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
  68                 if ( $this->ticket ) {
  69                         // Make sure all links update threads see the changes of each other.
  70                         // This handles the case when updates have to batched into several COMMITs.
  71                         $scopedLock = LinksUpdate::acquirePageLock( $this->getDB(), $id );
  72                 }
  73
  74                 $title = $this->page->getTitle();
  75                 $dbw = $this->getDB(); // convenience
  76
  77                 // Delete restrictions for it
  78                 $dbw->delete( 'page_restrictions', [ 'pr_page' => $id ], __METHOD__ );
  79
  80                 // Fix category table counts
  81                 $cats = $dbw->selectFieldValues(
  82                         'categorylinks',
  83                         'cl_to',
  84                         [ 'cl_from' => $id ],
  85                         __METHOD__
  86                 );
  87                 $catBatches = array_chunk( $cats, $batchSize );
  88                 foreach ( $catBatches as $catBatch ) {
  89                         $this->page->updateCategoryCounts( [], $catBatch, $id );
  90                         if ( count( $catBatches ) > 1 ) {
  91                                 $lbFactory->commitAndWaitForReplication(
  92                                         __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
  93                                 );
  94                         }
  95                 }
  96
  97                 // Refresh the category table entry if it seems to have no pages. Check
  98                 // master for the most up-to-date cat_pages count.
  99                 if ( $title->getNamespace() === NS_CATEGORY ) {
 100                         $row = $dbw->selectRow(
 101                                 'category',
 102                                 [ 'cat_id', 'cat_title', 'cat_pages', 'cat_subcats', 'cat_files' ],
 103                                 [ 'cat_title' => $title->getDBkey(), 'cat_pages <= 0' ],
 104                                 __METHOD__
 105                         );
 106                         if ( $row ) {
 107                                 Category::newFromRow( $row, $title )->refreshCounts();
 108                         }
 109                 }
 110
 111                 // If using cascading deletes, we can skip some explicit deletes
 112                 if ( !$dbw->cascadingDeletes() ) {
 113                         // Delete outgoing links
 114                         $this->batchDeleteByPK(
 115                                 'pagelinks',
 116                                 [ 'pl_from' => $id ],
 117                                 [ 'pl_from', 'pl_namespace', 'pl_title' ],
 118                                 $batchSize
 119                         );
 120                         $this->batchDeleteByPK(
 121                                 'imagelinks',
 122                                 [ 'il_from' => $id ],
 123                                 [ 'il_from', 'il_to' ],
 124                                 $batchSize
 125                         );
 126                         $this->batchDeleteByPK(
 127                                 'categorylinks',
 128                                 [ 'cl_from' => $id ],
 129                                 [ 'cl_from', 'cl_to' ],
 130                                 $batchSize
 131                         );
 132                         $this->batchDeleteByPK(
 133                                 'templatelinks',
 134                                 [ 'tl_from' => $id ],
 135                                 [ 'tl_from', 'tl_namespace', 'tl_title' ],
 136                                 $batchSize
 137                         );
 138                         $this->batchDeleteByPK(
 139                                 'externallinks',
 140                                 [ 'el_from' => $id ],
 141                                 [ 'el_id' ],
 142                                 $batchSize
 143                         );
 144                         $this->batchDeleteByPK(
 145                                 'langlinks',
 146                                 [ 'll_from' => $id ],
 147                                 [ 'll_from', 'll_lang' ],
 148                                 $batchSize
 149                         );
 150                         $this->batchDeleteByPK(
 151                                 'iwlinks',
 152                                 [ 'iwl_from' => $id ],
 153                                 [ 'iwl_from', 'iwl_prefix', 'iwl_title' ],
 154                                 $batchSize
 155                         );
 156                         // Delete any redirect entry or page props entries
 157                         $dbw->delete( 'redirect', [ 'rd_from' => $id ], __METHOD__ );
 158                         $dbw->delete( 'page_props', [ 'pp_page' => $id ], __METHOD__ );
 159                 }
 160
 161                 // If using cleanup triggers, we can skip some manual deletes
 162                 if ( !$dbw->cleanupTriggers() ) {
 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, [ 'wiki' => $dbw->getWikiID() ]
 190                                         );
 191                                 }
 192                         }
 193                 }
 194
 195                 // Commit and release the lock (if set)
 196                 ScopedCallback::consume( $scopedLock );
 197         }
 198
 199         private function batchDeleteByPK( $table, array $conds, array $pk, $bSize ) {
 200                 $services = MediaWikiServices::getInstance();
 201                 $lbFactory = $services->getDBLoadBalancerFactory();
 202                 $dbw = $this->getDB(); // convenience
 203
 204                 $res = $dbw->select( $table, $pk, $conds, __METHOD__ );
 205
 206                 $pkDeleteConds = [];
 207                 foreach ( $res as $row ) {
 208                         $pkDeleteConds[] = $dbw->makeList( (array)$row, LIST_AND );
 209                         if ( count( $pkDeleteConds ) >= $bSize ) {
 210                                 $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 211                                 $lbFactory->commitAndWaitForReplication(
 212                                         __METHOD__, $this->ticket, [ 'wiki' => $dbw->getWikiID() ]
 213                                 );
 214                                 $pkDeleteConds = [];
 215                         }
 216                 }
 217
 218                 if ( $pkDeleteConds ) {
 219                         $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 220                 }
 221         }
 222
 223         protected function getDB() {
 224                 if ( !$this->db ) {
 225                         $this->db = wfGetDB( DB_MASTER );
 226                 }
 227
 228                 return $this->db;
 229         }
 230
 231         public function getAsJobSpecification() {
 232                 return [
 233                         'wiki' => $this->getDB()->getWikiID(),
 234                         'job'  => new JobSpecification(
 235                                 'deleteLinks',
 236                                 [ 'pageId' => $this->pageId, 'timestamp' => $this->timestamp ],
 237                                 [ 'removeDuplicates' => true ],
 238                                 $this->page->getTitle()
 239                         )
 240                 ];
 241         }
 242 }