1<?php
2/**
3 * Performs fuzz-style testing of MediaWiki's preprocessor.
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 * @ingroup Maintenance
22 */
23
24use MediaWiki\MediaWikiServices;
25
26use Wikimedia\TestingAccessWrapper;
27
28$optionsWithoutArgs = [ 'verbose' ];
29require_once __DIR__ . '/CommandLineInc.php';
30
31class PPFuzzTester {
32	public $hairs = [
33		'[[', ']]', '{{', '{{', '}}', '}}', '{{{', '}}}',
34		'<', '>', '<nowiki', '<gallery', '</nowiki>', '</gallery>', '<nOwIkI>', '</NoWiKi>',
35		'<!--', '-->',
36		"\n==", "==\n",
37		'|', '=', "\n", ' ', "\t", "\x7f",
38		'~~', '~~~', '~~~~', 'subst:',
39		'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j',
40		'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't',
41
42		// extensions
43		// '<ref>', '</ref>', '<references/>',
44	];
45	public $minLength = 0;
46	public $maxLength = 20;
47	public $maxTemplates = 5;
48	// public $outputTypes = [ 'OT_HTML', 'OT_WIKI', 'OT_PREPROCESS' ];
49	public $entryPoints = [ 'fuzzTestSrvus', 'fuzzTestPst', 'fuzzTestPreprocess' ];
50	public $verbose = false;
51
52	/**
53	 * @var bool|PPFuzzTest
54	 */
55	private static $currentTest = false;
56
57	public function execute() {
58		if ( !file_exists( 'results' ) ) {
59			mkdir( 'results' );
60		}
61		if ( !is_dir( 'results' ) ) {
62			echo "Unable to create 'results' directory\n";
63			exit( 1 );
64		}
65		$overallStart = microtime( true );
66		$reportInterval = 1000;
67		// @phan-suppress-next-line PhanInfiniteLoop
68		for ( $i = 1; true; $i++ ) {
69			$t = -microtime( true );
70			try {
71				self::$currentTest = new PPFuzzTest( $this );
72				self::$currentTest->execute();
73				$passed = 'passed';
74			} catch ( Exception $e ) {
75				$testReport = self::$currentTest->getReport();
76				$exceptionReport = $e instanceof MWException ? $e->getText() : (string)$e;
77				$hash = md5( $testReport );
78				// @phan-suppress-next-line SecurityCheck-PathTraversal False positive T268920
79				file_put_contents( "results/ppft-$hash.in", serialize( self::$currentTest ) );
80				// @phan-suppress-next-line SecurityCheck-PathTraversal False positive T268920
81				file_put_contents( "results/ppft-$hash.fail",
82					"Input:\n$testReport\n\nException report:\n$exceptionReport\n" );
83				print "Test $hash failed\n";
84				$passed = 'failed';
85			}
86			$t += microtime( true );
87
88			if ( $this->verbose ) {
89				printf( "Test $passed in %.3f seconds\n", $t );
90				print self::$currentTest->getReport();
91			}
92
93			$reportMetric = ( microtime( true ) - $overallStart ) / $i * $reportInterval;
94			if ( $reportMetric > 25 ) {
95				if ( substr( $reportInterval, 0, 1 ) === '1' ) {
96					$reportInterval /= 2;
97				} else {
98					$reportInterval /= 5;
99				}
100			} elseif ( $reportMetric < 4 ) {
101				if ( substr( $reportInterval, 0, 1 ) === '1' ) {
102					$reportInterval *= 5;
103				} else {
104					$reportInterval *= 2;
105				}
106			}
107			if ( $i % $reportInterval == 0 ) {
108				print "$i tests done\n";
109				/*
110				$testReport = self::$currentTest->getReport();
111				$filename = 'results/ppft-' . md5( $testReport ) . '.pass';
112				file_put_contents( $filename, "Input:\n$testReport\n" );*/
113			}
114		}
115	}
116
117	public function makeInputText( $max = false ) {
118		if ( $max === false ) {
119			$max = $this->maxLength;
120		}
121		$length = mt_rand( $this->minLength, $max );
122		$s = '';
123		for ( $i = 0; $i < $length; $i++ ) {
124			$hairIndex = mt_rand( 0, count( $this->hairs ) - 1 );
125			$s .= $this->hairs[$hairIndex];
126		}
127		// Send through the UTF-8 normaliser
128		// This resolves a few differences between the old preprocessor and the
129		// XML-based one, which doesn't like illegals and converts line endings.
130		// It's done by the MW UI, so it's a reasonably legitimate thing to do.
131		$s = MediaWikiServices::getInstance()->getContentLanguage()->normalize( $s );
132
133		return $s;
134	}
135
136	public function makeTitle() {
137		return Title::newFromText( mt_rand( 0, 1000000 ), mt_rand( 0, 10 ) );
138	}
139
140	/*
141	public function pickOutputType() {
142		$count = count( $this->outputTypes );
143		return $this->outputTypes[ mt_rand( 0, $count - 1 ) ];
144	}*/
145
146	public function pickEntryPoint() {
147		$count = count( $this->entryPoints );
148
149		return $this->entryPoints[mt_rand( 0, $count - 1 )];
150	}
151}
152
153class PPFuzzTest {
154	/**
155	 * @var array[]
156	 * @phan-var array<string,array{text:string|false,finalTitle:Title}>
157	 */
158	public $templates;
159	public $mainText, $title, $entryPoint, $output;
160
161	/** @var PPFuzzTester */
162	private $parent;
163	/** @var string */
164	public $nickname;
165	/** @var bool */
166	public $fancySig;
167
168	/**
169	 * @param PPFuzzTester $tester
170	 */
171	public function __construct( $tester ) {
172		global $wgMaxSigChars;
173		$this->parent = $tester;
174		$this->mainText = $tester->makeInputText();
175		$this->title = $tester->makeTitle();
176		// $this->outputType = $tester->pickOutputType();
177		$this->entryPoint = $tester->pickEntryPoint();
178		$this->nickname = $tester->makeInputText( $wgMaxSigChars + 10 );
179		$this->fancySig = (bool)mt_rand( 0, 1 );
180		$this->templates = [];
181	}
182
183	/**
184	 * @param Title $title
185	 * @return array
186	 */
187	public function templateHook( $title ) {
188		$titleText = $title->getPrefixedDBkey();
189
190		if ( !isset( $this->templates[$titleText] ) ) {
191			$finalTitle = $title;
192			if ( count( $this->templates ) >= $this->parent->maxTemplates ) {
193				// Too many templates
194				$text = false;
195			} else {
196				if ( !mt_rand( 0, 1 ) ) {
197					// Redirect
198					$finalTitle = $this->parent->makeTitle();
199				}
200				if ( !mt_rand( 0, 5 ) ) {
201					// Doesn't exist
202					$text = false;
203				} else {
204					$text = $this->parent->makeInputText();
205				}
206			}
207			$this->templates[$titleText] = [
208				'text' => $text,
209				'finalTitle' => $finalTitle ];
210		}
211
212		return $this->templates[$titleText];
213	}
214
215	public function execute() {
216		global $wgUser;
217
218		$user = new PPFuzzUser;
219		$user->mName = 'Fuzz';
220		$user->mFrom = 'name';
221		$user->ppfz_test = $this;
222
223		$wgUser = $user;
224
225		$options = ParserOptions::newFromUser( $user );
226		$options->setTemplateCallback( [ $this, 'templateHook' ] );
227		$options->setTimestamp( wfTimestampNow() );
228		$this->output = call_user_func(
229			[ TestingAccessWrapper::newFromObject(
230				MediaWikiServices::getInstance()->getParser()
231			), $this->entryPoint ],
232			$this->mainText,
233			$this->title,
234			$options
235		);
236
237		return $this->output;
238	}
239
240	public function getReport() {
241		$s = "Title: " . $this->title->getPrefixedDBkey() . "\n" .
242			// "Output type: {$this->outputType}\n" .
243			"Entry point: {$this->entryPoint}\n" .
244			"User: " . ( $this->fancySig ? 'fancy' : 'no-fancy' ) .
245			' ' . var_export( $this->nickname, true ) . "\n" .
246			"Main text: " . var_export( $this->mainText, true ) . "\n";
247		foreach ( $this->templates as $titleText => $template ) {
248			$finalTitle = $template['finalTitle'];
249			if ( $finalTitle != $titleText ) {
250				$s .= "[[$titleText]] -> [[$finalTitle]]: " . var_export( $template['text'], true ) . "\n";
251			} else {
252				$s .= "[[$titleText]]: " . var_export( $template['text'], true ) . "\n";
253			}
254		}
255		$s .= "Output: " . var_export( $this->output, true ) . "\n";
256
257		return $s;
258	}
259}
260
261class PPFuzzUser extends User {
262	public $ppfz_test, $mDataLoaded;
263
264	public function load( $flags = null ) {
265		if ( $this->mDataLoaded ) {
266			return;
267		}
268		$this->mDataLoaded = true;
269		$this->loadDefaults( $this->mName );
270	}
271
272	public function getOption( $oname, $defaultOverride = null, $ignoreHidden = false ) {
273		if ( $oname === 'fancysig' ) {
274			return $this->ppfz_test->fancySig;
275		} elseif ( $oname === 'nickname' ) {
276			return $this->ppfz_test->nickname;
277		} else {
278			return parent::getOption( $oname, $defaultOverride, $ignoreHidden );
279		}
280	}
281}
282
283ini_set( 'memory_limit', '50M' );
284if ( isset( $args[0] ) ) {
285	$testText = file_get_contents( $args[0] );
286	if ( !$testText ) {
287		print "File not found\n";
288		exit( 1 );
289	}
290	$test = unserialize( $testText );
291	$result = $test->execute();
292	print "Test passed.\n";
293} else {
294	$tester = new PPFuzzTester;
295	$tester->verbose = isset( $options['verbose'] );
296	$tester->execute();
297}
298