maintenance/compareParsers.php

   1 <?php
   2 /**
   3  * Take page text out of an XML dump file and render basic HTML out to files.
   4  * This is *NOT* suitable for publishing or offline use; it's intended for
   5  * running comparative tests of parsing behavior using real-world data.
   6  *
   7  * Templates etc are pulled from the local wiki database, not from the dump.
   8  *
   9  * Copyright © 2011 Platonides
  10  * https://www.mediawiki.org/
  11  *
  12  * This program is free software; you can redistribute it and/or modify
  13  * it under the terms of the GNU General Public License as published by
  14  * the Free Software Foundation; either version 2 of the License, or
  15  * (at your option) any later version.
  16  *
  17  * This program is distributed in the hope that it will be useful,
  18  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  19  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  20  * GNU General Public License for more details.
  21  *
  22  * You should have received a copy of the GNU General Public License along
  23  * with this program; if not, write to the Free Software Foundation, Inc.,
  24  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  25  * http://www.gnu.org/copyleft/gpl.html
  26  *
  27  * @file
  28  * @ingroup Maintenance
  29  */
  30
  31 use MediaWiki\Content\WikitextContent;
  32 use MediaWiki\Parser\ParserOptions;
  33 use MediaWiki\User\User;
  34 use Wikimedia\Diff\Diff;
  35 use Wikimedia\Diff\UnifiedDiffFormatter;
  36
  37 // @codeCoverageIgnoreStart
  38 require_once __DIR__ . '/dumpIterator.php';
  39 // @codeCoverageIgnoreEnd
  40
  41 /**
  42  * Maintenance script to take page text out of an XML dump file and render
  43  * basic HTML out to files.
  44  *
  45  * @ingroup Maintenance
  46  */
  47 class CompareParsers extends DumpIterator {
  48
  49         /** @var int */
  50         private $count = 0;
  51         /** @var string|false */
  52         private $saveFailed = false;
  53         /** @var bool */
  54         private $stripParametersEnabled;
  55         /** @var bool */
  56         private $showParsedOutput;
  57         /** @var bool */
  58         private $showDiff;
  59         /** @var ParserOptions */
  60         private $options;
  61         /** @var int */
  62         private $failed;
  63
  64         public function __construct() {
  65                 parent::__construct();
  66                 $this->addDescription( 'Run a file or dump with several parsers' );
  67                 $this->addOption( 'parser1', 'The first parser to compare.', true, true );
  68                 $this->addOption( 'parser2', 'The second parser to compare.', true, true );
  69                 $this->addOption(
  70                         'save-failed',
  71                         'Folder in which articles which differ will be stored.',
  72                         false,
  73                         true
  74                 );
  75                 $this->addOption( 'show-diff', 'Show a diff of the two renderings.', false, false );
  76                 $this->addOption(
  77                         'diff-bin',
  78                         'Binary to use for diffing (can also be provided by DIFF env var).',
  79                         false,
  80                         false
  81                 );
  82                 $this->addOption(
  83                         'strip-parameters',
  84                         'Remove parameters of html tags to increase readability.',
  85                         false,
  86                         false
  87                 );
  88                 $this->addOption(
  89                         'show-parsed-output',
  90                         'Show the parsed html if both Parsers give the same output.',
  91                         false,
  92                         false
  93                 );
  94         }
  95
  96         public function checkOptions() {
  97                 if ( $this->hasOption( 'save-failed' ) ) {
  98                         $this->saveFailed = $this->getOption( 'save-failed' );
  99                 }
 100
 101                 $this->stripParametersEnabled = $this->hasOption( 'strip-parameters' );
 102                 $this->showParsedOutput = $this->hasOption( 'show-parsed-output' );
 103
 104                 $this->showDiff = $this->hasOption( 'show-diff' );
 105                 if ( $this->showDiff ) {
 106                         $bin = $this->getOption( 'diff-bin', getenv( 'DIFF' ) );
 107                         if ( $bin != '' ) {
 108                                 global $wgDiff;
 109                                 $wgDiff = $bin;
 110                         }
 111                 }
 112
 113                 $user = new User();
 114                 $this->options = ParserOptions::newFromUser( $user );
 115
 116                 $this->failed = 0;
 117         }
 118
 119         public function conclusions() {
 120                 $this->error( "{$this->failed} failed revisions out of {$this->count}" );
 121                 if ( $this->count > 0 ) {
 122                         $this->output( " (" . ( $this->failed / $this->count ) . "%)\n" );
 123                 }
 124         }
 125
 126         private function stripParameters( $text ) {
 127                 if ( !$this->stripParametersEnabled ) {
 128                         return $text;
 129                 }
 130
 131                 return preg_replace( '/(<a) [^>]+>/', '$1>', $text );
 132         }
 133
 134         /**
 135          * Callback function for each revision, parse with both parsers and compare
 136          * @param WikiRevision $rev
 137          */
 138         public function processRevision( WikiRevision $rev ) {
 139                 $title = $rev->getTitle();
 140
 141                 $parser1Name = $this->getOption( 'parser1' );
 142                 $parser2Name = $this->getOption( 'parser2' );
 143
 144                 self::checkParserLocally( $parser1Name );
 145                 self::checkParserLocally( $parser2Name );
 146
 147                 $parser1 = new $parser1Name();
 148                 $parser2 = new $parser2Name();
 149
 150                 $content = $rev->getContent();
 151
 152                 if ( $content->getModel() !== CONTENT_MODEL_WIKITEXT ) {
 153                         $this->error( "Page {$title->getPrefixedText()} does not contain wikitext "
 154                                 . "but {$content->getModel()}\n" );
 155
 156                         return;
 157                 }
 158
 159                 /** @var WikitextContent $content */
 160                 '@phan-var WikitextContent $content';
 161                 $text = strval( $content->getText() );
 162
 163                 $output1 = $parser1->parse( $text, $title, $this->options );
 164                 $output2 = $parser2->parse( $text, $title, $this->options );
 165
 166                 if ( $output1->getText() != $output2->getText() ) {
 167                         $this->failed++;
 168                         $this->error( "Parsing for {$title->getPrefixedText()} differs\n" );
 169
 170                         if ( $this->saveFailed ) {
 171                                 file_put_contents(
 172                                         $this->saveFailed . '/' . rawurlencode( $title->getPrefixedText() ) . ".txt",
 173                                         $text
 174                                 );
 175                         }
 176                         if ( $this->showDiff ) {
 177                                 $diffs = new Diff(
 178                                         explode( "\n", $this->stripParameters( $output1->getText() ) ),
 179                                         explode( "\n", $this->stripParameters( $output2->getText() ) )
 180                                 );
 181                                 $formatter = new UnifiedDiffFormatter();
 182                                 $unifiedDiff = $formatter->format( $diffs );
 183
 184                                 $this->output( $unifiedDiff );
 185                         }
 186                 } else {
 187                         $this->output( $title->getPrefixedText() . "\tOK\n" );
 188
 189                         if ( $this->showParsedOutput ) {
 190                                 $this->output( $this->stripParameters( $output1->getText() ) );
 191                         }
 192                 }
 193         }
 194
 195         private static function checkParserLocally( $parserName ) {
 196                 /* Look for the parser in a file appropriately named in the current folder */
 197                 if ( !class_exists( $parserName ) && file_exists( "$parserName.php" ) ) {
 198                         global $wgAutoloadClasses;
 199                         $wgAutoloadClasses[$parserName] = realpath( '.' ) . "/$parserName.php";
 200                 }
 201         }
 202 }
 203
 204 // @codeCoverageIgnoreStart
 205 $maintClass = CompareParsers::class;
 206 require_once RUN_MAINTENANCE_IF_MAIN;
 207 // @codeCoverageIgnoreEnd