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                 wfProfileIn( __METHOD__ );
  41
  42                 $linkCache = LinkCache::singleton();
  43                 $linkCache->clear();
  44
  45                 if ( is_null( $this->title ) ) {
  46                         $this->error = "refreshLinks: Invalid title";
  47                         wfProfileOut( __METHOD__ );
  48                         return false;
  49                 }
  50
  51                 # Wait for the DB of the current/next slave DB handle to catch up to the master.
  52                 # This way, we get the correct page_latest for templates or files that just changed
  53                 # milliseconds ago, having triggered this job to begin with.
  54                 if ( isset( $this->params['masterPos'] ) ) {
  55                         wfGetLB()->waitFor( $this->params['masterPos'] );
  56                 }
  57
  58                 $revision = Revision::newFromTitle( $this->title, false, Revision::READ_NORMAL );
  59                 if ( !$revision ) {
  60                         $this->error = 'refreshLinks: Article not found "' .
  61                                 $this->title->getPrefixedDBkey() . '"';
  62                         wfProfileOut( __METHOD__ );
  63                         return false; // XXX: what if it was just deleted?
  64                 }
  65
  66                 self::runForTitleInternal( $this->title, $revision, __METHOD__ );
  67
  68                 wfProfileOut( __METHOD__ );
  69                 return true;
  70         }
  71
  72         /**
  73          * @return Array
  74          */
  75         public function getDeduplicationInfo() {
  76                 $info = parent::getDeduplicationInfo();
  77                 // Don't let highly unique "masterPos" values ruin duplicate detection
  78                 if ( is_array( $info['params'] ) ) {
  79                         unset( $info['params']['masterPos'] );
  80                 }
  81                 return $info;
  82         }
  83
  84         /**
  85          * @param $title Title
  86          * @param $revision Revision
  87          * @param $fname string
  88          * @return void
  89          */
  90         public static function runForTitleInternal( Title $title, Revision $revision, $fname ) {
  91                 wfProfileIn( $fname );
  92                 $content = $revision->getContent( Revision::RAW );
  93
  94                 if ( !$content ) {
  95                         // if there is no content, pretend the content is empty
  96                         $content = $revision->getContentHandler()->makeEmptyContent();
  97                 }
  98
  99                 // Revision ID must be passed to the parser output to get revision variables correct
 100                 $parserOutput = $content->getParserOutput( $title, $revision->getId(), null, false );
 101
 102                 $updates = $content->getSecondaryDataUpdates( $title, null, false, $parserOutput );
 103                 DataUpdate::runUpdates( $updates );
 104                 wfProfileOut( $fname );
 105         }
 106 }
 107
 108 /**
 109  * Background job to update links for a given title.
 110  * Newer version for high use templates.
 111  *
 112  * @ingroup JobQueue
 113  */
 114 class RefreshLinksJob2 extends Job {
 115         function __construct( $title, $params, $id = 0 ) {
 116                 parent::__construct( 'refreshLinks2', $title, $params, $id );
 117         }
 118
 119         /**
 120          * Run a refreshLinks2 job
 121          * @return boolean success
 122          */
 123         function run() {
 124                 global $wgUpdateRowsPerJob;
 125
 126                 wfProfileIn( __METHOD__ );
 127
 128                 $linkCache = LinkCache::singleton();
 129                 $linkCache->clear();
 130
 131                 if ( is_null( $this->title ) ) {
 132                         $this->error = "refreshLinks2: Invalid title";
 133                         wfProfileOut( __METHOD__ );
 134                         return false;
 135                 }
 136
 137                 // Back compat for pre-r94435 jobs
 138                 $table = isset( $this->params['table'] ) ? $this->params['table'] : 'templatelinks';
 139
 140                 // Avoid slave lag when fetching templates.
 141                 // When the outermost job is run, we know that the caller that enqueued it must have
 142                 // committed the relevant changes to the DB by now. At that point, record the master
 143                 // position and pass it along as the job recursively breaks into smaller range jobs.
 144                 // Hopefully, when leaf jobs are popped, the slaves will have reached that position.
 145                 if ( isset( $this->params['masterPos'] ) ) {
 146                         $masterPos = $this->params['masterPos'];
 147                 } elseif ( wfGetLB()->getServerCount() > 1  ) {
 148                         $masterPos = wfGetLB()->getMasterPos();
 149                 } else {
 150                         $masterPos = false;
 151                 }
 152
 153                 $tbc = $this->title->getBacklinkCache();
 154
 155                 $jobs = array(); // jobs to insert
 156                 if ( isset( $this->params['start'] ) && isset( $this->params['end'] ) ) {
 157                         # This is a partition job to trigger the insertion of leaf jobs...
 158                         $jobs = array_merge( $jobs, $this->getSingleTitleJobs( $table, $masterPos ) );
 159                 } else {
 160                         # This is a base job to trigger the insertion of partitioned jobs...
 161                         if ( $tbc->getNumLinks( $table ) <= $wgUpdateRowsPerJob ) {
 162                                 # Just directly insert the single per-title jobs
 163                                 $jobs = array_merge( $jobs, $this->getSingleTitleJobs( $table, $masterPos ) );
 164                         } else {
 165                                 # Insert the partition jobs to make per-title jobs
 166                                 foreach ( $tbc->partition( $table, $wgUpdateRowsPerJob ) as $batch ) {
 167                                         list( $start, $end ) = $batch;
 168                                         $jobs[] = new RefreshLinksJob2( $this->title,
 169                                                 array(
 170                                                         'table'            => $table,
 171                                                         'start'            => $start,
 172                                                         'end'              => $end,
 173                                                         'masterPos'        => $masterPos,
 174                                                 ) + $this->getRootJobParams() // carry over information for de-duplication
 175                                         );
 176                                 }
 177                         }
 178                 }
 179
 180                 if ( count( $jobs ) ) {
 181                         JobQueueGroup::singleton()->push( $jobs );
 182                 }
 183
 184                 wfProfileOut( __METHOD__ );
 185                 return true;
 186         }
 187
 188         /**
 189          * @param $table string
 190          * @param $masterPos mixed
 191          * @return Array
 192          */
 193         protected function getSingleTitleJobs( $table, $masterPos ) {
 194                 # The "start"/"end" fields are not set for the base jobs
 195                 $start = isset( $this->params['start'] ) ? $this->params['start'] : false;
 196                 $end = isset( $this->params['end'] ) ? $this->params['end'] : false;
 197                 $titles = $this->title->getBacklinkCache()->getLinks( $table, $start, $end );
 198                 # Convert into single page refresh links jobs.
 199                 # This handles well when in sapi mode and is useful in any case for job
 200                 # de-duplication. If many pages use template A, and that template itself
 201                 # uses template B, then an edit to both will create many duplicate jobs.
 202                 # Roughly speaking, for each page, one of the "RefreshLinksJob" jobs will
 203                 # get run first, and when it does, it will remove the duplicates. Of course,
 204                 # one page could have its job popped when the other page's job is still
 205                 # buried within the logic of a refreshLinks2 job.
 206                 $jobs = array();
 207                 foreach ( $titles as $title ) {
 208                         $jobs[] = new RefreshLinksJob( $title,
 209                                 array( 'masterPos' => $masterPos ) + $this->getRootJobParams()
 210                         ); // carry over information for de-duplication
 211                 }
 212                 return $jobs;
 213         }
 214
 215         /**
 216          * @return Array
 217          */
 218         public function getDeduplicationInfo() {
 219                 $info = parent::getDeduplicationInfo();
 220                 // Don't let highly unique "masterPos" values ruin duplicate detection
 221                 if ( is_array( $info['params'] ) ) {
 222                         unset( $info['params']['masterPos'] );
 223                 }
 224                 return $info;
 225         }
 226 }