From 06e3d0e37777a221559c52d28de479e0a21a8f72 Mon Sep 17 00:00:00 2001
From: Brion Vibber <brion@users.mediawiki.org>
Date: Tue, 19 Aug 2008 20:32:30 +0000
Subject: [PATCH] * $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.
---
 RELEASE-NOTES                |  4 ++++
 includes/DefaultSettings.php | 14 +++++++++++---
 includes/EditPage.php        | 32 +++++++++++++++++++++++++++-----
 includes/Title.php           |  6 ++++++
 4 files changed, 48 insertions(+), 8 deletions(-)

diff --git a/RELEASE-NOTES b/RELEASE-NOTES
index 683880826e..f4bb65bf26 100644
--- a/RELEASE-NOTES
+++ b/RELEASE-NOTES
@@ -37,6 +37,7 @@ it from source control: http://www.mediawiki.org/wiki/Download_from_SVN
 * $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 ===
 
@@ -134,6 +135,9 @@ it from source control: http://www.mediawiki.org/wiki/Download_from_SVN
 * 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 ===
 
diff --git a/includes/DefaultSettings.php b/includes/DefaultSettings.php
index c78a772207..e3a48d401f 100644
--- a/includes/DefaultSettings.php
+++ b/includes/DefaultSettings.php
@@ -2095,9 +2095,17 @@ $wgExportMaxHistory = 0;
 $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
diff --git a/includes/EditPage.php b/includes/EditPage.php
index 1b86871937..614019ed0b 100644
--- a/includes/EditPage.php
+++ b/includes/EditPage.php
@@ -733,7 +733,7 @@ class EditPage {
 	 * @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 @@ class EditPage {
 		$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 @@ class EditPage {
 		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
diff --git a/includes/Title.php b/includes/Title.php
index 0f1a18939b..67cd6e6051 100644
--- a/includes/Title.php
+++ b/includes/Title.php
@@ -2500,6 +2500,12 @@ class Title {
 					$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 ) ) ) {
-- 
2.20.1