Performs updates and deletes on the Elasticsearch index. More...

Inheritance diagram for CirrusSearch\Updater:

Collaboration diagram for CirrusSearch\Updater:

Public Member Functions
	__construct (Connection $readConnection, $writeToClusterName=null)

	updateFromTitle ( $title, ?string $updateKind, ?int $rootEventTime)
	Update a single page.

	traceRedirects ( $title)
	Trace redirects from the title to the destination.

	updatePages ( $pages, $flags, string $updateKind=null, int $rootEventTime=null)
	This updates pages in elasticsearch.

	updateWeightedTags (ProperPageIdentity $page, string $tagField, string $tagPrefix, $tagNames=null, $tagWeights=null)

	resetWeightedTags (ProperPageIdentity $page, string $tagField, string $tagPrefix)

	deletePages ( $titles, $docIds, $indexSuffix=null, array $writeJobParams=[])
	Delete pages from the elasticsearch index.

	archivePages ( $archived)
	Add documents to archive index.

	updateLinkedArticles ( $titles)
	Update the search index for newly linked or unlinked articles.

Public Member Functions inherited from CirrusSearch\ElasticsearchIntermediary
	start (RequestLog $log)
	Mark the start of a request to Elasticsearch.

	success ( $result=null, Connection $connection=null)
	Log a successful request and return the provided result in a good Status.

	successViaCache (RequestLog $log)
	Log a successful request when the response comes from a cache outside elasticsearch.

	failure (ExceptionInterface $exception=null, Connection $connection=null)
	Log a failure and return an appropriate status.

	getSearchMetrics ()
	Get the search metrics we have.

Static Public Member Functions
static	build (SearchConfig $config, $cluster)

Static Public Member Functions inherited from CirrusSearch\ElasticsearchIntermediary
static	setResultPages (array $matches)
	This is set externally because we don't have complete control, from the SearchEngine interface, of what is actually sent to the user.

static	getQueryTypesUsed ()
	Report the types of queries that were issued within the current request.

static	hasQueryLogs ()

static	appendLastLogPayload ( $key, $value)

static	isMSearchResultSetOK (MultiResultSet $multiResultSet)
	check validity of the multisearch response

Protected Member Functions
	pushElasticaWriteJobs (string $updateGroup, array $items, $factory, int $batchSize=10)

	newLog ( $description, $queryType, array $extra=[])

Protected Member Functions inherited from CirrusSearch\ElasticsearchIntermediary
	__construct (Connection $connection, UserIdentity $user=null, $slowSeconds=null, $extraBackendLatency=0)

	startNewLog ( $description, $queryType, array $extra=[])

	getTimeout ( $searchType='default')

	getClientTimeout ( $searchType='default')

	appendMetrics (SearchMetricsProvider $provider)

	runMSearch (Search $search, RequestLog $log, Connection $connection=null, callable $resultsTransformer=null)

Protected Attributes
	$writeToClusterName

Protected Attributes inherited from CirrusSearch\ElasticsearchIntermediary
	$connection

	$user

	$currentRequestLog = null

Additional Inherited Members
Static Protected Attributes inherited from CirrusSearch\ElasticsearchIntermediary
static	$requestLogger

Detailed Description

Performs updates and deletes on the Elasticsearch index.

Called by CirrusSearch.php (our SearchEngine implementation), forceSearchIndex (for bulk updates), and CirrusSearch's jobs.

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. http://www.gnu.org/copyleft/gpl.html

Constructor & Destructor Documentation

◆ __construct()

CirrusSearch\Updater::__construct	(	Connection	$readConnection,
			$writeToClusterName = null )

Parameters

Connection	$readConnection	connection used to pull data out of elasticsearch
string \| null	$writeToClusterName

Member Function Documentation

◆ archivePages()

CirrusSearch\Updater::archivePages ( $archived )

Add documents to archive index.

Parameters

array $archived

Returns: bool

◆ build()

static CirrusSearch\Updater::build	(	SearchConfig	$config,
			$cluster )

static

Parameters

SearchConfig	$config
string \| null	$cluster	cluster to read from and write to, null to read from the default cluster and write to all

Returns: Updater

◆ deletePages()

CirrusSearch\Updater::deletePages	(		$titles,
			$docIds,
			$indexSuffix = null,
		array	$writeJobParams = [] )

Delete pages from the elasticsearch index.

$titles and $docIds must point to the same pages and should point to them in the same order.

Parameters

Title[]	$titles	List of titles to delete. If empty then skipped other index maintenance is skipped.
int[] \| string[]	$docIds	List of elasticsearch document ids to delete
string \| null	$indexSuffix	index from which to delete. null means all.
array	$writeJobParams	Parameters passed on to ElasticaWriteJob

◆ newLog()

CirrusSearch\Updater::newLog	(		$description,
			$queryType,
		array	$extra = [] )

protected

Parameters

string	$description
string	$queryType
string[]	$extra

Returns: SearchRequestLog

Reimplemented from CirrusSearch\ElasticsearchIntermediary.

◆ pushElasticaWriteJobs()

CirrusSearch\Updater::pushElasticaWriteJobs	(	string	$updateGroup,
		array	$items,
			$factory,
		int	$batchSize = 10 )

protected

Parameters

string	$updateGroup	UpdateGroup::* constant
mixed[]	$items
callable	$factory
int	$batchSize

◆ resetWeightedTags()

CirrusSearch\Updater::resetWeightedTags	(	ProperPageIdentity	$page,
		string	$tagField,
		string	$tagPrefix )

Parameters

ProperPageIdentity	$page
string	$tagField
string	$tagPrefix

◆ traceRedirects()

CirrusSearch\Updater::traceRedirects ( $title )

Trace redirects from the title to the destination.

Also registers the title in the memory of titles updated and detects special pages.

Parameters

Title $title title to trace

Returns

array with keys: target, redirects

target is WikiPage|null wikipage if the $title either isn't a redirect or resolves to an updatable page that hasn't been updated yet. Null if the page has been updated, is a special page, or the redirects enter a loop.
redirects is an array of WikiPages, one per redirect in the chain. If title isn't a redirect then this will be an empty array

◆ updateFromTitle()

CirrusSearch\Updater::updateFromTitle	(		$title,
		?string	$updateKind,
		?int	$rootEventTime )

Update a single page.

Parameters

Title	$title
string \| null	$updateKind	kind of update to perform (used for monitoring)
int \| null	$rootEventTime	the time of MW event that caused this update (used for monitoring)

◆ updateLinkedArticles()

CirrusSearch\Updater::updateLinkedArticles ( $titles )

Update the search index for newly linked or unlinked articles.

Parameters

Title[] $titles titles to update

◆ updatePages()

CirrusSearch\Updater::updatePages	(		$pages,
			$flags,
		string	$updateKind = null,
		int	$rootEventTime = null )

This updates pages in elasticsearch.

$flags includes: INDEX_EVERYTHING Cirrus will parse the page and count the links and send the document to Elasticsearch as an index so if it doesn't exist it'll be created. SKIP_PARSE Cirrus will skip parsing the page when building the document. It makes sense to do this when you know the page hasn't changed like when it is newly linked from another page. SKIP_LINKS Cirrus will skip collecting links information. It makes sense to do this when you know the link counts aren't yet available like during the first phase of the two phase index build. INDEX_ON_SKIP Cirrus will send an update if SKIP_PARSE or SKIP_LINKS rather than an index. Indexing with any portion of the document skipped is dangerous because it can put half created pages in the index. This is only a good idea during the first half of the two phase index build.

Parameters

WikiPage[]	$pages	pages to update
int	$flags	Bit field containing instructions about how the document should be built and sent to Elasticsearch.
string \| null	$updateKind	kind of update to perform (used for monitoring)
int \| null	$rootEventTime	the time of MW event that caused this update (used for monitoring)

Returns: int Number of documents updated

◆ updateWeightedTags()

CirrusSearch\Updater::updateWeightedTags	(	ProperPageIdentity	$page,
		string	$tagField,
		string	$tagPrefix,
			$tagNames = null,
			$tagWeights = null )

Parameters

ProperPageIdentity	$page
string	$tagField
string	$tagPrefix
string \| string[] \| null	$tagNames
int \| int[] \| null	$tagWeights

The documentation for this class was generated from the following file:

includes/Updater.php

Public Member Functions

Static Public Member Functions

Protected Member Functions

Protected Attributes

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ __construct()

Member Function Documentation

◆ archivePages()

◆ build()

◆ deletePages()

◆ newLog()

◆ pushElasticaWriteJobs()

◆ resetWeightedTags()

◆ traceRedirects()

◆ updateFromTitle()

◆ updateLinkedArticles()

◆ updatePages()

◆ updateWeightedTags()