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->mDescription = "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( 'hostname', 'Hostname that was spamming, single * wildcard in the beginning allowed' );
  39         }
  40
  41         public function execute() {
  42                 global $wgLocalDatabases, $wgUser;
  43
  44                 $username = wfMessage( 'spambot_username' )->text();
  45                 $wgUser = User::newFromName( $username );
  46                 if ( !$wgUser ) {
  47                         $this->error( "Invalid username specified in 'spambot_username' message: $username", true );
  48                 }
  49                 // Create the user if necessary
  50                 if ( !$wgUser->getId() ) {
  51                         $wgUser->addToDatabase();
  52                 }
  53                 $spec = $this->getArg();
  54                 $like = LinkFilter::makeLikeArray( $spec );
  55                 if ( !$like ) {
  56                         $this->error( "Not a valid hostname specification: $spec", true );
  57                 }
  58
  59                 if ( $this->hasOption( 'all' ) ) {
  60                         // Clean up spam on all wikis
  61                         $this->output( "Finding spam on " . count( $wgLocalDatabases ) . " wikis\n" );
  62                         $found = false;
  63                         foreach ( $wgLocalDatabases as $wikiID ) {
  64                                 $dbr = wfGetDB( DB_SLAVE, array(), $wikiID );
  65
  66                                 $count = $dbr->selectField( 'externallinks', 'COUNT(*)',
  67                                         array( 'el_index' . $dbr->buildLike( $like ) ), __METHOD__ );
  68                                 if ( $count ) {
  69                                         $found = true;
  70                                         passthru( "php cleanupSpam.php --wiki='$wikiID' $spec | sed 's/^/$wikiID:  /'" );
  71                                 }
  72                         }
  73                         if ( $found ) {
  74                                 $this->output( "All done\n" );
  75                         } else {
  76                                 $this->output( "None found\n" );
  77                         }
  78                 } else {
  79                         // Clean up spam on this wiki
  80
  81                         $dbr = wfGetDB( DB_SLAVE );
  82                         $res = $dbr->select( 'externallinks', array( 'DISTINCT el_from' ),
  83                                 array( 'el_index' . $dbr->buildLike( $like ) ), __METHOD__ );
  84                         $count = $dbr->numRows( $res );
  85                         $this->output( "Found $count articles containing $spec\n" );
  86                         foreach ( $res as $row ) {
  87                                 $this->cleanupArticle( $row->el_from, $spec );
  88                         }
  89                         if ( $count ) {
  90                                 $this->output( "Done\n" );
  91                         }
  92                 }
  93         }
  94
  95         private function cleanupArticle( $id, $domain ) {
  96                 $title = Title::newFromID( $id );
  97                 if ( !$title ) {
  98                         $this->error( "Internal error: no page for ID $id" );
  99                         return;
 100                 }
 101
 102                 $this->output( $title->getPrefixedDBkey() . " ..." );
 103                 $rev = Revision::newFromTitle( $title );
 104                 $currentRevId = $rev->getId();
 105
 106                 while ( $rev && ( $rev->isDeleted( Revision::DELETED_TEXT )
 107                                                 || LinkFilter::matchEntry( $rev->getContent( Revision::RAW ), $domain ) ) ) {
 108                         $rev = $rev->getPrevious();
 109                 }
 110
 111                 if ( $rev && $rev->getId() == $currentRevId ) {
 112                         // The regex didn't match the current article text
 113                         // This happens e.g. when a link comes from a template rather than the page itself
 114                         $this->output( "False match\n" );
 115                 } else {
 116                         $dbw = wfGetDB( DB_MASTER );
 117                         $dbw->begin( __METHOD__ );
 118                         $page = WikiPage::factory( $title );
 119                         if ( $rev ) {
 120                                 // Revert to this revision
 121                                 $content = $rev->getContent( Revision::RAW );
 122
 123                                 $this->output( "reverting\n" );
 124                                 $page->doEditContent( $content, wfMessage( 'spam_reverting', $domain )->inContentLanguage()->text(),
 125                                         EDIT_UPDATE, $rev->getId() );
 126                         } elseif ( $this->hasOption( 'delete' ) ) {
 127                                 // Didn't find a non-spammy revision, blank the page
 128                                 $this->output( "deleting\n" );
 129                                 $page->doDeleteArticle( wfMessage( 'spam_deleting', $domain )->inContentLanguage()->text() );
 130                         } else {
 131                                 // Didn't find a non-spammy revision, blank the page
 132                                 $handler = ContentHandler::getForTitle( $title );
 133                                 $content = $handler->makeEmptyContent();
 134
 135                                 $this->output( "blanking\n" );
 136                                 $page->doEditContent( $content, wfMessage( 'spam_blanking', $domain )->inContentLanguage()->text() );
 137                         }
 138                         $dbw->commit( __METHOD__ );
 139                 }
 140         }
 141 }
 142
 143 $maintClass = "CleanupSpam";
 144 require_once( RUN_MAINTENANCE_IF_MAIN );