maintenance/orphans.php

   1 <?php
   2 /**
   3  * Look for 'orphan' revisions hooked to pages which don't exist
   4  * And 'childless' pages with no revisions.
   5  * Then, kill the poor widows and orphans.
   6  * Man this is depressing.
   7  *
   8  * Copyright (C) 2005 Brion Vibber <brion@pobox.com>
   9  * http://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  * @author <brion@pobox.com>
  27  * @ingroup Maintenance
  28  */
  29
  30 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  31
  32 class Orphans extends Maintenance {
  33         public function __construct() {
  34                 parent::__construct();
  35                 $this->mDescription = "Look for 'orphan' revisions hooked to pages which don't exist\n" .
  36                                                                 "And 'childless' pages with no revisions\n" .
  37                                                                 "Then, kill the poor widows and orphans\n" .
  38                                                                 "Man this is depressing";
  39                 $this->addOption( 'fix', 'Actually fix broken entries' );
  40         }
  41
  42         public function execute() {
  43                 global $wgTitle;
  44                 $wgTitle = Title::newFromText( 'Orphan revision cleanup script' );
  45                 $this->checkOrphans( $this->hasOption( 'fix' ) );
  46                 $this->checkSeparation( $this->hasOption( 'fix' ) );
  47                 # Does not work yet, do not use
  48                 # $this->checkWidows( $this->hasOption( 'fix' ) );
  49         }
  50
  51         /**
  52          * Lock the appropriate tables for the script
  53          * @param $db Database object
  54          * @param $extraTable String The name of any extra tables to lock (eg: text)
  55          */
  56         private function lockTables( &$db, $extraTable = null ) {
  57                 $tbls = array( 'page', 'revision', 'redirect' );
  58                 if ( $extraTable )
  59                         $tbls[] = $extraTable;
  60                 $db->lockTables( array(), $tbls, __METHOD__, false );
  61         }
  62
  63         /**
  64          * Check for orphan revisions
  65          * @param $fix bool Whether to fix broken revisions when found
  66          */
  67         private function checkOrphans( $fix ) {
  68                 $dbw = wfGetDB( DB_MASTER );
  69                 $page = $dbw->tableName( 'page' );
  70                 $revision = $dbw->tableName( 'revision' );
  71
  72                 if ( $fix ) {
  73                         $this->lockTables( $dbw );
  74                 }
  75
  76                 $this->output( "Checking for orphan revision table entries... (this may take a while on a large wiki)\n" );
  77                 $result = $dbw->query( "
  78                         SELECT *
  79                         FROM $revision LEFT OUTER JOIN $page ON rev_page=page_id
  80                         WHERE page_id IS NULL
  81                 " );
  82                 $orphans = $dbw->numRows( $result );
  83                 if ( $orphans > 0 ) {
  84                         global $wgContLang;
  85                         $this->output( "$orphans orphan revisions...\n" );
  86                         $this->output( sprintf( "%10s %10s %14s %20s %s\n", 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text', 'rev_comment' ) );
  87                         foreach ( $result as $row ) {
  88                                 $comment = ( $row->rev_comment == '' )
  89                                         ? ''
  90                                         : '(' . $wgContLang->truncate( $row->rev_comment, 40 ) . ')';
  91                                 $this->output( sprintf( "%10d %10d %14s %20s %s\n",
  92                                         $row->rev_id,
  93                                         $row->rev_page,
  94                                         $row->rev_timestamp,
  95                                         $wgContLang->truncate( $row->rev_user_text, 17 ),
  96                                         $comment ) );
  97                                 if ( $fix ) {
  98                                         $dbw->delete( 'revision', array( 'rev_id' => $row->rev_id ) );
  99                                 }
 100                         }
 101                         if ( !$fix ) {
 102                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 103                         }
 104                 } else {
 105                         $this->output( "No orphans! Yay!\n" );
 106                 }
 107
 108                 if ( $fix ) {
 109                         $dbw->unlockTables();
 110                 }
 111         }
 112
 113         /**
 114          * @param $fix bool
 115          * @todo DON'T USE THIS YET! It will remove entries which have children,
 116          *       but which aren't properly attached (eg if page_latest is bogus
 117          *       but valid revisions do exist)
 118          */
 119         private function checkWidows( $fix ) {
 120                 $dbw = wfGetDB( DB_MASTER );
 121                 $page = $dbw->tableName( 'page' );
 122                 $revision = $dbw->tableName( 'revision' );
 123
 124                 if ( $fix ) {
 125                         $this->lockTables( $dbw );
 126                 }
 127
 128                 $this->output( "\nChecking for childless page table entries... (this may take a while on a large wiki)\n" );
 129                 $result = $dbw->query( "
 130                         SELECT *
 131                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 132                         WHERE rev_id IS NULL
 133                 " );
 134                 $widows = $dbw->numRows( $result );
 135                 if ( $widows > 0 ) {
 136                         $this->output( "$widows childless pages...\n" );
 137                         $this->output( sprintf( "%10s %11s %2s %s\n", 'page_id', 'page_latest', 'ns', 'page_title' ) );
 138                         foreach ( $result as $row ) {
 139                                 printf( "%10d %11d %2d %s\n",
 140                                         $row->page_id,
 141                                         $row->page_latest,
 142                                         $row->page_namespace,
 143                                         $row->page_title );
 144                                 if ( $fix ) {
 145                                         $dbw->delete( 'page', array( 'page_id' => $row->page_id ) );
 146                                 }
 147                         }
 148                         if ( !$fix ) {
 149                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 150                         }
 151                 } else {
 152                         $this->output( "No childless pages! Yay!\n" );
 153                 }
 154
 155                 if ( $fix ) {
 156                         $dbw->unlockTables();
 157                 }
 158         }
 159
 160         /**
 161          * Check for pages where page_latest is wrong
 162          * @param $fix bool Whether to fix broken entries
 163          */
 164         private function checkSeparation( $fix ) {
 165                 $dbw = wfGetDB( DB_MASTER );
 166                 $page     = $dbw->tableName( 'page' );
 167                 $revision = $dbw->tableName( 'revision' );
 168                 $text     = $dbw->tableName( 'text' );
 169
 170                 if ( $fix ) {
 171                         $dbw->lockTables( $dbw, 'text' );
 172                 }
 173
 174                 $this->output( "\nChecking for pages whose page_latest links are incorrect... (this may take a while on a large wiki)\n" );
 175                 $result = $dbw->query( "
 176                         SELECT *
 177                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 178                 " );
 179                 $found = 0;
 180                 foreach ( $result as $row ) {
 181                         $result2 = $dbw->query( "
 182                                 SELECT MAX(rev_timestamp) as max_timestamp
 183                                 FROM $revision
 184                                 WHERE rev_page=$row->page_id
 185                         " );
 186                         $row2 = $dbw->fetchObject( $result2 );
 187                         $dbw->freeResult( $result2 );
 188                         if ( $row2 ) {
 189                                 if ( $row->rev_timestamp != $row2->max_timestamp ) {
 190                                         if ( $found == 0 ) {
 191                                                 $this->output( sprintf( "%10s %10s %14s %14s\n",
 192                                                         'page_id', 'rev_id', 'timestamp', 'max timestamp' ) );
 193                                         }
 194                                         ++$found;
 195                                         $this->output( sprintf( "%10d %10d %14s %14s\n",
 196                                                 $row->page_id,
 197                                                 $row->page_latest,
 198                                                 $row->rev_timestamp,
 199                                                 $row2->max_timestamp ) );
 200                                         if ( $fix ) {
 201                                                 # ...
 202                                                 $maxId = $dbw->selectField(
 203                                                         'revision',
 204                                                         'rev_id',
 205                                                         array(
 206                                                                 'rev_page'      => $row->page_id,
 207                                                                 'rev_timestamp' => $row2->max_timestamp ) );
 208                                                 $this->output( "... updating to revision $maxId\n" );
 209                                                 $maxRev = Revision::newFromId( $maxId );
 210                                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 211                                                 $article = new Article( $title );
 212                                                 $article->updateRevisionOn( $dbw, $maxRev );
 213                                         }
 214                                 }
 215                         } else {
 216                                 $this->output( "wtf\n" );
 217                         }
 218                 }
 219
 220                 if ( $found ) {
 221                         $this->output( "Found $found pages with incorrect latest revision.\n" );
 222                 } else {
 223                         $this->output( "No pages with incorrect latest revision. Yay!\n" );
 224                 }
 225                 if ( !$fix && $found > 0 ) {
 226                         $this->output( "Run again with --fix to remove these entries automatically.\n" );
 227                 }
 228
 229                 if ( $fix ) {
 230                         $dbw->unlockTables();
 231                 }
 232         }
 233 }
 234
 235 $maintClass = "Orphans";
 236 require_once( DO_MAINTENANCE );