includes/job/jobs/RefreshLinksJob.php

   1 <?php
   2 /**
   3  * Job to update link tables for pages
   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  * @ingroup JobQueue
  22  */
  23
  24 /**
  25  * Job to update link tables for pages
  26  *
  27  * This job comes in a few variants:
  28  *   - a) Recursive jobs to update links for backlink pages for a given title
  29  *   - b) Jobs to update links for a set of titles (the job title is ignored)
  30  *   - c) Jobs to update links for a single title (the job title)
  31  *
  32  * @ingroup JobQueue
  33  */
  34 class RefreshLinksJob extends Job {
  35         const VERSION = 1;
  36
  37         function __construct( $title, $params = '', $id = 0 ) {
  38                 parent::__construct( 'refreshLinks', $title, $params, $id );
  39                 $this->params['version'] = self::VERSION;
  40                 // Base backlink update jobs and per-title update jobs can be de-duplicated.
  41                 // If template A changes twice before any jobs run, a clean queue will have:
  42                 //              (A base, A base)
  43                 // The second job is ignored by the queue on insertion.
  44                 // Suppose, many pages use template A, and that template itself uses template B.
  45                 // An edit to both will first create two base jobs. A clean FIFO queue will have:
  46                 //              (A base, B base)
  47                 // When these jobs run, the queue will have per-title and remnant partition jobs:
  48                 //              (titleX,titleY,titleZ,...,A remnant,titleM,titleN,titleO,...,B remnant)
  49                 // Some these jobs will be the same, and will automatically be ignored by
  50                 // the queue upon insertion. Some title jobs will run before the duplicate is
  51                 // inserted, so the work will still be done twice in those cases. More titles
  52                 // can be de-duplicated as the remnant jobs continue to be broken down. This
  53                 // works best when $wgUpdateRowsPerJob, and either the pages have few backlinks
  54                 // and/or the backlink sets for pages A and B are almost identical.
  55                 $this->removeDuplicates = !isset( $params['range'] )
  56                         && ( !isset( $params['pages'] ) || count( $params['pages'] ) == 1 );
  57         }
  58
  59         function run() {
  60                 global $wgUpdateRowsPerJob;
  61
  62                 if ( is_null( $this->title ) ) {
  63                         $this->setLastError( "Invalid page title" );
  64                         return false;
  65                 }
  66
  67                 // Job to update all (or a range of) backlink pages for a page
  68                 if ( isset( $this->params['recursive'] ) ) {
  69                         // Carry over information for de-duplication
  70                         $extraParams = $this->getRootJobParams();
  71                         // Avoid slave lag when fetching templates.
  72                         // When the outermost job is run, we know that the caller that enqueued it must have
  73                         // committed the relevant changes to the DB by now. At that point, record the master
  74                         // position and pass it along as the job recursively breaks into smaller range jobs.
  75                         // Hopefully, when leaf jobs are popped, the slaves will have reached that position.
  76                         if ( isset( $this->params['masterPos'] ) ) {
  77                                 $extraParams['masterPos'] = $this->params['masterPos'];
  78                         } elseif ( wfGetLB()->getServerCount() > 1 ) {
  79                                 $extraParams['masterPos'] = wfGetLB()->getMasterPos();
  80                         } else {
  81                                 $extraParams['masterPos'] = false;
  82                         }
  83                         // Convert this into no more than $wgUpdateRowsPerJob RefreshLinks per-title
  84                         // jobs and possibly a recursive RefreshLinks job for the rest of the backlinks
  85                         $jobs = BacklinkJobUtils::partitionBacklinkJob(
  86                                 $this,
  87                                 $wgUpdateRowsPerJob,
  88                                 1, // job-per-title
  89                                 array( 'params' => $extraParams )
  90                         );
  91                         JobQueueGroup::singleton()->push( $jobs );
  92                 // Job to update link tables for for a set of titles
  93                 } elseif ( isset( $this->params['pages'] ) ) {
  94                         foreach ( $this->params['pages'] as $pageId => $nsAndKey ) {
  95                                 list( $ns, $dbKey ) = $nsAndKey;
  96                                 $this->runForTitle( Title::makeTitleSafe( $ns, $dbKey ) );
  97                         }
  98                 // Job to update link tables for a given title
  99                 } else {
 100                         $this->runForTitle( $this->title );
 101                 }
 102
 103                 return true;
 104         }
 105
 106         protected function runForTitle( Title $title = null ) {
 107                 $linkCache = LinkCache::singleton();
 108                 $linkCache->clear();
 109
 110                 if ( is_null( $title ) ) {
 111                         $this->setLastError( "refreshLinks: Invalid title" );
 112                         return false;
 113                 }
 114
 115                 // Wait for the DB of the current/next slave DB handle to catch up to the master.
 116                 // This way, we get the correct page_latest for templates or files that just changed
 117                 // milliseconds ago, having triggered this job to begin with.
 118                 if ( isset( $this->params['masterPos'] ) && $this->params['masterPos'] !== false ) {
 119                         wfGetLB()->waitFor( $this->params['masterPos'] );
 120                 }
 121
 122                 // Fetch the current revision...
 123                 $revision = Revision::newFromTitle( $title, false, Revision::READ_NORMAL );
 124                 if ( !$revision ) {
 125                         $this->setLastError( "refreshLinks: Article not found {$title->getPrefixedDBkey()}" );
 126                         return false; // XXX: what if it was just deleted?
 127                 }
 128                 $content = $revision->getContent( Revision::RAW );
 129                 if ( !$content ) {
 130                         // If there is no content, pretend the content is empty
 131                         $content = $revision->getContentHandler()->makeEmptyContent();
 132                 }
 133
 134                 $parserOutput = false;
 135                 // If page_touched changed after this root job (with a good slave lag skew factor),
 136                 // then it is likely that any views of the pages already resulted in re-parses which
 137                 // are now in cache. This can be reused to avoid expensive parsing in some cases.
 138                 if ( isset( $this->params['rootJobTimestamp'] ) ) {
 139                         $page = WikiPage::factory( $title );
 140                         $skewedTimestamp = wfTimestamp( TS_UNIX, $this->params['rootJobTimestamp'] ) + 5;
 141                         if ( $page->getTouched() > wfTimestamp( TS_MW, $skewedTimestamp ) ) {
 142                                 $parserOptions = $page->makeParserOptions( 'canonical' );
 143                                 $parserOutput = ParserCache::singleton()->getDirty( $page, $parserOptions );
 144                                 if ( $parserOutput && $parserOutput->getCacheTime() <= $skewedTimestamp ) {
 145                                         $parserOutput = false; // too stale
 146                                 }
 147                         }
 148                 }
 149                 // Fetch the current revision and parse it if necessary...
 150                 if ( $parserOutput == false ) {
 151                         // Revision ID must be passed to the parser output to get revision variables correct
 152                         $parserOutput = $content->getParserOutput( $title, $revision->getId(), null, false );
 153                 }
 154
 155                 $updates = $content->getSecondaryDataUpdates( $title, null, false, $parserOutput );
 156                 DataUpdate::runUpdates( $updates );
 157
 158                 InfoAction::invalidateCache( $title );
 159
 160                 return true;
 161         }
 162
 163         public function getDeduplicationInfo() {
 164                 $info = parent::getDeduplicationInfo();
 165                 if ( is_array( $info['params'] ) ) {
 166                         // Don't let highly unique "masterPos" values ruin duplicate detection
 167                         unset( $info['params']['masterPos'] );
 168                         // For per-pages jobs, the job title is that of the template that changed
 169                         // (or similar), so remove that since it ruins duplicate detection
 170                         if ( isset( $info['pages'] ) ) {
 171                                 unset( $info['namespace'] );
 172                                 unset( $info['title'] );
 173                         }
 174                 }
 175
 176                 return $info;
 177         }
 178 }