Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
0.00% |
0 / 68 |
|
0.00% |
0 / 3 |
CRAP | |
0.00% |
0 / 1 |
| Cleanup | |
0.00% |
0 / 62 |
|
0.00% |
0 / 3 |
420 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
2 | |||
| execute | |
0.00% |
0 / 34 |
|
0.00% |
0 / 1 |
132 | |||
| cleanupArticle | |
0.00% |
0 / 22 |
|
0.00% |
0 / 1 |
72 | |||
| 1 | <?php |
| 2 | /** |
| 3 | * An aggressive spam cleanup script. |
| 4 | * Searches the database for matching pages, and reverts them to |
| 5 | * the last non-spammed revision. |
| 6 | * If all revisions contain spam, blanks the page |
| 7 | */ |
| 8 | |
| 9 | use MediaWiki\Extension\SpamBlacklist\BaseBlacklist; |
| 10 | use MediaWiki\MediaWikiServices; |
| 11 | use MediaWiki\Page\WikiPageFactory; |
| 12 | use MediaWiki\Revision\RevisionLookup; |
| 13 | use MediaWiki\Revision\RevisionRecord; |
| 14 | use MediaWiki\Revision\SlotRecord; |
| 15 | use MediaWiki\Title\Title; |
| 16 | use MediaWiki\Title\TitleFormatter; |
| 17 | use MediaWiki\User\User; |
| 18 | |
| 19 | $IP = getenv( 'MW_INSTALL_PATH' ); |
| 20 | if ( $IP === false ) { |
| 21 | $IP = __DIR__ . '/../../..'; |
| 22 | } |
| 23 | require_once "$IP/maintenance/Maintenance.php"; |
| 24 | |
| 25 | class Cleanup extends Maintenance { |
| 26 | /** @var RevisionLookup */ |
| 27 | private $revisionLookup; |
| 28 | /** @var TitleFormatter */ |
| 29 | private $titleFormatter; |
| 30 | /** @var WikiPageFactory */ |
| 31 | private $wikiPageFactory; |
| 32 | |
| 33 | public function __construct() { |
| 34 | parent::__construct(); |
| 35 | $this->revisionLookup = MediaWikiServices::getInstance()->getRevisionLookup(); |
| 36 | $this->titleFormatter = MediaWikiServices::getInstance()->getTitleFormatter(); |
| 37 | $this->wikiPageFactory = MediaWikiServices::getInstance()->getWikiPageFactory(); |
| 38 | |
| 39 | $this->requireExtension( 'SpamBlacklist' ); |
| 40 | $this->addOption( 'dry-run', 'Only do a dry run' ); |
| 41 | } |
| 42 | |
| 43 | public function execute() { |
| 44 | $user = User::newSystemUser( 'Spam cleanup script', [ 'steal' => true ] ); |
| 45 | |
| 46 | $sb = BaseBlacklist::getSpamBlacklist(); |
| 47 | $regexes = $sb->getBlacklists(); |
| 48 | if ( !$regexes ) { |
| 49 | $this->fatalError( "Invalid regex, can't clean up spam" ); |
| 50 | } |
| 51 | $dryRun = $this->hasOption( 'dry-run' ); |
| 52 | |
| 53 | $dbr = $this->getReplicaDB(); |
| 54 | $maxID = (int)$dbr->newSelectQueryBuilder() |
| 55 | ->select( 'MAX(page_id)' ) |
| 56 | ->from( 'page' ) |
| 57 | ->caller( __METHOD__ ) |
| 58 | ->fetchField(); |
| 59 | $reportingInterval = 100; |
| 60 | |
| 61 | $this->output( "Regexes are " . implode( ', ', array_map( 'strlen', $regexes ) ) . " bytes\n" ); |
| 62 | $this->output( "Searching for spam in $maxID pages...\n" ); |
| 63 | if ( $dryRun ) { |
| 64 | $this->output( "Dry run only\n" ); |
| 65 | } |
| 66 | |
| 67 | for ( $id = 1; $id <= $maxID; $id++ ) { |
| 68 | if ( $id % $reportingInterval == 0 ) { |
| 69 | printf( "%-8d %-5.2f%%\r", $id, $id / $maxID * 100 ); |
| 70 | } |
| 71 | $revision = $this->revisionLookup->getRevisionByPageId( $id ); |
| 72 | if ( $revision ) { |
| 73 | $content = $revision->getContent( SlotRecord::MAIN ); |
| 74 | $text = ( $content instanceof TextContent ) ? $content->getText() : null; |
| 75 | if ( $text ) { |
| 76 | foreach ( $regexes as $regex ) { |
| 77 | if ( preg_match( $regex, $text, $matches ) ) { |
| 78 | $titleText = $this->titleFormatter->getPrefixedText( $revision->getPageAsLinkTarget() ); |
| 79 | if ( $dryRun ) { |
| 80 | $this->output( "Found spam in [[$titleText]]\n" ); |
| 81 | } else { |
| 82 | $this->output( "Cleaning up links to {$matches[0]} in [[$titleText]]\n" ); |
| 83 | $match = str_replace( 'http://', '', $matches[0] ); |
| 84 | $this->cleanupArticle( $revision, $regexes, $match, $user ); |
| 85 | } |
| 86 | } |
| 87 | } |
| 88 | } |
| 89 | } |
| 90 | } |
| 91 | // Just for satisfaction |
| 92 | printf( "%-8d %-5.2f%%\n", $id - 1, ( $id - 1 ) / $maxID * 100 ); |
| 93 | } |
| 94 | |
| 95 | /** |
| 96 | * Find the latest revision of the article that does not contain spam and revert to it |
| 97 | * @param RevisionRecord $rev |
| 98 | * @param array $regexes |
| 99 | * @param string $match |
| 100 | * @param User $user |
| 101 | */ |
| 102 | private function cleanupArticle( RevisionRecord $rev, $regexes, $match, User $user ) { |
| 103 | $title = Title::newFromLinkTarget( $rev->getPageAsLinkTarget() ); |
| 104 | while ( $rev ) { |
| 105 | $matches = false; |
| 106 | $content = $rev->getContent( SlotRecord::MAIN ); |
| 107 | foreach ( $regexes as $regex ) { |
| 108 | $matches = $matches |
| 109 | || preg_match( |
| 110 | $regex, |
| 111 | ( $content instanceof TextContent ) ? $content->getText() : null |
| 112 | ); |
| 113 | } |
| 114 | if ( !$matches ) { |
| 115 | // Didn't find any spam |
| 116 | break; |
| 117 | } |
| 118 | |
| 119 | $rev = $this->revisionLookup->getPreviousRevision( $rev ); |
| 120 | } |
| 121 | if ( !$rev ) { |
| 122 | // Didn't find a non-spammy revision, blank the page |
| 123 | $this->output( "All revisions are spam, blanking...\n" ); |
| 124 | $content = ContentHandler::makeContent( '', $title ); |
| 125 | $comment = "All revisions matched the spam blacklist ($match), blanking"; |
| 126 | } else { |
| 127 | // Revert to this revision |
| 128 | $content = $rev->getContent( SlotRecord::MAIN ) ?: |
| 129 | ContentHandler::makeContent( '', $title ); |
| 130 | $comment = "Cleaning up links to $match"; |
| 131 | } |
| 132 | $wikiPage = $this->wikiPageFactory->newFromTitle( $title ); |
| 133 | $wikiPage->doUserEditContent( $content, $user, $comment ); |
| 134 | } |
| 135 | } |
| 136 | |
| 137 | $maintClass = Cleanup::class; |
| 138 | require_once RUN_MAINTENANCE_IF_MAIN; |