brion at svn
Aug 19, 2008, 1:32 PM
Post #1 of 1
(109 views)
Permalink
|
|
SVN: [39660] trunk/phase3
|
|
Revision: 39660
Author: brion
Date: 2008-08-19 20:32:30 +0000 (Tue, 19 Aug 2008)
Log Message:
-----------
* $wgSpamRegex now matches the edit summary and page move descriptions in
addition to body text.
Could use some cleanup to the error display on page moves; it recycles the general edit notice for spam hits currently, which is pretty big on English Wikipedia, and doesn't say what match you hit.
Modified Paths:
--------------
trunk/phase3/RELEASE-NOTES
trunk/phase3/includes/DefaultSettings.php
trunk/phase3/includes/EditPage.php
trunk/phase3/includes/Title.php
Modified: trunk/phase3/RELEASE-NOTES
===================================================================
--- trunk/phase3/RELEASE-NOTES 2008-08-19 19:44:10 UTC (rev 39659)
+++ trunk/phase3/RELEASE-NOTES 2008-08-19 20:32:30 UTC (rev 39660)
@@ -37,6 +37,7 @@
* $wgRestrictDisplayTitle controls if the use of the {{DISPLAYTITLE}} magic
word is restricted to titles equivalent to the actual page title. This
is true per default, but can be set to false to allow any title.
+* $wgSpamRegex may now be an array of multiple regular expressions.
=== New features in 1.14 ===
@@ -134,7 +135,10 @@
* Avoid recursive crazy expansions in section edit comments for pages which
contain '/*' in the title
* Fix excessive memory usage when parsing pages with lots of links
+* $wgSpamRegex now matches the edit summary and page move descriptions in
+ addition to body text.
+
=== API changes in 1.14 ===
* Registration time of users registered before the DB field was created is now
Modified: trunk/phase3/includes/DefaultSettings.php
===================================================================
--- trunk/phase3/includes/DefaultSettings.php 2008-08-19 19:44:10 UTC (rev 39659)
+++ trunk/phase3/includes/DefaultSettings.php 2008-08-19 20:32:30 UTC (rev 39660)
@@ -2095,9 +2095,17 @@
$wgExportAllowListContributors = false ;
-/** Text matching this regular expression will be recognised as spam
- * See http://en.wikipedia.org/wiki/Regular_expression */
-$wgSpamRegex = false;
+/**
+ * Edits matching these regular expressions in body text or edit summary
+ * will be recognised as spam and rejected automatically.
+ *
+ * There's no administrator override on-wiki, so be careful what you set. :)
+ * May be an array of regexes or a single string for backwards compatibility.
+ *
+ * See http://en.wikipedia.org/wiki/Regular_expression
+ */
+$wgSpamRegex = array();
+
/** Similarly you can get a function to do the job. The function will be given
* the following args:
* - a Title object for the article the edit is made on
Modified: trunk/phase3/includes/EditPage.php
===================================================================
--- trunk/phase3/includes/EditPage.php 2008-08-19 19:44:10 UTC (rev 39659)
+++ trunk/phase3/includes/EditPage.php 2008-08-19 20:32:30 UTC (rev 39660)
@@ -733,7 +733,7 @@
* @return one of the constants describing the result
*/
function internalAttemptSave( &$result, $bot = false ) {
- global $wgSpamRegex, $wgFilterCallback, $wgUser, $wgOut, $wgParser;
+ global $wgFilterCallback, $wgUser, $wgOut, $wgParser;
global $wgMaxArticleSize;
$fname = 'EditPage::attemptSave';
@@ -762,12 +762,15 @@
$this->mMetaData = '' ;
# Check for spam
- $matches = array();
- if ( $wgSpamRegex && preg_match( $wgSpamRegex, $this->textbox1, $matches ) ) {
- $result['spam'] = $matches[0];
+ $match = self::matchSpamRegex( $this->summary );
+ if( $match === false ) {
+ $match = self::matchSpamRegex( $this->textbox1 );
+ }
+ if( $match !== false ) {
+ $result['spam'] = $match;
$ip = wfGetIP();
$pdbk = $this->mTitle->getPrefixedDBkey();
- $match = str_replace( "\n", '', $matches[0] );
+ $match = str_replace( "\n", '', $match );
wfDebugLog( 'SpamRegex', "$ip spam regex hit [[$pdbk]]: \"$match\"" );
wfProfileOut( "$fname-checks" );
wfProfileOut( $fname );
@@ -1022,6 +1025,25 @@
wfProfileOut( $fname );
return self::AS_END;
}
+
+ /**
+ * Check given input text against $wgSpamRegex, and return the text of the first match.
+ * @return mixed -- matching string or false
+ */
+ public static function matchSpamRegex( $text ) {
+ global $wgSpamRegex;
+ if( $wgSpamRegex ) {
+ // For back compatibility, $wgSpamRegex may be a single string or an array of regexes.
+ $regexes = (array)$wgSpamRegex;
+ foreach( $regexes as $regex ) {
+ $matches = array();
+ if ( preg_match( $regex, $text, $matches ) ) {
+ return $matches[0];
+ }
+ }
+ }
+ return false;
+ }
/**
* Initialise form fields in the object
Modified: trunk/phase3/includes/Title.php
===================================================================
--- trunk/phase3/includes/Title.php 2008-08-19 19:44:10 UTC (rev 39659)
+++ trunk/phase3/includes/Title.php 2008-08-19 20:32:30 UTC (rev 39660)
@@ -2500,6 +2500,12 @@
$nt->getUserPermissionsErrors('edit', $wgUser));
}
+ $match = EditPage::matchSpamRegex( $reason );
+ if( $match !== false ) {
+ // This is kind of lame, won't display nice
+ $errors[] = array('spamprotectiontext');
+ }
+
global $wgUser;
$err = null;
if( !wfRunHooks( 'AbortMove', array( $this, $nt, $wgUser, &$err, $reason ) ) ) {
_______________________________________________
MediaWiki-CVS mailing list
MediaWiki-CVS[at]lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-cvs
|
