maintenance/recountCategories.php

   1 <?php
   2 /**
   3  * Refreshes category counts.
   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 use MediaWiki\MediaWikiServices;
  27
  28 /**
  29  * Maintenance script that refreshes category membership counts in the category
  30  * table.
  31  *
  32  * (The populateCategory.php script will also recalculate counts, but
  33  * recountCategories only updates rows that need to be updated, making it more
  34  * efficient.)
  35  *
  36  * @ingroup Maintenance
  37  */
  38 class RecountCategories extends Maintenance {
  39         public function __construct() {
  40                 parent::__construct();
  41                 $this->addDescription( <<<'TEXT'
  42 This script refreshes the category membership counts stored in the category
  43 table. As time passes, these counts often drift from the actual number of
  44 category members. The script identifies rows where the value in the category
  45 table does not match the number of categorylinks rows for that category, and
  46 updates the category table accordingly.
  47
  48 To fully refresh the data in the category table, you need to run this script
  49 three times: once in each mode. Alternatively, just one mode can be run if
  50 required.
  51 TEXT
  52                 );
  53                 $this->addOption(
  54                         'mode',
  55                         '(REQUIRED) Which category count column to recompute: "pages", "subcats" or "files".',
  56                         true,
  57                         true
  58                 );
  59                 $this->addOption(
  60                         'begin',
  61                         'Only recount categories with cat_id greater than the given value',
  62                         false,
  63                         true
  64                 );
  65                 $this->addOption(
  66                         'throttle',
  67                         'Wait this many milliseconds after each batch. Default: 0',
  68                         false,
  69                         true
  70                 );
  71
  72                 $this->setBatchSize( 500 );
  73         }
  74
  75         public function execute() {
  76                 $this->mode = $this->getOption( 'mode' );
  77                 if ( !in_array( $this->mode, [ 'pages', 'subcats', 'files' ] ) ) {
  78                         $this->error( 'Please specify a valid mode: one of "pages", "subcats" or "files".', 1 );
  79                 }
  80
  81                 $this->minimumId = intval( $this->getOption( 'begin', 0 ) );
  82
  83                 // do the work, batch by batch
  84                 $affectedRows = 0;
  85                 while ( ( $result = $this->doWork() ) !== false ) {
  86                         $affectedRows += $result;
  87                         usleep( $this->getOption( 'throttle', 0 ) * 1000 );
  88                 }
  89
  90                 $this->output( "Done! Updated the {$this->mode} counts of $affectedRows categories.\n" .
  91                         "Now run the script using the other --mode options if you haven't already.\n" );
  92                 if ( $this->mode === 'pages' ) {
  93                         $this->output(
  94                                 "Also run 'php cleanupEmptyCategories.php --mode remove' to remove empty,\n" .
  95                                 "nonexistent categories from the category table.\n\n" );
  96                 }
  97         }
  98
  99         protected function doWork() {
 100                 $this->output( "Finding up to {$this->mBatchSize} drifted rows " .
 101                         "starting at cat_id {$this->minimumId}...\n" );
 102
 103                 $countingConds = [ 'cl_to = cat_title' ];
 104                 if ( $this->mode === 'subcats' ) {
 105                         $countingConds['cl_type'] = 'subcat';
 106                 } elseif ( $this->mode === 'files' ) {
 107                         $countingConds['cl_type'] = 'file';
 108                 }
 109
 110                 $dbr = $this->getDB( DB_REPLICA, 'vslow' );
 111                 $countingSubquery = $dbr->selectSQLText( 'categorylinks',
 112                         'COUNT(*)',
 113                         $countingConds,
 114                         __METHOD__ );
 115
 116                 // First, let's find out which categories have drifted and need to be updated.
 117                 // The query counts the categorylinks for each category on the replica DB,
 118                 // but this data can't be used for updating the master, so we don't include it
 119                 // in the results.
 120                 $idsToUpdate = $dbr->selectFieldValues( 'category',
 121                         'cat_id',
 122                         [
 123                                 'cat_id > ' . $this->minimumId,
 124                                 "cat_{$this->mode} != ($countingSubquery)"
 125                         ],
 126                         __METHOD__,
 127                         [ 'LIMIT' => $this->mBatchSize ]
 128                 );
 129                 if ( !$idsToUpdate ) {
 130                         return false;
 131                 }
 132                 $this->output( "Updating cat_{$this->mode} field on " .
 133                         count( $idsToUpdate ) . " rows...\n" );
 134
 135                 // In the next batch, start where this query left off. The rows selected
 136                 // in this iteration shouldn't be selected again after being updated, but
 137                 // we still keep track of where we are up to, as extra protection against
 138                 // infinite loops.
 139                 $this->minimumId = end( $idsToUpdate );
 140
 141                 // Now, on master, find the correct counts for these categories.
 142                 $dbw = $this->getDB( DB_MASTER );
 143                 $res = $dbw->select( 'category',
 144                         [ 'cat_id', 'count' => "($countingSubquery)" ],
 145                         [ 'cat_id' => $idsToUpdate ],
 146                         __METHOD__ );
 147
 148                 // Update the category counts on the rows we just identified.
 149                 // This logic is equivalent to Category::refreshCounts, except here, we
 150                 // don't remove rows when cat_pages is zero and the category description page
 151                 // doesn't exist - instead we print a suggestion to run
 152                 // cleanupEmptyCategories.php.
 153                 $affectedRows = 0;
 154                 foreach ( $res as $row ) {
 155                         $dbw->update( 'category',
 156                                 [ "cat_{$this->mode}" => $row->count ],
 157                                 [
 158                                         'cat_id' => $row->cat_id,
 159                                         "cat_{$this->mode} != {$row->count}",
 160                                 ],
 161                                 __METHOD__ );
 162                         $affectedRows += $dbw->affectedRows();
 163                 }
 164
 165                 MediaWikiServices::getInstance()->getDBLoadBalancerFactory()->waitForReplication();
 166
 167                 return $affectedRows;
 168         }
 169 }
 170
 171 $maintClass = 'RecountCategories';
 172 require_once RUN_MAINTENANCE_IF_MAIN;