1 # ********************************************************************
3 # File : $Source: /cvsroot/ijbswa/current/re_filterfile,v $
5 # Purpose : Rules to process the content of web pages
7 # Copyright : Written by and Copyright (C) 2001 the SourceForge
8 # IJBSWA team. http://ijbswa.sourceforge.net
10 # This program is free software; you can redistribute it
11 # and/or modify it under the terms of the GNU General
12 # Public License as published by the Free Software
13 # Foundation; either version 2 of the License, or (at
14 # your option) any later version.
16 # This program is distributed in the hope that it will
17 # be useful, but WITHOUT ANY WARRANTY; without even the
18 # implied warranty of MERCHANTABILITY or FITNESS FOR A
19 # PARTICULAR PURPOSE. See the GNU General Public
20 # License for more details.
22 # The GNU General Public License should be included with
23 # this file. If not, you can view it at
24 # http://www.gnu.org/copyleft/gpl.html
25 # or write to the Free Software Foundation, Inc., 59
26 # Temple Place - Suite 330, Boston, MA 02111-1307, USA.
29 # $Log: re_filterfile,v $
30 # Revision 1.7 2001/06/19 14:21:56 oes
31 # Fixed microsuck line
33 # Revision 1.6 2001/06/09 14:01:57 swa
34 # header. cosmetics. default: no messing ala microsuck.
38 # ********************************************************************/
40 # Syntax: One Perl-Style substitution per line.
41 # For Details see the perlre, perlop and pcre manpages.
42 # Note that you are free to choose the delimter as you see fit.
44 # Note: In addidion to the Perl options egimosx, the following nonstandard
45 # options are supported:
47 # 'U' turns the default to ungreedy matching. Add ? to quantifiers to
48 # switch back to greedy.
49 # 'T' (trivial) prevents parsing for backreferences in the substitute.
50 # Use if you want to include text like '$&' in your substitute without
54 # ********************************************************************/
56 # Kill OnUnload popups. Yummy.
57 # check it out on http://www.zdnet.com/zdsubs/yahoo/tree/yfs.html
59 # ********************************************************************/
60 s/(<body .*?)onunload(.*?>)/$1never$2/i
62 # ********************************************************************/
64 # Kill refresh tags. I like to refresh myself. Manually.
65 # check it out on http://www.airport-cgn.de/ and go to the arrivals page.
67 # ********************************************************************/
68 s/<meta[^>]*http-equiv[^>]*refresh.*URL=([^>]*?)"?>/<link rev="x-refresh" href=$1>/i
69 s/<meta[^>]*http-equiv="?page-enter"?[^>]*content=[^>]*>/<!--no page enter for me-->/i
71 # ********************************************************************/
73 # If I allow popups, I want them to be resizeable and have a location
74 # and status bar: check it out on http://www.airport-cgn.de/ and go to
77 # ********************************************************************/
78 # s/resizable="?(no|0)"?/resizable=1/ig s/noresize/yesresize/ig
79 # s/location="?(no|0)"?/location=1/ig s/status="?(no|0)"?/status=1/ig
80 # s/scrolling="?(no|0|Auto)"?/scrolling=1/ig
81 # s/menubar="?(no|0)"?/menubar=1/ig #s/framespacing="?(no|0)"?//ig
82 # #s/margin(height|width)=[0-9]*//gi
84 # ********************************************************************/
86 # Remove frameborder=0 and border=0 from framesets
88 # ********************************************************************/
89 s/(<frameset[^>]+?)border=['"]?(no|0)['"]?/$1/ig
90 s/(<frameset[^>]+?)frameborder=['"]?(no|0)['"]?/$1/ig
92 # ********************************************************************/
94 # The status bar is for displaying link targets, not pointless buzzwords.
95 # Again, check it out on http://www.airport-cgn.de/
97 # ********************************************************************/
100 # ********************************************************************/
102 # Kill *all* popups a la popup.c. (But for *all* sites, so I wouldn't do that.)
104 # JavaScript: s/window\.open\(/who_wants_this_to.open(/ig
105 # HTML : s/target=['"]?_blank['"]?/target_who/g
107 # Kill banners by size:
108 # (Sizes from http://www.iab.net/iab_banner_standards/bannersizes.html)
109 # Note: Use http://i.j.b/send-banner?type=trans for a transparent 1x1 gif
110 # ********************************************************************/
111 s|<img\s+[^>]*?(width=['"]?468\D)[^>]*(height=['"]?60)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
112 s|<img\s+[^>]*?(width=['"]?234\D)[^>]*(height=['"]?60)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
113 s|<img\s+[^>]*?(width=['"]?88\D)[^>]*(height=['"]?31)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
114 s|<img\s+[^>]*?(width=['"]?120\D)[^>]*(height=['"]?90)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
115 s|<img\s+[^>]*?(width=['"]?120\D)[^>]*(height=['"]?600)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
116 s|<img\s+[^>]*?(width=['"]?120\D)[^>]*(height=['"]?60)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
117 s|<img\s+[^>]*?(width=['"]?160\D)[^>]*(height=['"]?600)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
118 s|<img\s+[^>]*?(width=['"]?125\D)[^>]*(height=['"]?125)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
119 s|<img\s+[^>]*?(width=['"]?120\D)[^>]*(height=['"]?240)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
120 s|<img\s+[^>]*?(width=['"]?180\D)[^>]*(height=['"]?150)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
121 s|<img\s+[^>]*?(width=['"]?300\D)[^>]*(height=['"]?250)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
122 s|<img\s+[^>]*?(width=['"]?250\D)[^>]*(height=['"]?250)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
123 s|<img\s+[^>]*?(width=['"]?240\D)[^>]*(height=['"]?400)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
124 s|<img\s+[^>]*?(width=['"]?336\D)[^>]*(height=['"]?280)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
126 # Where is that from?
127 s|<img\s+[^>]*?(width=['"]?200\D)[^>]*(height=['"]?50)[^>]*?>|<img src=http://i.j.b/send-banner $1 $2>|sig
129 # ********************************************************************/
133 # ********************************************************************/
134 s/<img\s+[^>]*?(width|height)\s+=\s+['"]?1\D[^>]*?(width|height)\s+=\s+['"]?1\D[^>]*?>/<!-- Squished WebBug -->/sig
136 # ********************************************************************/
140 # ********************************************************************/
141 s/microsoft(?!.com)/MicroSuck/ig
143 # ********************************************************************/
145 # Crude parental filtering? (Use along with a suitable blocklist).
146 # Shows how to deny access to whole page based on a keyword.
148 # (Note: Middlesex, Sussex and Essex are counties in the UK, not rude words)
149 # (Note #2: Is 'sex' a rude word?!)
151 #s%^.*(?<!middle)(?<!sus)(?<!es)sex.*$%<html><head><title>Blocked</title></head><body><h3>Blocked due to possible adult content. Please see <a href="http://dmoz.org/Kids_and_Teens/">this site</a>.</h3></body></html>%is
152 #s+^.*warez.*$+<html><head><title>No Warez</title></head><body><h3>You're not searching for illegal stuff, are you?</h3></body></html>+is
154 # ********************************************************************/
156 # http://www.farscapezone.com/wwwboard/messages/1451.html
158 # ********************************************************************/
159 s/(\w+) was tired/<b>$1 needed more coffee<\/b>/ig