includes/LinkFilter.php

   1 <?php
   2 /**
   3  * Functions to help implement an external link filter for spam control.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  */
  22
  23
  24 /**
  25  * Some functions to help implement an external link filter for spam control.
  26  *
  27  * @todo implement the filter. Currently these are just some functions to help
  28  * maintenance/cleanupSpam.php remove links to a single specified domain. The
  29  * next thing is to implement functions for checking a given page against a big
  30  * list of domains.
  31  *
  32  * Another cool thing to do would be a web interface for fast spam removal.
  33  */
  34 class LinkFilter {
  35
  36         /**
  37          * Check whether $text contains a link to $filterEntry
  38          *
  39          * @param $text String: text to check
  40          * @param $filterEntry String: domainparts, see makeRegex() for more details
  41          * @return Integer: 0 if no match or 1 if there's at least one match
  42          */
  43         static function matchEntry( $text, $filterEntry ) {
  44                 $regex = LinkFilter::makeRegex( $filterEntry );
  45                 return preg_match( $regex, $text );
  46         }
  47
  48         /**
  49          * Builds a regex pattern for $filterEntry.
  50          *
  51          * @param $filterEntry String: URL, if it begins with "*.", it'll be
  52          *        replaced to match any subdomain
  53          * @return String: regex pattern, for preg_match()
  54          */
  55         private static function makeRegex( $filterEntry ) {
  56                 $regex = '!http://';
  57                 if ( substr( $filterEntry, 0, 2 ) == '*.' ) {
  58                         $regex .= '(?:[A-Za-z0-9.-]+\.|)';
  59                         $filterEntry = substr( $filterEntry, 2 );
  60                 }
  61                 $regex .= preg_quote( $filterEntry, '!' ) . '!Si';
  62                 return $regex;
  63         }
  64
  65         /**
  66          * Make an array to be used for calls to DatabaseBase::buildLike(), which
  67          * will match the specified string. There are several kinds of filter entry:
  68          *     *.domain.com    -  Produces http://com.domain.%, matches domain.com
  69          *                        and www.domain.com
  70          *     domain.com      -  Produces http://com.domain./%, matches domain.com
  71          *                        or domain.com/ but not www.domain.com
  72          *     *.domain.com/x  -  Produces http://com.domain.%/x%, matches
  73          *                        www.domain.com/xy
  74          *     domain.com/x    -  Produces http://com.domain./x%, matches
  75          *                        domain.com/xy but not www.domain.com/xy
  76          *
  77          * Asterisks in any other location are considered invalid.
  78          *
  79          * @param $filterEntry String: domainparts
  80          * @param $prot        String: protocol
  81          * @return Array to be passed to DatabaseBase::buildLike() or false on error
  82          */
  83          public static function makeLikeArray( $filterEntry , $prot = 'http://' ) {
  84                 $db = wfGetDB( DB_MASTER );
  85                 if ( substr( $filterEntry, 0, 2 ) == '*.' ) {
  86                         $subdomains = true;
  87                         $filterEntry = substr( $filterEntry, 2 );
  88                         if ( $filterEntry == '' ) {
  89                                 // We don't want to make a clause that will match everything,
  90                                 // that could be dangerous
  91                                 return false;
  92                         }
  93                 } else {
  94                         $subdomains = false;
  95                 }
  96                 // No stray asterisks, that could cause confusion
  97                 // It's not simple or efficient to handle it properly so we don't
  98                 // handle it at all.
  99                 if ( strpos( $filterEntry, '*' ) !== false ) {
 100                         return false;
 101                 }
 102                 $slash = strpos( $filterEntry, '/' );
 103                 if ( $slash !== false ) {
 104                         $path = substr( $filterEntry, $slash );
 105                         $host = substr( $filterEntry, 0, $slash );
 106                 } else {
 107                         $path = '/';
 108                         $host = $filterEntry;
 109                 }
 110                 // Reverse the labels in the hostname, convert to lower case
 111                 // For emails reverse domainpart only
 112                 if ( $prot == 'mailto:' && strpos($host, '@') ) {
 113                         // complete email adress
 114                         $mailparts = explode( '@', $host );
 115                         $domainpart = strtolower( implode( '.', array_reverse( explode( '.', $mailparts[1] ) ) ) );
 116                         $host = $domainpart . '@' . $mailparts[0];
 117                         $like = array( "$prot$host", $db->anyString() );
 118                 } elseif ( $prot == 'mailto:' ) {
 119                         // domainpart of email adress only. do not add '.'
 120                         $host = strtolower( implode( '.', array_reverse( explode( '.', $host ) ) ) );
 121                         $like = array( "$prot$host", $db->anyString() );
 122                 } else {
 123                         $host = strtolower( implode( '.', array_reverse( explode( '.', $host ) ) ) );
 124                         if ( substr( $host, -1, 1 ) !== '.' ) {
 125                                 $host .= '.';
 126                         }
 127                         $like = array( "$prot$host" );
 128
 129                         if ( $subdomains ) {
 130                                 $like[] = $db->anyString();
 131                         }
 132                         if ( !$subdomains || $path !== '/' ) {
 133                                 $like[] = $path;
 134                                 $like[] = $db->anyString();
 135                         }
 136                 }
 137                 return $like;
 138         }
 139
 140         /**
 141          * Filters an array returned by makeLikeArray(), removing everything past first pattern placeholder.
 142          *
 143          * @param $arr array: array to filter
 144          * @return array filtered array
 145          */
 146         public static function keepOneWildcard( $arr ) {
 147                 if( !is_array( $arr ) ) {
 148                         return $arr;
 149                 }
 150
 151                 foreach( $arr as $key => $value ) {
 152                         if ( $value instanceof LikeMatch ) {
 153                                 return array_slice( $arr, 0, $key + 1 );
 154                         }
 155                 }
 156
 157                 return $arr;
 158         }
 159 }