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 Wikimedia\Rdbms\IDatabase;
  25
  26 require_once __DIR__ . '/Maintenance.php';
  27
  28 /**
  29  * Maintenance script that populates the rev_len and ar_len fields when they are NULL.
  30  * This is the case for all revisions created before MW 1.10, as well as those affected
  31  * by T18748 (MW 1.10-1.13) and those affected by T135414 (MW 1.21-1.24).
  32  *
  33  * @ingroup Maintenance
  34  */
  35 class PopulateRevisionLength extends LoggedUpdateMaintenance {
  36         public function __construct() {
  37                 parent::__construct();
  38                 $this->addDescription( 'Populates the rev_len and ar_len fields' );
  39                 $this->setBatchSize( 200 );
  40         }
  41
  42         protected function getUpdateKey() {
  43                 return 'populate rev_len and ar_len';
  44         }
  45
  46         public function doDBUpdates() {
  47                 $dbw = $this->getDB( DB_PRIMARY );
  48                 if ( !$dbw->tableExists( 'revision', __METHOD__ ) ) {
  49                         $this->fatalError( "revision table does not exist" );
  50                 } elseif ( !$dbw->tableExists( 'archive', __METHOD__ ) ) {
  51                         $this->fatalError( "archive table does not exist" );
  52                 } elseif ( !$dbw->fieldExists( 'revision', 'rev_len', __METHOD__ ) ) {
  53                         $this->output( "rev_len column does not exist\n\n", true );
  54
  55                         return false;
  56                 }
  57
  58                 $revisionStore = $this->getServiceContainer()->getRevisionStore();
  59
  60                 $this->output( "Populating rev_len column\n" );
  61                 $rev = $this->doLenUpdates(
  62                         'revision',
  63                         'rev_id',
  64                         'rev',
  65                         $revisionStore->getQueryInfo()
  66                 );
  67
  68                 $this->output( "Populating ar_len column\n" );
  69                 $ar = $this->doLenUpdates(
  70                         'archive',
  71                         'ar_id',
  72                         'ar',
  73                         $revisionStore->getArchiveQueryInfo()
  74                 );
  75
  76                 $this->output( "rev_len and ar_len population complete "
  77                         . "[$rev revision rows, $ar archive rows].\n" );
  78
  79                 return true;
  80         }
  81
  82         /**
  83          * @param string $table
  84          * @param string $idCol
  85          * @param string $prefix
  86          * @param array $queryInfo
  87          * @return int
  88          */
  89         protected function doLenUpdates( $table, $idCol, $prefix, $queryInfo ) {
  90                 $dbr = $this->getDB( DB_REPLICA );
  91                 $dbw = $this->getDB( DB_PRIMARY );
  92                 $batchSize = $this->getBatchSize();
  93                 $start = $dbw->newSelectQueryBuilder()
  94                         ->select( "MIN($idCol)" )
  95                         ->from( $table )
  96                         ->caller( __METHOD__ )->fetchField();
  97                 $end = $dbw->newSelectQueryBuilder()
  98                         ->select( "MAX($idCol)" )
  99                         ->from( $table )
 100                         ->caller( __METHOD__ )->fetchField();
 101                 if ( !$start || !$end ) {
 102                         $this->output( "...$table table seems to be empty.\n" );
 103
 104                         return 0;
 105                 }
 106
 107                 # Do remaining chunks
 108                 $blockStart = intval( $start );
 109                 $blockEnd = intval( $start ) + $batchSize - 1;
 110                 $count = 0;
 111
 112                 while ( $blockStart <= $end ) {
 113                         $this->output( "...doing $idCol from $blockStart to $blockEnd\n" );
 114                         $res = $dbr->select(
 115                                 $queryInfo['tables'],
 116                                 $queryInfo['fields'],
 117                                 [
 118                                         "$idCol >= $blockStart",
 119                                         "$idCol <= $blockEnd",
 120                                         $dbr->makeList( [
 121                                                 "{$prefix}_len" => null,
 122                                                 $dbr->makeList( [
 123                                                         "{$prefix}_len" => 0,
 124                                                         // sha1( "" )
 125                                                         "{$prefix}_sha1 != " . $dbr->addQuotes( 'phoiac9h4m842xq45sp7s6u21eteeq1' ),
 126                                                 ], IDatabase::LIST_AND )
 127                                         ], IDatabase::LIST_OR )
 128                                 ],
 129                                 __METHOD__,
 130                                 [],
 131                                 $queryInfo['joins']
 132                         );
 133
 134                         if ( $res->numRows() > 0 ) {
 135                                 $this->beginTransaction( $dbw, __METHOD__ );
 136                                 # Go through and update rev_len from these rows.
 137                                 foreach ( $res as $row ) {
 138                                         if ( $this->upgradeRow( $row, $table, $idCol, $prefix ) ) {
 139                                                 $count++;
 140                                         }
 141                                 }
 142                                 $this->commitTransaction( $dbw, __METHOD__ );
 143                         }
 144
 145                         $blockStart += $batchSize;
 146                         $blockEnd += $batchSize;
 147                 }
 148
 149                 return $count;
 150         }
 151
 152         /**
 153          * @param stdClass $row
 154          * @param string $table
 155          * @param string $idCol
 156          * @param string $prefix
 157          * @return bool
 158          */
 159         protected function upgradeRow( $row, $table, $idCol, $prefix ) {
 160                 $dbw = $this->getDB( DB_PRIMARY );
 161
 162                 $revFactory = $this->getServiceContainer()->getRevisionFactory();
 163                 if ( $table === 'archive' ) {
 164                         $revRecord = $revFactory->newRevisionFromArchiveRow( $row );
 165                 } else {
 166                         $revRecord = $revFactory->newRevisionFromRow( $row );
 167                 }
 168
 169                 if ( !$revRecord ) {
 170                         # This should not happen, but sometimes does (T22757)
 171                         $id = $row->$idCol;
 172                         $this->output( "RevisionRecord of $table $id unavailable!\n" );
 173
 174                         return false;
 175                 }
 176
 177                 # Update the row...
 178                 $dbw->update( $table,
 179                         [ "{$prefix}_len" => $revRecord->getSize() ],
 180                         [ $idCol => $row->$idCol ],
 181                         __METHOD__
 182                 );
 183
 184                 return true;
 185         }
 186 }
 187
 188 $maintClass = PopulateRevisionLength::class;
 189 require_once RUN_MAINTENANCE_IF_MAIN;