maintenance/resetPageRandom.php

   1 <?php
   2 /**
   3  * Resets the page_random field for articles in the provided time range.
   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\Maintenance\Maintenance;
  25
  26 // @codeCoverageIgnoreStart
  27 require_once __DIR__ . '/Maintenance.php';
  28 // @codeCoverageIgnoreEnd
  29
  30 /**
  31  * Maintenance script that resets page_random over a time range.
  32  *
  33  * @ingroup Maintenance
  34  */
  35 class ResetPageRandom extends Maintenance {
  36         public function __construct() {
  37                 parent::__construct();
  38                 $this->addDescription( 'Reset the page_random for articles within given date range' );
  39                 $this->addOption( 'from',
  40                         'From date range selector to select articles to update, ex: 20041011000000', true, true );
  41                 $this->addOption( 'to',
  42                         'To date range selector to select articles to update, ex: 20050708000000', true, true );
  43                 $this->addOption( 'dry', 'Do not update column' );
  44                 $this->addOption( 'batch-start',
  45                         'Optional: Use when you need to restart the reset process from a given page ID offset'
  46                         . ' in case a previous reset failed or was stopped'
  47                 );
  48                 // Initialize batch size to a good default value and enable the batch size option.
  49                 $this->setBatchSize( 200 );
  50         }
  51
  52         public function execute() {
  53                 $batchSize = $this->getBatchSize();
  54                 $dbw = $this->getPrimaryDB();
  55                 $dbr = $this->getReplicaDB();
  56                 $from = wfTimestampOrNull( TS_MW, $this->getOption( 'from' ) );
  57                 $to = wfTimestampOrNull( TS_MW, $this->getOption( 'to' ) );
  58
  59                 if ( $from === null || $to === null ) {
  60                         $this->output( "--from and --to have to be provided" . PHP_EOL );
  61                         return false;
  62                 }
  63                 if ( $from >= $to ) {
  64                         $this->output( "--from has to be smaller than --to" . PHP_EOL );
  65                         return false;
  66                 }
  67                 $batchStart = (int)$this->getOption( 'batch-start', 0 );
  68                 $changed = 0;
  69                 $dry = (bool)$this->getOption( 'dry' );
  70
  71                 $message = "Resetting page_random column within date range from $from to $to";
  72                 if ( $batchStart > 0 ) {
  73                         $message .= " starting from page ID $batchStart";
  74                 }
  75                 $message .= $dry ? ". dry run" : '.';
  76
  77                 $this->output( $message . PHP_EOL );
  78                 do {
  79                         $this->output( "  ...doing chunk of $batchSize from $batchStart " . PHP_EOL );
  80
  81                         // Find the oldest page revision associated with each page_id. Iff it falls in the given
  82                         // time range AND it's greater than $batchStart, yield the page ID. If it falls outside the
  83                         // time range, it was created before or after the occurrence of T208909 and its page_random
  84                         // is considered valid. The replica is used for this read since page_id and the rev_timestamp
  85                         // will not change between queries.
  86                         $queryBuilder = $dbr->newSelectQueryBuilder()
  87                                 ->select( 'page_id' )
  88                                 ->from( 'page' )
  89                                 ->where( $dbr->expr( 'page_id', '>', $batchStart ) )
  90                                 ->limit( $batchSize )
  91                                 ->orderBy( 'page_id' );
  92                         $subquery = $queryBuilder->newSubquery()
  93                                 ->select( 'MIN(rev_timestamp)' )
  94                                 ->from( 'revision' )
  95                                 ->where( 'rev_page=page_id' );
  96                         $queryBuilder->andWhere(
  97                                 '(' . $subquery->getSQL() . ') BETWEEN ' .
  98                                 $dbr->addQuotes( $dbr->timestamp( $from ) ) . ' AND ' . $dbr->addQuotes( $dbr->timestamp( $to ) )
  99                         );
 100
 101                         $res = $queryBuilder->caller( __METHOD__ )->fetchResultSet();
 102                         $row = null;
 103                         foreach ( $res as $row ) {
 104                                 if ( !$dry ) {
 105                                         # Update the row...
 106                                         $dbw->newUpdateQueryBuilder()
 107                                                 ->update( 'page' )
 108                                                 ->set( [ 'page_random' => wfRandom() ] )
 109                                                 ->where( [ 'page_id' => $row->page_id ] )
 110                                                 ->caller( __METHOD__ )
 111                                                 ->execute();
 112                                         $changed += $dbw->affectedRows();
 113                                 } else {
 114                                         $changed++;
 115                                 }
 116                         }
 117                         if ( $row ) {
 118                                 $batchStart = $row->page_id;
 119                         } else {
 120                                 // We don't need to set the $batchStart as $res is empty,
 121                                 // and we don't need to do another loop
 122                                 // the while() condition will evaluate to false and
 123                                 // we will leave the do{}while() block.
 124                         }
 125
 126                         $this->waitForReplication();
 127                 } while ( $res->numRows() === $batchSize );
 128                 $this->output( "page_random reset complete ... changed $changed rows" . PHP_EOL );
 129
 130                 return true;
 131         }
 132 }
 133
 134 // @codeCoverageIgnoreStart
 135 $maintClass = ResetPageRandom::class;
 136 require_once RUN_MAINTENANCE_IF_MAIN;
 137 // @codeCoverageIgnoreEnd