includes/job/jobs/RefreshLinksJob.php

   1 <?php
   2 /**
   3  * Job to update links for a given title.
   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  * Background job to update links for a given title.
  26  *
  27  * @ingroup JobQueue
  28  */
  29 class RefreshLinksJob extends Job {
  30         function __construct( $title, $params = '', $id = 0 ) {
  31                 parent::__construct( 'refreshLinks', $title, $params, $id );
  32                 $this->removeDuplicates = true; // job is expensive
  33         }
  34
  35         /**
  36          * Run a refreshLinks job
  37          * @return boolean success
  38          */
  39         function run() {
  40                 $linkCache = LinkCache::singleton();
  41                 $linkCache->clear();
  42
  43                 if ( is_null( $this->title ) ) {
  44                         $this->error = "refreshLinks: Invalid title";
  45                         return false;
  46                 }
  47
  48                 # Wait for the DB of the current/next slave DB handle to catch up to the master.
  49                 # This way, we get the correct page_latest for templates or files that just changed
  50                 # milliseconds ago, having triggered this job to begin with.
  51                 if ( isset( $this->params['masterPos'] ) ) {
  52                         wfGetLB()->waitFor( $this->params['masterPos'] );
  53                 }
  54
  55                 $revision = Revision::newFromTitle( $this->title, false, Revision::READ_NORMAL );
  56                 if ( !$revision ) {
  57                         $this->error = 'refreshLinks: Article not found "' .
  58                                 $this->title->getPrefixedDBkey() . '"';
  59                         return false; // XXX: what if it was just deleted?
  60                 }
  61
  62                 self::runForTitleInternal( $this->title, $revision, __METHOD__ );
  63
  64                 return true;
  65         }
  66
  67         /**
  68          * @return Array
  69          */
  70         public function getDeduplicationInfo() {
  71                 $info = parent::getDeduplicationInfo();
  72                 // Don't let highly unique "masterPos" values ruin duplicate detection
  73                 if ( is_array( $info['params'] ) ) {
  74                         unset( $info['params']['masterPos'] );
  75                 }
  76                 return $info;
  77         }
  78
  79         /**
  80          * @param $title Title
  81          * @param $revision Revision
  82          * @param $fname string
  83          * @return void
  84          */
  85         public static function runForTitleInternal( Title $title, Revision $revision, $fname ) {
  86                 wfProfileIn( $fname );
  87                 $content = $revision->getContent( Revision::RAW );
  88
  89                 if ( !$content ) {
  90                         // if there is no content, pretend the content is empty
  91                         $content = $revision->getContentHandler()->makeEmptyContent();
  92                 }
  93
  94                 // Revision ID must be passed to the parser output to get revision variables correct
  95                 $parserOutput = $content->getParserOutput( $title, $revision->getId(), null, false );
  96
  97                 $updates = $content->getSecondaryDataUpdates( $title, null, false, $parserOutput );
  98                 DataUpdate::runUpdates( $updates );
  99
 100                 InfoAction::invalidateCache( $title );
 101
 102                 wfProfileOut( $fname );
 103         }
 104 }
 105
 106 /**
 107  * Background job to update links for a given title.
 108  * Newer version for high use templates.
 109  *
 110  * @ingroup JobQueue
 111  */
 112 class RefreshLinksJob2 extends Job {
 113         function __construct( $title, $params, $id = 0 ) {
 114                 parent::__construct( 'refreshLinks2', $title, $params, $id );
 115         }
 116
 117         /**
 118          * Run a refreshLinks2 job
 119          * @return boolean success
 120          */
 121         function run() {
 122                 global $wgUpdateRowsPerJob;
 123
 124                 $linkCache = LinkCache::singleton();
 125                 $linkCache->clear();
 126
 127                 if ( is_null( $this->title ) ) {
 128                         $this->error = "refreshLinks2: Invalid title";
 129                         return false;
 130                 }
 131
 132                 // Back compat for pre-r94435 jobs
 133                 $table = isset( $this->params['table'] ) ? $this->params['table'] : 'templatelinks';
 134
 135                 // Avoid slave lag when fetching templates.
 136                 // When the outermost job is run, we know that the caller that enqueued it must have
 137                 // committed the relevant changes to the DB by now. At that point, record the master
 138                 // position and pass it along as the job recursively breaks into smaller range jobs.
 139                 // Hopefully, when leaf jobs are popped, the slaves will have reached that position.
 140                 if ( isset( $this->params['masterPos'] ) ) {
 141                         $masterPos = $this->params['masterPos'];
 142                 } elseif ( wfGetLB()->getServerCount() > 1 ) {
 143                         $masterPos = wfGetLB()->getMasterPos();
 144                 } else {
 145                         $masterPos = false;
 146                 }
 147
 148                 $tbc = $this->title->getBacklinkCache();
 149
 150                 $jobs = array(); // jobs to insert
 151                 if ( isset( $this->params['start'] ) && isset( $this->params['end'] ) ) {
 152                         # This is a partition job to trigger the insertion of leaf jobs...
 153                         $jobs = array_merge( $jobs, $this->getSingleTitleJobs( $table, $masterPos ) );
 154                 } else {
 155                         # This is a base job to trigger the insertion of partitioned jobs...
 156                         if ( $tbc->getNumLinks( $table, $wgUpdateRowsPerJob + 1 ) <= $wgUpdateRowsPerJob ) {
 157                                 # Just directly insert the single per-title jobs
 158                                 $jobs = array_merge( $jobs, $this->getSingleTitleJobs( $table, $masterPos ) );
 159                         } else {
 160                                 # Insert the partition jobs to make per-title jobs
 161                                 foreach ( $tbc->partition( $table, $wgUpdateRowsPerJob ) as $batch ) {
 162                                         list( $start, $end ) = $batch;
 163                                         $jobs[] = new RefreshLinksJob2( $this->title,
 164                                                 array(
 165                                                         'table' => $table,
 166                                                         'start' => $start,
 167                                                         'end' => $end,
 168                                                         'masterPos' => $masterPos,
 169                                                 ) + $this->getRootJobParams() // carry over information for de-duplication
 170                                         );
 171                                 }
 172                         }
 173                 }
 174
 175                 if ( count( $jobs ) ) {
 176                         JobQueueGroup::singleton()->push( $jobs );
 177                 }
 178
 179                 return true;
 180         }
 181
 182         /**
 183          * @param $table string
 184          * @param $masterPos mixed
 185          * @return Array
 186          */
 187         protected function getSingleTitleJobs( $table, $masterPos ) {
 188                 # The "start"/"end" fields are not set for the base jobs
 189                 $start = isset( $this->params['start'] ) ? $this->params['start'] : false;
 190                 $end = isset( $this->params['end'] ) ? $this->params['end'] : false;
 191                 $titles = $this->title->getBacklinkCache()->getLinks( $table, $start, $end );
 192                 # Convert into single page refresh links jobs.
 193                 # This handles well when in sapi mode and is useful in any case for job
 194                 # de-duplication. If many pages use template A, and that template itself
 195                 # uses template B, then an edit to both will create many duplicate jobs.
 196                 # Roughly speaking, for each page, one of the "RefreshLinksJob" jobs will
 197                 # get run first, and when it does, it will remove the duplicates. Of course,
 198                 # one page could have its job popped when the other page's job is still
 199                 # buried within the logic of a refreshLinks2 job.
 200                 $jobs = array();
 201                 foreach ( $titles as $title ) {
 202                         $jobs[] = new RefreshLinksJob( $title,
 203                                 array( 'masterPos' => $masterPos ) + $this->getRootJobParams()
 204                         ); // carry over information for de-duplication
 205                 }
 206                 return $jobs;
 207         }
 208
 209         /**
 210          * @return Array
 211          */
 212         public function getDeduplicationInfo() {
 213                 $info = parent::getDeduplicationInfo();
 214                 // Don't let highly unique "masterPos" values ruin duplicate detection
 215                 if ( is_array( $info['params'] ) ) {
 216                         unset( $info['params']['masterPos'] );
 217                 }
 218                 return $info;
 219         }
 220 }