BacklinkJobUtils.php

Go to the documentation of this file.

<?php
class BacklinkJobUtils {
    public static function partitionBacklinkJob( Job $job, $bSize, $cSize, $opts = [] ) {
        $class = get_class( $job );
        $title = $job->getTitle();
        $params = $job->getParams();
 
        if ( isset( $params['pages'] ) || empty( $params['recursive'] ) ) {
            $ranges = []; // sanity; this is a leaf node
            $realBSize = 0;
            wfWarn( __METHOD__ . " called on {$job->getType()} leaf job (explosive recursion)." );
        } elseif ( isset( $params['range'] ) ) {
            // This is a range job to trigger the insertion of partitioned/title jobs...
            $ranges = $params['range']['subranges'];
            $realBSize = $params['range']['batchSize'];
        } else {
            // This is a base job to trigger the insertion of partitioned jobs...
            $ranges = $title->getBacklinkCache()->partition( $params['table'], $bSize );
            $realBSize = $bSize;
        }
 
        $extraParams = isset( $opts['params'] ) ? $opts['params'] : [];
 
        $jobs = [];
        // Combine the first range (of size $bSize) backlinks into leaf jobs
        if ( isset( $ranges[0] ) ) {
            list( $start, $end ) = $ranges[0];
            $iter = $title->getBacklinkCache()->getLinks( $params['table'], $start, $end );
            $titles = iterator_to_array( $iter );
            foreach ( array_chunk( $titles, $cSize ) as $titleBatch ) {
                $pages = [];
                foreach ( $titleBatch as $tl ) {
                    $pages[$tl->getArticleID()] = [ $tl->getNamespace(), $tl->getDBkey() ];
                }
                $jobs[] = new $class(
                    $title, // maintain parent job title
                    [ 'pages' => $pages ] + $extraParams
                );
            }
        }
        // Take all of the remaining ranges and build a partition job from it
        if ( isset( $ranges[1] ) ) {
            $jobs[] = new $class(
                $title, // maintain parent job title
                [
                    'recursive'     => true,
                    'table'         => $params['table'],
                    'range'         => [
                        'start'     => $ranges[1][0],
                        'end'       => $ranges[count( $ranges ) - 1][1],
                        'batchSize' => $realBSize,
                        'subranges' => array_slice( $ranges, 1 )
                    ],
                    // Track how many times the base job divided for debugging
                    'division'      => isset( $params['division'] )
                        ? ( $params['division'] + 1 )
                        : 1
                ] + $extraParams
            );
        }
 
        return $jobs;
    }
}