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 bool 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
  46                         return false;
  47                 }
  48
  49                 # Wait for the DB of the current/next slave DB handle to catch up to the master.
  50                 # This way, we get the correct page_latest for templates or files that just changed
  51                 # milliseconds ago, having triggered this job to begin with.
  52                 if ( isset( $this->params['masterPos'] ) && $this->params['masterPos'] !== false ) {
  53                         wfGetLB()->waitFor( $this->params['masterPos'] );
  54                 }
  55
  56                 $revision = Revision::newFromTitle( $this->title, false, Revision::READ_NORMAL );
  57                 if ( !$revision ) {
  58                         $this->error = 'refreshLinks: Article not found "' .
  59                                 $this->title->getPrefixedDBkey() . '"';
  60
  61                         return false; // XXX: what if it was just deleted?
  62                 }
  63
  64                 self::runForTitleInternal( $this->title, $revision, __METHOD__ );
  65
  66                 return true;
  67         }
  68
  69         /**
  70          * @return array
  71          */
  72         public function getDeduplicationInfo() {
  73                 $info = parent::getDeduplicationInfo();
  74                 // Don't let highly unique "masterPos" values ruin duplicate detection
  75                 if ( is_array( $info['params'] ) ) {
  76                         unset( $info['params']['masterPos'] );
  77                 }
  78
  79                 return $info;
  80         }
  81
  82         /**
  83          * @param Title $title
  84          * @param Revision $revision
  85          * @param string $fname
  86          * @return void
  87          */
  88         public static function runForTitleInternal( Title $title, Revision $revision, $fname ) {
  89                 wfProfileIn( $fname );
  90                 $content = $revision->getContent( Revision::RAW );
  91
  92                 if ( !$content ) {
  93                         // if there is no content, pretend the content is empty
  94                         $content = $revision->getContentHandler()->makeEmptyContent();
  95                 }
  96
  97                 // Revision ID must be passed to the parser output to get revision variables correct
  98                 $parserOutput = $content->getParserOutput( $title, $revision->getId(), null, false );
  99
 100                 $updates = $content->getSecondaryDataUpdates( $title, null, false, $parserOutput );
 101                 DataUpdate::runUpdates( $updates );
 102
 103                 InfoAction::invalidateCache( $title );
 104
 105                 wfProfileOut( $fname );
 106         }
 107 }
 108
 109 /**
 110  * Background job to update links for a given title.
 111  * Newer version for high use templates.
 112  *
 113  * @ingroup JobQueue
 114  */
 115 class RefreshLinksJob2 extends Job {
 116         function __construct( $title, $params, $id = 0 ) {
 117                 parent::__construct( 'refreshLinks2', $title, $params, $id );
 118                 // Base jobs for large templates can easily be de-duplicated
 119                 $this->removeDuplicates = !isset( $params['start'] ) && !isset( $params['end'] );
 120         }
 121
 122         /**
 123          * Run a refreshLinks2 job
 124          * @return bool success
 125          */
 126         function run() {
 127                 global $wgUpdateRowsPerJob;
 128
 129                 $linkCache = LinkCache::singleton();
 130                 $linkCache->clear();
 131
 132                 if ( is_null( $this->title ) ) {
 133                         $this->error = "refreshLinks2: Invalid title";
 134
 135                         return false;
 136                 }
 137
 138                 // Back compat for pre-r94435 jobs
 139                 $table = isset( $this->params['table'] ) ? $this->params['table'] : 'templatelinks';
 140
 141                 // Avoid slave lag when fetching templates.
 142                 // When the outermost job is run, we know that the caller that enqueued it must have
 143                 // committed the relevant changes to the DB by now. At that point, record the master
 144                 // position and pass it along as the job recursively breaks into smaller range jobs.
 145                 // Hopefully, when leaf jobs are popped, the slaves will have reached that position.
 146                 if ( isset( $this->params['masterPos'] ) ) {
 147                         $masterPos = $this->params['masterPos'];
 148                 } elseif ( wfGetLB()->getServerCount() > 1 ) {
 149                         $masterPos = wfGetLB()->getMasterPos();
 150                 } else {
 151                         $masterPos = false;
 152                 }
 153
 154                 $tbc = $this->title->getBacklinkCache();
 155
 156                 $jobs = array(); // jobs to insert
 157                 if ( isset( $this->params['start'] ) && isset( $this->params['end'] ) ) {
 158                         # This is a partition job to trigger the insertion of leaf jobs...
 159                         $jobs = array_merge( $jobs, $this->getSingleTitleJobs( $table, $masterPos ) );
 160                 } else {
 161                         # This is a base job to trigger the insertion of partitioned jobs...
 162                         if ( $tbc->getNumLinks( $table, $wgUpdateRowsPerJob + 1 ) <= $wgUpdateRowsPerJob ) {
 163                                 # Just directly insert the single per-title jobs
 164                                 $jobs = array_merge( $jobs, $this->getSingleTitleJobs( $table, $masterPos ) );
 165                         } else {
 166                                 # Insert the partition jobs to make per-title jobs
 167                                 foreach ( $tbc->partition( $table, $wgUpdateRowsPerJob ) as $batch ) {
 168                                         list( $start, $end ) = $batch;
 169                                         $jobs[] = new RefreshLinksJob2( $this->title,
 170                                                 array(
 171                                                         'table' => $table,
 172                                                         'start' => $start,
 173                                                         'end' => $end,
 174                                                         'masterPos' => $masterPos,
 175                                                 ) + $this->getRootJobParams() // carry over information for de-duplication
 176                                         );
 177                                 }
 178                         }
 179                 }
 180
 181                 if ( count( $jobs ) ) {
 182                         JobQueueGroup::singleton()->push( $jobs );
 183                 }
 184
 185                 return true;
 186         }
 187
 188         /**
 189          * @param string $table
 190          * @param mixed $masterPos
 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
 213                 return $jobs;
 214         }
 215
 216         /**
 217          * @return array
 218          */
 219         public function getDeduplicationInfo() {
 220                 $info = parent::getDeduplicationInfo();
 221                 // Don't let highly unique "masterPos" values ruin duplicate detection
 222                 if ( is_array( $info['params'] ) ) {
 223                         unset( $info['params']['masterPos'] );
 224                 }
 225
 226                 return $info;
 227         }
 228 }