<?php
/**
- * Cleanup all spam from a given hostname
+ * Cleanup all spam from a given hostname.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* @ingroup Maintenance
*/
-require_once( dirname( __FILE__ ) . '/Maintenance.php' );
+require_once __DIR__ . '/Maintenance.php';
+/**
+ * Maintenance script to cleanup all spam from a given hostname.
+ *
+ * @ingroup Maintenance
+ */
class CleanupSpam extends Maintenance {
+
public function __construct() {
parent::__construct();
- $this->mDescription = "Cleanup all spam from a given hostname";
+ $this->addDescription( 'Cleanup all spam from a given hostname' );
$this->addOption( 'all', 'Check all wikis in $wgLocalDatabases' );
- $this->addArg( 'hostname', 'Hostname that was spamming' );
+ $this->addOption( 'delete', 'Delete pages containing only spam instead of blanking them' );
+ $this->addArg(
+ 'hostname',
+ 'Hostname that was spamming, single * wildcard in the beginning allowed'
+ );
}
public function execute() {
- global $wgLocalDatabases, $wgUser;
+ global $IP, $wgLocalDatabases, $wgUser;
- $username = wfMsg( 'spambot_username' );
- $wgUser = User::newFromName( $username );
+ $username = wfMessage( 'spambot_username' )->text();
+ $wgUser = User::newSystemUser( $username );
+ if ( !$wgUser ) {
+ $this->error( "Invalid username specified in 'spambot_username' message: $username", true );
+ }
// Create the user if necessary
if ( !$wgUser->getId() ) {
$wgUser->addToDatabase();
$this->output( "Finding spam on " . count( $wgLocalDatabases ) . " wikis\n" );
$found = false;
foreach ( $wgLocalDatabases as $wikiID ) {
- $dbr = wfGetDB( DB_SLAVE, array(), $wikiID );
+ $dbr = $this->getDB( DB_REPLICA, [], $wikiID );
$count = $dbr->selectField( 'externallinks', 'COUNT(*)',
- array( 'el_index' . $dbr->buildLike( $like ) ), __METHOD__ );
+ [ 'el_index' . $dbr->buildLike( $like ) ], __METHOD__ );
if ( $count ) {
$found = true;
- passthru( "php cleanupSpam.php --wiki='$wikiID' $spec | sed 's/^/$wikiID: /'" );
+ $cmd = wfShellWikiCmd( "$IP/maintenance/cleanupSpam.php",
+ [ '--wiki', $wikiID, $spec ] );
+ passthru( "$cmd | sed 's/^/$wikiID: /'" );
}
}
if ( $found ) {
} else {
// Clean up spam on this wiki
- $dbr = wfGetDB( DB_SLAVE );
- $res = $dbr->select( 'externallinks', array( 'DISTINCT el_from' ),
- array( 'el_index' . $dbr->buildLike( $like ) ), __METHOD__ );
+ $dbr = $this->getDB( DB_REPLICA );
+ $res = $dbr->select( 'externallinks', [ 'DISTINCT el_from' ],
+ [ 'el_index' . $dbr->buildLike( $like ) ], __METHOD__ );
$count = $dbr->numRows( $res );
$this->output( "Found $count articles containing $spec\n" );
foreach ( $res as $row ) {
$title = Title::newFromID( $id );
if ( !$title ) {
$this->error( "Internal error: no page for ID $id" );
+
return;
}
$this->output( $title->getPrefixedDBkey() . " ..." );
$rev = Revision::newFromTitle( $title );
- $revId = $rev->getId();
- $currentRevId = $revId;
-
- while ( $rev && LinkFilter::matchEntry( $rev->getText() , $domain ) ) {
- # Revision::getPrevious can't be used in this way before MW 1.6 (Revision.php 1.26)
- # $rev = $rev->getPrevious();
- $revId = $title->getPreviousRevisionID( $revId );
- if ( $revId ) {
- $rev = Revision::newFromTitle( $title, $revId );
- } else {
- $rev = false;
- }
+ $currentRevId = $rev->getId();
+
+ while ( $rev && ( $rev->isDeleted( Revision::DELETED_TEXT )
+ || LinkFilter::matchEntry( $rev->getContent( Revision::RAW ), $domain ) )
+ ) {
+ $rev = $rev->getPrevious();
}
- if ( $revId == $currentRevId ) {
+
+ if ( $rev && $rev->getId() == $currentRevId ) {
// The regex didn't match the current article text
// This happens e.g. when a link comes from a template rather than the page itself
$this->output( "False match\n" );
} else {
- $dbw = wfGetDB( DB_MASTER );
- $dbw->begin();
- if ( !$rev ) {
- // Didn't find a non-spammy revision, blank the page
- $this->output( "blanking\n" );
- $article = new Article( $title );
- $article->doEdit( '', wfMsg( 'spam_blanking', $domain ) );
- } else {
+ $dbw = $this->getDB( DB_MASTER );
+ $this->beginTransaction( $dbw, __METHOD__ );
+ $page = WikiPage::factory( $title );
+ if ( $rev ) {
// Revert to this revision
+ $content = $rev->getContent( Revision::RAW );
+
$this->output( "reverting\n" );
- $article = new Article( $title );
- $article->doEdit( $rev->getText(), wfMsg( 'spam_reverting', $domain ), EDIT_UPDATE );
+ $page->doEditContent(
+ $content,
+ wfMessage( 'spam_reverting', $domain )->inContentLanguage()->text(),
+ EDIT_UPDATE,
+ $rev->getId()
+ );
+ } elseif ( $this->hasOption( 'delete' ) ) {
+ // Didn't find a non-spammy revision, blank the page
+ $this->output( "deleting\n" );
+ $page->doDeleteArticle(
+ wfMessage( 'spam_deleting', $domain )->inContentLanguage()->text()
+ );
+ } else {
+ // Didn't find a non-spammy revision, blank the page
+ $handler = ContentHandler::getForTitle( $title );
+ $content = $handler->makeEmptyContent();
+
+ $this->output( "blanking\n" );
+ $page->doEditContent(
+ $content,
+ wfMessage( 'spam_blanking', $domain )->inContentLanguage()->text()
+ );
}
- $dbw->commit();
- wfDoUpdates();
+ $this->commitTransaction( $dbw, __METHOD__ );
}
}
}
$maintClass = "CleanupSpam";
-require_once( RUN_MAINTENANCE_IF_MAIN );
+require_once RUN_MAINTENANCE_IF_MAIN;