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          * @param string $protocol 'http://' or 'https://'
  42          * @return int 0 if no match or 1 if there's at least one match
  43          */
  44         public static function matchEntry( Content $content, $filterEntry, $protocol = 'http://' ) {
  45                 if ( !( $content instanceof TextContent ) ) {
  46                         // TODO: handle other types of content too.
  47                         //      Maybe create ContentHandler::matchFilter( LinkFilter ).
  48                         //      Think about a common base class for LinkFilter and MagicWord.
  49                         return 0;
  50                 }
  51
  52                 $text = $content->getNativeData();
  53
  54                 $regex = self::makeRegex( $filterEntry, $protocol );
  55                 return preg_match( $regex, $text );
  56         }
  57
  58         /**
  59          * Builds a regex pattern for $filterEntry.
  60          *
  61          * @param string $filterEntry URL, if it begins with "*.", it'll be
  62          *        replaced to match any subdomain
  63          * @param string $protocol 'http://' or 'https://'
  64          *
  65          * @return string Regex pattern, for preg_match()
  66          */
  67         private static function makeRegex( $filterEntry, $protocol ) {
  68                 $regex = '!' . preg_quote( $protocol, '!' );
  69                 if ( substr( $filterEntry, 0, 2 ) == '*.' ) {
  70                         $regex .= '(?:[A-Za-z0-9.-]+\.|)';
  71                         $filterEntry = substr( $filterEntry, 2 );
  72                 }
  73                 $regex .= preg_quote( $filterEntry, '!' ) . '!Si';
  74                 return $regex;
  75         }
  76
  77         /**
  78          * Make an array to be used for calls to Database::buildLike(), which
  79          * will match the specified string. There are several kinds of filter entry:
  80          *     *.domain.com    -  Produces http://com.domain.%, matches domain.com
  81          *                        and www.domain.com
  82          *     domain.com      -  Produces http://com.domain./%, matches domain.com
  83          *                        or domain.com/ but not www.domain.com
  84          *     *.domain.com/x  -  Produces http://com.domain.%/x%, matches
  85          *                        www.domain.com/xy
  86          *     domain.com/x    -  Produces http://com.domain./x%, matches
  87          *                        domain.com/xy but not www.domain.com/xy
  88          *
  89          * Asterisks in any other location are considered invalid.
  90          *
  91          * This function does the same as wfMakeUrlIndexes(), except it also takes care
  92          * of adding wildcards
  93          *
  94          * @param string $filterEntry Domainparts
  95          * @param string $protocol Protocol (default http://)
  96          * @return array|bool Array to be passed to Database::buildLike() or false on error
  97          */
  98         public static function makeLikeArray( $filterEntry, $protocol = 'http://' ) {
  99                 $db = wfGetDB( DB_REPLICA );
 100
 101                 $target = $protocol . $filterEntry;
 102                 $bits = wfParseUrl( $target );
 103
 104                 if ( $bits == false ) {
 105                         // Unknown protocol?
 106                         return false;
 107                 }
 108
 109                 if ( substr( $bits['host'], 0, 2 ) == '*.' ) {
 110                         $subdomains = true;
 111                         $bits['host'] = substr( $bits['host'], 2 );
 112                         if ( $bits['host'] == '' ) {
 113                                 // We don't want to make a clause that will match everything,
 114                                 // that could be dangerous
 115                                 return false;
 116                         }
 117                 } else {
 118                         $subdomains = false;
 119                 }
 120
 121                 // Reverse the labels in the hostname, convert to lower case
 122                 // For emails reverse domainpart only
 123                 if ( $bits['scheme'] === 'mailto' && strpos( $bits['host'], '@' ) ) {
 124                         // complete email address
 125                         $mailparts = explode( '@', $bits['host'] );
 126                         $domainpart = strtolower( implode( '.', array_reverse( explode( '.', $mailparts[1] ) ) ) );
 127                         $bits['host'] = $domainpart . '@' . $mailparts[0];
 128                 } elseif ( $bits['scheme'] === 'mailto' ) {
 129                         // domainpart of email address only, do not add '.'
 130                         $bits['host'] = strtolower( implode( '.', array_reverse( explode( '.', $bits['host'] ) ) ) );
 131                 } else {
 132                         $bits['host'] = strtolower( implode( '.', array_reverse( explode( '.', $bits['host'] ) ) ) );
 133                         if ( substr( $bits['host'], -1, 1 ) !== '.' ) {
 134                                 $bits['host'] .= '.';
 135                         }
 136                 }
 137
 138                 $like[] = $bits['scheme'] . $bits['delimiter'] . $bits['host'];
 139
 140                 if ( $subdomains ) {
 141                         $like[] = $db->anyString();
 142                 }
 143
 144                 if ( isset( $bits['port'] ) ) {
 145                         $like[] = ':' . $bits['port'];
 146                 }
 147                 if ( isset( $bits['path'] ) ) {
 148                         $like[] = $bits['path'];
 149                 } elseif ( !$subdomains ) {
 150                         $like[] = '/';
 151                 }
 152                 if ( isset( $bits['query'] ) ) {
 153                         $like[] = '?' . $bits['query'];
 154                 }
 155                 if ( isset( $bits['fragment'] ) ) {
 156                         $like[] = '#' . $bits['fragment'];
 157                 }
 158
 159                 // Check for stray asterisks: asterisk only allowed at the start of the domain
 160                 foreach ( $like as $likepart ) {
 161                         if ( !( $likepart instanceof LikeMatch ) && strpos( $likepart, '*' ) !== false ) {
 162                                 return false;
 163                         }
 164                 }
 165
 166                 if ( !( $like[count( $like ) - 1] instanceof LikeMatch ) ) {
 167                         // Add wildcard at the end if there isn't one already
 168                         $like[] = $db->anyString();
 169                 }
 170
 171                 return $like;
 172         }
 173
 174         /**
 175          * Filters an array returned by makeLikeArray(), removing everything past first
 176          * pattern placeholder.
 177          *
 178          * @param array $arr Array to filter
 179          * @return array Filtered array
 180          */
 181         public static function keepOneWildcard( $arr ) {
 182                 if ( !is_array( $arr ) ) {
 183                         return $arr;
 184                 }
 185
 186                 foreach ( $arr as $key => $value ) {
 187                         if ( $value instanceof LikeMatch ) {
 188                                 return array_slice( $arr, 0, $key + 1 );
 189                         }
 190                 }
 191
 192                 return $arr;
 193         }
 194 }