includes/deferred/LinksDeletionUpdate.php

   1 <?php
   2 /**
   3  * Updater for link tracking tables after a page edit.
   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  */
  22 use MediaWiki\MediaWikiServices;
  23 use Wikimedia\ScopedCallback;
  24 use Wikimedia\Rdbms\IDatabase;
  25
  26 /**
  27  * Update object handling the cleanup of links tables after a page was deleted.
  28  */
  29 class LinksDeletionUpdate extends DataUpdate implements EnqueueableDataUpdate {
  30         /** @var WikiPage */
  31         protected $page;
  32         /** @var int */
  33         protected $pageId;
  34         /** @var string */
  35         protected $timestamp;
  36
  37         /** @var IDatabase */
  38         private $db;
  39
  40         /**
  41          * @param WikiPage $page Page we are updating
  42          * @param int|null $pageId ID of the page we are updating [optional]
  43          * @param string|null $timestamp TS_MW timestamp of deletion
  44          * @throws MWException
  45          */
  46         function __construct( WikiPage $page, $pageId = null, $timestamp = null ) {
  47                 parent::__construct();
  48
  49                 $this->page = $page;
  50                 if ( $pageId ) {
  51                         $this->pageId = $pageId; // page ID at time of deletion
  52                 } elseif ( $page->exists() ) {
  53                         $this->pageId = $page->getId();
  54                 } else {
  55                         throw new InvalidArgumentException( "Page ID not known. Page doesn't exist?" );
  56                 }
  57
  58                 $this->timestamp = $timestamp ?: wfTimestampNow();
  59         }
  60
  61         public function doUpdate() {
  62                 $services = MediaWikiServices::getInstance();
  63                 $config = $services->getMainConfig();
  64                 $lbFactory = $services->getDBLoadBalancerFactory();
  65                 $batchSize = $config->get( 'UpdateRowsPerQuery' );
  66
  67                 // Page may already be deleted, so don't just getId()
  68                 $id = $this->pageId;
  69
  70                 if ( $this->ticket ) {
  71                         // Make sure all links update threads see the changes of each other.
  72                         // This handles the case when updates have to batched into several COMMITs.
  73                         $scopedLock = LinksUpdate::acquirePageLock( $this->getDB(), $id );
  74                 }
  75
  76                 $title = $this->page->getTitle();
  77                 $dbw = $this->getDB(); // convenience
  78
  79                 // Delete restrictions for it
  80                 $dbw->delete( 'page_restrictions', [ 'pr_page' => $id ], __METHOD__ );
  81
  82                 // Fix category table counts
  83                 $cats = $dbw->selectFieldValues(
  84                         'categorylinks',
  85                         'cl_to',
  86                         [ 'cl_from' => $id ],
  87                         __METHOD__
  88                 );
  89                 $catBatches = array_chunk( $cats, $batchSize );
  90                 foreach ( $catBatches as $catBatch ) {
  91                         $this->page->updateCategoryCounts( [], $catBatch, $id );
  92                         if ( count( $catBatches ) > 1 ) {
  93                                 $lbFactory->commitAndWaitForReplication(
  94                                         __METHOD__, $this->ticket, [ 'domain' => $dbw->getDomainID() ]
  95                                 );
  96                         }
  97                 }
  98
  99                 // Refresh counts on categories that should be empty now
 100                 if ( $title->getNamespace() === NS_CATEGORY ) {
 101                         $row = $dbw->selectRow(
 102                                 'category',
 103                                 [ 'cat_id', 'cat_title', 'cat_pages', 'cat_subcats', 'cat_files' ],
 104                                 [ 'cat_title' => $title->getDBkey(), 'cat_pages <= 100' ],
 105                                 __METHOD__
 106                         );
 107                         if ( $row ) {
 108                                 $cat = Category::newFromRow( $row, $title );
 109                                 // T166757: do the update after the main job DB commit
 110                                 DeferredUpdates::addCallableUpdate( function () use ( $cat ) {
 111                                         $cat->refreshCounts();
 112                                 } );
 113                         }
 114                 }
 115
 116                 $this->batchDeleteByPK(
 117                         'pagelinks',
 118                         [ 'pl_from' => $id ],
 119                         [ 'pl_from', 'pl_namespace', 'pl_title' ],
 120                         $batchSize
 121                 );
 122                 $this->batchDeleteByPK(
 123                         'imagelinks',
 124                         [ 'il_from' => $id ],
 125                         [ 'il_from', 'il_to' ],
 126                         $batchSize
 127                 );
 128                 $this->batchDeleteByPK(
 129                         'categorylinks',
 130                         [ 'cl_from' => $id ],
 131                         [ 'cl_from', 'cl_to' ],
 132                         $batchSize
 133                 );
 134                 $this->batchDeleteByPK(
 135                         'templatelinks',
 136                         [ 'tl_from' => $id ],
 137                         [ 'tl_from', 'tl_namespace', 'tl_title' ],
 138                         $batchSize
 139                 );
 140                 $this->batchDeleteByPK(
 141                         'externallinks',
 142                         [ 'el_from' => $id ],
 143                         [ 'el_id' ],
 144                         $batchSize
 145                 );
 146                 $this->batchDeleteByPK(
 147                         'langlinks',
 148                         [ 'll_from' => $id ],
 149                         [ 'll_from', 'll_lang' ],
 150                         $batchSize
 151                 );
 152                 $this->batchDeleteByPK(
 153                         'iwlinks',
 154                         [ 'iwl_from' => $id ],
 155                         [ 'iwl_from', 'iwl_prefix', 'iwl_title' ],
 156                         $batchSize
 157                 );
 158
 159                 // Delete any redirect entry or page props entries
 160                 $dbw->delete( 'redirect', [ 'rd_from' => $id ], __METHOD__ );
 161                 $dbw->delete( 'page_props', [ 'pp_page' => $id ], __METHOD__ );
 162
 163                 // Find recentchanges entries to clean up...
 164                 $rcIdsForTitle = $dbw->selectFieldValues(
 165                         'recentchanges',
 166                         'rc_id',
 167                         [
 168                                 'rc_type != ' . RC_LOG,
 169                                 'rc_namespace' => $title->getNamespace(),
 170                                 'rc_title' => $title->getDBkey(),
 171                                 'rc_timestamp < ' .
 172                                         $dbw->addQuotes( $dbw->timestamp( $this->timestamp ) )
 173                         ],
 174                         __METHOD__
 175                 );
 176                 $rcIdsForPage = $dbw->selectFieldValues(
 177                         'recentchanges',
 178                         'rc_id',
 179                         [ 'rc_type != ' . RC_LOG, 'rc_cur_id' => $id ],
 180                         __METHOD__
 181                 );
 182
 183                 // T98706: delete by PK to avoid lock contention with RC delete log insertions
 184                 $rcIdBatches = array_chunk( array_merge( $rcIdsForTitle, $rcIdsForPage ), $batchSize );
 185                 foreach ( $rcIdBatches as $rcIdBatch ) {
 186                         $dbw->delete( 'recentchanges', [ 'rc_id' => $rcIdBatch ], __METHOD__ );
 187                         if ( count( $rcIdBatches ) > 1 ) {
 188                                 $lbFactory->commitAndWaitForReplication(
 189                                         __METHOD__, $this->ticket, [ 'domain' => $dbw->getDomainID() ]
 190                                 );
 191                         }
 192                 }
 193
 194                 // Commit and release the lock (if set)
 195                 ScopedCallback::consume( $scopedLock );
 196         }
 197
 198         private function batchDeleteByPK( $table, array $conds, array $pk, $bSize ) {
 199                 $services = MediaWikiServices::getInstance();
 200                 $lbFactory = $services->getDBLoadBalancerFactory();
 201                 $dbw = $this->getDB(); // convenience
 202
 203                 $res = $dbw->select( $table, $pk, $conds, __METHOD__ );
 204
 205                 $pkDeleteConds = [];
 206                 foreach ( $res as $row ) {
 207                         $pkDeleteConds[] = $dbw->makeList( (array)$row, LIST_AND );
 208                         if ( count( $pkDeleteConds ) >= $bSize ) {
 209                                 $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 210                                 $lbFactory->commitAndWaitForReplication(
 211                                         __METHOD__, $this->ticket, [ 'domain' => $dbw->getDomainID() ]
 212                                 );
 213                                 $pkDeleteConds = [];
 214                         }
 215                 }
 216
 217                 if ( $pkDeleteConds ) {
 218                         $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 219                 }
 220         }
 221
 222         protected function getDB() {
 223                 if ( !$this->db ) {
 224                         $this->db = wfGetDB( DB_MASTER );
 225                 }
 226
 227                 return $this->db;
 228         }
 229
 230         public function getAsJobSpecification() {
 231                 return [
 232                         'wiki' => WikiMap::getWikiIdFromDomain( $this->getDB()->getDomainID() ),
 233                         'job'  => new JobSpecification(
 234                                 'deleteLinks',
 235                                 [ 'pageId' => $this->pageId, 'timestamp' => $this->timestamp ],
 236                                 [ 'removeDuplicates' => true ],
 237                                 $this->page->getTitle()
 238                         )
 239                 ];
 240         }
 241 }