linksupdate.php

来自「php 开发的内容管理系统」· PHP 代码 · 共 602 行 · 第 1/2 页

<?php
/**
 * See deferred.txt
 * @package MediaWiki
 */

/**
 * @todo document
 * @package MediaWiki
 */
class LinksUpdate {

	/**@{{
	 * @private
	 */
	var $mId,            //!< Page ID of the article linked from
		$mTitle,         //!< Title object of the article linked from
		$mLinks,         //!< Map of title strings to IDs for the links in the document
		$mImages,        //!< DB keys of the images used, in the array key only
		$mTemplates,     //!< Map of title strings to IDs for the template references, including broken ones
		$mExternals,     //!< URLs of external links, array key only
		$mCategories,    //!< Map of category names to sort keys
		$mInterlangs,    //!< Map of language codes to titles
		$mDb,            //!< Database connection reference
		$mOptions,       //!< SELECT options to be used (array)
		$mRecursive;     //!< Whether to queue jobs for recursive updates
	/**@}}*/

	/**
	 * Constructor
	 * Initialize private variables
	 * @param $title Integer: FIXME
	 * @param $parserOutput FIXME
	 * @param $recursive Boolean: FIXME, default 'true'.
	 */
	function LinksUpdate( $title, $parserOutput, $recursive = true ) {
		global $wgAntiLockFlags;

		if ( $wgAntiLockFlags & ALF_NO_LINK_LOCK ) {
			$this->mOptions = array();
		} else {
			$this->mOptions = array( 'FOR UPDATE' );
		}
		$this->mDb =& wfGetDB( DB_MASTER );

		if ( !is_object( $title ) ) {
			throw new MWException( "The calling convention to LinksUpdate::LinksUpdate() has changed. " .
				"Please see Article::editUpdates() for an invocation example.\n" );
		}
		$this->mTitle = $title;
		$this->mId = $title->getArticleID();

		$this->mLinks = $parserOutput->getLinks();
		$this->mImages = $parserOutput->getImages();
		$this->mTemplates = $parserOutput->getTemplates();
		$this->mExternals = $parserOutput->getExternalLinks();
		$this->mCategories = $parserOutput->getCategories();

		# Convert the format of the interlanguage links
		# I didn't want to change it in the ParserOutput, because that array is passed all 
		# the way back to the skin, so either a skin API break would be required, or an 
		# inefficient back-conversion.
		$ill = $parserOutput->getLanguageLinks();
		$this->mInterlangs = array();
		foreach ( $ill as $link ) {
			list( $key, $title ) = explode( ':', $link, 2 );
			$this->mInterlangs[$key] = $title;
		}

		$this->mRecursive = $recursive;
	}

	/**
	 * Update link tables with outgoing links from an updated article
	 */
	function doUpdate() {
		global $wgUseDumbLinkUpdate;
		if ( $wgUseDumbLinkUpdate ) {
			$this->doDumbUpdate();
		} else {
			$this->doIncrementalUpdate();
		}
	}

	function doIncrementalUpdate() {
		$fname = 'LinksUpdate::doIncrementalUpdate';
		wfProfileIn( $fname );
		
		# Page links
		$existing = $this->getExistingLinks();
		$this->incrTableUpdate( 'pagelinks', 'pl', $this->getLinkDeletions( $existing ),
			$this->getLinkInsertions( $existing ) );

		# Image links
		$existing = $this->getExistingImages();
		$this->incrTableUpdate( 'imagelinks', 'il', $this->getImageDeletions( $existing ),
			$this->getImageInsertions( $existing ) );

		# Invalidate all image description pages which had links added or removed
		$imageUpdates = array_diff_key( $existing, $this->mImages ) + array_diff_key( $this->mImages, $existing );
		$this->invalidateImageDescriptions( $imageUpdates );

		# External links
		$existing = $this->getExistingExternals();
		$this->incrTableUpdate( 'externallinks', 'el', $this->getExternalDeletions( $existing ),
	        $this->getExternalInsertions( $existing ) );

		# Language links
		$existing = $this->getExistingInterlangs();
		$this->incrTableUpdate( 'langlinks', 'll', $this->getInterlangDeletions( $existing ),
			$this->getInterlangInsertions( $existing ) );

		# Template links
		$existing = $this->getExistingTemplates();
		$this->incrTableUpdate( 'templatelinks', 'tl', $this->getTemplateDeletions( $existing ),
			$this->getTemplateInsertions( $existing ) );

		# Category links
		$existing = $this->getExistingCategories();
		$this->incrTableUpdate( 'categorylinks', 'cl', $this->getCategoryDeletions( $existing ),
			$this->getCategoryInsertions( $existing ) );

		# Invalidate all categories which were added, deleted or changed (set symmetric difference)
		$categoryUpdates = array_diff_assoc( $existing, $this->mCategories ) + array_diff_assoc( $this->mCategories, $existing );
		$this->invalidateCategories( $categoryUpdates );

		# Refresh links of all pages including this page
		# This will be in a separate transaction
		if ( $this->mRecursive ) {
			$this->queueRecursiveJobs();
		}
		
		wfProfileOut( $fname );
	}

	/**
	 * Link update which clears the previous entries and inserts new ones
	 * May be slower or faster depending on level of lock contention and write speed of DB
	 * Also useful where link table corruption needs to be repaired, e.g. in refreshLinks.php
	 */
	function doDumbUpdate() {
		$fname = 'LinksUpdate::doDumbUpdate';
		wfProfileIn( $fname );

		# Refresh category pages and image description pages
		$existing = $this->getExistingCategories();
		$categoryUpdates = array_diff_assoc( $existing, $this->mCategories ) + array_diff_assoc( $this->mCategories, $existing );
		$existing = $this->getExistingImages();
		$imageUpdates = array_diff_key( $existing, $this->mImages ) + array_diff_key( $this->mImages, $existing );

		$this->dumbTableUpdate( 'pagelinks',     $this->getLinkInsertions(),     'pl_from' );
		$this->dumbTableUpdate( 'imagelinks',    $this->getImageInsertions(),    'il_from' );
		$this->dumbTableUpdate( 'categorylinks', $this->getCategoryInsertions(), 'cl_from' );
		$this->dumbTableUpdate( 'templatelinks', $this->getTemplateInsertions(), 'tl_from' );
		$this->dumbTableUpdate( 'externallinks', $this->getExternalInsertions(), 'el_from' );
		$this->dumbTableUpdate( 'langlinks',     $this->getInterlangInsertions(), 'll_from' );

		# Update the cache of all the category pages and image description pages which were changed
		$this->invalidateCategories( $categoryUpdates );
		$this->invalidateImageDescriptions( $imageUpdates );

		# Refresh links of all pages including this page
		# This will be in a separate transaction
		if ( $this->mRecursive ) {
			$this->queueRecursiveJobs();
		}

		wfProfileOut( $fname );
	}

	function queueRecursiveJobs() {
		wfProfileIn( __METHOD__ );
		
		$batchSize = 100;
		$dbr =& wfGetDB( DB_SLAVE );
		$res = $dbr->select( array( 'templatelinks', 'page' ), 
			array( 'page_namespace', 'page_title' ),
			array( 
				'page_id=tl_from', 
				'tl_namespace' => $this->mTitle->getNamespace(),
				'tl_title' => $this->mTitle->getDBkey()
			), __METHOD__
		);

		$done = false;
		while ( !$done ) {
			$jobs = array();
			for ( $i = 0; $i < $batchSize; $i++ ) {
				$row = $dbr->fetchObject( $res );
				if ( !$row ) {
					$done = true;
					break;
				}
				$title = Title::makeTitle( $row->page_namespace, $row->page_title );
				$jobs[] = Job::factory( 'refreshLinks', $title );
			}
			Job::batchInsert( $jobs );
		}
		$dbr->freeResult( $res );
		wfProfileOut( __METHOD__ );
	}
	
	/**
	 * Invalidate the cache of a list of pages from a single namespace
	 *
	 * @param integer $namespace
	 * @param array $dbkeys
	 */
	function invalidatePages( $namespace, $dbkeys ) {
		$fname = 'LinksUpdate::invalidatePages';
		
		if ( !count( $dbkeys ) ) {
			return;
		}
		
		/**
		 * Determine which pages need to be updated
		 * This is necessary to prevent the job queue from smashing the DB with
		 * large numbers of concurrent invalidations of the same page
		 */
		$now = $this->mDb->timestamp();
		$ids = array();
		$res = $this->mDb->select( 'page', array( 'page_id' ), 
			array( 
				'page_namespace' => $namespace,
				'page_title IN (' . $this->mDb->makeList( $dbkeys ) . ')',
				'page_touched < ' . $this->mDb->addQuotes( $now )
			), $fname
		);
		while ( $row = $this->mDb->fetchObject( $res ) ) {
			$ids[] = $row->page_id;
		}
		if ( !count( $ids ) ) {
			return;
		}
		
		/**
		 * Do the update
		 * We still need the page_touched condition, in case the row has changed since 
		 * the non-locking select above.
		 */
		$this->mDb->update( 'page', array( 'page_touched' => $now ), 
			array( 
				'page_id IN (' . $this->mDb->makeList( $ids ) . ')',
				'page_touched < ' . $this->mDb->addQuotes( $now )
			), $fname
		);
	}

	function invalidateCategories( $cats ) {
		$this->invalidatePages( NS_CATEGORY, array_keys( $cats ) );
	}

	function invalidateImageDescriptions( $images ) {
		$this->invalidatePages( NS_IMAGE, array_keys( $images ) );
	}

	function dumbTableUpdate( $table, $insertions, $fromField ) {
		$fname = 'LinksUpdate::dumbTableUpdate';
		$this->mDb->delete( $table, array( $fromField => $this->mId ), $fname );
		if ( count( $insertions ) ) {
			# The link array was constructed without FOR UPDATE, so there may be collisions
			# This may cause minor link table inconsistencies, which is better than
			# crippling the site with lock contention.
			$this->mDb->insert( $table, $insertions, $fname, array( 'IGNORE' ) );
		}
	}

	/**
	 * Make a WHERE clause from a 2-d NS/dbkey array
	 *
	 * @param array $arr 2-d array indexed by namespace and DB key
	 * @param string $prefix Field name prefix, without the underscore
	 */
	function makeWhereFrom2d( &$arr, $prefix ) {
		$lb = new LinkBatch;
		$lb->setArray( $arr );
		return $lb->constructSet( $prefix, $this->mDb );
	}

	/**
	 * Update a table by doing a delete query then an insert query
	 * @private
	 */
	function incrTableUpdate( $table, $prefix, $deletions, $insertions ) {
		$fname = 'LinksUpdate::incrTableUpdate';
		$where = array( "{$prefix}_from" => $this->mId );
		if ( $table == 'pagelinks' || $table == 'templatelinks' ) {
			$clause = $this->makeWhereFrom2d( $deletions, $prefix );
			if ( $clause ) {
				$where[] = $clause;
			} else {
				$where = false;
			}
		} else {
			if ( $table == 'langlinks' ) {
				$toField = 'll_lang';
			} else {
				$toField = $prefix . '_to';
			}
			if ( count( $deletions ) ) {