1<?php
2/**
3 * Benchmark script for parse operations
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 * @author Tim Starling <tstarling@wikimedia.org>
22 * @ingroup Benchmark
23 */
24
25require_once __DIR__ . '/../Maintenance.php';
26
27use MediaWiki\Linker\LinkTarget;
28use MediaWiki\MediaWikiServices;
29use MediaWiki\Revision\RevisionRecord;
30use MediaWiki\Revision\SlotRecord;
31
32/**
33 * Maintenance script to benchmark how long it takes to parse a given title at an optionally
34 * specified timestamp
35 *
36 * @since 1.23
37 */
38class BenchmarkParse extends Maintenance {
39	/** @var string MediaWiki concatenated string timestamp (YYYYMMDDHHMMSS) */
40	private $templateTimestamp = null;
41
42	private $clearLinkCache = false;
43
44	/**
45	 * @var LinkCache
46	 */
47	private $linkCache;
48
49	/** @var array Cache that maps a Title DB key to revision ID for the requested timestamp */
50	private $idCache = [];
51
52	public function __construct() {
53		parent::__construct();
54		$this->addDescription( 'Benchmark parse operation' );
55		$this->addArg( 'title', 'The name of the page to parse' );
56		$this->addOption( 'warmup', 'Repeat the parse operation this number of times to warm the cache',
57			false, true );
58		$this->addOption( 'loops', 'Number of times to repeat parse operation post-warmup',
59			false, true );
60		$this->addOption( 'page-time',
61			'Use the version of the page which was current at the given time',
62			false, true );
63		$this->addOption( 'tpl-time',
64			'Use templates which were current at the given time (except that moves and ' .
65			'deletes are not handled properly)',
66			false, true );
67		$this->addOption( 'reset-linkcache', 'Reset the LinkCache after every parse.',
68			false, false );
69	}
70
71	public function execute() {
72		if ( $this->hasOption( 'tpl-time' ) ) {
73			$this->templateTimestamp = wfTimestamp( TS_MW, strtotime( $this->getOption( 'tpl-time' ) ) );
74			Hooks::register( 'BeforeParserFetchTemplateRevisionRecord', [ $this, 'onFetchTemplate' ] );
75		}
76
77		$this->clearLinkCache = $this->hasOption( 'reset-linkcache' );
78		// Set as a member variable to avoid function calls when we're timing the parse
79		$this->linkCache = MediaWikiServices::getInstance()->getLinkCache();
80
81		$title = Title::newFromText( $this->getArg( 0 ) );
82		if ( !$title ) {
83			$this->fatalError( "Invalid title" );
84		}
85
86		$revLookup = MediaWikiServices::getInstance()->getRevisionLookup();
87		if ( $this->hasOption( 'page-time' ) ) {
88			$pageTimestamp = wfTimestamp( TS_MW, strtotime( $this->getOption( 'page-time' ) ) );
89			$id = $this->getRevIdForTime( $title, $pageTimestamp );
90			if ( !$id ) {
91				$this->fatalError( "The page did not exist at that time" );
92			}
93
94			$revision = $revLookup->getRevisionById( $id );
95		} else {
96			$revision = $revLookup->getRevisionByTitle( $title );
97		}
98
99		if ( !$revision ) {
100			$this->fatalError( "Unable to load revision, incorrect title?" );
101		}
102
103		$warmup = $this->getOption( 'warmup', 1 );
104		for ( $i = 0; $i < $warmup; $i++ ) {
105			$this->runParser( $revision );
106		}
107
108		$loops = $this->getOption( 'loops', 1 );
109		if ( $loops < 1 ) {
110			$this->fatalError( 'Invalid number of loops specified' );
111		}
112		$startUsage = getrusage();
113		$startTime = microtime( true );
114		for ( $i = 0; $i < $loops; $i++ ) {
115			$this->runParser( $revision );
116		}
117		$endUsage = getrusage();
118		$endTime = microtime( true );
119
120		printf( "CPU time = %.3f s, wall clock time = %.3f s\n",
121			// CPU time
122			( $endUsage['ru_utime.tv_sec'] + $endUsage['ru_utime.tv_usec'] * 1e-6
123			- $startUsage['ru_utime.tv_sec'] - $startUsage['ru_utime.tv_usec'] * 1e-6 ) / $loops,
124			// Wall clock time
125			( $endTime - $startTime ) / $loops
126		);
127	}
128
129	/**
130	 * Fetch the ID of the revision of a Title that occurred
131	 *
132	 * @param Title $title
133	 * @param string $timestamp
134	 * @return bool|string Revision ID, or false if not found or error
135	 */
136	private function getRevIdForTime( Title $title, $timestamp ) {
137		$dbr = $this->getDB( DB_REPLICA );
138
139		$id = $dbr->selectField(
140			[ 'revision', 'page' ],
141			'rev_id',
142			[
143				'page_namespace' => $title->getNamespace(),
144				'page_title' => $title->getDBkey(),
145				'rev_timestamp <= ' . $dbr->addQuotes( $timestamp )
146			],
147			__METHOD__,
148			[ 'ORDER BY' => 'rev_timestamp DESC' ],
149			[ 'revision' => [ 'JOIN', 'rev_page=page_id' ] ]
150		);
151
152		return $id;
153	}
154
155	/**
156	 * Parse the text from a given RevisionRecord
157	 *
158	 * @param RevisionRecord $revision
159	 */
160	private function runParser( RevisionRecord $revision ) {
161		$content = $revision->getContent( SlotRecord::MAIN );
162		$title = Title::newFromLinkTarget( $revision->getPageAsLinkTarget() );
163
164		$content->getParserOutput( $title, $revision->getId() );
165		if ( $this->clearLinkCache ) {
166			$this->linkCache->clear();
167		}
168	}
169
170	/**
171	 * Hook into the parser's revision ID fetcher. Make sure that the parser only
172	 * uses revisions around the specified timestamp.
173	 *
174	 * @param ?LinkTarget $contextTitle
175	 * @param LinkTarget $titleTarget
176	 * @param bool &$skip
177	 * @param ?RevisionRecord &$revRecord
178	 * @return bool
179	 */
180	private function onFetchTemplate(
181		?LinkTarget $contextTitle,
182		LinkTarget $titleTarget,
183		bool &$skip,
184		?RevisionRecord &$revRecord
185	): bool {
186		$title = Title::castFromLinkTarget( $titleTarget );
187
188		$pdbk = $title->getPrefixedDBkey();
189		if ( !isset( $this->idCache[$pdbk] ) ) {
190			$proposedId = $this->getRevIdForTime( $title, $this->templateTimestamp );
191			$this->idCache[$pdbk] = $proposedId;
192		}
193		if ( $this->idCache[$pdbk] !== false ) {
194			$revLookup = MediaWikiServices::getInstance()->getRevisionLookup();
195			$revRecord = $revLookup->getRevisionById( $this->idCache[$pdbk] );
196		}
197
198		return true;
199	}
200}
201
202$maintClass = BenchmarkParse::class;
203require_once RUN_MAINTENANCE_IF_MAIN;
204