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 use Wikimedia\Rdbms\LikeMatch;
  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 $content contains a link to $filterEntry
  38          *
  39          * @param Content $content Content to check
  40          * @param string $filterEntry Domainparts, see makeRegex() for more details
  41          * @return int 0 if no match or 1 if there's at least one match
  42          */
  43         static function matchEntry( Content $content, $filterEntry ) {
  44                 if ( !( $content instanceof TextContent ) ) {
  45                         // TODO: handle other types of content too.
  46                         //      Maybe create ContentHandler::matchFilter( LinkFilter ).
  47                         //      Think about a common base class for LinkFilter and MagicWord.
  48                         return 0;
  49                 }
  50
  51                 $text = $content->getNativeData();
  52
  53                 $regex = LinkFilter::makeRegex( $filterEntry );
  54                 return preg_match( $regex, $text );
  55         }
  56
  57         /**
  58          * Builds a regex pattern for $filterEntry.
  59          *
  60          * @param string $filterEntry URL, if it begins with "*.", it'll be
  61          *        replaced to match any subdomain
  62          * @return string Regex pattern, for preg_match()
  63          */
  64         private static function makeRegex( $filterEntry ) {
  65                 $regex = '!http://';
  66                 if ( substr( $filterEntry, 0, 2 ) == '*.' ) {
  67                         $regex .= '(?:[A-Za-z0-9.-]+\.|)';
  68                         $filterEntry = substr( $filterEntry, 2 );
  69                 }
  70                 $regex .= preg_quote( $filterEntry, '!' ) . '!Si';
  71                 return $regex;
  72         }
  73
  74         /**
  75          * Make an array to be used for calls to Database::buildLike(), which
  76          * will match the specified string. There are several kinds of filter entry:
  77          *     *.domain.com    -  Produces http://com.domain.%, matches domain.com
  78          *                        and www.domain.com
  79          *     domain.com      -  Produces http://com.domain./%, matches domain.com
  80          *                        or domain.com/ but not www.domain.com
  81          *     *.domain.com/x  -  Produces http://com.domain.%/x%, matches
  82          *                        www.domain.com/xy
  83          *     domain.com/x    -  Produces http://com.domain./x%, matches
  84          *                        domain.com/xy but not www.domain.com/xy
  85          *
  86          * Asterisks in any other location are considered invalid.
  87          *
  88          * This function does the same as wfMakeUrlIndexes(), except it also takes care
  89          * of adding wildcards
  90          *
  91          * @param string $filterEntry Domainparts
  92          * @param string $protocol Protocol (default http://)
  93          * @return array|bool Array to be passed to Database::buildLike() or false on error
  94          */
  95         public static function makeLikeArray( $filterEntry, $protocol = 'http://' ) {
  96                 $db = wfGetDB( DB_REPLICA );
  97
  98                 $target = $protocol . $filterEntry;
  99                 $bits = wfParseUrl( $target );
 100
 101                 if ( $bits == false ) {
 102                         // Unknown protocol?
 103                         return false;
 104                 }
 105
 106                 if ( substr( $bits['host'], 0, 2 ) == '*.' ) {
 107                         $subdomains = true;
 108                         $bits['host'] = substr( $bits['host'], 2 );
 109                         if ( $bits['host'] == '' ) {
 110                                 // We don't want to make a clause that will match everything,
 111                                 // that could be dangerous
 112                                 return false;
 113                         }
 114                 } else {
 115                         $subdomains = false;
 116                 }
 117
 118                 // Reverse the labels in the hostname, convert to lower case
 119                 // For emails reverse domainpart only
 120                 if ( $bits['scheme'] === 'mailto' && strpos( $bits['host'], '@' ) ) {
 121                         // complete email address
 122                         $mailparts = explode( '@', $bits['host'] );
 123                         $domainpart = strtolower( implode( '.', array_reverse( explode( '.', $mailparts[1] ) ) ) );
 124                         $bits['host'] = $domainpart . '@' . $mailparts[0];
 125                 } elseif ( $bits['scheme'] === 'mailto' ) {
 126                         // domainpart of email address only, do not add '.'
 127                         $bits['host'] = strtolower( implode( '.', array_reverse( explode( '.', $bits['host'] ) ) ) );
 128                 } else {
 129                         $bits['host'] = strtolower( implode( '.', array_reverse( explode( '.', $bits['host'] ) ) ) );
 130                         if ( substr( $bits['host'], -1, 1 ) !== '.' ) {
 131                                 $bits['host'] .= '.';
 132                         }
 133                 }
 134
 135                 $like[] = $bits['scheme'] . $bits['delimiter'] . $bits['host'];
 136
 137                 if ( $subdomains ) {
 138                         $like[] = $db->anyString();
 139                 }
 140
 141                 if ( isset( $bits['port'] ) ) {
 142                         $like[] = ':' . $bits['port'];
 143                 }
 144                 if ( isset( $bits['path'] ) ) {
 145                         $like[] = $bits['path'];
 146                 } elseif ( !$subdomains ) {
 147                         $like[] = '/';
 148                 }
 149                 if ( isset( $bits['query'] ) ) {
 150                         $like[] = '?' . $bits['query'];
 151                 }
 152                 if ( isset( $bits['fragment'] ) ) {
 153                         $like[] = '#' . $bits['fragment'];
 154                 }
 155
 156                 // Check for stray asterisks: asterisk only allowed at the start of the domain
 157                 foreach ( $like as $likepart ) {
 158                         if ( !( $likepart instanceof LikeMatch ) && strpos( $likepart, '*' ) !== false ) {
 159                                 return false;
 160                         }
 161                 }
 162
 163                 if ( !( $like[count( $like ) - 1] instanceof LikeMatch ) ) {
 164                         // Add wildcard at the end if there isn't one already
 165                         $like[] = $db->anyString();
 166                 }
 167
 168                 return $like;
 169         }
 170
 171         /**
 172          * Filters an array returned by makeLikeArray(), removing everything past first
 173          * pattern placeholder.
 174          *
 175          * @param array $arr Array to filter
 176          * @return array Filtered array
 177          */
 178         public static function keepOneWildcard( $arr ) {
 179                 if ( !is_array( $arr ) ) {
 180                         return $arr;
 181                 }
 182
 183                 foreach ( $arr as $key => $value ) {
 184                         if ( $value instanceof LikeMatch ) {
 185                                 return array_slice( $arr, 0, $key + 1 );
 186                         }
 187                 }
 188
 189                 return $arr;
 190         }
 191 }