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                 parent::__construct();
  46
  47                 $this->page = $page;
  48                 if ( $pageId ) {
  49                         $this->pageId = $pageId; // page ID at time of deletion
  50                 } elseif ( $page->exists() ) {
  51                         $this->pageId = $page->getId();
  52                 } else {
  53                         throw new InvalidArgumentException( "Page ID not known. Page doesn't exist?" );
  54                 }
  55
  56                 $this->timestamp = $timestamp ?: wfTimestampNow();
  57         }
  58
  59         public function doUpdate() {
  60                 $services = MediaWikiServices::getInstance();
  61                 $config = $services->getMainConfig();
  62                 $lbFactory = $services->getDBLoadBalancerFactory();
  63                 $batchSize = $config->get( 'UpdateRowsPerQuery' );
  64
  65                 // Page may already be deleted, so don't just getId()
  66                 $id = $this->pageId;
  67                 // Make sure all links update threads see the changes of each other.
  68                 // This handles the case when updates have to batched into several COMMITs.
  69                 $scopedLock = LinksUpdate::acquirePageLock( $this->getDB(), $id );
  70
  71                 $title = $this->page->getTitle();
  72                 $dbw = $this->getDB(); // convenience
  73
  74                 // Delete restrictions for it
  75                 $dbw->delete( 'page_restrictions', [ 'pr_page' => $id ], __METHOD__ );
  76
  77                 // Fix category table counts
  78                 $cats = $dbw->selectFieldValues(
  79                         'categorylinks',
  80                         'cl_to',
  81                         [ 'cl_from' => $id ],
  82                         __METHOD__
  83                 );
  84                 $catBatches = array_chunk( $cats, $batchSize );
  85                 foreach ( $catBatches as $catBatch ) {
  86                         $this->page->updateCategoryCounts( [], $catBatch, $id );
  87                         if ( count( $catBatches ) > 1 ) {
  88                                 $lbFactory->commitAndWaitForReplication(
  89                                         __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
  90                                 );
  91                         }
  92                 }
  93
  94                 // Refresh the category table entry if it seems to have no pages. Check
  95                 // master for the most up-to-date cat_pages count.
  96                 if ( $title->getNamespace() === NS_CATEGORY ) {
  97                         $row = $dbw->selectRow(
  98                                 'category',
  99                                 [ 'cat_id', 'cat_title', 'cat_pages', 'cat_subcats', 'cat_files' ],
 100                                 [ 'cat_title' => $title->getDBkey(), 'cat_pages <= 0' ],
 101                                 __METHOD__
 102                         );
 103                         if ( $row ) {
 104                                 Category::newFromRow( $row, $title )->refreshCounts();
 105                         }
 106                 }
 107
 108                 // If using cascading deletes, we can skip some explicit deletes
 109                 if ( !$dbw->cascadingDeletes() ) {
 110                         // Delete outgoing links
 111                         $this->batchDeleteByPK(
 112                                 'pagelinks',
 113                                 [ 'pl_from' => $id ],
 114                                 [ 'pl_from', 'pl_namespace', 'pl_title' ],
 115                                 $batchSize
 116                         );
 117                         $this->batchDeleteByPK(
 118                                 'imagelinks',
 119                                 [ 'il_from' => $id ],
 120                                 [ 'il_from', 'il_to' ],
 121                                 $batchSize
 122                         );
 123                         $this->batchDeleteByPK(
 124                                 'categorylinks',
 125                                 [ 'cl_from' => $id ],
 126                                 [ 'cl_from', 'cl_to' ],
 127                                 $batchSize
 128                         );
 129                         $this->batchDeleteByPK(
 130                                 'templatelinks',
 131                                 [ 'tl_from' => $id ],
 132                                 [ 'tl_from', 'tl_namespace', 'tl_title' ],
 133                                 $batchSize
 134                         );
 135                         $this->batchDeleteByPK(
 136                                 'externallinks',
 137                                 [ 'el_from' => $id ],
 138                                 [ 'el_id' ],
 139                                 $batchSize
 140                         );
 141                         $this->batchDeleteByPK(
 142                                 'langlinks',
 143                                 [ 'll_from' => $id ],
 144                                 [ 'll_from', 'll_lang' ],
 145                                 $batchSize
 146                         );
 147                         $this->batchDeleteByPK(
 148                                 'iwlinks',
 149                                 [ 'iwl_from' => $id ],
 150                                 [ 'iwl_from', 'iwl_prefix', 'iwl_title' ],
 151                                 $batchSize
 152                         );
 153                         // Delete any redirect entry or page props entries
 154                         $dbw->delete( 'redirect', [ 'rd_from' => $id ], __METHOD__ );
 155                         $dbw->delete( 'page_props', [ 'pp_page' => $id ], __METHOD__ );
 156                 }
 157
 158                 // If using cleanup triggers, we can skip some manual deletes
 159                 if ( !$dbw->cleanupTriggers() ) {
 160                         // Find recentchanges entries to clean up...
 161                         $rcIdsForTitle = $dbw->selectFieldValues(
 162                                 'recentchanges',
 163                                 'rc_id',
 164                                 [
 165                                         'rc_type != ' . RC_LOG,
 166                                         'rc_namespace' => $title->getNamespace(),
 167                                         'rc_title' => $title->getDBkey(),
 168                                         'rc_timestamp < ' .
 169                                                 $dbw->addQuotes( $dbw->timestamp( $this->timestamp ) )
 170                                 ],
 171                                 __METHOD__
 172                         );
 173                         $rcIdsForPage = $dbw->selectFieldValues(
 174                                 'recentchanges',
 175                                 'rc_id',
 176                                 [ 'rc_type != ' . RC_LOG, 'rc_cur_id' => $id ],
 177                                 __METHOD__
 178                         );
 179
 180                         // T98706: delete by PK to avoid lock contention with RC delete log insertions
 181                         $rcIdBatches = array_chunk( array_merge( $rcIdsForTitle, $rcIdsForPage ), $batchSize );
 182                         foreach ( $rcIdBatches as $rcIdBatch ) {
 183                                 $dbw->delete( 'recentchanges', [ 'rc_id' => $rcIdBatch ], __METHOD__ );
 184                                 if ( count( $rcIdBatches ) > 1 ) {
 185                                         $lbFactory->commitAndWaitForReplication(
 186                                                 __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
 187                                         );
 188                                 }
 189                         }
 190                 }
 191
 192                 // Commit and release the lock
 193                 ScopedCallback::consume( $scopedLock );
 194         }
 195
 196         private function batchDeleteByPK( $table, array $conds, array $pk, $bSize ) {
 197                 $services = MediaWikiServices::getInstance();
 198                 $lbFactory = $services->getDBLoadBalancerFactory();
 199                 $dbw = $this->getDB(); // convenience
 200
 201                 $res = $dbw->select( $table, $pk, $conds, __METHOD__ );
 202
 203                 $pkDeleteConds = [];
 204                 foreach ( $res as $row ) {
 205                         $pkDeleteConds[] = $dbw->makeList( (array)$row, LIST_AND );
 206                         if ( count( $pkDeleteConds ) >= $bSize ) {
 207                                 $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 208                                 $lbFactory->commitAndWaitForReplication(
 209                                         __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
 210                                 );
 211                                 $pkDeleteConds = [];
 212                         }
 213                 }
 214
 215                 if ( $pkDeleteConds ) {
 216                         $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 217                 }
 218         }
 219
 220         protected function getDB() {
 221                 if ( !$this->db ) {
 222                         $this->db = wfGetDB( DB_MASTER );
 223                 }
 224
 225                 return $this->db;
 226         }
 227
 228         public function getAsJobSpecification() {
 229                 return [
 230                         'wiki' => $this->getDB()->getWikiID(),
 231                         'job'  => new JobSpecification(
 232                                 'deleteLinks',
 233                                 [ 'pageId' => $this->pageId, 'timestamp' => $this->timestamp ],
 234                                 [ 'removeDuplicates' => true ],
 235                                 $this->page->getTitle()
 236                         )
 237                 ];
 238         }
 239 }