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  * 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  * @file
  27  * @author <brion@pobox.com>
  28  * @ingroup Maintenance
  29  */
  30
  31 require_once( __DIR__ . '/Maintenance.php' );
  32
  33 /**
  34  * Maintenance script that looks for 'orphan' revisions hooked to pages which
  35  * don't exist and 'childless' pages with no revisions.
  36  *
  37  * @ingroup Maintenance
  38  */
  39 class Orphans extends Maintenance {
  40         public function __construct() {
  41                 parent::__construct();
  42                 $this->mDescription = "Look for 'orphan' revisions hooked to pages which don't exist\n" .
  43                                                                 "and 'childless' pages with no revisions\n" .
  44                                                                 "Then, kill the poor widows and orphans\n" .
  45                                                                 "Man this is depressing";
  46                 $this->addOption( 'fix', 'Actually fix broken entries' );
  47         }
  48
  49         public function execute() {
  50                 global $wgTitle;
  51                 $wgTitle = Title::newFromText( 'Orphan revision cleanup script' );
  52                 $this->checkOrphans( $this->hasOption( 'fix' ) );
  53                 $this->checkSeparation( $this->hasOption( 'fix' ) );
  54                 # Does not work yet, do not use
  55                 # $this->checkWidows( $this->hasOption( 'fix' ) );
  56         }
  57
  58         /**
  59          * Lock the appropriate tables for the script
  60          * @param $db DatabaseBase object
  61          * @param $extraTable String The name of any extra tables to lock (eg: text)
  62          */
  63         private function lockTables( $db, $extraTable = array() ) {
  64                 $tbls = array( 'page', 'revision', 'redirect' );
  65                 if ( $extraTable ) {
  66                         $tbls = array_merge( $tbls, $extraTable );
  67                 }
  68                 $db->lockTables( array(), $tbls, __METHOD__, false );
  69         }
  70
  71         /**
  72          * Check for orphan revisions
  73          * @param $fix bool Whether to fix broken revisions when found
  74          */
  75         private function checkOrphans( $fix ) {
  76                 $dbw = wfGetDB( DB_MASTER );
  77                 $page = $dbw->tableName( 'page' );
  78                 $revision = $dbw->tableName( 'revision' );
  79
  80                 if ( $fix ) {
  81                         $this->lockTables( $dbw );
  82                 }
  83
  84                 $this->output( "Checking for orphan revision table entries... (this may take a while on a large wiki)\n" );
  85                 $result = $dbw->query( "
  86                         SELECT *
  87                         FROM $revision LEFT OUTER JOIN $page ON rev_page=page_id
  88                         WHERE page_id IS NULL
  89                 " );
  90                 $orphans = $result->numRows();
  91                 if ( $orphans > 0 ) {
  92                         global $wgContLang;
  93                         $this->output( "$orphans orphan revisions...\n" );
  94                         $this->output( sprintf( "%10s %10s %14s %20s %s\n", 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text', 'rev_comment' ) );
  95                         foreach ( $result as $row ) {
  96                                 $comment = ( $row->rev_comment == '' )
  97                                         ? ''
  98                                         : '(' . $wgContLang->truncate( $row->rev_comment, 40 ) . ')';
  99                                 $this->output( sprintf( "%10d %10d %14s %20s %s\n",
 100                                         $row->rev_id,
 101                                         $row->rev_page,
 102                                         $row->rev_timestamp,
 103                                         $wgContLang->truncate( $row->rev_user_text, 17 ),
 104                                         $comment ) );
 105                                 if ( $fix ) {
 106                                         $dbw->delete( 'revision', array( 'rev_id' => $row->rev_id ) );
 107                                 }
 108                         }
 109                         if ( !$fix ) {
 110                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 111                         }
 112                 } else {
 113                         $this->output( "No orphans! Yay!\n" );
 114                 }
 115
 116                 if ( $fix ) {
 117                         $dbw->unlockTables( __METHOD__ );
 118                 }
 119         }
 120
 121         /**
 122          * @param $fix bool
 123          * @todo DON'T USE THIS YET! It will remove entries which have children,
 124          *       but which aren't properly attached (eg if page_latest is bogus
 125          *       but valid revisions do exist)
 126          */
 127         private function checkWidows( $fix ) {
 128                 $dbw = wfGetDB( DB_MASTER );
 129                 $page = $dbw->tableName( 'page' );
 130                 $revision = $dbw->tableName( 'revision' );
 131
 132                 if ( $fix ) {
 133                         $this->lockTables( $dbw );
 134                 }
 135
 136                 $this->output( "\nChecking for childless page table entries... (this may take a while on a large wiki)\n" );
 137                 $result = $dbw->query( "
 138                         SELECT *
 139                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 140                         WHERE rev_id IS NULL
 141                 " );
 142                 $widows = $result->numRows();
 143                 if ( $widows > 0 ) {
 144                         $this->output( "$widows childless pages...\n" );
 145                         $this->output( sprintf( "%10s %11s %2s %s\n", 'page_id', 'page_latest', 'ns', 'page_title' ) );
 146                         foreach ( $result as $row ) {
 147                                 printf( "%10d %11d %2d %s\n",
 148                                         $row->page_id,
 149                                         $row->page_latest,
 150                                         $row->page_namespace,
 151                                         $row->page_title );
 152                                 if ( $fix ) {
 153                                         $dbw->delete( 'page', array( 'page_id' => $row->page_id ) );
 154                                 }
 155                         }
 156                         if ( !$fix ) {
 157                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 158                         }
 159                 } else {
 160                         $this->output( "No childless pages! Yay!\n" );
 161                 }
 162
 163                 if ( $fix ) {
 164                         $dbw->unlockTables( __METHOD__ );
 165                 }
 166         }
 167
 168         /**
 169          * Check for pages where page_latest is wrong
 170          * @param $fix bool Whether to fix broken entries
 171          */
 172         private function checkSeparation( $fix ) {
 173                 $dbw = wfGetDB( DB_MASTER );
 174                 $page = $dbw->tableName( 'page' );
 175                 $revision = $dbw->tableName( 'revision' );
 176
 177                 if ( $fix ) {
 178                         $this->lockTables( $dbw, array( 'user', 'text' ) );
 179                 }
 180
 181                 $this->output( "\nChecking for pages whose page_latest links are incorrect... (this may take a while on a large wiki)\n" );
 182                 $result = $dbw->query( "
 183                         SELECT *
 184                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 185                 " );
 186                 $found = 0;
 187                 foreach ( $result as $row ) {
 188                         $result2 = $dbw->query( "
 189                                 SELECT MAX(rev_timestamp) as max_timestamp
 190                                 FROM $revision
 191                                 WHERE rev_page=$row->page_id
 192                         " );
 193                         $row2 = $dbw->fetchObject( $result2 );
 194                         if ( $row2 ) {
 195                                 if ( $row->rev_timestamp != $row2->max_timestamp ) {
 196                                         if ( $found == 0 ) {
 197                                                 $this->output( sprintf( "%10s %10s %14s %14s\n",
 198                                                         'page_id', 'rev_id', 'timestamp', 'max timestamp' ) );
 199                                         }
 200                                         ++$found;
 201                                         $this->output( sprintf( "%10d %10d %14s %14s\n",
 202                                                 $row->page_id,
 203                                                 $row->page_latest,
 204                                                 $row->rev_timestamp,
 205                                                 $row2->max_timestamp ) );
 206                                         if ( $fix ) {
 207                                                 # ...
 208                                                 $maxId = $dbw->selectField(
 209                                                         'revision',
 210                                                         'rev_id',
 211                                                         array(
 212                                                                 'rev_page' => $row->page_id,
 213                                                                 'rev_timestamp' => $row2->max_timestamp ) );
 214                                                 $this->output( "... updating to revision $maxId\n" );
 215                                                 $maxRev = Revision::newFromId( $maxId );
 216                                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 217                                                 $article = WikiPage::factory( $title );
 218                                                 $article->updateRevisionOn( $dbw, $maxRev );
 219                                         }
 220                                 }
 221                         } else {
 222                                 $this->output( "wtf\n" );
 223                         }
 224                 }
 225
 226                 if ( $found ) {
 227                         $this->output( "Found $found pages with incorrect latest revision.\n" );
 228                 } else {
 229                         $this->output( "No pages with incorrect latest revision. Yay!\n" );
 230                 }
 231                 if ( !$fix && $found > 0 ) {
 232                         $this->output( "Run again with --fix to remove these entries automatically.\n" );
 233                 }
 234
 235                 if ( $fix ) {
 236                         $dbw->unlockTables( __METHOD__ );
 237                 }
 238         }
 239 }
 240
 241 $maintClass = "Orphans";
 242 require_once( RUN_MAINTENANCE_IF_MAIN );