maintenance/cleanupSpam.php

   1 <?php
   2 /**
   3  * Cleanup all spam from a given hostname.
   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  * @ingroup Maintenance
  22  */
  23
  24 require_once __DIR__ . '/Maintenance.php';
  25
  26 /**
  27  * Maintenance script to cleanup all spam from a given hostname.
  28  *
  29  * @ingroup Maintenance
  30  */
  31 class CleanupSpam extends Maintenance {
  32
  33         public function __construct() {
  34                 parent::__construct();
  35                 $this->addDescription( 'Cleanup all spam from a given hostname' );
  36                 $this->addOption( 'all', 'Check all wikis in $wgLocalDatabases' );
  37                 $this->addOption( 'delete', 'Delete pages containing only spam instead of blanking them' );
  38                 $this->addArg(
  39                         'hostname',
  40                         'Hostname that was spamming, single * wildcard in the beginning allowed'
  41                 );
  42         }
  43
  44         public function execute() {
  45                 global $IP, $wgLocalDatabases, $wgUser;
  46
  47                 $username = wfMessage( 'spambot_username' )->text();
  48                 $wgUser = User::newSystemUser( $username );
  49                 if ( !$wgUser ) {
  50                         $this->error( "Invalid username specified in 'spambot_username' message: $username", true );
  51                 }
  52                 // Create the user if necessary
  53                 if ( !$wgUser->getId() ) {
  54                         $wgUser->addToDatabase();
  55                 }
  56                 $spec = $this->getArg();
  57                 $like = LinkFilter::makeLikeArray( $spec );
  58                 if ( !$like ) {
  59                         $this->error( "Not a valid hostname specification: $spec", true );
  60                 }
  61
  62                 if ( $this->hasOption( 'all' ) ) {
  63                         // Clean up spam on all wikis
  64                         $this->output( "Finding spam on " . count( $wgLocalDatabases ) . " wikis\n" );
  65                         $found = false;
  66                         foreach ( $wgLocalDatabases as $wikiID ) {
  67                                 $dbr = $this->getDB( DB_REPLICA, [], $wikiID );
  68
  69                                 $count = $dbr->selectField( 'externallinks', 'COUNT(*)',
  70                                         [ 'el_index' . $dbr->buildLike( $like ) ], __METHOD__ );
  71                                 if ( $count ) {
  72                                         $found = true;
  73                                         $cmd = wfShellWikiCmd( "$IP/maintenance/cleanupSpam.php",
  74                                                 [ '--wiki', $wikiID, $spec ] );
  75                                         passthru( "$cmd | sed 's/^/$wikiID:  /'" );
  76                                 }
  77                         }
  78                         if ( $found ) {
  79                                 $this->output( "All done\n" );
  80                         } else {
  81                                 $this->output( "None found\n" );
  82                         }
  83                 } else {
  84                         // Clean up spam on this wiki
  85
  86                         $dbr = $this->getDB( DB_REPLICA );
  87                         $res = $dbr->select( 'externallinks', [ 'DISTINCT el_from' ],
  88                                 [ 'el_index' . $dbr->buildLike( $like ) ], __METHOD__ );
  89                         $count = $dbr->numRows( $res );
  90                         $this->output( "Found $count articles containing $spec\n" );
  91                         foreach ( $res as $row ) {
  92                                 $this->cleanupArticle( $row->el_from, $spec );
  93                         }
  94                         if ( $count ) {
  95                                 $this->output( "Done\n" );
  96                         }
  97                 }
  98         }
  99
 100         private function cleanupArticle( $id, $domain ) {
 101                 $title = Title::newFromID( $id );
 102                 if ( !$title ) {
 103                         $this->error( "Internal error: no page for ID $id" );
 104
 105                         return;
 106                 }
 107
 108                 $this->output( $title->getPrefixedDBkey() . " ..." );
 109                 $rev = Revision::newFromTitle( $title );
 110                 $currentRevId = $rev->getId();
 111
 112                 while ( $rev && ( $rev->isDeleted( Revision::DELETED_TEXT )
 113                         || LinkFilter::matchEntry( $rev->getContent( Revision::RAW ), $domain ) )
 114                 ) {
 115                         $rev = $rev->getPrevious();
 116                 }
 117
 118                 if ( $rev && $rev->getId() == $currentRevId ) {
 119                         // The regex didn't match the current article text
 120                         // This happens e.g. when a link comes from a template rather than the page itself
 121                         $this->output( "False match\n" );
 122                 } else {
 123                         $dbw = $this->getDB( DB_MASTER );
 124                         $this->beginTransaction( $dbw, __METHOD__ );
 125                         $page = WikiPage::factory( $title );
 126                         if ( $rev ) {
 127                                 // Revert to this revision
 128                                 $content = $rev->getContent( Revision::RAW );
 129
 130                                 $this->output( "reverting\n" );
 131                                 $page->doEditContent(
 132                                         $content,
 133                                         wfMessage( 'spam_reverting', $domain )->inContentLanguage()->text(),
 134                                         EDIT_UPDATE,
 135                                         $rev->getId()
 136                                 );
 137                         } elseif ( $this->hasOption( 'delete' ) ) {
 138                                 // Didn't find a non-spammy revision, blank the page
 139                                 $this->output( "deleting\n" );
 140                                 $page->doDeleteArticle(
 141                                         wfMessage( 'spam_deleting', $domain )->inContentLanguage()->text()
 142                                 );
 143                         } else {
 144                                 // Didn't find a non-spammy revision, blank the page
 145                                 $handler = ContentHandler::getForTitle( $title );
 146                                 $content = $handler->makeEmptyContent();
 147
 148                                 $this->output( "blanking\n" );
 149                                 $page->doEditContent(
 150                                         $content,
 151                                         wfMessage( 'spam_blanking', $domain )->inContentLanguage()->text()
 152                                 );
 153                         }
 154                         $this->commitTransaction( $dbw, __METHOD__ );
 155                 }
 156         }
 157 }
 158
 159 $maintClass = "CleanupSpam";
 160 require_once RUN_MAINTENANCE_IF_MAIN;