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                 $page = $dbw->tableName( 'page' );
  79                 $revision = $dbw->tableName( 'revision' );
  80
  81                 if ( $fix ) {
  82                         $this->lockTables( $dbw );
  83                 }
  84
  85                 $this->output( "Checking for orphan revision table entries... "
  86                         . "(this may take a while on a large wiki)\n" );
  87                 $result = $dbw->query( "
  88                         SELECT *
  89                         FROM $revision LEFT OUTER JOIN $page ON rev_page=page_id
  90                         WHERE page_id IS NULL
  91                 " );
  92                 $orphans = $result->numRows();
  93                 if ( $orphans > 0 ) {
  94                         global $wgContLang;
  95
  96                         $this->output( "$orphans orphan revisions...\n" );
  97                         $this->output( sprintf(
  98                                 "%10s %10s %14s %20s %s\n",
  99                                 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text', 'rev_comment'
 100                         ) );
 101
 102                         foreach ( $result as $row ) {
 103                                 $comment = ( $row->rev_comment == '' )
 104                                         ? ''
 105                                         : '(' . $wgContLang->truncate( $row->rev_comment, 40 ) . ')';
 106                                 $this->output( sprintf( "%10d %10d %14s %20s %s\n",
 107                                         $row->rev_id,
 108                                         $row->rev_page,
 109                                         $row->rev_timestamp,
 110                                         $wgContLang->truncate( $row->rev_user_text, 17 ),
 111                                         $comment ) );
 112                                 if ( $fix ) {
 113                                         $dbw->delete( 'revision', [ 'rev_id' => $row->rev_id ] );
 114                                 }
 115                         }
 116                         if ( !$fix ) {
 117                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 118                         }
 119                 } else {
 120                         $this->output( "No orphans! Yay!\n" );
 121                 }
 122
 123                 if ( $fix ) {
 124                         $dbw->unlockTables( __METHOD__ );
 125                 }
 126         }
 127
 128         /**
 129          * @param bool $fix
 130          * @todo DON'T USE THIS YET! It will remove entries which have children,
 131          *       but which aren't properly attached (eg if page_latest is bogus
 132          *       but valid revisions do exist)
 133          */
 134         private function checkWidows( $fix ) {
 135                 $dbw = $this->getDB( DB_MASTER );
 136                 $page = $dbw->tableName( 'page' );
 137                 $revision = $dbw->tableName( 'revision' );
 138
 139                 if ( $fix ) {
 140                         $this->lockTables( $dbw );
 141                 }
 142
 143                 $this->output( "\nChecking for childless page table entries... "
 144                         . "(this may take a while on a large wiki)\n" );
 145                 $result = $dbw->query( "
 146                         SELECT *
 147                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 148                         WHERE rev_id IS NULL
 149                 " );
 150                 $widows = $result->numRows();
 151                 if ( $widows > 0 ) {
 152                         $this->output( "$widows childless pages...\n" );
 153                         $this->output( sprintf( "%10s %11s %2s %s\n", 'page_id', 'page_latest', 'ns', 'page_title' ) );
 154                         foreach ( $result as $row ) {
 155                                 printf( "%10d %11d %2d %s\n",
 156                                         $row->page_id,
 157                                         $row->page_latest,
 158                                         $row->page_namespace,
 159                                         $row->page_title );
 160                                 if ( $fix ) {
 161                                         $dbw->delete( 'page', [ 'page_id' => $row->page_id ] );
 162                                 }
 163                         }
 164                         if ( !$fix ) {
 165                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 166                         }
 167                 } else {
 168                         $this->output( "No childless pages! Yay!\n" );
 169                 }
 170
 171                 if ( $fix ) {
 172                         $dbw->unlockTables( __METHOD__ );
 173                 }
 174         }
 175
 176         /**
 177          * Check for pages where page_latest is wrong
 178          * @param bool $fix Whether to fix broken entries
 179          */
 180         private function checkSeparation( $fix ) {
 181                 $dbw = $this->getDB( DB_MASTER );
 182                 $page = $dbw->tableName( 'page' );
 183                 $revision = $dbw->tableName( 'revision' );
 184
 185                 if ( $fix ) {
 186                         $this->lockTables( $dbw, [ 'user', 'text' ] );
 187                 }
 188
 189                 $this->output( "\nChecking for pages whose page_latest links are incorrect... "
 190                         . "(this may take a while on a large wiki)\n" );
 191                 $result = $dbw->query( "
 192                         SELECT *
 193                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 194                 " );
 195                 $found = 0;
 196                 foreach ( $result as $row ) {
 197                         $result2 = $dbw->query( "
 198                                 SELECT MAX(rev_timestamp) as max_timestamp
 199                                 FROM $revision
 200                                 WHERE rev_page=$row->page_id
 201                         " );
 202                         $row2 = $dbw->fetchObject( $result2 );
 203                         if ( $row2 ) {
 204                                 if ( $row->rev_timestamp != $row2->max_timestamp ) {
 205                                         if ( $found == 0 ) {
 206                                                 $this->output( sprintf( "%10s %10s %14s %14s\n",
 207                                                         'page_id', 'rev_id', 'timestamp', 'max timestamp' ) );
 208                                         }
 209                                         ++$found;
 210                                         $this->output( sprintf( "%10d %10d %14s %14s\n",
 211                                                 $row->page_id,
 212                                                 $row->page_latest,
 213                                                 $row->rev_timestamp,
 214                                                 $row2->max_timestamp ) );
 215                                         if ( $fix ) {
 216                                                 # ...
 217                                                 $maxId = $dbw->selectField(
 218                                                         'revision',
 219                                                         'rev_id',
 220                                                         [
 221                                                                 'rev_page' => $row->page_id,
 222                                                                 'rev_timestamp' => $row2->max_timestamp ] );
 223                                                 $this->output( "... updating to revision $maxId\n" );
 224                                                 $maxRev = Revision::newFromId( $maxId );
 225                                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 226                                                 $article = WikiPage::factory( $title );
 227                                                 $article->updateRevisionOn( $dbw, $maxRev );
 228                                         }
 229                                 }
 230                         } else {
 231                                 $this->output( "wtf\n" );
 232                         }
 233                 }
 234
 235                 if ( $found ) {
 236                         $this->output( "Found $found pages with incorrect latest revision.\n" );
 237                 } else {
 238                         $this->output( "No pages with incorrect latest revision. Yay!\n" );
 239                 }
 240                 if ( !$fix && $found > 0 ) {
 241                         $this->output( "Run again with --fix to remove these entries automatically.\n" );
 242                 }
 243
 244                 if ( $fix ) {
 245                         $dbw->unlockTables( __METHOD__ );
 246                 }
 247         }
 248 }
 249
 250 $maintClass = "Orphans";
 251 require_once RUN_MAINTENANCE_IF_MAIN;