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