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
  24 /**
  25  * Update object handling the cleanup of links tables after a page was deleted.
  26  **/
  27 class LinksDeletionUpdate extends DataUpdate implements EnqueueableDataUpdate {
  28         /** @var WikiPage */
  29         protected $page;
  30         /** @var integer */
  31         protected $pageId;
  32         /** @var string */
  33         protected $timestamp;
  34
  35         /** @var IDatabase */
  36         private $db;
  37
  38         /**
  39          * @param WikiPage $page Page we are updating
  40          * @param integer|null $pageId ID of the page we are updating [optional]
  41          * @param string|null $timestamp TS_MW timestamp of deletion
  42          * @throws MWException
  43          */
  44         function __construct( WikiPage $page, $pageId = null, $timestamp = null ) {
  45                 $this->page = $page;
  46                 if ( $pageId ) {
  47                         $this->pageId = $pageId; // page ID at time of deletion
  48                 } elseif ( $page->exists() ) {
  49                         $this->pageId = $page->getId();
  50                 } else {
  51                         throw new InvalidArgumentException( "Page ID not known. Page doesn't exist?" );
  52                 }
  53
  54                 $this->timestamp = $timestamp ?: wfTimestampNow();
  55         }
  56
  57         public function doUpdate() {
  58                 $services = MediaWikiServices::getInstance();
  59                 $config = $services->getMainConfig();
  60                 $lbFactory = $services->getDBLoadBalancerFactory();
  61                 $batchSize = $config->get( 'UpdateRowsPerQuery' );
  62
  63                 // Page may already be deleted, so don't just getId()
  64                 $id = $this->pageId;
  65                 // Make sure all links update threads see the changes of each other.
  66                 // This handles the case when updates have to batched into several COMMITs.
  67                 $scopedLock = LinksUpdate::acquirePageLock( $this->getDB(), $id );
  68
  69                 $title = $this->page->getTitle();
  70                 $dbw = $this->getDB(); // convenience
  71
  72                 // Delete restrictions for it
  73                 $dbw->delete( 'page_restrictions', [ 'pr_page' => $id ], __METHOD__ );
  74
  75                 // Fix category table counts
  76                 $cats = $dbw->selectFieldValues(
  77                         'categorylinks',
  78                         'cl_to',
  79                         [ 'cl_from' => $id ],
  80                         __METHOD__
  81                 );
  82                 $catBatches = array_chunk( $cats, $batchSize );
  83                 foreach ( $catBatches as $catBatch ) {
  84                         $this->page->updateCategoryCounts( [], $catBatch, $id );
  85                         if ( count( $catBatches ) > 1 ) {
  86                                 $lbFactory->commitAndWaitForReplication(
  87                                         __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
  88                                 );
  89                         }
  90                 }
  91
  92                 // Refresh the category table entry if it seems to have no pages. Check
  93                 // master for the most up-to-date cat_pages count.
  94                 if ( $title->getNamespace() === NS_CATEGORY ) {
  95                         $row = $dbw->selectRow(
  96                                 'category',
  97                                 [ 'cat_id', 'cat_title', 'cat_pages', 'cat_subcats', 'cat_files' ],
  98                                 [ 'cat_title' => $title->getDBkey(), 'cat_pages <= 0' ],
  99                                 __METHOD__
 100                         );
 101                         if ( $row ) {
 102                                 Category::newFromRow( $row, $title )->refreshCounts();
 103                         }
 104                 }
 105
 106                 // If using cascading deletes, we can skip some explicit deletes
 107                 if ( !$dbw->cascadingDeletes() ) {
 108                         // Delete outgoing links
 109                         $this->batchDeleteByPK(
 110                                 'pagelinks',
 111                                 [ 'pl_from' => $id ],
 112                                 [ 'pl_from', 'pl_namespace', 'pl_title' ],
 113                                 $batchSize
 114                         );
 115                         $this->batchDeleteByPK(
 116                                 'imagelinks',
 117                                 [ 'il_from' => $id ],
 118                                 [ 'il_from', 'il_to' ],
 119                                 $batchSize
 120                         );
 121                         $this->batchDeleteByPK(
 122                                 'categorylinks',
 123                                 [ 'cl_from' => $id ],
 124                                 [ 'cl_from', 'cl_to' ],
 125                                 $batchSize
 126                         );
 127                         $this->batchDeleteByPK(
 128                                 'templatelinks',
 129                                 [ 'tl_from' => $id ],
 130                                 [ 'tl_from', 'tl_namespace', 'tl_title' ],
 131                                 $batchSize
 132                         );
 133                         $this->batchDeleteByPK(
 134                                 'externallinks',
 135                                 [ 'el_from' => $id ],
 136                                 [ 'el_id' ],
 137                                 $batchSize
 138                         );
 139                         $this->batchDeleteByPK(
 140                                 'langlinks',
 141                                 [ 'll_from' => $id ],
 142                                 [ 'll_from', 'll_lang' ],
 143                                 $batchSize
 144                         );
 145                         $this->batchDeleteByPK(
 146                                 'iwlinks',
 147                                 [ 'iwl_from' => $id ],
 148                                 [ 'iwl_from', 'iwl_prefix', 'iwl_title' ],
 149                                 $batchSize
 150                         );
 151                         // Delete any redirect entry or page props entries
 152                         $dbw->delete( 'redirect', [ 'rd_from' => $id ], __METHOD__ );
 153                         $dbw->delete( 'page_props', [ 'pp_page' => $id ], __METHOD__ );
 154                 }
 155
 156                 // If using cleanup triggers, we can skip some manual deletes
 157                 if ( !$dbw->cleanupTriggers() ) {
 158                         // Find recentchanges entries to clean up...
 159                         $rcIdsForTitle = $dbw->selectFieldValues(
 160                                 'recentchanges',
 161                                 'rc_id',
 162                                 [
 163                                         'rc_type != ' . RC_LOG,
 164                                         'rc_namespace' => $title->getNamespace(),
 165                                         'rc_title' => $title->getDBkey(),
 166                                         'rc_timestamp < ' .
 167                                                 $dbw->addQuotes( $dbw->timestamp( $this->timestamp ) )
 168                                 ],
 169                                 __METHOD__
 170                         );
 171                         $rcIdsForPage = $dbw->selectFieldValues(
 172                                 'recentchanges',
 173                                 'rc_id',
 174                                 [ 'rc_type != ' . RC_LOG, 'rc_cur_id' => $id ],
 175                                 __METHOD__
 176                         );
 177
 178                         // T98706: delete by PK to avoid lock contention with RC delete log insertions
 179                         $rcIdBatches = array_chunk( array_merge( $rcIdsForTitle, $rcIdsForPage ), $batchSize );
 180                         foreach ( $rcIdBatches as $rcIdBatch ) {
 181                                 $dbw->delete( 'recentchanges', [ 'rc_id' => $rcIdBatch ], __METHOD__ );
 182                                 if ( count( $rcIdBatches ) > 1 ) {
 183                                         $lbFactory->commitAndWaitForReplication(
 184                                                 __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
 185                                         );
 186                                 }
 187                         }
 188                 }
 189
 190                 // Commit and release the lock
 191                 ScopedCallback::consume( $scopedLock );
 192         }
 193
 194         private function batchDeleteByPK( $table, array $conds, array $pk, $bSize ) {
 195                 $services = MediaWikiServices::getInstance();
 196                 $lbFactory = $services->getDBLoadBalancerFactory();
 197                 $dbw = $this->getDB(); // convenience
 198
 199                 $res = $dbw->select( $table, $pk, $conds, __METHOD__ );
 200
 201                 $pkDeleteConds = [];
 202                 foreach ( $res as $row ) {
 203                         $pkDeleteConds[] = $dbw->makeList( (array)$row, LIST_AND );
 204                         if ( count( $pkDeleteConds ) >= $bSize ) {
 205                                 $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 206                                 $lbFactory->commitAndWaitForReplication(
 207                                         __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
 208                                 );
 209                                 $pkDeleteConds = [];
 210                         }
 211                 }
 212
 213                 if ( $pkDeleteConds ) {
 214                         $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 215                 }
 216         }
 217
 218         protected function getDB() {
 219                 if ( !$this->db ) {
 220                         $this->db = wfGetDB( DB_MASTER );
 221                 }
 222
 223                 return $this->db;
 224         }
 225
 226         public function getAsJobSpecification() {
 227                 return [
 228                         'wiki' => $this->getDB()->getWikiID(),
 229                         'job'  => new JobSpecification(
 230                                 'deleteLinks',
 231                                 [ 'pageId' => $this->pageId, 'timestamp' => $this->timestamp ],
 232                                 [ 'removeDuplicates' => true ],
 233                                 $this->page->getTitle()
 234                         )
 235                 ];
 236         }
 237 }