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