Changeset 1037

Timestamp:

05/15/07 22:16:00 (19 months ago)

Author:

kake

Message:

Added experimental support for local spam detection modules.

Location:

trunk

Files:

• Build.PL (modified) (2 diffs)
• Changes (modified) (1 diff)
• INSTALL (modified) (1 diff)
• MANIFEST (modified) (2 diffs)
• lib/OpenGuides.pm (modified) (3 diffs)
• lib/OpenGuides/Config.pm (modified) (4 diffs)
• t/76_detect_spam.t (added)
• templates/spam_detected.tt (added)

trunk/Build.PL

@@ -105 +105 @@
    licence_name licence_url licence_info_url moderation_requires_password
    enable_node_image enable_common_categories enable_common_locales
+   spam_detector_module
   ) ) {
     my $q_method = $var . "__qu";
@@ -370 +371 @@
                       "site_index.tt",
                       "search.tt",
+                      "spam_detected.tt",
                       "userstats.tt",
                       "wanted_pages.tt"

trunk/Changes

@@ -5 +5 @@
 
 0.61
+        Added experimental support for local spam detection modules; see the
+          commit_node method in "perldoc OpenGuides" for details.
         Added extra "edit this page" link next to the node name; if you don't
           want it, add div#title_edit_link {display:none;} to your stylesheet.

trunk/INSTALL

@@ -278 +278 @@
 policy, but you can leave this blank for now if you don't have one yet.
 
+  "What module would you like to use for spam detection? (optional)"
+
+The module you choose should have a method called "looks_like_spam", which
+accepts a hash with content and metadata as keys, and returns true or false
+to the question of whether the edit should be considered to be spam. 
+
   "Distance calculation methods available are:
       1) British National Grid

trunk/MANIFEST

@@ -74 +74 @@
 templates/site_index.tt
 templates/search.tt
+templates/spam_detected.tt
 templates/userstats.tt
 templates/wanted_pages.tt
@@ -137 +138 @@
 t/74_ping_plugin.t
 t/75_revert_user.t
+t/76_detect_spam.t
 t/templates/15_test.tt
 wiki.cgi

trunk/lib/OpenGuides.pm

@@ -1241 +1241 @@
 printing the output to STDOUT.
 
+If you have specified the C<spam_detector_module> option in your
+C<wiki.conf>, this method will attempt to call the <looks_like_spam>
+method of that module to determine whether the edit is spam.  If this
+method returns true, then the C<spam_detected.tt> template will be
+used to display an error message.
+
+The C<looks_like_spam> method will be passed a datastructure containing
+content and metadata.
+
 The geographical data that you should provide in the L<CGI> object
 depends on the handler you chose in C<wiki.conf>.
@@ -1295 +1304 @@
     $new_metadata{longitude} = delete $new_metadata{longitude_unmunged}
         if $new_metadata{longitude_unmunged};
+
+    foreach my $var ( qw( summary username comment edit_type ) ) {
+        $new_metadata{$var} = $q->param($var) || "";
+    }
+    $new_metadata{host} = $ENV{REMOTE_ADDR};
+
+    # Wiki::Toolkit::Plugin::RSS::ModWiki wants "major_change" to be set.
+    $new_metadata{major_change} = ( $new_metadata{edit_type} eq "Normal edit" )
+                                    ? 1
+                                    : 0;
+
+    # If we can, check to see if this edit looks like spam.
+    my $spam_detector = $config->spam_detector_module;
+    my $is_spam;
+    if ( $spam_detector ) {
+        eval {
+            eval "require $spam_detector";
+            $is_spam = $spam_detector->looks_like_spam(
+                node    => $node,
+                content => $content,
+                metadata => \%new_metadata,
+            );
+        };
+    }
+
+    if ( $is_spam ) {
+        my $output = OpenGuides::Template->output(
+            wiki     => $self->wiki,
+            config   => $config,
+            template => "spam_detected.tt",
+            vars     => {
+                          not_editable => 1,
+                        },
+        );
+        return $output if $return_output;
+        print $output;
+        return;
+    }
 
     # Check to make sure all the indexable nodes are created
@@ -1306 +1353 @@
     }
     
-    foreach my $var ( qw( summary username comment edit_type ) ) {
-        $new_metadata{$var} = $q->param($var) || "";
-    }
-    $new_metadata{host} = $ENV{REMOTE_ADDR};
-
-    # Wiki::Toolkit::Plugin::RSS::ModWiki wants "major_change" to be set.
-    $new_metadata{major_change} = ( $new_metadata{edit_type} eq "Normal edit" )
-                                    ? 1
-                                    : 0;
-
     my $written = $wiki->write_node( $node, $content, $checksum,
                                      \%new_metadata );

trunk/lib/OpenGuides/Config.pm

@@ -22 +22 @@
    licence_name licence_url licence_info_url moderation_requires_password
    enable_node_image enable_common_categories enable_common_locales
+   spam_detector_module
 );
 my @questions = map { $_ . "__qu" } @variables;
@@ -106 +107 @@
                      licence_name => "",
                      licence_url => "",
-                     licence_info_url => ""
+                     licence_info_url => "",
+                     spam_detector_module => "",
                    );
 
@@ -186 +188 @@
         licence_name => "What licence will you use for the guide?",
         licence_url => "What is the URL to your licence?",
-        licence_info_url => "What is the URL to your local page about your licensing policy?"
+        licence_info_url => "What is the URL to your local page about your licensing policy?",
+        spam_detector_module => "What module would you like to use for spam detection? (optional)",
     );
 
@@ -311 +314 @@
 =item * licence_info_url
 
+=item * spam_detector_module
+
 =back