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