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