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  * http://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 require_once __DIR__ . '/dumpIterator.php';
  32
  33 /**
  34  * Maintenance script to take page text out of an XML dump file and render
  35  * basic HTML out to files.
  36  *
  37  * @ingroup Maintenance
  38  */
  39 class CompareParsers extends DumpIterator {
  40
  41         private $count = 0;
  42
  43         public function __construct() {
  44                 parent::__construct();
  45                 $this->saveFailed = false;
  46                 $this->mDescription = "Run a file or dump with several parsers";
  47                 $this->addOption( 'parser1', 'The first parser to compare.', true, true );
  48                 $this->addOption( 'parser2', 'The second parser to compare.', true, true );
  49                 $this->addOption( 'tidy', 'Run tidy on the articles.', false, false );
  50                 $this->addOption( 'save-failed', 'Folder in which articles which differ will be stored.', false, true );
  51                 $this->addOption( 'show-diff', 'Show a diff of the two renderings.', false, false );
  52                 $this->addOption( 'diff-bin', 'Binary to use for diffing (can also be provided by DIFF env var).', false, false );
  53                 $this->addOption( 'strip-parameters', 'Remove parameters of html tags to increase readability.', false, false );
  54                 $this->addOption( 'show-parsed-output', 'Show the parsed html if both Parsers give the same output.', false, false );
  55         }
  56
  57         public function checkOptions() {
  58                 if ( $this->hasOption( 'save-failed' ) ) {
  59                         $this->saveFailed = $this->getOption( 'save-failed' );
  60                 }
  61
  62                 $this->stripParametersEnabled = $this->hasOption( 'strip-parameters' );
  63                 $this->showParsedOutput = $this->hasOption( 'show-parsed-output' );
  64
  65                 $this->showDiff = $this->hasOption( 'show-diff' );
  66                 if ( $this->showDiff ) {
  67                         $bin = $this->getOption( 'diff-bin', getenv( 'DIFF' ) );
  68                         if ( $bin != '' ) {
  69                                 global $wgDiff;
  70                                 $wgDiff = $bin;
  71                         }
  72                 }
  73
  74                 $user = new User();
  75                 $this->options = ParserOptions::newFromUser( $user );
  76
  77                 if ( $this->hasOption( 'tidy' ) ) {
  78                         global $wgUseTidy;
  79                         if ( !$wgUseTidy ) {
  80                                 $this->error( 'Tidy was requested but $wgUseTidy is not set in LocalSettings.php', true );
  81                         }
  82                         $this->options->setTidy( true );
  83                 }
  84
  85                 $this->failed = 0;
  86         }
  87
  88         public function conclusions() {
  89                 $this->error( "{$this->failed} failed revisions out of {$this->count}" );
  90                 if ( $this->count > 0 ) {
  91                         $this->output( " (" . ( $this->failed / $this->count ) . "%)\n" );
  92                 }
  93         }
  94
  95         function stripParameters( $text ) {
  96                 if ( !$this->stripParametersEnabled ) {
  97                         return $text;
  98                 }
  99                 return preg_replace( '/(<a) [^>]+>/', '$1>', $text );
 100         }
 101
 102         /**
 103          * Callback function for each revision, parse with both parsers and compare
 104          * @param $rev Revision
 105          */
 106         public function processRevision( $rev ) {
 107                 $title = $rev->getTitle();
 108
 109                 $parser1Name = $this->getOption( 'parser1' );
 110                 $parser2Name = $this->getOption( 'parser2' );
 111
 112                 self::checkParserLocally( $parser1Name );
 113                 self::checkParserLocally( $parser2Name );
 114
 115                 $parser1 = new $parser1Name();
 116                 $parser2 = new $parser2Name();
 117
 118                 $content = $rev->getContent();
 119
 120                 if ( $content->getModel() !== CONTENT_MODEL_WIKITEXT ) {
 121                         $this->error( "Page {$title->getPrefixedText()} does not contain wikitext but {$content->getModel()}\n" );
 122                         return;
 123                 }
 124
 125                 $text = strval( $content->getNativeData() );
 126
 127                 $output1 = $parser1->parse( $text, $title, $this->options );
 128                 $output2 = $parser2->parse( $text, $title, $this->options );
 129
 130                 if ( $output1->getText() != $output2->getText() ) {
 131                         $this->failed++;
 132                         $this->error( "Parsing for {$title->getPrefixedText()} differs\n" );
 133
 134                         if ( $this->saveFailed ) {
 135                                 file_put_contents( $this->saveFailed . '/' . rawurlencode( $title->getPrefixedText() ) . ".txt", $text );
 136                         }
 137                         if ( $this->showDiff ) {
 138                                 $this->output( wfDiff( $this->stripParameters( $output1->getText() ), $this->stripParameters( $output2->getText() ), '' ) );
 139                         }
 140                 } else {
 141                         $this->output( $title->getPrefixedText() . "\tOK\n" );
 142                         if ( $this->showParsedOutput ) {
 143                                 $this->output( $this->stripParameters( $output1->getText() ) );
 144                         }
 145                 }
 146         }
 147
 148         private static function checkParserLocally( $parserName ) {
 149                 /* Look for the parser in a file appropiately named in the current folder */
 150                 if ( !class_exists( $parserName ) && file_exists( "$parserName.php" ) ) {
 151                         global $wgAutoloadClasses;
 152                         $wgAutoloadClasses[ $parserName ] = realpath( '.' ) . "/$parserName.php";
 153                 }
 154         }
 155
 156 }
 157
 158 $maintClass = "CompareParsers";
 159 require_once RUN_MAINTENANCE_IF_MAIN;