maintenance/preprocessDump.php

   1 <?php
   2 /**
   3  * Take page text out of an XML dump file and preprocess it to obj.
   4  * It may be useful for getting preprocessor statistics or filling the
   5  * preprocessor cache.
   6  *
   7  * Copyright (C) 2011 Platonides - http://www.mediawiki.org/
   8  *
   9  * This program is free software; you can redistribute it and/or modify
  10  * it under the terms of the GNU General Public License as published by
  11  * the Free Software Foundation; either version 2 of the License, or
  12  * (at your option) any later version.
  13  *
  14  * This program is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  17  * GNU General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU General Public License along
  20  * with this program; if not, write to the Free Software Foundation, Inc.,
  21  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  22  * http://www.gnu.org/copyleft/gpl.html
  23  *
  24  * @file
  25  * @ingroup Maintenance
  26  */
  27
  28 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  29
  30 class PreprocessDump extends Maintenance {
  31
  32         private $count = 0;
  33         private $outputDirectory, $startTime;
  34
  35         /* Variables for dressing up as a parser */
  36         public $mTitle = 'PreprocessDump';
  37         public $mPPNodeCount = 0;
  38
  39         public function getStripList() {
  40                 global $wgParser;
  41                 return $wgParser->getStripList();
  42         }
  43
  44         public function __construct() {
  45                 parent::__construct();
  46                 $this->saveFailed = false;
  47                 $this->mDescription = "Run a file or dump with a preprocessor";
  48                 $this->addOption( 'file',  'File with text to run.', false, true );
  49                 $this->addOption( 'dump',  'XML dump to execute all revisions.', false, true );
  50                 $this->addOption( 'from',  'Article from XML dump to start from.', false, true );
  51                 $this->addOption( 'cache', 'Use and populate the preprocessor cache.', false, false );
  52                 $this->addOption( 'preprocessor', 'Preprocessor to use.', false, false );
  53         }
  54
  55         public function getDbType() {
  56                 return Maintenance::DB_NONE;
  57         }
  58
  59         public function execute() {
  60                 global $wgParser, $wgParserConf, $wgPreprocessorCacheThreshold;
  61
  62                 if (! ( $this->hasOption( 'file' ) ^ $this->hasOption( 'dump' ) ) ) {
  63                         $this->error("You must provide a file or dump", true);
  64                 }
  65
  66                 if ( !$this->hasOption( 'cache' ) ) {
  67                         $wgPreprocessorCacheThreshold = false;
  68                         $this->saveFailed = $this->getOption('save-failed');
  69                 }
  70
  71                 if ( $this->hasOption( 'preprocessor' ) ) {
  72                         $name = $this->getOption( 'preprocessor' );
  73                 } elseif ( isset( $wgParserConf['preprocessorClass'] ) ) {
  74                         $name = $wgParserConf['preprocessorClass'];
  75                 } else {
  76                         $name = 'Preprocessor_DOM';
  77                 }
  78
  79                 $wgParser->firstCallInit();
  80                 $this->mPreprocessor = new $name( $this );
  81
  82                 if ( $this->hasOption( 'file' ) ) {
  83                         $revision = new WikiRevision;
  84
  85                         $revision->setText( file_get_contents( $this->getOption( 'file' ) ) );
  86                         $revision->setTitle( Title::newFromText( rawurldecode( basename( $this->getOption('file'), '.txt' ) ) ) );
  87                         $this->handleRevision( $revision );
  88                         return;
  89                 }
  90
  91                 $this->startTime = wfTime();
  92
  93                 if ( $this->getOption('dump') == '-' ) {
  94                         $source = new ImportStreamSource( $this->getStdin() );
  95                 } else {
  96                         $this->error("Sorry, I don't support dump filenames yet. Use - and provide it on stdin on the meantime.", true);
  97                 }
  98                 $importer = new WikiImporter( $source );
  99
 100                 $importer->setRevisionCallback(
 101                         array( &$this, 'handleRevision' ) );
 102
 103                 $this->from = $this->getOption( 'from', null );
 104                 $this->count = 0;
 105                 $importer->doImport();
 106
 107                 $delta = wfTime() - $this->startTime;
 108                 $this->error( "{$this->count} revisions preprocessed in " . round($delta, 2) . " seconds " );
 109                 if ($delta > 0)
 110                         $this->error( round($this->count / $delta, 2) . " pages/sec" );
 111
 112                 # Perform the memory_get_peak_usage() when all the other data has been output so there's no damage if it dies.
 113                 # It is only available since 5.2.0 (since 5.2.1 if you haven't compiled with --enable-memory-limit)
 114                 $this->error( "Memory peak usage of " . memory_get_peak_usage() . " bytes\n" );
 115         }
 116
 117         /**
 118          * Callback function for each revision, preprocessToObj()
 119          * @param $rev Revision
 120          */
 121         public function handleRevision( $rev ) {
 122                 $title = $rev->getTitle();
 123                 if ( !$title ) {
 124                         $this->error( "Got bogus revision with null title!" );
 125                         return;
 126                 }
 127
 128                 $this->count++;
 129                 if ( isset( $this->from ) ) {
 130                         if ( $this->from != $title )
 131                                 return;
 132                         $this->output( "Skipped " . ($this->count - 1) . " pages\n" );
 133
 134                         $this->count = 1;
 135                         $this->from = null;
 136                 }
 137                 try {
 138                         $this->mPreprocessor->preprocessToObj( $rev->getText(), 0 );
 139                 }
 140                 catch(Exception $e) {
 141                         $this->error("Caught exception " . $e->getMessage() . " in " . $title-> getPrefixedText() );
 142                 }
 143         }
 144 }
 145
 146 $maintClass = "PreprocessDump";
 147 require_once( DO_MAINTENANCE );
 148