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                                 // Only sacrifice atomicity if necessary due to size
  94                                 $lbFactory->commitAndWaitForReplication(
  95                                         __METHOD__, $this->ticket, [ 'domain' => $dbw->getDomainID() ]
  96                                 );
  97                         }
  98                 }
  99
 100                 // Refresh counts on categories that should be empty now
 101                 if ( $title->getNamespace() === NS_CATEGORY ) {
 102                         // T166757: do the update after the main job DB commit
 103                         DeferredUpdates::addCallableUpdate( function () use ( $title ) {
 104                                 $this->refreshCategoryIfEmpty( $title );
 105                         } );
 106                 }
 107
 108                 $this->batchDeleteByPK(
 109                         'pagelinks',
 110                         [ 'pl_from' => $id ],
 111                         [ 'pl_from', 'pl_namespace', 'pl_title' ],
 112                         $batchSize
 113                 );
 114                 $this->batchDeleteByPK(
 115                         'imagelinks',
 116                         [ 'il_from' => $id ],
 117                         [ 'il_from', 'il_to' ],
 118                         $batchSize
 119                 );
 120                 $this->batchDeleteByPK(
 121                         'categorylinks',
 122                         [ 'cl_from' => $id ],
 123                         [ 'cl_from', 'cl_to' ],
 124                         $batchSize
 125                 );
 126                 $this->batchDeleteByPK(
 127                         'templatelinks',
 128                         [ 'tl_from' => $id ],
 129                         [ 'tl_from', 'tl_namespace', 'tl_title' ],
 130                         $batchSize
 131                 );
 132                 $this->batchDeleteByPK(
 133                         'externallinks',
 134                         [ 'el_from' => $id ],
 135                         [ 'el_id' ],
 136                         $batchSize
 137                 );
 138                 $this->batchDeleteByPK(
 139                         'langlinks',
 140                         [ 'll_from' => $id ],
 141                         [ 'll_from', 'll_lang' ],
 142                         $batchSize
 143                 );
 144                 $this->batchDeleteByPK(
 145                         'iwlinks',
 146                         [ 'iwl_from' => $id ],
 147                         [ 'iwl_from', 'iwl_prefix', 'iwl_title' ],
 148                         $batchSize
 149                 );
 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                 // Find recentchanges entries to clean up...
 156                 $rcIdsForTitle = $dbw->selectFieldValues(
 157                         'recentchanges',
 158                         'rc_id',
 159                         [
 160                                 'rc_type != ' . RC_LOG,
 161                                 'rc_namespace' => $title->getNamespace(),
 162                                 'rc_title' => $title->getDBkey(),
 163                                 'rc_timestamp < ' .
 164                                         $dbw->addQuotes( $dbw->timestamp( $this->timestamp ) )
 165                         ],
 166                         __METHOD__
 167                 );
 168                 $rcIdsForPage = $dbw->selectFieldValues(
 169                         'recentchanges',
 170                         'rc_id',
 171                         [ 'rc_type != ' . RC_LOG, 'rc_cur_id' => $id ],
 172                         __METHOD__
 173                 );
 174
 175                 // T98706: delete by PK to avoid lock contention with RC delete log insertions
 176                 $rcIdBatches = array_chunk( array_merge( $rcIdsForTitle, $rcIdsForPage ), $batchSize );
 177                 foreach ( $rcIdBatches as $rcIdBatch ) {
 178                         $dbw->delete( 'recentchanges', [ 'rc_id' => $rcIdBatch ], __METHOD__ );
 179                         if ( count( $rcIdBatches ) > 1 ) {
 180                                 $lbFactory->commitAndWaitForReplication(
 181                                         __METHOD__, $this->ticket, [ 'domain' => $dbw->getDomainID() ]
 182                                 );
 183                         }
 184                 }
 185
 186                 // Commit and release the lock (if set)
 187                 ScopedCallback::consume( $scopedLock );
 188         }
 189
 190         /**
 191          * @param Title $title
 192          */
 193         private function refreshCategoryIfEmpty( Title $title ) {
 194                 $dbw = $this->getDB();
 195
 196                 $row = $dbw->selectRow(
 197                         'category',
 198                         [ 'cat_id', 'cat_title', 'cat_pages', 'cat_subcats', 'cat_files' ],
 199                         [ 'cat_title' => $title->getDBkey(), 'cat_pages <= 100' ],
 200                         __METHOD__
 201                 );
 202
 203                 if ( !$row ) {
 204                         return; // nothing to delete
 205                 }
 206
 207                 $cat = Category::newFromRow( $row, $title );
 208                 $hasLink = $dbw->selectField(
 209                         'categorylinks',
 210                         '1',
 211                         [ 'cl_to' => $title->getDBkey() ],
 212                         __METHOD__
 213                 );
 214                 if ( !$hasLink ) {
 215                         $cat->refreshCounts(); // delete the category table entry
 216                 }
 217         }
 218
 219         private function batchDeleteByPK( $table, array $conds, array $pk, $bSize ) {
 220                 $services = MediaWikiServices::getInstance();
 221                 $lbFactory = $services->getDBLoadBalancerFactory();
 222                 $dbw = $this->getDB(); // convenience
 223
 224                 $res = $dbw->select( $table, $pk, $conds, __METHOD__ );
 225
 226                 $pkDeleteConds = [];
 227                 foreach ( $res as $row ) {
 228                         $pkDeleteConds[] = $dbw->makeList( (array)$row, LIST_AND );
 229                         if ( count( $pkDeleteConds ) >= $bSize ) {
 230                                 $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 231                                 $lbFactory->commitAndWaitForReplication(
 232                                         __METHOD__, $this->ticket, [ 'domain' => $dbw->getDomainID() ]
 233                                 );
 234                                 $pkDeleteConds = [];
 235                         }
 236                 }
 237
 238                 if ( $pkDeleteConds ) {
 239                         $dbw->delete( $table, $dbw->makeList( $pkDeleteConds, LIST_OR ), __METHOD__ );
 240                 }
 241         }
 242
 243         protected function getDB() {
 244                 if ( !$this->db ) {
 245                         $this->db = wfGetDB( DB_MASTER );
 246                 }
 247
 248                 return $this->db;
 249         }
 250
 251         public function getAsJobSpecification() {
 252                 return [
 253                         'wiki' => WikiMap::getWikiIdFromDomain( $this->getDB()->getDomainID() ),
 254                         'job'  => new JobSpecification(
 255                                 'deleteLinks',
 256                                 [ 'pageId' => $this->pageId, 'timestamp' => $this->timestamp ],
 257                                 [ 'removeDuplicates' => true ],
 258                                 $this->page->getTitle()
 259                         )
 260                 ];
 261         }
 262 }