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 use MediaWiki\Revision\RevisionRecord;
  25
  26 require_once __DIR__ . '/Maintenance.php';
  27
  28 /**
  29  * Maintenance script to cleanup all spam from a given hostname.
  30  *
  31  * @ingroup Maintenance
  32  */
  33 class CleanupSpam extends Maintenance {
  34
  35         public function __construct() {
  36                 parent::__construct();
  37                 $this->addDescription( 'Cleanup all spam from a given hostname' );
  38                 $this->addOption( 'all', 'Check all wikis in $wgLocalDatabases' );
  39                 $this->addOption( 'delete', 'Delete pages containing only spam instead of blanking them' );
  40                 $this->addArg(
  41                         'hostname',
  42                         'Hostname that was spamming, single * wildcard in the beginning allowed'
  43                 );
  44         }
  45
  46         public function execute() {
  47                 global $IP, $wgLocalDatabases, $wgUser;
  48
  49                 $username = wfMessage( 'spambot_username' )->text();
  50                 $wgUser = User::newSystemUser( $username );
  51                 if ( !$wgUser ) {
  52                         $this->fatalError( "Invalid username specified in 'spambot_username' message: $username" );
  53                 }
  54                 // Hack: Grant bot rights so we don't flood RecentChanges
  55                 $wgUser->addGroup( 'bot' );
  56
  57                 $spec = $this->getArg( 0 );
  58
  59                 $protConds = [];
  60                 foreach ( [ 'http://', 'https://' ] as $prot ) {
  61                         $conds = LinkFilter::getQueryConditions( $spec, [ 'protocol' => $prot ] );
  62                         if ( !$conds ) {
  63                                 $this->fatalError( "Not a valid hostname specification: $spec" );
  64                         }
  65                         $protConds[$prot] = $conds;
  66                 }
  67
  68                 if ( $this->hasOption( 'all' ) ) {
  69                         // Clean up spam on all wikis
  70                         $this->output( "Finding spam on " . count( $wgLocalDatabases ) . " wikis\n" );
  71                         $found = false;
  72                         foreach ( $wgLocalDatabases as $wikiId ) {
  73                                 /** @var Database $dbr */
  74                                 $dbr = $this->getDB( DB_REPLICA, [], $wikiId );
  75
  76                                 foreach ( $protConds as $conds ) {
  77                                         $count = $dbr->selectField(
  78                                                 'externallinks',
  79                                                 'COUNT(*)',
  80                                                 $conds,
  81                                                 __METHOD__
  82                                         );
  83                                         if ( $count ) {
  84                                                 $found = true;
  85                                                 $cmd = wfShellWikiCmd(
  86                                                         "$IP/maintenance/cleanupSpam.php",
  87                                                         [ '--wiki', $wikiId, $spec ]
  88                                                 );
  89                                                 passthru( "$cmd | sed 's/^/$wikiId:  /'" );
  90                                         }
  91                                 }
  92                         }
  93                         if ( $found ) {
  94                                 $this->output( "All done\n" );
  95                         } else {
  96                                 $this->output( "None found\n" );
  97                         }
  98                 } else {
  99                         // Clean up spam on this wiki
 100
 101                         $count = 0;
 102                         /** @var Database $dbr */
 103                         $dbr = $this->getDB( DB_REPLICA );
 104                         foreach ( $protConds as $prot => $conds ) {
 105                                 $res = $dbr->select(
 106                                         'externallinks',
 107                                         [ 'DISTINCT el_from' ],
 108                                         $conds,
 109                                         __METHOD__
 110                                 );
 111                                 $count = $dbr->numRows( $res );
 112                                 $this->output( "Found $count articles containing $spec\n" );
 113                                 foreach ( $res as $row ) {
 114                                         $this->cleanupArticle( $row->el_from, $spec, $prot );
 115                                 }
 116                         }
 117                         if ( $count ) {
 118                                 $this->output( "Done\n" );
 119                         }
 120                 }
 121         }
 122
 123         /**
 124          * @param int $id
 125          * @param string $domain
 126          * @param string $protocol
 127          * @throws MWException
 128          */
 129         private function cleanupArticle( $id, $domain, $protocol ) {
 130                 $title = Title::newFromID( $id );
 131                 if ( !$title ) {
 132                         $this->error( "Internal error: no page for ID $id" );
 133
 134                         return;
 135                 }
 136
 137                 $this->output( $title->getPrefixedDBkey() . " ..." );
 138                 $rev = Revision::newFromTitle( $title );
 139                 $currentRevId = $rev->getId();
 140
 141                 while ( $rev && ( $rev->isDeleted( RevisionRecord::DELETED_TEXT )
 142                         || LinkFilter::matchEntry( $rev->getContent( RevisionRecord::RAW ), $domain, $protocol ) )
 143                 ) {
 144                         $rev = $rev->getPrevious();
 145                 }
 146
 147                 if ( $rev && $rev->getId() == $currentRevId ) {
 148                         // The regex didn't match the current article text
 149                         // This happens e.g. when a link comes from a template rather than the page itself
 150                         $this->output( "False match\n" );
 151                 } else {
 152                         $dbw = $this->getDB( DB_MASTER );
 153                         $this->beginTransaction( $dbw, __METHOD__ );
 154                         $page = WikiPage::factory( $title );
 155                         if ( $rev ) {
 156                                 // Revert to this revision
 157                                 $content = $rev->getContent( RevisionRecord::RAW );
 158
 159                                 $this->output( "reverting\n" );
 160                                 $page->doEditContent(
 161                                         $content,
 162                                         wfMessage( 'spam_reverting', $domain )->inContentLanguage()->text(),
 163                                         EDIT_UPDATE | EDIT_FORCE_BOT,
 164                                         $rev->getId()
 165                                 );
 166                         } elseif ( $this->hasOption( 'delete' ) ) {
 167                                 // Didn't find a non-spammy revision, blank the page
 168                                 $this->output( "deleting\n" );
 169                                 $page->doDeleteArticle(
 170                                         wfMessage( 'spam_deleting', $domain )->inContentLanguage()->text()
 171                                 );
 172                         } else {
 173                                 // Didn't find a non-spammy revision, blank the page
 174                                 $handler = ContentHandler::getForTitle( $title );
 175                                 $content = $handler->makeEmptyContent();
 176
 177                                 $this->output( "blanking\n" );
 178                                 $page->doEditContent(
 179                                         $content,
 180                                         wfMessage( 'spam_blanking', $domain )->inContentLanguage()->text(),
 181                                         EDIT_UPDATE | EDIT_FORCE_BOT
 182                                 );
 183                         }
 184                         $this->commitTransaction( $dbw, __METHOD__ );
 185                 }
 186         }
 187 }
 188
 189 $maintClass = CleanupSpam::class;
 190 require_once RUN_MAINTENANCE_IF_MAIN;