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 (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 abstract class DumpIterator extends Maintenance {
  31
  32         private $count = 0;
  33         private $startTime;
  34
  35         public function __construct() {
  36                 parent::__construct();
  37                 $this->mDescription = "Does something with a dump";
  38                 $this->addOption( 'file', 'File with text to run.', false, true );
  39                 $this->addOption( 'dump', 'XML dump to execute all revisions.', false, true );
  40                 $this->addOption( 'from', 'Article from XML dump to start from.', false, true );
  41         }
  42
  43         public function execute() {
  44                 if (! ( $this->hasOption('file') ^ $this->hasOption('dump') ) ) {
  45                         $this->error("You must provide a file or dump", true);
  46                 }
  47
  48                 $this->checkOptions();
  49
  50                 if ( $this->hasOption('file') ) {
  51                         $revision = new WikiRevision;
  52
  53                         $revision->setText( file_get_contents( $this->getOption( 'file' ) ) );
  54                         $revision->setTitle( Title::newFromText( rawurldecode( basename( $this->getOption( 'file' ), '.txt' ) ) ) );
  55                         $this->handleRevision( $revision );
  56                         return;
  57                 }
  58
  59                 $this->startTime = wfTime();
  60
  61                 if ( $this->getOption('dump') == '-' ) {
  62                         $source = new ImportStreamSource( $this->getStdin() );
  63                 } else {
  64                         $this->error("Sorry, I don't support dump filenames yet. Use - and provide it on stdin on the meantime.", true);
  65                 }
  66                 $importer = new WikiImporter( $source );
  67
  68                 $importer->setRevisionCallback(
  69                         array( &$this, 'handleRevision' ) );
  70
  71                 $this->from = $this->getOption( 'from', null );
  72                 $this->count = 0;
  73                 $importer->doImport();
  74
  75                 $this->conclusions();
  76
  77                 $delta = wfTime() - $this->startTime;
  78                 $this->error( "Done {$this->count} revisions in " . round($delta, 2) . " seconds " );
  79                 if ($delta > 0)
  80                         $this->error( round($this->count / $delta, 2) . " pages/sec" );
  81
  82                 # Perform the memory_get_peak_usage() when all the other data has been output so there's no damage if it dies.
  83                 # It is only available since 5.2.0 (since 5.2.1 if you haven't compiled with --enable-memory-limit)
  84                 $this->error( "Memory peak usage of " . memory_get_peak_usage() . " bytes\n" );
  85         }
  86
  87         function stripParameters( $text ) {
  88                 if ( !$this->stripParametersEnabled ) {
  89                         return $text;
  90                 }
  91                 return preg_replace( '/(<a) [^>]+>/', '$1>', $text );
  92         }
  93
  94         /**
  95          * Callback function for each revision, child classes should override
  96          * processRevision instead.
  97          * @param $rev Revision
  98          */
  99         public function handleRevision( $rev ) {
 100                 $title = $rev->getTitle();
 101                 if ( !$title ) {
 102                         $this->error( "Got bogus revision with null title!" );
 103                         return;
 104                 }
 105
 106                 $this->count++;
 107                 if ( isset( $this->from ) ) {
 108                         if ( $this->from != $title )
 109                                 return;
 110                         $this->output( "Skipped " . ($this->count - 1) . " pages\n" );
 111
 112                         $this->count = 1;
 113                         $this->from = null;
 114                 }
 115
 116                 $this->processRevision( $rev );
 117         }
 118
 119         /* Stub function for processing additional options */
 120         public function checkOptions() {
 121                 return;
 122         }
 123
 124         /* Stub function for giving data about what was computed */
 125         public function conclusions() {
 126                 return;
 127         }
 128
 129         /* Core function which does whatever the maintenance script is designed to do */
 130         abstract public function processRevision( $rev );
 131 }
 132
 133 class SearchDump extends DumpIterator {
 134
 135         public function __construct() {
 136                 parent::__construct();
 137                 $this->mDescription = "Runs a regex in the revisions from a dump";
 138                 $this->addOption( 'regex', 'Searching regex', true, true );
 139         }
 140
 141         public function processRevision( $rev ) {
 142                 if ( preg_match( $this->getOption( 'regex' ), $rev->getText() ) ) {
 143                         $this->output( $rev->getTitle() . " matches at edit from " . $rev->getTimestamp() . "\n" );
 144                 }
 145         }
 146 }
 147
 148 $maintClass = "SearchDump";
 149 require_once( RUN_MAINTENANCE_IF_MAIN );