From f7b2cfcf50f12ad6ce7358f74e8db9d6083aa284 Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Fri, 16 Jan 2009 20:46:55 -0500 Subject: checkcontent: New hook, can be used to implement arbitrary content filters, including spam filters. --- doc/plugins/write.mdwn | 22 +++++++++++++++++++++- 1 file changed, 21 insertions(+), 1 deletion(-) (limited to 'doc/plugins') diff --git a/doc/plugins/write.mdwn b/doc/plugins/write.mdwn index 405876d58..99eea3d16 100644 --- a/doc/plugins/write.mdwn +++ b/doc/plugins/write.mdwn @@ -303,7 +303,7 @@ can check if the session object has a "name" parameter set. ### canedit - hook(type => "canedit", id => "foo", call => \&pagelocked); + hook(type => "canedit", id => "foo", call => \&canedit); This hook can be used to implement arbitrary access methods to control when a page can be edited using the web interface (commits from revision control @@ -321,6 +321,26 @@ This hook should avoid directly redirecting the user to a signin page, since it's sometimes used to test to see which pages in a set of pages a user can edit. +### checkcontent + + hook(type => "checkcontent", id => "foo", call => \&checkcontent); + +This hook is called to check the content a user has entered on a page, +before it is saved, and decide if it should be allowed. + +It is passed named parameters: `content`, `page`, `cgi`, and `session`. If +the content the user has entered is a comment, it may also be passed some +additional parameters: `author`, `url`, and `subject`. The `subject` +parameter may also be filled with the user's comment about the change. + +Note: When the user edits an existing wiki page, the passed `content` will +include only the lines that they added to the page, or modified. + +The hook should return `undef` on success. If the content is disallowed, it +should return a message stating what the problem is, or a function +that can be run to perform whatever action is necessary to allow the user +to post the content. + ### editcontent hook(type => "editcontent", id => "foo", call => \&editcontent); -- cgit v1.2.3 From 7ee92cab4095fb22d77a0b0100a472c6901eb519 Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Fri, 16 Jan 2009 22:39:11 -0500 Subject: blogspam: New plugin, adding spam filtering for page editing / comment posting using the BlogSpam.net API. --- IkiWiki/Plugin/blogspam.pm | 111 +++++++++++++++++++++++++++++++++++++ debian/changelog | 2 + doc/plugins/blogspam.mdwn | 23 ++++++++ doc/todo/anti-spam_protection.mdwn | 11 ++++ t/syntax.t | 4 +- 5 files changed, 149 insertions(+), 2 deletions(-) create mode 100644 IkiWiki/Plugin/blogspam.pm create mode 100644 doc/plugins/blogspam.mdwn (limited to 'doc/plugins') diff --git a/IkiWiki/Plugin/blogspam.pm b/IkiWiki/Plugin/blogspam.pm new file mode 100644 index 000000000..6e68a9856 --- /dev/null +++ b/IkiWiki/Plugin/blogspam.pm @@ -0,0 +1,111 @@ +#!/usr/bin/perl +package IkiWiki::Plugin::blogspam; + +use warnings; +use strict; +use IkiWiki 3.00; +require RPC::XML; +require RPC::XML::Client; + +my $defaulturl='http://test.blogspam.net:8888/'; + +sub import { + hook(type => "getsetup", id => "blogspam", call => \&getsetup); + hook(type => "checkcontent", id => "blogspam", call => \&checkcontent); +} + +sub getsetup () { + return + plugin => { + safe => 1, + rebuild => 0, + }, + blogspam_pagespec => { + type => 'pagespec', + example => 'postcomment(*)', + description => 'PageSpec of pages to check for spam', + link => 'ikiwiki/PageSpec', + safe => 1, + rebuild => 0, + }, + blogspam_options => { + type => "string", + example => "blacklist=1.2.3.4,blacklist=8.7.6.5,max-links=10", + description => "options to send to blogspam server", + link => "http://blogspam.net/api/testComment.html#options", + safe => 1, + rebuild => 0, + }, + blogspam_server => { + type => "string", + default => $defaulturl, + description => "blogspam server XML-RPC url", + safe => 1, + rebuild => 0, + }, +} + +sub checkcontent (@) { + my %params=@_; + + if (exists $config{blogspam_pagespec}) { + return undef + if ! pagespec_match($params{page}, $config{blogspam_pagespec}, + location => $params{page}); + } + + my $url=$defaulturl; + $url = $params{blogspam_server} if exists $params{blogspam_server}; + my $client = RPC::XML::Client->new($url); + + my @options = split(",", $params{blogspam_options}) + if exists $params{blogspam_options}; + + # Allow short comments and whitespace-only edits, unless the user + # has overridden min-words themselves. + push @options, "min-words=0" + unless grep /^min-words=/i, @options; + # Wiki pages can have a lot of urls, unless the user specifically + # wants to limit them. + push @options, "exclude=lotsaurls" + unless grep /^max-links/i, @options; + # Unless the user specified a size check, disable such checking. + push @options, "exclude=size" + unless grep /^(?:max|min)-size/i, @options; + # This test has absurd false positives on words like "alpha" + # and "buy". + push @options, "exclude=stopwords"; + + # blogspam API does not have a field for author url, so put it in + # the content to be checked. + if (exists $params{url}) { + $params{content}.="\n".$params{url}; + } + + my $res = $client->send_request('testComment', { + ip => $ENV{REMOTE_ADDR}, + comment => $params{content}, + subject => defined $params{subject} ? $params{subject} : "", + name => defined $params{author} ? $params{author} : "", + options => join(",", @options), + site => $config{url}, + version => "ikiwiki ".$IkiWiki::version, + }); + + if (! ref $res || ! defined $res->value) { + debug("failed to get response from blogspam server ($url)"); + return undef; + } + elsif ($res->value =~ /^SPAM:(.*)/) { + return gettext("Sorry, but that looks like spam to blogspam: ").$1; + } + elsif ($res->value ne 'OK') { + debug(gettext("blogspam server failure: ").$res->value); + return undef; + } + else { + return undef; + } +} + +1 diff --git a/debian/changelog b/debian/changelog index b75fe6afc..cb362d6e2 100644 --- a/debian/changelog +++ b/debian/changelog @@ -20,6 +20,8 @@ ikiwiki (3.02) UNRELEASED; urgency=low in blogging. * checkcontent: New hook, can be used to implement arbitrary content filters, including spam filters. + * blogspam: New plugin, adding spam filtering for page editing / comment + posting using the BlogSpam.net API. -- Joey Hess Tue, 06 Jan 2009 15:02:52 -0500 diff --git a/doc/plugins/blogspam.mdwn b/doc/plugins/blogspam.mdwn new file mode 100644 index 000000000..307d4646d --- /dev/null +++ b/doc/plugins/blogspam.mdwn @@ -0,0 +1,23 @@ +[[!template id=plugin name=blogspam author="[[Joey]]"]] +[[!tag type/auth]] + +This plugin adds antispam support to ikiwiki, using the +[blogspam.net](http://blogspam.net/) API. Both page edits and +[[comment|comments]] postings can be checked for spam. Currently, +detected spam is not saved for human review, it is just rejected. + +You can control how content is tested via the `blogspam_options` +setting. By default, the options are configured in a way that is +appropriate for wiki content. This includes turning off some of the +more problimatic tests. + +The `blogspam_pagespec` setting is a [[ikiwiki/PageSpec]] that can be +used to configure which pages are checked for spam. The default is to check +all edits. If you only want to check [[comments]] (not wiki page edits), +set it to "postcomment(*)". + +By default, the blogspam.net server is used to do the spam checking. To +change this, the `blogspam_server` option can be set to the url for a +different server implementing the same API. Note that content is sent +unencrypted over the internet to the server, and the server sees +the full text of the content. diff --git a/doc/todo/anti-spam_protection.mdwn b/doc/todo/anti-spam_protection.mdwn index cb45faee5..b0524be5f 100644 --- a/doc/todo/anti-spam_protection.mdwn +++ b/doc/todo/anti-spam_protection.mdwn @@ -17,3 +17,14 @@ Cheers, You might look at the Wikipedia page on "Spam\_in\_blogs" for more ideas. In particular, would it be possible to force a subset of the pages (by regex, but you'd choose the regex to match those pages which are publicly writable) to use rel="nofollow" in all links. > I just wanted to leave a link here to the [[todo/require_CAPTCHA_to_edit]] plugin patch. Unfortunately that plugin currently interacts badly with the openid plugin. -- [[Will]] + + +--- + +Ikiwiki now has a checkcontent hook that plugins can use to see content +that is being entered and check it for spam/whatever. + +There is a blogspam plugin that uses the blogspam.org service +to check for common spam signatures. --[[Joey]] + +[[done]] diff --git a/t/syntax.t b/t/syntax.t index d09d17f7f..9d5cbc373 100755 --- a/t/syntax.t +++ b/t/syntax.t @@ -5,8 +5,8 @@ use Test::More; my @progs="ikiwiki.in"; my @libs="IkiWiki.pm"; -# monotone, external, amazon_s3 skipped since they need perl modules -push @libs, map { chomp; $_ } `find IkiWiki -type f -name \\*.pm | grep -v monotone.pm | grep -v external.pm | grep -v amazon_s3.pm`; +# monotone, external, blogspam, amazon_s3 skipped since they need perl modules +push @libs, map { chomp; $_ } `find IkiWiki -type f -name \\*.pm | grep -v monotone.pm | grep -v external.pm | grep -v blogspam.pm | grep -v amazon_s3.pm`; push @libs, 'IkiWiki/Plugin/skeleton.pm.example'; plan(tests => (@progs + @libs)); -- cgit v1.2.3 From 26fc2dde448151a4c83857738e0bcb8f8169c228 Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Fri, 16 Jan 2009 22:46:10 -0500 Subject: add link to options list --- doc/plugins/blogspam.mdwn | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'doc/plugins') diff --git a/doc/plugins/blogspam.mdwn b/doc/plugins/blogspam.mdwn index 307d4646d..fd854b02e 100644 --- a/doc/plugins/blogspam.mdwn +++ b/doc/plugins/blogspam.mdwn @@ -6,10 +6,10 @@ This plugin adds antispam support to ikiwiki, using the [[comment|comments]] postings can be checked for spam. Currently, detected spam is not saved for human review, it is just rejected. -You can control how content is tested via the `blogspam_options` -setting. By default, the options are configured in a way that is -appropriate for wiki content. This includes turning off some of the -more problimatic tests. +You can control how content is tested via the `blogspam_options` setting. +The list of options is [here](http://blogspam.net/api/testComment.html#options). +By default, the options are configured in a way that is appropriate for +wiki content. This includes turning off some of the more problimatic tests. The `blogspam_pagespec` setting is a [[ikiwiki/PageSpec]] that can be used to configure which pages are checked for spam. The default is to check -- cgit v1.2.3 From a214d537b300a9277a6786a01488ba201e8bc8bb Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Fri, 16 Jan 2009 23:07:02 -0500 Subject: mention dep --- doc/plugins/blogspam.mdwn | 2 ++ 1 file changed, 2 insertions(+) (limited to 'doc/plugins') diff --git a/doc/plugins/blogspam.mdwn b/doc/plugins/blogspam.mdwn index fd854b02e..a090f9c60 100644 --- a/doc/plugins/blogspam.mdwn +++ b/doc/plugins/blogspam.mdwn @@ -6,6 +6,8 @@ This plugin adds antispam support to ikiwiki, using the [[comment|comments]] postings can be checked for spam. Currently, detected spam is not saved for human review, it is just rejected. +The plugin requires the [[!cpan RPC::XML]] perl module. + You can control how content is tested via the `blogspam_options` setting. The list of options is [here](http://blogspam.net/api/testComment.html#options). By default, the options are configured in a way that is appropriate for -- cgit v1.2.3