maintenance/cleanupEmptyCategories.php

   1 <?php
   2 /**
   3  * Clean up empty categories in the category table.
   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\Category\Category;
  25 use MediaWiki\Maintenance\LoggedUpdateMaintenance;
  26
  27 // @codeCoverageIgnoreStart
  28 require_once __DIR__ . '/Maintenance.php';
  29 // @codeCoverageIgnoreEnd
  30
  31 /**
  32  * Maintenance script to clean up empty categories in the category table.
  33  *
  34  * @ingroup Maintenance
  35  * @since 1.28
  36  */
  37 class CleanupEmptyCategories extends LoggedUpdateMaintenance {
  38
  39         public function __construct() {
  40                 parent::__construct();
  41                 $this->addDescription(
  42                         <<<TEXT
  43 This script will clean up the category table by removing entries for empty
  44 categories without a description page and adding entries for empty categories
  45 with a description page. It will print out progress indicators every batch. The
  46 script is perfectly safe to run on large, live wikis, and running it multiple
  47 times is harmless. You may want to use the throttling options if it's causing
  48 too much load; they will not affect correctness.
  49
  50 If the script is stopped and later resumed, you can use the --mode and --begin
  51 options with the last printed progress indicator to pick up where you left off.
  52
  53 When the script has finished, it will make a note of this in the database, and
  54 will not run again without the --force option.
  55 TEXT
  56                 );
  57
  58                 $this->addOption(
  59                         'mode',
  60                         '"add" empty categories with description pages, "remove" empty categories '
  61                         . 'without description pages, or "both"',
  62                         false,
  63                         true
  64                 );
  65                 $this->addOption(
  66                         'begin',
  67                         'Only do categories whose names are alphabetically after the provided name',
  68                         false,
  69                         true
  70                 );
  71                 $this->addOption(
  72                         'throttle',
  73                         'Wait this many milliseconds after each batch. Default: 0',
  74                         false,
  75                         true
  76                 );
  77         }
  78
  79         protected function getUpdateKey() {
  80                 return 'cleanup empty categories';
  81         }
  82
  83         protected function doDBUpdates() {
  84                 $mode = $this->getOption( 'mode', 'both' );
  85                 $begin = $this->getOption( 'begin', '' );
  86                 $throttle = $this->getOption( 'throttle', 0 );
  87
  88                 if ( !in_array( $mode, [ 'add', 'remove', 'both' ] ) ) {
  89                         $this->output( "--mode must be 'add', 'remove', or 'both'.\n" );
  90                         return false;
  91                 }
  92
  93                 $dbw = $this->getPrimaryDB();
  94
  95                 $throttle = intval( $throttle );
  96
  97                 if ( $mode === 'add' || $mode === 'both' ) {
  98                         if ( $begin !== '' ) {
  99                                 $where = [ $dbw->expr( 'page_title', '>', $begin ) ];
 100                         } else {
 101                                 $where = [];
 102                         }
 103
 104                         $this->output( "Adding empty categories with description pages...\n" );
 105                         while ( true ) {
 106                                 # Find which category to update
 107                                 $rows = $dbw->newSelectQueryBuilder()
 108                                         ->select( 'page_title' )
 109                                         ->from( 'page' )
 110                                         ->leftJoin( 'category', null, 'page_title = cat_title' )
 111                                         ->where( $where )
 112                                         ->andWhere( [ 'page_namespace' => NS_CATEGORY, 'cat_title' => null ] )
 113                                         ->orderBy( 'page_title' )
 114                                         ->limit( $this->getBatchSize() )
 115                                         ->caller( __METHOD__ )->fetchResultSet();
 116                                 if ( !$rows || $rows->numRows() <= 0 ) {
 117                                         break;
 118                                 }
 119
 120                                 foreach ( $rows as $row ) {
 121                                         $name = $row->page_title;
 122                                         $where = [ $dbw->expr( 'page_title', '>', $name ) ];
 123
 124                                         # Use the row to update the category count
 125                                         $cat = Category::newFromName( $name );
 126                                         if ( !is_object( $cat ) ) {
 127                                                 $this->output( "The category named $name is not valid?!\n" );
 128                                         } else {
 129                                                 $cat->refreshCounts();
 130                                         }
 131                                 }
 132                                 // @phan-suppress-next-line PhanPossiblyUndeclaredVariable $rows has at at least one item
 133                                 $this->output( "--mode=$mode --begin=$name\n" );
 134
 135                                 $this->waitForReplication();
 136                                 usleep( $throttle * 1000 );
 137                         }
 138
 139                         $begin = '';
 140                 }
 141
 142                 if ( $mode === 'remove' || $mode === 'both' ) {
 143                         if ( $begin !== '' ) {
 144                                 $where = [ $dbw->expr( 'cat_title', '>', $begin ) ];
 145                         } else {
 146                                 $where = [];
 147                         }
 148
 149                         $this->output( "Removing empty categories without description pages...\n" );
 150                         while ( true ) {
 151                                 # Find which category to update
 152                                 $rows = $dbw->newSelectQueryBuilder()
 153                                         ->select( 'cat_title' )
 154                                         ->from( 'category' )
 155                                         ->leftJoin( 'page', null, [ 'page_namespace' => NS_CATEGORY, 'page_title = cat_title' ] )
 156                                         ->where( $where )
 157                                         ->andWhere( [ 'page_title' => null, 'cat_pages' => 0 ] )
 158                                         ->orderBy( 'cat_title' )
 159                                         ->limit( $this->getBatchSize() )
 160                                         ->caller( __METHOD__ )->fetchResultSet();
 161                                 if ( !$rows || $rows->numRows() <= 0 ) {
 162                                         break;
 163                                 }
 164                                 foreach ( $rows as $row ) {
 165                                         $name = $row->cat_title;
 166                                         $where = [ $dbw->expr( 'cat_title', '>', $name ) ];
 167
 168                                         # Use the row to update the category count
 169                                         $cat = Category::newFromName( $name );
 170                                         if ( !is_object( $cat ) ) {
 171                                                 $this->output( "The category named $name is not valid?!\n" );
 172                                         } else {
 173                                                 $cat->refreshCounts();
 174                                         }
 175                                 }
 176
 177                                 // @phan-suppress-next-line PhanPossiblyUndeclaredVariable rows contains at least one item
 178                                 $this->output( "--mode=remove --begin=$name\n" );
 179
 180                                 $this->waitForReplication();
 181                                 usleep( $throttle * 1000 );
 182                         }
 183                 }
 184
 185                 $this->output( "Category cleanup complete.\n" );
 186
 187                 return true;
 188         }
 189 }
 190
 191 // @codeCoverageIgnoreStart
 192 $maintClass = CleanupEmptyCategories::class;
 193 require_once RUN_MAINTENANCE_IF_MAIN;
 194 // @codeCoverageIgnoreEnd