maintenance/orphans.php

   1 <?php
   2 # Copyright (C) 2005 Brion Vibber <brion@pobox.com>
   3 # http://www.mediawiki.org/
   4 #
   5 # This program is free software; you can redistribute it and/or modify
   6 # it under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 2 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # This program is distributed in the hope that it will be useful,
  11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License along
  16 # with this program; if not, write to the Free Software Foundation, Inc.,
  17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18 # http://www.gnu.org/copyleft/gpl.html
  19
  20 /**
  21  * Look for 'orphan' revisions hooked to pages which don't exist
  22  * And 'childless' pages with no revisions.
  23  * Then, kill the poor widows and orphans.
  24  * Man this is depressing.
  25  *
  26  * @author <brion@pobox.com>
  27  * @addtogroup Maintenance
  28  */
  29
  30 $options = array( 'fix' );
  31
  32 /** */
  33 require_once( 'commandLine.inc' );
  34 $wgTitle = Title::newFromText( 'Orphan revision cleanup script' );
  35
  36 checkOrphans( isset( $options['fix'] ) );
  37 checkSeparation( isset( $options['fix'] ) );
  38 #checkWidows( isset( $options['fix'] ) );
  39
  40 # ------
  41
  42 function checkOrphans( $fix ) {
  43         $dbw = wfGetDB( DB_MASTER );
  44         $page = $dbw->tableName( 'page' );
  45         $revision = $dbw->tableName( 'revision' );
  46
  47         if( $fix ) {
  48                 $dbw->query( "LOCK TABLES $page WRITE, $revision WRITE" );
  49         }
  50
  51         echo "Checking for orphan revision table entries... (this may take a while on a large wiki)\n";
  52         $result = $dbw->query( "
  53                 SELECT *
  54                 FROM $revision LEFT OUTER JOIN $page ON rev_page=page_id
  55                 WHERE page_id IS NULL
  56         ");
  57         $orphans = $dbw->numRows( $result );
  58         if( $orphans > 0 ) {
  59                 global $wgContLang;
  60                 echo "$orphans orphan revisions...\n";
  61                 printf( "%10s %10s %14s %20s %s\n", 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text', 'rev_comment' );
  62                 while( $row = $dbw->fetchObject( $result ) ) {
  63                         $comment = ( $row->rev_comment == '' )
  64                                 ? ''
  65                                 : '(' . $wgContLang->truncate( $row->rev_comment, 40, '...' ) . ')';
  66                         printf( "%10d %10d %14s %20s %s\n",
  67                                 $row->rev_id,
  68                                 $row->rev_page,
  69                                 $row->rev_timestamp,
  70                                 $wgContLang->truncate( $row->rev_user_text, 17, '...' ),
  71                                 $comment );
  72                         if( $fix ) {
  73                                 $dbw->delete( 'revision', array( 'rev_id' => $row->rev_id ) );
  74                         }
  75                 }
  76                 if( !$fix ) {
  77                         echo "Run again with --fix to remove these entries automatically.\n";
  78                 }
  79         } else {
  80                 echo "No orphans! Yay!\n";
  81         }
  82
  83         if( $fix ) {
  84                 $dbw->query( "UNLOCK TABLES" );
  85         }
  86 }
  87
  88 /**
  89  * @todo DON'T USE THIS YET! It will remove entries which have children,
  90  *       but which aren't properly attached (eg if page_latest is bogus
  91  *       but valid revisions do exist)
  92  */
  93 function checkWidows( $fix ) {
  94         $dbw = wfGetDB( DB_MASTER );
  95         $page = $dbw->tableName( 'page' );
  96         $revision = $dbw->tableName( 'revision' );
  97
  98         if( $fix ) {
  99                 $dbw->query( "LOCK TABLES $page WRITE, $revision WRITE" );
 100         }
 101
 102         echo "\nChecking for childless page table entries... (this may take a while on a large wiki)\n";
 103         $result = $dbw->query( "
 104                 SELECT *
 105                 FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 106                 WHERE rev_id IS NULL
 107         ");
 108         $widows = $dbw->numRows( $result );
 109         if( $widows > 0 ) {
 110                 global $wgContLang;
 111                 echo "$widows childless pages...\n";
 112                 printf( "%10s %11s %2s %s\n", 'page_id', 'page_latest', 'ns', 'page_title' );
 113                 while( $row = $dbw->fetchObject( $result ) ) {
 114                         printf( "%10d %11d %2d %s\n",
 115                                 $row->page_id,
 116                                 $row->page_latest,
 117                                 $row->page_namespace,
 118                                 $row->page_title );
 119                         if( $fix ) {
 120                                 $dbw->delete( 'page', array( 'page_id' => $row->page_id ) );
 121                         }
 122                 }
 123                 if( !$fix ) {
 124                         echo "Run again with --fix to remove these entries automatically.\n";
 125                 }
 126         } else {
 127                 echo "No childless pages! Yay!\n";
 128         }
 129
 130         if( $fix ) {
 131                 $dbw->query( "UNLOCK TABLES" );
 132         }
 133 }
 134
 135
 136 function checkSeparation( $fix ) {
 137         $dbw = wfGetDB( DB_MASTER );
 138         $page     = $dbw->tableName( 'page' );
 139         $revision = $dbw->tableName( 'revision' );
 140         $text     = $dbw->tableName( 'text' );
 141
 142         if( $fix ) {
 143                 $dbw->query( "LOCK TABLES $page WRITE, $revision WRITE, $text WRITE" );
 144         }
 145
 146         echo "\nChecking for pages whose page_latest links are incorrect... (this may take a while on a large wiki)\n";
 147         $result = $dbw->query( "
 148                 SELECT *
 149                 FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 150         ");
 151         $found = 0;
 152         while( $row = $dbw->fetchObject( $result ) ) {
 153                 $result2 = $dbw->query( "
 154                         SELECT MAX(rev_timestamp) as max_timestamp
 155                         FROM $revision
 156                         WHERE rev_page=$row->page_id
 157                 " );
 158                 $row2 = $dbw->fetchObject( $result2 );
 159                 $dbw->freeResult( $result2 );
 160                 if( $row2 ) {
 161                         if( $row->rev_timestamp != $row2->max_timestamp ) {
 162                                 if( $found == 0 ) {
 163                                         printf( "%10s %10s %14s %14s\n",
 164                                                 'page_id', 'rev_id', 'timestamp', 'max timestamp' );
 165                                 }
 166                                 ++$found;
 167                                 printf( "%10d %10d %14s %14s\n",
 168                                         $row->page_id,
 169                                         $row->page_latest,
 170                                         $row->rev_timestamp,
 171                                         $row2->max_timestamp );
 172                                 if( $fix ) {
 173                                         # ...
 174                                         $maxId = $dbw->selectField(
 175                                                 'revision',
 176                                                 'rev_id',
 177                                                 array(
 178                                                         'rev_page'      => $row->page_id,
 179                                                         'rev_timestamp' => $row2->max_timestamp ) );
 180                                         echo "... updating to revision $maxId\n";
 181                                         $maxRev = Revision::newFromId( $maxId );
 182                                         $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 183                                         $article = new Article( $title );
 184                                         $article->updateRevisionOn( $dbw, $maxRev );
 185                                 }
 186                         }
 187                 } else {
 188                         echo "wtf\n";
 189                 }
 190         }
 191
 192         if( $found ) {
 193                 echo "Found $found pages with incorrect latest revision.\n";
 194         } else {
 195                 echo "No pages with incorrect latest revision. Yay!\n";
 196         }
 197         if( !$fix && $found > 0 ) {
 198                 echo "Run again with --fix to remove these entries automatically.\n";
 199         }
 200
 201         if( $fix ) {
 202                 $dbw->query( "UNLOCK TABLES" );
 203         }
 204 }
 205