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  * Some functions to help implement an external link filter for spam control.
  25  *
  26  * @todo implement the filter. Currently these are just some functions to help
  27  * maintenance/cleanupSpam.php remove links to a single specified domain. The
  28  * next thing is to implement functions for checking a given page against a big
  29  * list of domains.
  30  *
  31  * Another cool thing to do would be a web interface for fast spam removal.
  32  */
  33 class LinkFilter {
  34
  35         /**
  36          * Check whether $content contains a link to $filterEntry
  37          *
  38          * @param Content $content Content to check
  39          * @param string $filterEntry Domainparts, see makeRegex() for more details
  40          * @return int 0 if no match or 1 if there's at least one match
  41          */
  42         static function matchEntry( Content $content, $filterEntry ) {
  43                 if ( !( $content instanceof TextContent ) ) {
  44                         //TODO: handle other types of content too.
  45                         //      Maybe create ContentHandler::matchFilter( LinkFilter ).
  46                         //      Think about a common base class for LinkFilter and MagicWord.
  47                         return 0;
  48                 }
  49
  50                 $text = $content->getNativeData();
  51
  52                 $regex = LinkFilter::makeRegex( $filterEntry );
  53                 return preg_match( $regex, $text );
  54         }
  55
  56         /**
  57          * Builds a regex pattern for $filterEntry.
  58          *
  59          * @param string $filterEntry URL, if it begins with "*.", it'll be
  60          *        replaced to match any subdomain
  61          * @return string Regex pattern, for preg_match()
  62          */
  63         private static function makeRegex( $filterEntry ) {
  64                 $regex = '!http://';
  65                 if ( substr( $filterEntry, 0, 2 ) == '*.' ) {
  66                         $regex .= '(?:[A-Za-z0-9.-]+\.|)';
  67                         $filterEntry = substr( $filterEntry, 2 );
  68                 }
  69                 $regex .= preg_quote( $filterEntry, '!' ) . '!Si';
  70                 return $regex;
  71         }
  72
  73         /**
  74          * Make an array to be used for calls to DatabaseBase::buildLike(), which
  75          * will match the specified string. There are several kinds of filter entry:
  76          *     *.domain.com    -  Produces http://com.domain.%, matches domain.com
  77          *                        and www.domain.com
  78          *     domain.com      -  Produces http://com.domain./%, matches domain.com
  79          *                        or domain.com/ but not www.domain.com
  80          *     *.domain.com/x  -  Produces http://com.domain.%/x%, matches
  81          *                        www.domain.com/xy
  82          *     domain.com/x    -  Produces http://com.domain./x%, matches
  83          *                        domain.com/xy but not www.domain.com/xy
  84          *
  85          * Asterisks in any other location are considered invalid.
  86          *
  87          * This function does the same as wfMakeUrlIndexes(), except it also takes care
  88          * of adding wildcards
  89          *
  90          * @param string $filterEntry Domainparts
  91          * @param string $protocol Protocol (default http://)
  92          * @return array Array to be passed to DatabaseBase::buildLike() or false on error
  93          */
  94         public static function makeLikeArray( $filterEntry, $protocol = 'http://' ) {
  95                 $db = wfGetDB( DB_MASTER );
  96
  97                 $target = $protocol . $filterEntry;
  98                 $bits = wfParseUrl( $target );
  99
 100                 if ( $bits == false ) {
 101                         // Unknown protocol?
 102                         return false;
 103                 }
 104
 105                 if ( substr( $bits['host'], 0, 2 ) == '*.' ) {
 106                         $subdomains = true;
 107                         $bits['host'] = substr( $bits['host'], 2 );
 108                         if ( $bits['host'] == '' ) {
 109                                 // We don't want to make a clause that will match everything,
 110                                 // that could be dangerous
 111                                 return false;
 112                         }
 113                 } else {
 114                         $subdomains = false;
 115                 }
 116
 117                 // Reverse the labels in the hostname, convert to lower case
 118                 // For emails reverse domainpart only
 119                 if ( $bits['scheme'] === 'mailto' && strpos( $bits['host'], '@' ) ) {
 120                         // complete email address
 121                         $mailparts = explode( '@', $bits['host'] );
 122                         $domainpart = strtolower( implode( '.', array_reverse( explode( '.', $mailparts[1] ) ) ) );
 123                         $bits['host'] = $domainpart . '@' . $mailparts[0];
 124                 } elseif ( $bits['scheme'] === 'mailto' ) {
 125                         // domainpart of email address only, do not add '.'
 126                         $bits['host'] = strtolower( implode( '.', array_reverse( explode( '.', $bits['host'] ) ) ) );
 127                 } else {
 128                         $bits['host'] = strtolower( implode( '.', array_reverse( explode( '.', $bits['host'] ) ) ) );
 129                         if ( substr( $bits['host'], -1, 1 ) !== '.' ) {
 130                                 $bits['host'] .= '.';
 131                         }
 132                 }
 133
 134                 $like[] = $bits['scheme'] . $bits['delimiter'] . $bits['host'];
 135
 136                 if ( $subdomains ) {
 137                         $like[] = $db->anyString();
 138                 }
 139
 140                 if ( isset( $bits['port'] ) ) {
 141                         $like[] = ':' . $bits['port'];
 142                 }
 143                 if ( isset( $bits['path'] ) ) {
 144                         $like[] = $bits['path'];
 145                 } elseif ( !$subdomains ) {
 146                         $like[] = '/';
 147                 }
 148                 if ( isset( $bits['query'] ) ) {
 149                         $like[] = '?' . $bits['query'];
 150                 }
 151                 if ( isset( $bits['fragment'] ) ) {
 152                         $like[] = '#' . $bits['fragment'];
 153                 }
 154
 155                 // Check for stray asterisks: asterisk only allowed at the start of the domain
 156                 foreach ( $like as $likepart ) {
 157                         if ( !( $likepart instanceof LikeMatch ) && strpos( $likepart, '*' ) !== false ) {
 158                                 return false;
 159                         }
 160                 }
 161
 162                 if ( !( $like[count( $like ) - 1] instanceof LikeMatch ) ) {
 163                         // Add wildcard at the end if there isn't one already
 164                         $like[] = $db->anyString();
 165                 }
 166
 167                 return $like;
 168         }
 169
 170         /**
 171          * Filters an array returned by makeLikeArray(), removing everything past first
 172          * pattern placeholder.
 173          *
 174          * @param array $arr Array to filter
 175          * @return array Filtered array
 176          */
 177         public static function keepOneWildcard( $arr ) {
 178                 if ( !is_array( $arr ) ) {
 179                         return $arr;
 180                 }
 181
 182                 foreach ( $arr as $key => $value ) {
 183                         if ( $value instanceof LikeMatch ) {
 184                                 return array_slice( $arr, 0, $key + 1 );
 185                         }
 186                 }
 187
 188                 return $arr;
 189         }
 190 }