X-Git-Url: https://scripts.mit.edu/gitweb/autoinstalls/mediawiki.git/blobdiff_plain/51ff73d1c23dde7d9149040d018c42ae27b20a0c..HEAD:/maintenance/cleanupTitles.php diff --git a/maintenance/cleanupTitles.php b/maintenance/cleanupTitles.php index 4d76ac7a..50e17d8d 100644 --- a/maintenance/cleanupTitles.php +++ b/maintenance/cleanupTitles.php @@ -1,13 +1,9 @@ - * http://www.mediawiki.org/ + * Copyright © 2005 Brion Vibber + * https://www.mediawiki.org/ * * 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 @@ -29,125 +25,169 @@ * @ingroup Maintenance */ -require_once( 'commandLine.inc' ); -require_once( 'cleanupTable.inc' ); +use MediaWiki\MediaWikiServices; + +require_once __DIR__ . '/cleanupTable.inc'; /** + * Maintenance script to clean up broken, unparseable titles. + * * @ingroup Maintenance */ class TitleCleanup extends TableCleanup { - function __construct( $dryrun = false ) { - parent::__construct( 'page', $dryrun ); + public function __construct() { + parent::__construct(); + $this->addDescription( 'Script to clean up broken, unparseable titles' ); } - function processPage( $row ) { - $current = Title::makeTitle( $row->page_namespace, $row->page_title ); - $display = $current->getPrefixedText(); - - $verified = UtfNormal::cleanUp( $display ); - + /** + * @param object $row + */ + protected function processRow( $row ) { + global $wgContLang; + $display = Title::makeName( $row->page_namespace, $row->page_title ); + $verified = $wgContLang->normalize( $display ); $title = Title::newFromText( $verified ); - if( !is_null( $title ) && $title->equals( $current ) && $title->canExist() ) { - return $this->progress( 0 ); // all is fine + if ( !is_null( $title ) + && $title->canExist() + && $title->getNamespace() == $row->page_namespace + && $title->getDBkey() === $row->page_title + ) { + $this->progress( 0 ); // all is fine + + return; } - if( $row->page_namespace == NS_FILE && $this->fileExists( $row->page_title ) ) { - $this->log( "file $row->page_title needs cleanup, please run cleanupImages.php." ); - return $this->progress( 0 ); - } elseif( is_null( $title ) ) { - $this->log( "page $row->page_id ($display) is illegal." ); + if ( $row->page_namespace == NS_FILE && $this->fileExists( $row->page_title ) ) { + $this->output( "file $row->page_title needs cleanup, please run cleanupImages.php.\n" ); + $this->progress( 0 ); + } elseif ( is_null( $title ) ) { + $this->output( "page $row->page_id ($display) is illegal.\n" ); $this->moveIllegalPage( $row ); - return $this->progress( 1 ); + $this->progress( 1 ); } else { - $this->log( "page $row->page_id ($display) doesn't match self." ); + $this->output( "page $row->page_id ($display) doesn't match self.\n" ); $this->moveInconsistentPage( $row, $title ); - return $this->progress( 1 ); + $this->progress( 1 ); } } - function fileExists( $name ) { + /** + * @param string $name + * @return bool + */ + protected function fileExists( $name ) { // XXX: Doesn't actually check for file existence, just presence of image record. // This is reasonable, since cleanupImages.php only iterates over the image table. - $dbr = wfGetDB( DB_SLAVE ); - $row = $dbr->selectRow( 'image', array( 'img_name' ), array( 'img_name' => $name ), __METHOD__ ); + $dbr = $this->getDB( DB_REPLICA ); + $row = $dbr->selectRow( 'image', [ 'img_name' ], [ 'img_name' => $name ], __METHOD__ ); + return $row !== false; } - function moveIllegalPage( $row ) { + /** + * @param object $row + */ + protected function moveIllegalPage( $row ) { $legal = 'A-Za-z0-9_/\\\\-'; $legalized = preg_replace_callback( "!([^$legal])!", - array( &$this, 'hexChar' ), + [ $this, 'hexChar' ], $row->page_title ); - if( $legalized == '.' ) $legalized = '(dot)'; - if( $legalized == '_' ) $legalized = '(space)'; + if ( $legalized == '.' ) { + $legalized = '(dot)'; + } + if ( $legalized == '_' ) { + $legalized = '(space)'; + } $legalized = 'Broken/' . $legalized; $title = Title::newFromText( $legalized ); - if( is_null( $title ) ) { + if ( is_null( $title ) ) { $clean = 'Broken/id:' . $row->page_id; - $this->log( "Couldn't legalize; form '$legalized' still invalid; using '$clean'" ); + $this->output( "Couldn't legalize; form '$legalized' still invalid; using '$clean'\n" ); $title = Title::newFromText( $clean ); - } elseif( $title->exists() ) { + } elseif ( $title->exists() ) { $clean = 'Broken/id:' . $row->page_id; - $this->log( "Legalized for '$legalized' exists; using '$clean'" ); + $this->output( "Legalized for '$legalized' exists; using '$clean'\n" ); $title = Title::newFromText( $clean ); } $dest = $title->getDBkey(); - if( $this->dryrun ) { - $this->log( "DRY RUN: would rename $row->page_id ($row->page_namespace,'$row->page_title') to ($row->page_namespace,'$dest')" ); + if ( $this->dryrun ) { + $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace," . + "'$row->page_title') to ($row->page_namespace,'$dest')\n" ); } else { - $this->log( "renaming $row->page_id ($row->page_namespace,'$row->page_title') to ($row->page_namespace,'$dest')" ); - $dbw = wfGetDB( DB_MASTER ); + $this->output( "renaming $row->page_id ($row->page_namespace," . + "'$row->page_title') to ($row->page_namespace,'$dest')\n" ); + $dbw = $this->getDB( DB_MASTER ); $dbw->update( 'page', - array( 'page_title' => $dest ), - array( 'page_id' => $row->page_id ), - 'cleanupTitles::moveInconsistentPage' ); + [ 'page_title' => $dest ], + [ 'page_id' => $row->page_id ], + __METHOD__ ); } } - function moveInconsistentPage( $row, $title ) { - if( $title->exists() || $title->getInterwiki() ) { - if( $title->getInterwiki() ) { - $prior = $title->getPrefixedDbKey(); + /** + * @param object $row + * @param Title $title + */ + protected function moveInconsistentPage( $row, Title $title ) { + if ( $title->exists( Title::GAID_FOR_UPDATE ) + || $title->getInterwiki() + || !$title->canExist() + ) { + if ( $title->getInterwiki() || !$title->canExist() ) { + $prior = $title->getPrefixedDBkey(); } else { $prior = $title->getDBkey(); } + + # Old cleanupTitles could move articles there. See T25147. + $ns = $row->page_namespace; + if ( $ns < 0 ) { + $ns = 0; + } + + # Namespace which no longer exists. Put the page in the main namespace + # since we don't have any idea of the old namespace name. See T70501. + if ( !MWNamespace::exists( $ns ) ) { + $ns = 0; + } + $clean = 'Broken/' . $prior; - $verified = Title::makeTitleSafe( $row->page_namespace, $clean ); - if( $verified->exists() ) { + $verified = Title::makeTitleSafe( $ns, $clean ); + if ( !$verified || $verified->exists() ) { $blah = "Broken/id:" . $row->page_id; - $this->log( "Couldn't legalize; form '$clean' exists; using '$blah'" ); - $verified = Title::makeTitleSafe( $row->page_namespace, $blah ); + $this->output( "Couldn't legalize; form '$clean' exists; using '$blah'\n" ); + $verified = Title::makeTitleSafe( $ns, $blah ); } $title = $verified; } - if( is_null( $title ) ) { - wfDie( "Something awry; empty title.\n" ); + if ( is_null( $title ) ) { + $this->error( "Something awry; empty title.", true ); } $ns = $title->getNamespace(); $dest = $title->getDBkey(); - if( $this->dryrun ) { - $this->log( "DRY RUN: would rename $row->page_id ($row->page_namespace,'$row->page_title') to ($row->page_namespace,'$dest')" ); + + if ( $this->dryrun ) { + $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace," . + "'$row->page_title') to ($ns,'$dest')\n" ); } else { - $this->log( "renaming $row->page_id ($row->page_namespace,'$row->page_title') to ($ns,'$dest')" ); - $dbw = wfGetDB( DB_MASTER ); + $this->output( "renaming $row->page_id ($row->page_namespace," . + "'$row->page_title') to ($ns,'$dest')\n" ); + $dbw = $this->getDB( DB_MASTER ); $dbw->update( 'page', - array( + [ 'page_namespace' => $ns, 'page_title' => $dest - ), - array( 'page_id' => $row->page_id ), - 'cleanupTitles::moveInconsistentPage' ); - $linkCache = LinkCache::singleton(); - $linkCache->clear(); + ], + [ 'page_id' => $row->page_id ], + __METHOD__ ); + MediaWikiServices::getInstance()->getLinkCache()->clear(); } } } -$wgUser->setName( 'Conversion script' ); -$caps = new TitleCleanup( !isset( $options['fix'] ) ); -$caps->cleanup(); - - +$maintClass = "TitleCleanup"; +require_once RUN_MAINTENANCE_IF_MAIN;