maintenance/populateRevisionLength.php

   1 <?php
   2 /**
   3  * Populates the rev_len and ar_len fields when they are NULL.
   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  * @file
  21  * @ingroup Maintenance
  22  */
  23
  24 use MediaWiki\Revision\RevisionRecord;
  25 use Wikimedia\Rdbms\IDatabase;
  26
  27 require_once __DIR__ . '/Maintenance.php';
  28
  29 /**
  30  * Maintenance script that populates the rev_len and ar_len fields when they are NULL.
  31  * This is the case for all revisions created before MW 1.10, as well as those affected
  32  * by T18748 (MW 1.10-1.13) and those affected by T135414 (MW 1.21-1.24).
  33  *
  34  * @ingroup Maintenance
  35  */
  36 class PopulateRevisionLength extends LoggedUpdateMaintenance {
  37         public function __construct() {
  38                 parent::__construct();
  39                 $this->addDescription( 'Populates the rev_len and ar_len fields' );
  40                 $this->setBatchSize( 200 );
  41         }
  42
  43         protected function getUpdateKey() {
  44                 return 'populate rev_len and ar_len';
  45         }
  46
  47         public function doDBUpdates() {
  48                 $dbw = $this->getDB( DB_MASTER );
  49                 if ( !$dbw->tableExists( 'revision' ) ) {
  50                         $this->fatalError( "revision table does not exist" );
  51                 } elseif ( !$dbw->tableExists( 'archive' ) ) {
  52                         $this->fatalError( "archive table does not exist" );
  53                 } elseif ( !$dbw->fieldExists( 'revision', 'rev_len', __METHOD__ ) ) {
  54                         $this->output( "rev_len column does not exist\n\n", true );
  55
  56                         return false;
  57                 }
  58
  59                 $this->output( "Populating rev_len column\n" );
  60                 $rev = $this->doLenUpdates( 'revision', 'rev_id', 'rev', Revision::getQueryInfo() );
  61
  62                 $this->output( "Populating ar_len column\n" );
  63                 $ar = $this->doLenUpdates( 'archive', 'ar_id', 'ar', Revision::getArchiveQueryInfo() );
  64
  65                 $this->output( "rev_len and ar_len population complete "
  66                         . "[$rev revision rows, $ar archive rows].\n" );
  67
  68                 return true;
  69         }
  70
  71         /**
  72          * @param string $table
  73          * @param string $idCol
  74          * @param string $prefix
  75          * @param array $queryInfo
  76          * @return int
  77          */
  78         protected function doLenUpdates( $table, $idCol, $prefix, $queryInfo ) {
  79                 $dbr = $this->getDB( DB_REPLICA );
  80                 $dbw = $this->getDB( DB_MASTER );
  81                 $batchSize = $this->getBatchSize();
  82                 $start = $dbw->selectField( $table, "MIN($idCol)", '', __METHOD__ );
  83                 $end = $dbw->selectField( $table, "MAX($idCol)", '', __METHOD__ );
  84                 if ( !$start || !$end ) {
  85                         $this->output( "...$table table seems to be empty.\n" );
  86
  87                         return 0;
  88                 }
  89
  90                 # Do remaining chunks
  91                 $blockStart = intval( $start );
  92                 $blockEnd = intval( $start ) + $batchSize - 1;
  93                 $count = 0;
  94
  95                 while ( $blockStart <= $end ) {
  96                         $this->output( "...doing $idCol from $blockStart to $blockEnd\n" );
  97                         $res = $dbr->select(
  98                                 $queryInfo['tables'],
  99                                 $queryInfo['fields'],
 100                                 [
 101                                         "$idCol >= $blockStart",
 102                                         "$idCol <= $blockEnd",
 103                                         $dbr->makeList( [
 104                                                 "{$prefix}_len IS NULL",
 105                                                 $dbr->makeList( [
 106                                                         "{$prefix}_len = 0",
 107                                                         "{$prefix}_sha1 != " . $dbr->addQuotes( 'phoiac9h4m842xq45sp7s6u21eteeq1' ), // sha1( "" )
 108                                                 ], IDatabase::LIST_AND )
 109                                         ], IDatabase::LIST_OR )
 110                                 ],
 111                                 __METHOD__,
 112                                 [],
 113                                 $queryInfo['joins']
 114                         );
 115
 116                         if ( $res->numRows() > 0 ) {
 117                                 $this->beginTransaction( $dbw, __METHOD__ );
 118                                 # Go through and update rev_len from these rows.
 119                                 foreach ( $res as $row ) {
 120                                         if ( $this->upgradeRow( $row, $table, $idCol, $prefix ) ) {
 121                                                 $count++;
 122                                         }
 123                                 }
 124                                 $this->commitTransaction( $dbw, __METHOD__ );
 125                         }
 126
 127                         $blockStart += $batchSize;
 128                         $blockEnd += $batchSize;
 129                 }
 130
 131                 return $count;
 132         }
 133
 134         /**
 135          * @param stdClass $row
 136          * @param string $table
 137          * @param string $idCol
 138          * @param string $prefix
 139          * @return bool
 140          */
 141         protected function upgradeRow( $row, $table, $idCol, $prefix ) {
 142                 $dbw = $this->getDB( DB_MASTER );
 143
 144                 $rev = ( $table === 'archive' )
 145                         ? Revision::newFromArchiveRow( $row )
 146                         : new Revision( $row );
 147
 148                 $content = $rev->getContent( RevisionRecord::RAW );
 149                 if ( !$content ) {
 150                         # This should not happen, but sometimes does (T22757)
 151                         $id = $row->$idCol;
 152                         $this->output( "Content of $table $id unavailable!\n" );
 153
 154                         return false;
 155                 }
 156
 157                 # Update the row...
 158                 $dbw->update( $table,
 159                         [ "{$prefix}_len" => $content->getSize() ],
 160                         [ $idCol => $row->$idCol ],
 161                         __METHOD__
 162                 );
 163
 164                 return true;
 165         }
 166 }
 167
 168 $maintClass = PopulateRevisionLength::class;
 169 require_once RUN_MAINTENANCE_IF_MAIN;