maintenance/orphans.php

   1 <?php
   2 /**
   3  * Look for 'orphan' revisions hooked to pages which don't exist and
   4  * 'childless' pages with no revisions.
   5  * Then, kill the poor widows and orphans.
   6  * Man this is depressing.
   7  *
   8  * Copyright © 2005 Brion Vibber <brion@pobox.com>
   9  * https://www.mediawiki.org/
  10  *
  11  * This program is free software; you can redistribute it and/or modify
  12  * it under the terms of the GNU General Public License as published by
  13  * the Free Software Foundation; either version 2 of the License, or
  14  * (at your option) any later version.
  15  *
  16  * This program is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  19  * GNU General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU General Public License along
  22  * with this program; if not, write to the Free Software Foundation, Inc.,
  23  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  24  * http://www.gnu.org/copyleft/gpl.html
  25  *
  26  * @file
  27  * @author <brion@pobox.com>
  28  * @ingroup Maintenance
  29  */
  30
  31 require_once __DIR__ . '/Maintenance.php';
  32
  33 use MediaWiki\MediaWikiServices;
  34 use Wikimedia\Rdbms\IMaintainableDatabase;
  35
  36 /**
  37  * Maintenance script that looks for 'orphan' revisions hooked to pages which
  38  * don't exist and 'childless' pages with no revisions.
  39  *
  40  * @ingroup Maintenance
  41  */
  42 class Orphans extends Maintenance {
  43         public function __construct() {
  44                 parent::__construct();
  45                 $this->addDescription( "Look for 'orphan' revisions hooked to pages which don't exist\n" .
  46                         "and 'childless' pages with no revisions\n" .
  47                         "Then, kill the poor widows and orphans\n" .
  48                         "Man this is depressing"
  49                 );
  50                 $this->addOption( 'fix', 'Actually fix broken entries' );
  51         }
  52
  53         public function execute() {
  54                 $this->checkOrphans( $this->hasOption( 'fix' ) );
  55                 $this->checkSeparation( $this->hasOption( 'fix' ) );
  56         }
  57
  58         /**
  59          * Lock the appropriate tables for the script
  60          * @param IMaintainableDatabase $db
  61          * @param string[] $extraTable The name of any extra tables to lock (eg: text)
  62          */
  63         private function lockTables( $db, $extraTable = [] ) {
  64                 $tbls = [ 'page', 'revision', 'redirect' ];
  65                 if ( $extraTable ) {
  66                         $tbls = array_merge( $tbls, $extraTable );
  67                 }
  68                 $db->lockTables( [], $tbls, __METHOD__ );
  69         }
  70
  71         /**
  72          * Check for orphan revisions
  73          * @param bool $fix Whether to fix broken revisions when found
  74          */
  75         private function checkOrphans( $fix ) {
  76                 $dbw = $this->getDB( DB_MASTER );
  77                 $commentStore = CommentStore::getStore();
  78
  79                 if ( $fix ) {
  80                         $this->lockTables( $dbw );
  81                 }
  82
  83                 $commentQuery = $commentStore->getJoin( 'rev_comment' );
  84                 $actorQuery = ActorMigration::newMigration()->getJoin( 'rev_user' );
  85
  86                 $this->output( "Checking for orphan revision table entries... "
  87                         . "(this may take a while on a large wiki)\n" );
  88                 $result = $dbw->select(
  89                         [ 'revision', 'page' ] + $commentQuery['tables'] + $actorQuery['tables'],
  90                         [ 'rev_id', 'rev_page', 'rev_timestamp' ] + $commentQuery['fields'] + $actorQuery['fields'],
  91                         [ 'page_id' => null ],
  92                         __METHOD__,
  93                         [],
  94                         [ 'page' => [ 'LEFT JOIN', [ 'rev_page=page_id' ] ] ] + $commentQuery['joins']
  95                                 + $actorQuery['joins']
  96                 );
  97                 $orphans = $result->numRows();
  98                 if ( $orphans > 0 ) {
  99                         $this->output( "$orphans orphan revisions...\n" );
 100                         $this->output( sprintf(
 101                                 "%10s %10s %14s %20s %s\n",
 102                                 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text', 'rev_comment'
 103                         ) );
 104
 105                         $contLang = MediaWikiServices::getInstance()->getContentLanguage();
 106                         foreach ( $result as $row ) {
 107                                 $comment = $commentStore->getComment( 'rev_comment', $row )->text;
 108                                 if ( $comment !== '' ) {
 109                                         $comment = '(' . $contLang->truncateForVisual( $comment, 40 ) . ')';
 110                                 }
 111                                 $rev_user_text = $contLang->truncateForVisual( $row->rev_user_text, 20 );
 112                                 # pad $rev_user_text to 20 characters.  Note that this may
 113                                 # yield poor results if $rev_user_text contains combining
 114                                 # or half-width characters.  Alas.
 115                                 if ( mb_strlen( $rev_user_text ) < 20 ) {
 116                                         $rev_user_text = str_repeat( ' ', 20 - mb_strlen( $rev_user_text ) );
 117                                 }
 118                                 $this->output( sprintf( "%10d %10d %14s %s %s\n",
 119                                         $row->rev_id,
 120                                         $row->rev_page,
 121                                         $row->rev_timestamp,
 122                                         $rev_user_text,
 123                                         $comment ) );
 124                                 if ( $fix ) {
 125                                         $dbw->delete( 'revision', [ 'rev_id' => $row->rev_id ] );
 126                                 }
 127                         }
 128                         if ( !$fix ) {
 129                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 130                         }
 131                 } else {
 132                         $this->output( "No orphans! Yay!\n" );
 133                 }
 134
 135                 if ( $fix ) {
 136                         $dbw->unlockTables( __METHOD__ );
 137                 }
 138         }
 139
 140         /**
 141          * Check for pages where page_latest is wrong
 142          * @param bool $fix Whether to fix broken entries
 143          */
 144         private function checkSeparation( $fix ) {
 145                 $dbw = $this->getDB( DB_MASTER );
 146                 $page = $dbw->tableName( 'page' );
 147                 $revision = $dbw->tableName( 'revision' );
 148
 149                 if ( $fix ) {
 150                         $this->lockTables( $dbw, [ 'user', 'text' ] );
 151                 }
 152
 153                 $this->output( "\nChecking for pages whose page_latest links are incorrect... "
 154                         . "(this may take a while on a large wiki)\n" );
 155                 $result = $dbw->query( "
 156                         SELECT *
 157                         FROM $page LEFT JOIN $revision ON page_latest=rev_id
 158                 " );
 159                 $found = 0;
 160                 foreach ( $result as $row ) {
 161                         $result2 = $dbw->query( "
 162                                 SELECT MAX(rev_timestamp) as max_timestamp
 163                                 FROM $revision
 164                                 WHERE rev_page=" . (int)( $row->page_id )
 165                         );
 166                         $row2 = $dbw->fetchObject( $result2 );
 167                         if ( $row2 ) {
 168                                 if ( $row->rev_timestamp != $row2->max_timestamp ) {
 169                                         if ( $found == 0 ) {
 170                                                 $this->output( sprintf( "%10s %10s %14s %14s\n",
 171                                                         'page_id', 'rev_id', 'timestamp', 'max timestamp' ) );
 172                                         }
 173                                         ++$found;
 174                                         $this->output( sprintf( "%10d %10d %14s %14s\n",
 175                                                 $row->page_id,
 176                                                 $row->page_latest,
 177                                                 $row->rev_timestamp,
 178                                                 $row2->max_timestamp ) );
 179                                         if ( $fix ) {
 180                                                 # ...
 181                                                 $maxId = $dbw->selectField(
 182                                                         'revision',
 183                                                         'rev_id',
 184                                                         [
 185                                                                 'rev_page' => $row->page_id,
 186                                                                 'rev_timestamp' => $row2->max_timestamp ] );
 187                                                 $this->output( "... updating to revision $maxId\n" );
 188                                                 $maxRev = Revision::newFromId( $maxId );
 189                                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 190                                                 $article = WikiPage::factory( $title );
 191                                                 $article->updateRevisionOn( $dbw, $maxRev );
 192                                         }
 193                                 }
 194                         } else {
 195                                 $this->output( "wtf\n" );
 196                         }
 197                 }
 198
 199                 if ( $found ) {
 200                         $this->output( "Found $found pages with incorrect latest revision.\n" );
 201                 } else {
 202                         $this->output( "No pages with incorrect latest revision. Yay!\n" );
 203                 }
 204                 if ( !$fix && $found > 0 ) {
 205                         $this->output( "Run again with --fix to remove these entries automatically.\n" );
 206                 }
 207
 208                 if ( $fix ) {
 209                         $dbw->unlockTables( __METHOD__ );
 210                 }
 211         }
 212 }
 213
 214 $maintClass = Orphans::class;
 215 require_once RUN_MAINTENANCE_IF_MAIN;