
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
31
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
32 ;;; Code:
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
33
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
34 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
35 ;; <Global variables>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
36 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
37
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
38 (eval-when-compile
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
39 (require 'cl))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
40
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
41 (defvar html2text-format-single-element-list '(("hr" . html2text-clean-hr)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
42
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
43 (defvar html2text-replace-list
85712
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
44 '(("&acute;" . "`")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
45 ("&amp;" . "&")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
46 ("&apos;" . "'")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
47 ("&brvbar;" . "|")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
48 ("&cent;" . "c")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
49 ("&circ;" . "^")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
50 ("&copy;" . "(C)")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
51 ("&curren;" . "(#)")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
52 ("&deg;" . "degree")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
53 ("&divide;" . "/")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
54 ("&euro;" . "e")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
55 ("&frac12;" . "1/2")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
56 ("&gt;" . ">")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
57 ("&iquest;" . "?")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
58 ("&laquo;" . "<<")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
59 ("&ldquo" . "\"")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
60 ("&lsaquo;" . "(")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
61 ("&lsquo;" . "`")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
62 ("&lt;" . "<")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
63 ("&mdash;" . "--")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
64 ("&nbsp;" . " ")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
65 ("&ndash;" . "-")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
66 ("&permil;" . "%%")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
67 ("&plusmn;" . "+-")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
68 ("&pound;" . "£")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
69 ("&quot;" . "\"")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
70 ("&raquo;" . ">>")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
71 ("&rdquo" . "\"")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
72 ("&reg;" . "(R)")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
73 ("&rsaquo;" . ")")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
74 ("&rsquo;" . "'")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
75 ("&sect;" . "§")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
76 ("&sup1;" . "^1")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
77 ("&sup2;" . "^2")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
78 ("&sup3;" . "^3")
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
79 ("&tilde;" . "~"))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
80 "The map of entity to text.
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
81
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
82 This is an alist were each element is a dotted pair consisting of an
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
83 old string, and a replacement string. This replacement is done by the
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
84 function `html2text-substitute' which basically performs a
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
85 `replace-string' operation for every element in the list. This is
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
86 completely verbatim - without any use of REGEXP.")
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
87
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
88 (defvar html2text-remove-tag-list
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
89 '("html" "body" "p" "img" "dir" "head" "div" "br" "font" "title" "meta")
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
90 "A list of removable tags.
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
91
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
92 This is a list of tags which should be removed, without any
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
93 formatting. Note that tags in the list are presented *without*
68932
dc25f1fd56f5 (html2text-remove-tag-list): Fix typo in docstring.
Juanma Barranquero <lekktu@gmail.com>
parents: 68633
diff changeset
94 any \"<\" or \">\". All occurrences of a tag appearing in this
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
95 list are removed, irrespective of whether it is a closing or
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
96 opening tag, or if the tag has additional attributes. The
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
97 deletion is done by the function `html2text-remove-tags'.
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
98
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
99 For instance the text:
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
100
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
101 \"Here comes something <font size\"+3\" face=\"Helvetica\"> big </font>.\"
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
102
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
103 will be reduced to:
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
104
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
105 \"Here comes something big.\"
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
106
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
107 If this list contains the element \"font\".")
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
108
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
109 (defvar html2text-format-tag-list
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
110 '(("b" . html2text-clean-bold)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
111 ("strong" . html2text-clean-bold)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
112 ("u" . html2text-clean-underline)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
113 ("i" . html2text-clean-italic)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
114 ("em" . html2text-clean-italic)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
115 ("blockquote" . html2text-clean-blockquote)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
116 ("a" . html2text-clean-anchor)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
117 ("ul" . html2text-clean-ul)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
118 ("ol" . html2text-clean-ol)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
119 ("dl" . html2text-clean-dl)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
120 ("center" . html2text-clean-center))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
121 "An alist of tags and processing functions.
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
122
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
123 This is an alist where each dotted pair consists of a tag, and then
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
124 the name of a function to be called when this tag is found. The
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
125 function is called with the arguments p1, p2, p3 and p4. These are
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
126 demontrated below:
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
127
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
128 \"<b> This is bold text </b>\"
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
129 ^ ^ ^ ^
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
130 | | | |
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
131 p1 p2 p3 p4
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
132
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
133 Then the called function will typically format the text somewhat and
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
134 remove the tags.")
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
135
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
136 (defvar html2text-remove-tag-list2 '("li" "dt" "dd" "meta")
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
137 "Another list of removable tags.
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
138
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
139 This is a list of tags which are removed similarly to the list
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
140 `html2text-remove-tag-list' - but these tags are retained for the
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
141 formatting, and then moved afterward.")
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
142
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
143 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
144 ;; </Global variables>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
145 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
146
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
147 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
148 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
149
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
150 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
151 ;; <Utility functions>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
152 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
153
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
154
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
155 (defun html2text-replace-string (from-string to-string min max)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
156 "Replace FROM-STRING with TO-STRING in region from MIN to MAX."
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
157 (goto-char min)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
158 (let ((delta (- (string-width to-string) (string-width from-string)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
159 (change 0))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
160 (while (search-forward from-string max t)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
161 (replace-match to-string)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
162 (setq change (+ change delta)))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
163 change))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
164
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
165 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
166 ;; </Utility functions>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
167 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
168
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
169 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
170 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
171
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
172 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
173 ;; <Functions related to attributes> i.e. <font size=+3>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
174 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
175
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
176 (defun html2text-attr-value (list attribute)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
177 "Get value of ATTRIBUTE from LIST."
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
178 (nth 1 (assoc attribute list)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
179
57923
d7def5572cf3 Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-668
Miles Bader <miles@gnu.org>
parents: 57856
diff changeset
180 (defun html2text-get-attr (p1 p2)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
181 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
182 (re-search-forward " +[^ ]" p2 t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
183 (let* ((attr-string (buffer-substring-no-properties (1- (point)) (1- p2)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
184 (tmp-list (split-string attr-string))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
185 (attr-list)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
186 (counter 0)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
187 (prev (car tmp-list))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
188 (this (nth 1 tmp-list))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
189 (next (nth 2 tmp-list))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
190 (index 1))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
191
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
192 (cond
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
193 ;; size=3
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
194 ((string-match "[^ ]=[^ ]" prev)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
195 (let ((attr (nth 0 (split-string prev "=")))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
196 (value (nth 1 (split-string prev "="))))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
197 (setq attr-list (cons (list attr value) attr-list))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
198 ;; size= 3
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
199 ((string-match "[^ ]=\\'" prev)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
200 (setq attr-list (cons (list (substring prev 0 -1) this) attr-list))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
201
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
202 (while (< index (length tmp-list))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
203 (cond
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
204 ;; size=3
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
205 ((string-match "[^ ]=[^ ]" this)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
206 (let ((attr (nth 0 (split-string this "=")))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
207 (value (nth 1 (split-string this "="))))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
208 (setq attr-list (cons (list attr value) attr-list))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
209 ;; size =3
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
210 ((string-match "\\`=[^ ]" this)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
211 (setq attr-list (cons (list prev (substring this 1)) attr-list)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
212 ;; size= 3
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
213 ((string-match "[^ ]=\\'" this)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
214 (setq attr-list (cons (list (substring this 0 -1) next) attr-list)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
215 ;; size = 3
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
216 ((string= "=" this)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
217 (setq attr-list (cons (list prev next) attr-list))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
218 (setq index (1+ index))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
219 (setq prev this)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
220 (setq this next)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
221 (setq next (nth (1+ index) tmp-list)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
222 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
223 ;; Tags with no accompanying "=" i.e. value=nil
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
224 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
225 (setq prev (car tmp-list))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
226 (setq this (nth 1 tmp-list))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
227 (setq next (nth 2 tmp-list))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
228 (setq index 1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
229
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
230 (when (and (not (string-match "=" prev))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
231 (not (string= (substring this 0 1) "=")))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
232 (setq attr-list (cons (list prev nil) attr-list)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
233 (while (< index (1- (length tmp-list)))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
234 (when (and (not (string-match "=" this))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
235 (not (or (string= (substring next 0 1) "=")
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
236 (string= (substring prev -1) "="))))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
237 (setq attr-list (cons (list this nil) attr-list)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
238 (setq index (1+ index))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
239 (setq prev this)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
240 (setq this next)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
241 (setq next (nth (1+ index) tmp-list)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
242
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
243 (when (and this
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
244 (not (string-match "=" this))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
245 (not (string= (substring prev -1) "=")))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
246 (setq attr-list (cons (list this nil) attr-list)))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
247 ;; return - value
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
248 attr-list))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
249
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
250 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
251 ;; </Functions related to attributes>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
252 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
253
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
254 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
255 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
256
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
257 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
258 ;; <Functions to be called to format a tag-pair>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
259 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
260 (defun html2text-clean-list-items (p1 p2 list-type)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
261 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
262 (let ((item-nr 0)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
263 (items 0))
85712
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
264 (while (search-forward "<li>" p2 t)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
265 (setq items (1+ items)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
266 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
267 (while (< item-nr items)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
268 (setq item-nr (1+ item-nr))
85712
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
269 (search-forward "<li>" (point-max) t)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
270 (cond
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
271 ((string= list-type "ul") (insert " o "))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
272 ((string= list-type "ol") (insert (format " %s: " item-nr)))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
273 (t (insert " x "))))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
274
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
275 (defun html2text-clean-dtdd (p1 p2)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
276 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
277 (let ((items 0)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
278 (item-nr 0))
85712
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
279 (while (search-forward "<dt>" p2 t)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
280 (setq items (1+ items)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
281 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
282 (while (< item-nr items)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
283 (setq item-nr (1+ item-nr))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
284 (re-search-forward "<dt>\\([ ]*\\)" (point-max) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
285 (when (match-string 1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
286 (delete-region (point) (- (point) (string-width (match-string 1)))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
287 (let ((def-p1 (point))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
288 (def-p2 0))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
289 (re-search-forward "\\([ ]*\\)\\(</dt>\\|<dd>\\)" (point-max) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
290 (if (match-string 1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
291 (progn
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
292 (let* ((mw1 (string-width (match-string 1)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
293 (mw2 (string-width (match-string 2)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
294 (mw (+ mw1 mw2)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
295 (goto-char (- (point) mw))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
296 (delete-region (point) (+ (point) mw1))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
297 (setq def-p2 (point))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
298 (setq def-p2 (- (point) (string-width (match-string 2)))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
299 (put-text-property def-p1 def-p2 'face 'bold)))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
300
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
301 (defun html2text-delete-tags (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
302 (delete-region p1 p2)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
303 (delete-region (- p3 (- p2 p1)) (- p4 (- p2 p1))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
304
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
305 (defun html2text-delete-single-tag (p1 p2)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
306 (delete-region p1 p2))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
307
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
308 (defun html2text-clean-hr (p1 p2)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
309 (html2text-delete-single-tag p1 p2)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
310 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
311 (newline 1)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
312 (insert (make-string fill-column ?-)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
313
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
314 (defun html2text-clean-ul (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
315 (html2text-delete-tags p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
316 (html2text-clean-list-items p1 (- p3 (- p1 p2)) "ul"))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
317
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
318 (defun html2text-clean-ol (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
319 (html2text-delete-tags p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
320 (html2text-clean-list-items p1 (- p3 (- p1 p2)) "ol"))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
321
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
322 (defun html2text-clean-dl (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
323 (html2text-delete-tags p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
324 (html2text-clean-dtdd p1 (- p3 (- p1 p2))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
325
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
326 (defun html2text-clean-center (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
327 (html2text-delete-tags p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
328 (center-region p1 (- p3 (- p2 p1))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
329
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
330 (defun html2text-clean-bold (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
331 (put-text-property p2 p3 'face 'bold)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
332 (html2text-delete-tags p1 p2 p3 p4))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
333
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
334 (defun html2text-clean-title (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
335 (put-text-property p2 p3 'face 'bold)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
336 (html2text-delete-tags p1 p2 p3 p4))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
337
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
338 (defun html2text-clean-underline (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
339 (put-text-property p2 p3 'face 'underline)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
340 (html2text-delete-tags p1 p2 p3 p4))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
341
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
342 (defun html2text-clean-italic (p1 p2 p3 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
343 (put-text-property p2 p3 'face 'italic)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
344 (html2text-delete-tags p1 p2 p3 p4))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
345
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
346 (defun html2text-clean-font (p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
347 (html2text-delete-tags p1 p2 p3 p4))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
348
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
349 (defun html2text-clean-blockquote (p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
350 (html2text-delete-tags p1 p2 p3 p4))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
351
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
352 (defun html2text-clean-anchor (p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
353 ;; If someone can explain how to make the URL clickable I will surely
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
354 ;; improve upon this.
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
355 ;; Maybe `goto-addr.el' can be used here.
57923
d7def5572cf3 Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-668
Miles Bader <miles@gnu.org>
parents: 57856
diff changeset
356 (let* ((attr-list (html2text-get-attr p1 p2))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
357 (href (html2text-attr-value attr-list "href")))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
358 (delete-region p1 p4)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
359 (when href
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
360 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
361 (insert (substring href 1 -1 ))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
362 (put-text-property p1 (point) 'face 'bold))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
363
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
364 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
365 ;; </Functions to be called to format a tag-pair>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
366 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
367
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
368 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
369 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
370
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
371 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
372 ;; <Functions to be called to fix up paragraphs>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
373 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
374
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
375 (defun html2text-fix-paragraph (p1 p2)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
376 (goto-char p1)
85712
a3c27999decb Update Gnus to No Gnus 0.7 from the Gnus CVS trunk
Miles Bader <miles@gnu.org>
parents: 79236
diff changeset
377 (let ((refill-start)
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
378 (refill-stop))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
379 (when (re-search-forward "<br>$" p2 t)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
380 (goto-char p1)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
381 (when (re-search-forward ".+[^<][^b][^r][^>]$" p2 t)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
382 (beginning-of-line)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
383 (setq refill-start (point))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
384 (goto-char p2)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
385 (re-search-backward ".+[^<][^b][^r][^>]$" refill-start t)
79236
9e2e45e4da24 Merge from gnus--rel--5.10
Miles Bader <miles@gnu.org>
parents: 78224
diff changeset
386 (forward-line 1)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
387 (end-of-line)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
388 ;; refill-stop should ideally be adjusted to
95238
2763e6e1fbe8 Spelling fix.
Glenn Morris <rgm@gnu.org>
parents: 94662
diff changeset
389 ;; accommodate the "<br>" strings which are removed
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
390 ;; between refill-start and refill-stop. Can simply
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
391 ;; be returned from my-replace-string
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
392 (setq refill-stop (+ (point)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
393 (html2text-replace-string
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
394 "<br>" ""
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
395 refill-start (point))))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
396 ;; (message "Point = %s refill-stop = %s" (point) refill-stop)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
397 ;; (sleep-for 4)
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
398 (fill-region refill-start refill-stop))))
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
399 (html2text-replace-string "<br>" "" p1 p2))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
400
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
401 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
402 ;; This one is interactive ...
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
403 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
404 (defun html2text-fix-paragraphs ()
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
405 "This _tries_ to fix up the paragraphs - this is done in quite a ad-hook
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
406 fashion, quite close to pure guess-work. It does work in some cases though."
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
407 (interactive)
57723
6a65cb24e1be Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-640
Miles Bader <miles@gnu.org>
parents: 57243
diff changeset
408 (goto-char (point-min))
62943
11d53dd5abd9 Revision: miles@gnu.org--gnu-2005/emacs--cvs-trunk--0--patch-345
Miles Bader <miles@gnu.org>
parents: 57923
diff changeset
409 (while (re-search-forward "^<br>$" nil t)
11d53dd5abd9 Revision: miles@gnu.org--gnu-2005/emacs--cvs-trunk--0--patch-345
Miles Bader <miles@gnu.org>
parents: 57923
diff changeset
410 (delete-region (match-beginning 0) (match-end 0)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
411 ;; Removing lonely <br> on a single line, if they are left intact we
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
412 ;; dont have any paragraphs at all.
57723
6a65cb24e1be Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-640
Miles Bader <miles@gnu.org>
parents: 57243
diff changeset
413 (goto-char (point-min))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
414 (while (not (eobp))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
415 (let ((p1 (point)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
416 (forward-paragraph 1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
417 ;;(message "Kaller fix med p1=%s p2=%s " p1 (1- (point))) (sleep-for 5)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
418 (html2text-fix-paragraph p1 (1- (point)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
419 (goto-char p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
420 (when (not (eobp))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
421 (forward-paragraph 1)))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
422
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
423 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
424 ;; </Functions to be called to fix up paragraphs>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
425 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
426
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
427 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
428 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
429
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
430 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
431 ;; <Interactive functions>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
432 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
433
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
434 (defun html2text-remove-tags (tag-list)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
435 "Removes the tags listed in the list `html2text-remove-tag-list'.
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
436 See the documentation for that variable."
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
437 (interactive)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
438 (dolist (tag tag-list)
57723
6a65cb24e1be Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-640
Miles Bader <miles@gnu.org>
parents: 57243
diff changeset
439 (goto-char (point-min))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
440 (while (re-search-forward (format "\\(</?%s[^>]*>\\)" tag) (point-max) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
441 (delete-region (match-beginning 0) (match-end 0)))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
442
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
443 (defun html2text-format-tags ()
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
444 "See the variable `html2text-format-tag-list' for documentation."
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
445 (interactive)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
446 (dolist (tag-and-function html2text-format-tag-list)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
447 (let ((tag (car tag-and-function))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
448 (function (cdr tag-and-function)))
57723
6a65cb24e1be Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-640
Miles Bader <miles@gnu.org>
parents: 57243
diff changeset
449 (goto-char (point-min))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
450 (while (re-search-forward (format "\\(<%s\\( [^>]*\\)?>\\)" tag)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
451 (point-max) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
452 (let ((p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
453 (p2 (point))
57923
d7def5572cf3 Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-668
Miles Bader <miles@gnu.org>
parents: 57856
diff changeset
454 (p3) (p4))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
455 (search-backward "<" (point-min) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
456 (setq p1 (point))
73627
e7c14a8e6688 Merge from gnus--rel--5.10
Miles Bader <miles@gnu.org>
parents: 68932
diff changeset
457 (unless (search-forward (format "</%s>" tag) (point-max) t)
e7c14a8e6688 Merge from gnus--rel--5.10
Miles Bader <miles@gnu.org>
parents: 68932
diff changeset
458 (goto-char p2)
e7c14a8e6688 Merge from gnus--rel--5.10
Miles Bader <miles@gnu.org>
parents: 68932
diff changeset
459 (insert (format "</%s>" tag)))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
460 (setq p4 (point))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
461 (search-backward "</" (point-min) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
462 (setq p3 (point))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
463 (funcall function p1 p2 p3 p4)
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
464 (goto-char p1))))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
465
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
466 (defun html2text-substitute ()
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
467 "See the variable `html2text-replace-list' for documentation."
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
468 (interactive)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
469 (dolist (e html2text-replace-list)
57723
6a65cb24e1be Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-640
Miles Bader <miles@gnu.org>
parents: 57243
diff changeset
470 (goto-char (point-min))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
471 (let ((old-string (car e))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
472 (new-string (cdr e)))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
473 (html2text-replace-string old-string new-string (point-min) (point-max)))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
474
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
475 (defun html2text-format-single-elements ()
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
476 (interactive)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
477 (dolist (tag-and-function html2text-format-single-element-list)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
478 (let ((tag (car tag-and-function))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
479 (function (cdr tag-and-function)))
57723
6a65cb24e1be Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-640
Miles Bader <miles@gnu.org>
parents: 57243
diff changeset
480 (goto-char (point-min))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
481 (while (re-search-forward (format "\\(<%s\\( [^>]*\\)?>\\)" tag)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
482 (point-max) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
483 (let ((p1)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
484 (p2 (point)))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
485 (search-backward "<" (point-min) t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
486 (setq p1 (point))
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
487 (funcall function p1 p2))))))
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
488
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
489 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
490 ;; Main function
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
491 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
492
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
493 ;;;###autoload
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
494 (defun html2text ()
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
495 "Convert HTML to plain text in the current buffer."
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
496 (interactive)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
497 (save-excursion
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
498 (let ((case-fold-search t)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
499 (buffer-read-only))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
500 (html2text-remove-tags html2text-remove-tag-list)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
501 (html2text-format-tags)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
502 (html2text-remove-tags html2text-remove-tag-list2)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
503 (html2text-substitute)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
504 (html2text-format-single-elements)
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
505 (html2text-fix-paragraphs))))
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
506
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
507 ;;
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
508 ;; </Interactive functions>
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
509 ;;
57856
df80d19d7a2e Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-660
Miles Bader <miles@gnu.org>
parents: 57723
diff changeset
510 (provide 'html2text)
93975
1e3a407766b9 Fix up comment convention on the arch-tag lines.
Stefan Monnier <monnier@iro.umontreal.ca>
parents: 87649
diff changeset
511 ;; arch-tag: e9e57b79-35d4-4de1-a647-e7e01fe56d1e
56927
55fd4f77387a Revision: miles@gnu.org--gnu-2004/emacs--cvs-trunk--0--patch-523
Miles Bader <miles@gnu.org>
parents:
diff changeset
512 ;;; html2text.el ends here