includes/job/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         public static function runForTitleInternal( Title $title, Revision $revision, $fname ) {
  73                 global $wgContLang;
  74
  75                 wfProfileIn( $fname . '-parse' );
  76                 $options = ParserOptions::newFromUserAndLang( new User, $wgContLang );
  77                 $content = $revision->getContent();
  78                 $parserOutput = $content->getParserOutput( $title, $revision->getId(), $options, false );
  79                 wfProfileOut( $fname . '-parse' );
  80
  81                 wfProfileIn( $fname . '-update' );
  82                 $updates = $content->getSecondaryDataUpdates( $title, null, false, $parserOutput  );
  83                 DataUpdate::runUpdates( $updates );
  84                 wfProfileOut( $fname . '-update' );
  85         }
  86 }
  87
  88 /**
  89  * Background job to update links for a given title.
  90  * Newer version for high use templates.
  91  *
  92  * @ingroup JobQueue
  93  */
  94 class RefreshLinksJob2 extends Job {
  95         const MAX_TITLES_RUN = 10;
  96
  97         function __construct( $title, $params, $id = 0 ) {
  98                 parent::__construct( 'refreshLinks2', $title, $params, $id );
  99         }
 100
 101         /**
 102          * Run a refreshLinks2 job
 103          * @return boolean success
 104          */
 105         function run() {
 106                 wfProfileIn( __METHOD__ );
 107
 108                 $linkCache = LinkCache::singleton();
 109                 $linkCache->clear();
 110
 111                 if ( is_null( $this->title ) ) {
 112                         $this->error = "refreshLinks2: Invalid title";
 113                         wfProfileOut( __METHOD__ );
 114                         return false;
 115                 } elseif ( !isset( $this->params['start'] ) || !isset( $this->params['end'] ) ) {
 116                         $this->error = "refreshLinks2: Invalid params";
 117                         wfProfileOut( __METHOD__ );
 118                         return false;
 119                 }
 120
 121                 // Back compat for pre-r94435 jobs
 122                 $table = isset( $this->params['table'] ) ? $this->params['table'] : 'templatelinks';
 123
 124                 // Avoid slave lag when fetching templates
 125                 if ( isset( $this->params['masterPos'] ) ) {
 126                         $masterPos = $this->params['masterPos'];
 127                 } elseif ( wfGetLB()->getServerCount() > 1  ) {
 128                         $masterPos = wfGetLB()->getMasterPos();
 129                 } else {
 130                         $masterPos = false;
 131                 }
 132
 133                 $titles = $this->title->getBacklinkCache()->getLinks(
 134                         $table, $this->params['start'], $this->params['end'] );
 135
 136                 if ( $titles->count() > self::MAX_TITLES_RUN ) {
 137                         # We don't want to parse too many pages per job as it can starve other jobs.
 138                         # If there are too many pages to parse, break this up into smaller jobs. By passing
 139                         # in the master position here we can cut down on the time spent waiting for slaves to
 140                         # catch up by the runners handling these jobs since time will have passed between now
 141                         # and when they pop these jobs off the queue.
 142                         $start = 0; // batch start
 143                         $end   = 0; // batch end
 144                         $bsize = 0; // batch size
 145                         $first = true; // first of batch
 146                         $jobs  = array();
 147                         foreach ( $titles as $title ) {
 148                                 $start = $first ? $title->getArticleId() : $start;
 149                                 $end   = $title->getArticleId();
 150                                 $first = false;
 151                                 if ( ++$bsize >= self::MAX_TITLES_RUN ) {
 152                                         $jobs[] = new RefreshLinksJob2( $this->title, array(
 153                                                 'table'     => $table,
 154                                                 'start'     => $start,
 155                                                 'end'       => $end,
 156                                                 'masterPos' => $masterPos
 157                                         ) );
 158                                         $first = true;
 159                                         $start = $end = $bsize = 0;
 160                                 }
 161                         }
 162                         if ( $bsize > 0 ) { // group remaining pages into a job
 163                                 $jobs[] = new RefreshLinksJob2( $this->title, array(
 164                                         'table'     => $table,
 165                                         'start'     => $start,
 166                                         'end'       => $end,
 167                                         'masterPos' => $masterPos
 168                                 ) );
 169                         }
 170                         Job::batchInsert( $jobs );
 171                 } elseif ( php_sapi_name() != 'cli' ) {
 172                         # Not suitable for page load triggered job running!
 173                         # Gracefully switch to refreshLinks jobs if this happens.
 174                         $jobs = array();
 175                         foreach ( $titles as $title ) {
 176                                 $jobs[] = new RefreshLinksJob( $title, array( 'masterPos' => $masterPos ) );
 177                         }
 178                         Job::batchInsert( $jobs );
 179                 } else {
 180                         # Wait for the DB of the current/next slave DB handle to catch up to the master.
 181                         # This way, we get the correct page_latest for templates or files that just changed
 182                         # milliseconds ago, having triggered this job to begin with.
 183                         if ( $masterPos ) {
 184                                 wfGetLB()->waitFor( $masterPos );
 185                         }
 186                         # Re-parse each page that transcludes this page and update their tracking links...
 187                         foreach ( $titles as $title ) {
 188                                 $revision = Revision::newFromTitle( $title, false, Revision::READ_NORMAL );
 189                                 if ( !$revision ) {
 190                                         $this->error = 'refreshLinks: Article not found "' .
 191                                                 $title->getPrefixedDBkey() . '"';
 192                                         continue; // skip this page
 193                                 }
 194                                 RefreshLinksJob::runForTitleInternal( $title, $revision, __METHOD__ );
 195                                 wfWaitForSlaves();
 196                         }
 197                 }
 198
 199                 wfProfileOut( __METHOD__ );
 200                 return true;
 201         }
 202 }