Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
10.91% |
6 / 55 |
|
10.00% |
1 / 10 |
CRAP | |
0.00% |
0 / 1 |
| LinksUpdate | |
10.91% |
6 / 55 |
|
10.00% |
1 / 10 |
274.27 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
6 | |||
| newPageChangeUpdate | |
0.00% |
0 / 9 |
|
0.00% |
0 / 1 |
12 | |||
| newPastRevisionVisibilityChange | |
100.00% |
6 / 6 |
|
100.00% |
1 / 1 |
1 | |||
| newPageRefreshUpdate | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
2 | |||
| newSaneitizerUpdate | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
2 | |||
| doJob | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
12 | |||
| saneitize | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| update | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| queueIncomingLinksJobs | |
0.00% |
0 / 14 |
|
0.00% |
0 / 1 |
20 | |||
| isPrioritized | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
6 | |||
| 1 | <?php |
| 2 | |
| 3 | namespace CirrusSearch\Job; |
| 4 | |
| 5 | use CirrusSearch\Updater; |
| 6 | use MediaWiki\MediaWikiServices; |
| 7 | use MediaWiki\Revision\RevisionRecord; |
| 8 | use MediaWiki\Title\Title; |
| 9 | use MediaWiki\Utils\MWTimestamp; |
| 10 | |
| 11 | /** |
| 12 | * Performs the appropriate updates to Elasticsearch after a LinksUpdate is |
| 13 | * completed. The page itself is updated first then a second copy of this job |
| 14 | * is queued to update linked articles if any links change. The job can be |
| 15 | * 'prioritized' via the 'prioritize' parameter which will switch it to a |
| 16 | * different queue then the non-prioritized jobs. Prioritized jobs will never |
| 17 | * be deduplicated with non-prioritized jobs which is good because we can't |
| 18 | * control which job is removed during deduplication. In our case it'd only be |
| 19 | * ok to remove the non-prioritized version. |
| 20 | * |
| 21 | * This program is free software; you can redistribute it and/or modify |
| 22 | * it under the terms of the GNU General Public License as published by |
| 23 | * the Free Software Foundation; either version 2 of the License, or |
| 24 | * (at your option) any later version. |
| 25 | * |
| 26 | * This program is distributed in the hope that it will be useful, |
| 27 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 28 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 29 | * GNU General Public License for more details. |
| 30 | * |
| 31 | * You should have received a copy of the GNU General Public License along |
| 32 | * with this program; if not, write to the Free Software Foundation, Inc., |
| 33 | * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. |
| 34 | * http://www.gnu.org/copyleft/gpl.html |
| 35 | */ |
| 36 | class LinksUpdate extends CirrusTitleJob { |
| 37 | /** |
| 38 | * param key to determine if the job should be "prioritized" |
| 39 | */ |
| 40 | private const PRIORITIZE = 'prioritize'; |
| 41 | |
| 42 | public function __construct( Title $title, array $params ) { |
| 43 | parent::__construct( $title, $params ); |
| 44 | |
| 45 | if ( $this->isPrioritized() ) { |
| 46 | $this->command .= 'Prioritized'; |
| 47 | } |
| 48 | // Note that we have to keep the prioritized param or else when the job |
| 49 | // is loaded it'll load under a different name/command/type which would |
| 50 | // be confusing. |
| 51 | } |
| 52 | |
| 53 | /** |
| 54 | * Prepare a page update for when this page is directly updated (new revision/delete/restore) |
| 55 | * |
| 56 | * @param Title $title |
| 57 | * @param RevisionRecord|null $revisionRecord |
| 58 | * @param array $params |
| 59 | * @return LinksUpdate |
| 60 | */ |
| 61 | public static function newPageChangeUpdate( Title $title, ?RevisionRecord $revisionRecord, array $params ): LinksUpdate { |
| 62 | if ( $revisionRecord !== null && $revisionRecord->getTimestamp() !== null ) { |
| 63 | $ts = (int)MWTimestamp::convert( TS_UNIX, $revisionRecord->getTimestamp() ); |
| 64 | } else { |
| 65 | $ts = MWTimestamp::time(); |
| 66 | } |
| 67 | $params += [ |
| 68 | self::PRIORITIZE => true, |
| 69 | self::UPDATE_KIND => self::PAGE_CHANGE, |
| 70 | self::ROOT_EVENT_TIME => $ts, |
| 71 | ]; |
| 72 | |
| 73 | return new self( $title, $params ); |
| 74 | } |
| 75 | |
| 76 | /** |
| 77 | * Prepare a cautionary update of a page that had some of its revision's visibility changed. |
| 78 | * (Theoretically not required because old revisions should not be part of the index) |
| 79 | * @param Title $title |
| 80 | * @return LinksUpdate |
| 81 | */ |
| 82 | public static function newPastRevisionVisibilityChange( Title $title ): LinksUpdate { |
| 83 | $params = [ |
| 84 | self::PRIORITIZE => true, |
| 85 | self::UPDATE_KIND => self::VISIBILITY_CHANGE, |
| 86 | self::ROOT_EVENT_TIME => MWTimestamp::time(), |
| 87 | ]; |
| 88 | |
| 89 | return new self( $title, $params ); |
| 90 | } |
| 91 | |
| 92 | /** |
| 93 | * Prepare a page update for when the rendered output of the page might have changed due to a |
| 94 | * change not directly related to this page (e.g. template update). |
| 95 | * |
| 96 | * @param Title $title |
| 97 | * @param array $params |
| 98 | * @return LinksUpdate |
| 99 | */ |
| 100 | public static function newPageRefreshUpdate( Title $title, array $params ): LinksUpdate { |
| 101 | $params += [ |
| 102 | self::PRIORITIZE => false, |
| 103 | self::UPDATE_KIND => self::PAGE_REFRESH, |
| 104 | self::ROOT_EVENT_TIME => MWTimestamp::time(), |
| 105 | ]; |
| 106 | return new self( $title, $params ); |
| 107 | } |
| 108 | |
| 109 | /** |
| 110 | * New change emitted from the saneitizer |
| 111 | * @param Title $title |
| 112 | * @param string|null $cluster optional target cluster, null for all clusters |
| 113 | * @return LinksUpdate |
| 114 | */ |
| 115 | public static function newSaneitizerUpdate( Title $title, ?string $cluster ): LinksUpdate { |
| 116 | $params = [ |
| 117 | self::PRIORITIZE => false, |
| 118 | self::UPDATE_KIND => self::SANEITIZER, |
| 119 | self::ROOT_EVENT_TIME => MWTimestamp::time(), |
| 120 | self::CLUSTER => $cluster |
| 121 | ]; |
| 122 | return new self( $title, $params ); |
| 123 | } |
| 124 | |
| 125 | /** |
| 126 | * @return bool |
| 127 | */ |
| 128 | protected function doJob() { |
| 129 | $updater = Updater::build( $this->getSearchConfig(), $this->params['cluster'] ?? null ); |
| 130 | if ( $this->params[self::UPDATE_KIND] === self::SANEITIZER ) { |
| 131 | $this->saneitize( $updater ); |
| 132 | } else { |
| 133 | $this->update( $updater ); |
| 134 | } |
| 135 | |
| 136 | if ( $this->getSearchConfig()->get( 'CirrusSearchEnableIncomingLinkCounting' ) ) { |
| 137 | $this->queueIncomingLinksJobs(); |
| 138 | } |
| 139 | |
| 140 | return true; |
| 141 | } |
| 142 | |
| 143 | /** |
| 144 | * Indirection doing technically nothing but help measure the impact of these jobs via flame graphs. |
| 145 | * @param Updater $updater |
| 146 | * @return void |
| 147 | */ |
| 148 | private function saneitize( Updater $updater ): void { |
| 149 | $this->update( $updater ); |
| 150 | } |
| 151 | |
| 152 | private function update( Updater $updater ): void { |
| 153 | $updater->updateFromTitle( $this->title, $this->params[self::UPDATE_KIND], $this->params[self::ROOT_EVENT_TIME] ); |
| 154 | } |
| 155 | |
| 156 | /** |
| 157 | * Queue IncomingLinkCount jobs when pages are newly linked or unlinked |
| 158 | */ |
| 159 | private function queueIncomingLinksJobs() { |
| 160 | $titleKeys = array_merge( $this->params[ 'addedLinks' ] ?? [], |
| 161 | $this->params[ 'removedLinks' ] ?? [] ); |
| 162 | $refreshInterval = $this->getSearchConfig()->get( 'CirrusSearchRefreshInterval' ); |
| 163 | $jobs = []; |
| 164 | $jobQueue = MediaWikiServices::getInstance()->getJobQueueGroup(); |
| 165 | foreach ( $titleKeys as $titleKey ) { |
| 166 | $title = Title::newFromDBkey( $titleKey ); |
| 167 | if ( !$title || !$title->canExist() ) { |
| 168 | continue; |
| 169 | } |
| 170 | // If possible, delay the job execution by a few seconds so Elasticsearch |
| 171 | // can refresh to contain what we just sent it. The delay should be long |
| 172 | // enough for Elasticsearch to complete the refresh cycle, which normally |
| 173 | // takes wgCirrusSearchRefreshInterval seconds but we double it and add |
| 174 | // one just in case. |
| 175 | $delay = 2 * $refreshInterval + 1; |
| 176 | $jobs[] = new IncomingLinkCount( $title, [ |
| 177 | 'cluster' => $this->params['cluster'], |
| 178 | ] + self::buildJobDelayOptions( IncomingLinkCount::class, $delay, $jobQueue ) ); |
| 179 | } |
| 180 | $jobQueue->push( $jobs ); |
| 181 | } |
| 182 | |
| 183 | /** |
| 184 | * @return bool Is this job prioritized? |
| 185 | */ |
| 186 | public function isPrioritized() { |
| 187 | return isset( $this->params[self::PRIORITIZE] ) && $this->params[self::PRIORITIZE]; |
| 188 | } |
| 189 | } |