Reverted r113177 per CR

[lhc/web/wiklou.git] / includes / LinkFilter.php

diff --git a/includes/LinkFilter.php b/includes/LinkFilter.php
index 5010009..f1e19f9 100644 (file)
--- a/includes/LinkFilter.php
+++ b/includes/LinkFilter.php
@@ -3,7 +3,7 @@
 /**
  * Some functions to help implement an external link filter for spam control.
  * 
- * TODO: implement the filter. Currently these are just some functions to help
+ * @todo implement the filter. Currently these are just some functions to help
  * maintenance/cleanupSpam.php remove links to a single specified domain. The
  * next thing is to implement functions for checking a given page against a big
  * list of domains.
@@ -11,21 +11,30 @@
  * Another cool thing to do would be a web interface for fast spam removal.
  */
 class LinkFilter {
+
        /**
-        * @static
+        * Check whether $text contains a link to $filterEntry
+        *
+        * @param $text String: text to check
+        * @param $filterEntry String: domainparts, see makeRegex() for more details
+        * @return Integer: 0 if no match or 1 if there's at least one match
         */
-       function matchEntry( $text, $filterEntry ) {
+       static function matchEntry( $text, $filterEntry ) {
                $regex = LinkFilter::makeRegex( $filterEntry );
                return preg_match( $regex, $text );
        }
 
        /**
-        * @static
+        * Builds a regex pattern for $filterEntry.
+        *
+        * @param $filterEntry String: URL, if it begins with "*.", it'll be
+        *        replaced to match any subdomain
+        * @return String: regex pattern, for preg_match()
         */
-       function makeRegex( $filterEntry ) {
+       private static function makeRegex( $filterEntry ) {
                $regex = '!http://';
                if ( substr( $filterEntry, 0, 2 ) == '*.' ) {
-                       $regex .= '([A-Za-z0-9.-]+\.|)';
+                       $regex .= '(?:[A-Za-z0-9.-]+\.|)';
                        $filterEntry = substr( $filterEntry, 2 );
                }
                $regex .= preg_quote( $filterEntry, '!' ) . '!Si';
@@ -33,8 +42,8 @@ class LinkFilter {
        }
 
        /**
-        * Make a string to go after an SQL LIKE, which will match the specified
-        * string. There are several kinds of filter entry:
+        * Make an array to be used for calls to DatabaseBase::buildLike(), which
+        * will match the specified string. There are several kinds of filter entry:
         *     *.domain.com    -  Produces http://com.domain.%, matches domain.com
         *                        and www.domain.com
         *     domain.com      -  Produces http://com.domain./%, matches domain.com
@@ -45,12 +54,13 @@ class LinkFilter {
         *                        domain.com/xy but not www.domain.com/xy
         *
         * Asterisks in any other location are considered invalid.
-        * 
-        * @static
+        *
         * @param $filterEntry String: domainparts
         * @param $prot        String: protocol
+        * @return Array to be passed to DatabaseBase::buildLike() or false on error
         */
-       function makeLike( $filterEntry , $prot = 'http' ) {
+        public static function makeLikeArray( $filterEntry , $prot = 'http://' ) {
+               $db = wfGetDB( DB_MASTER );
                if ( substr( $filterEntry, 0, 2 ) == '*.' ) {
                        $subdomains = true;
                        $filterEntry = substr( $filterEntry, 2 );
@@ -76,19 +86,53 @@ class LinkFilter {
                        $path = '/';
                        $host = $filterEntry;
                }
-               $host = strtolower( implode( '.', array_reverse( explode( '.', $host ) ) ) );
-               if ( substr( $host, -1, 1 ) !== '.' ) {
-                       $host .= '.';
+               // Reverse the labels in the hostname, convert to lower case
+               // For emails reverse domainpart only
+               if ( $prot == 'mailto:' && strpos($host, '@') ) {
+                       // complete email adress 
+                       $mailparts = explode( '@', $host );
+                       $domainpart = strtolower( implode( '.', array_reverse( explode( '.', $mailparts[1] ) ) ) );
+                       $host = $domainpart . '@' . $mailparts[0];
+                       $like = array( "$prot$host", $db->anyString() );
+               } elseif ( $prot == 'mailto:' ) {
+                       // domainpart of email adress only. do not add '.'
+                       $host = strtolower( implode( '.', array_reverse( explode( '.', $host ) ) ) );   
+                       $like = array( "$prot$host", $db->anyString() );                        
+               } else {
+                       $host = strtolower( implode( '.', array_reverse( explode( '.', $host ) ) ) );   
+                       if ( substr( $host, -1, 1 ) !== '.' ) {
+                               $host .= '.';
+                       }
+                       $like = array( "$prot$host" );
+
+                       if ( $subdomains ) {
+                               $like[] = $db->anyString();
+                       }
+                       if ( !$subdomains || $path !== '/' ) {
+                               $like[] = $path;
+                               $like[] = $db->anyString();
+                       }
                }
-               $like = "$prot://$host";
-               
-               if ( $subdomains ) {
-                       $like .= '%';
+               return $like;
+       }
+
+       /**
+        * Filters an array returned by makeLikeArray(), removing everything past first pattern placeholder.
+        *
+        * @param $arr array: array to filter
+        * @return array filtered array
+        */
+       public static function keepOneWildcard( $arr ) {
+               if( !is_array( $arr ) ) {
+                       return $arr;
                }
-               if ( !$subdomains || $path !== '/' ) {
-                       $like .= $path . '%';
+
+               foreach( $arr as $key => $value ) {
+                       if ( $value instanceof LikeMatch ) {
+                               return array_slice( $arr, 0, $key + 1 );
+                       }
                }
-               return $like;
+
+               return $arr;
        }
 }
-?>