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 Wikimedia\Rdbms\IMaintainableDatabase;
  34
  35 /**
  36  * Maintenance script that looks for 'orphan' revisions hooked to pages which
  37  * don't exist and 'childless' pages with no revisions.
  38  *
  39  * @ingroup Maintenance
  40  */
  41 class Orphans extends Maintenance {
  42         public function __construct() {
  43                 parent::__construct();
  44                 $this->addDescription( "Look for 'orphan' revisions hooked to pages which don't exist\n" .
  45                         "and 'childless' pages with no revisions\n" .
  46                         "Then, kill the poor widows and orphans\n" .
  47                         "Man this is depressing"
  48                 );
  49                 $this->addOption( 'fix', 'Actually fix broken entries' );
  50         }
  51
  52         public function execute() {
  53                 $this->checkOrphans( $this->hasOption( 'fix' ) );
  54                 $this->checkSeparation( $this->hasOption( 'fix' ) );
  55                 # Does not work yet, do not use
  56                 # $this->checkWidows( $this->hasOption( 'fix' ) );
  57         }
  58
  59         /**
  60          * Lock the appropriate tables for the script
  61          * @param IMaintainableDatabase $db
  62          * @param string[] $extraTable The name of any extra tables to lock (eg: text)
  63          */
  64         private function lockTables( $db, $extraTable = [] ) {
  65                 $tbls = [ 'page', 'revision', 'redirect' ];
  66                 if ( $extraTable ) {
  67                         $tbls = array_merge( $tbls, $extraTable );
  68                 }
  69                 $db->lockTables( [], $tbls, __METHOD__, false );
  70         }
  71
  72         /**
  73          * Check for orphan revisions
  74          * @param bool $fix Whether to fix broken revisions when found
  75          */
  76         private function checkOrphans( $fix ) {
  77                 $dbw = $this->getDB( DB_MASTER );
  78                 $commentStore = CommentStore::getStore();
  79
  80                 if ( $fix ) {
  81                         $this->lockTables( $dbw );
  82                 }
  83
  84                 $commentQuery = $commentStore->getJoin( 'rev_comment' );
  85                 $actorQuery = ActorMigration::newMigration()->getJoin( 'rev_user' );
  86
  87                 $this->output( "Checking for orphan revision table entries... "
  88                         . "(this may take a while on a large wiki)\n" );
  89                 $result = $dbw->select(
  90                         [ 'revision', 'page' ] + $commentQuery['tables'] + $actorQuery['tables'],
  91                         [ 'rev_id', 'rev_page', 'rev_timestamp' ] + $commentQuery['fields'] + $actorQuery['fields'],
  92                         [ 'page_id' => null ],
  93                         __METHOD__,
  94                         [],
  95                         [ 'page' => [ 'LEFT JOIN', [ 'rev_page=page_id' ] ] ] + $commentQuery['joins']
  96                                 + $actorQuery['joins']
  97                 );
  98                 $orphans = $result->numRows();
  99                 if ( $orphans > 0 ) {
 100                         global $wgContLang;
 101
 102                         $this->output( "$orphans orphan revisions...\n" );
 103                         $this->output( sprintf(
 104                                 "%10s %10s %14s %20s %s\n",
 105                                 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text', 'rev_comment'
 106                         ) );
 107
 108                         foreach ( $result as $row ) {
 109                                 $comment = $commentStore->getComment( 'rev_comment', $row )->text;
 110                                 if ( $comment !== '' ) {
 111                                         $comment = '(' . $wgContLang->truncate( $comment, 40 ) . ')';
 112                                 }
 113                                 $this->output( sprintf( "%10d %10d %14s %20s %s\n",
 114                                         $row->rev_id,
 115                                         $row->rev_page,
 116                                         $row->rev_timestamp,
 117                                         $wgContLang->truncate( $row->rev_user_text, 17 ),
 118                                         $comment ) );
 119                                 if ( $fix ) {
 120                                         $dbw->delete( 'revision', [ 'rev_id' => $row->rev_id ] );
 121                                 }
 122                         }
 123                         if ( !$fix ) {
 124                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 125                         }
 126                 } else {
 127                         $this->output( "No orphans! Yay!\n" );
 128                 }
 129
 130                 if ( $fix ) {
 131                         $dbw->unlockTables( __METHOD__ );
 132                 }
 133         }
 134
 135         /**
 136          * @param bool $fix
 137          * @todo DON'T USE THIS YET! It will remove entries which have children,
 138          *       but which aren't properly attached (eg if page_latest is bogus
 139          *       but valid revisions do exist)
 140          */
 141         private function checkWidows( $fix ) {
 142                 $dbw = $this->getDB( DB_MASTER );
 143                 $page = $dbw->tableName( 'page' );
 144                 $revision = $dbw->tableName( 'revision' );
 145
 146                 if ( $fix ) {
 147                         $this->lockTables( $dbw );
 148                 }
 149
 150                 $this->output( "\nChecking for childless page table entries... "
 151                         . "(this may take a while on a large wiki)\n" );
 152                 $result = $dbw->query( "
 153                         SELECT *
 154                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 155                         WHERE rev_id IS NULL
 156                 " );
 157                 $widows = $result->numRows();
 158                 if ( $widows > 0 ) {
 159                         $this->output( "$widows childless pages...\n" );
 160                         $this->output( sprintf( "%10s %11s %2s %s\n", 'page_id', 'page_latest', 'ns', 'page_title' ) );
 161                         foreach ( $result as $row ) {
 162                                 printf( "%10d %11d %2d %s\n",
 163                                         $row->page_id,
 164                                         $row->page_latest,
 165                                         $row->page_namespace,
 166                                         $row->page_title );
 167                                 if ( $fix ) {
 168                                         $dbw->delete( 'page', [ 'page_id' => $row->page_id ] );
 169                                 }
 170                         }
 171                         if ( !$fix ) {
 172                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 173                         }
 174                 } else {
 175                         $this->output( "No childless pages! Yay!\n" );
 176                 }
 177
 178                 if ( $fix ) {
 179                         $dbw->unlockTables( __METHOD__ );
 180                 }
 181         }
 182
 183         /**
 184          * Check for pages where page_latest is wrong
 185          * @param bool $fix Whether to fix broken entries
 186          */
 187         private function checkSeparation( $fix ) {
 188                 $dbw = $this->getDB( DB_MASTER );
 189                 $page = $dbw->tableName( 'page' );
 190                 $revision = $dbw->tableName( 'revision' );
 191
 192                 if ( $fix ) {
 193                         $this->lockTables( $dbw, [ 'user', 'text' ] );
 194                 }
 195
 196                 $this->output( "\nChecking for pages whose page_latest links are incorrect... "
 197                         . "(this may take a while on a large wiki)\n" );
 198                 $result = $dbw->query( "
 199                         SELECT *
 200                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 201                 " );
 202                 $found = 0;
 203                 foreach ( $result as $row ) {
 204                         $result2 = $dbw->query( "
 205                                 SELECT MAX(rev_timestamp) as max_timestamp
 206                                 FROM $revision
 207                                 WHERE rev_page=" . (int)( $row->page_id )
 208                         );
 209                         $row2 = $dbw->fetchObject( $result2 );
 210                         if ( $row2 ) {
 211                                 if ( $row->rev_timestamp != $row2->max_timestamp ) {
 212                                         if ( $found == 0 ) {
 213                                                 $this->output( sprintf( "%10s %10s %14s %14s\n",
 214                                                         'page_id', 'rev_id', 'timestamp', 'max timestamp' ) );
 215                                         }
 216                                         ++$found;
 217                                         $this->output( sprintf( "%10d %10d %14s %14s\n",
 218                                                 $row->page_id,
 219                                                 $row->page_latest,
 220                                                 $row->rev_timestamp,
 221                                                 $row2->max_timestamp ) );
 222                                         if ( $fix ) {
 223                                                 # ...
 224                                                 $maxId = $dbw->selectField(
 225                                                         'revision',
 226                                                         'rev_id',
 227                                                         [
 228                                                                 'rev_page' => $row->page_id,
 229                                                                 'rev_timestamp' => $row2->max_timestamp ] );
 230                                                 $this->output( "... updating to revision $maxId\n" );
 231                                                 $maxRev = Revision::newFromId( $maxId );
 232                                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 233                                                 $article = WikiPage::factory( $title );
 234                                                 $article->updateRevisionOn( $dbw, $maxRev );
 235                                         }
 236                                 }
 237                         } else {
 238                                 $this->output( "wtf\n" );
 239                         }
 240                 }
 241
 242                 if ( $found ) {
 243                         $this->output( "Found $found pages with incorrect latest revision.\n" );
 244                 } else {
 245                         $this->output( "No pages with incorrect latest revision. Yay!\n" );
 246                 }
 247                 if ( !$fix && $found > 0 ) {
 248                         $this->output( "Run again with --fix to remove these entries automatically.\n" );
 249                 }
 250
 251                 if ( $fix ) {
 252                         $dbw->unlockTables( __METHOD__ );
 253                 }
 254         }
 255 }
 256
 257 $maintClass = Orphans::class;
 258 require_once RUN_MAINTENANCE_IF_MAIN;