maintenance/dumpIterator.php

   1 <?php
   2 /**
   3  * Take page text out of an XML dump file and perform some operation on it.
   4  * Used as a base class for CompareParsers and PreprocessDump.
   5  * We implement below the simple task of searching inside a dump.
   6  *
   7  * Copyright © 2011 Platonides
   8  * http://www.mediawiki.org/
   9  *
  10  * This program is free software; you can redistribute it and/or modify
  11  * it under the terms of the GNU General Public License as published by
  12  * the Free Software Foundation; either version 2 of the License, or
  13  * (at your option) any later version.
  14  *
  15  * This program is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  18  * GNU General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU General Public License along
  21  * with this program; if not, write to the Free Software Foundation, Inc.,
  22  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  23  * http://www.gnu.org/copyleft/gpl.html
  24  *
  25  * @file
  26  * @ingroup Maintenance
  27  */
  28
  29 require_once( __DIR__ . '/Maintenance.php' );
  30
  31 /**
  32  * Base class for interating over a dump.
  33  *
  34  * @ingroup Maintenance
  35  */
  36 abstract class DumpIterator extends Maintenance {
  37
  38         private $count = 0;
  39         private $startTime;
  40
  41         public function __construct() {
  42                 parent::__construct();
  43                 $this->mDescription = "Does something with a dump";
  44                 $this->addOption( 'file', 'File with text to run.', false, true );
  45                 $this->addOption( 'dump', 'XML dump to execute all revisions.', false, true );
  46                 $this->addOption( 'from', 'Article from XML dump to start from.', false, true );
  47         }
  48
  49         public function execute() {
  50                 if ( !( $this->hasOption( 'file' ) ^ $this->hasOption( 'dump' ) ) ) {
  51                         $this->error( "You must provide a file or dump", true );
  52                 }
  53
  54                 $this->checkOptions();
  55
  56                 if ( $this->hasOption( 'file' ) ) {
  57                         $revision = new WikiRevision;
  58
  59                         $revision->setText( file_get_contents( $this->getOption( 'file' ) ) );
  60                         $revision->setTitle( Title::newFromText( rawurldecode( basename( $this->getOption( 'file' ), '.txt' ) ) ) );
  61                         $this->handleRevision( $revision );
  62                         return;
  63                 }
  64
  65                 $this->startTime = microtime( true );
  66
  67                 if ( $this->getOption( 'dump' ) == '-' ) {
  68                         $source = new ImportStreamSource( $this->getStdin() );
  69                 } else {
  70                         $this->error( "Sorry, I don't support dump filenames yet. Use - and provide it on stdin on the meantime.", true );
  71                 }
  72                 $importer = new WikiImporter( $source );
  73
  74                 $importer->setRevisionCallback(
  75                         array( &$this, 'handleRevision' ) );
  76
  77                 $this->from = $this->getOption( 'from', null );
  78                 $this->count = 0;
  79                 $importer->doImport();
  80
  81                 $this->conclusions();
  82
  83                 $delta = microtime( true ) - $this->startTime;
  84                 $this->error( "Done {$this->count} revisions in " . round( $delta, 2 ) . " seconds " );
  85                 if ( $delta > 0 ) {
  86                         $this->error( round( $this->count / $delta, 2 ) . " pages/sec" );
  87                 }
  88
  89                 # Perform the memory_get_peak_usage() when all the other data has been output so there's no damage if it dies.
  90                 # It is only available since 5.2.0 (since 5.2.1 if you haven't compiled with --enable-memory-limit)
  91                 $this->error( "Memory peak usage of " . memory_get_peak_usage() . " bytes\n" );
  92         }
  93
  94         public function finalSetup() {
  95                 parent::finalSetup();
  96
  97                 if ( $this->getDbType() == Maintenance::DB_NONE ) {
  98                         global $wgUseDatabaseMessages, $wgLocalisationCacheConf, $wgHooks;
  99                         $wgUseDatabaseMessages = false;
 100                         $wgLocalisationCacheConf['storeClass'] = 'LCStore_Null';
 101                         $wgHooks['InterwikiLoadPrefix'][] = 'DumpIterator::disableInterwikis';
 102                 }
 103         }
 104
 105         static function disableInterwikis( $prefix, &$data ) {
 106                 # Title::newFromText will check on each namespaced article if it's an interwiki.
 107                 # We always answer that it is not.
 108
 109                 return false;
 110         }
 111
 112         /**
 113          * Callback function for each revision, child classes should override
 114          * processRevision instead.
 115          * @param $rev Revision
 116          */
 117         public function handleRevision( $rev ) {
 118                 $title = $rev->getTitle();
 119                 if ( !$title ) {
 120                         $this->error( "Got bogus revision with null title!" );
 121                         return;
 122                 }
 123
 124                 $this->count++;
 125                 if ( isset( $this->from ) ) {
 126                         if ( $this->from != $title ) {
 127                                 return;
 128                         }
 129                         $this->output( "Skipped " . ( $this->count - 1 ) . " pages\n" );
 130
 131                         $this->count = 1;
 132                         $this->from = null;
 133                 }
 134
 135                 $this->processRevision( $rev );
 136         }
 137
 138         /* Stub function for processing additional options */
 139         public function checkOptions() {
 140                 return;
 141         }
 142
 143         /* Stub function for giving data about what was computed */
 144         public function conclusions() {
 145                 return;
 146         }
 147
 148         /* Core function which does whatever the maintenance script is designed to do */
 149         abstract public function processRevision( $rev );
 150 }
 151
 152 /**
 153  * Maintenance script that runs a regex in the revisions from a dump.
 154  *
 155  * @ingroup Maintenance
 156  */
 157 class SearchDump extends DumpIterator {
 158
 159         public function __construct() {
 160                 parent::__construct();
 161                 $this->mDescription = "Runs a regex in the revisions from a dump";
 162                 $this->addOption( 'regex', 'Searching regex', true, true );
 163         }
 164
 165         public function getDbType() {
 166                 return Maintenance::DB_NONE;
 167         }
 168
 169         /**
 170          * @param $rev Revision
 171          */
 172         public function processRevision( $rev ) {
 173                 if ( preg_match( $this->getOption( 'regex' ), $rev->getContent()->getTextForSearchIndex() ) ) {
 174                         $this->output( $rev->getTitle() . " matches at edit from " . $rev->getTimestamp() . "\n" );
 175                 }
 176         }
 177 }
 178
 179 $maintClass = "SearchDump";
 180 require_once( RUN_MAINTENANCE_IF_MAIN );