maintenance/backupPrefetch.inc

   1 <?php
   2 /**
   3  * Helper class for the --prefetch option of dumpTextPass.php
   4  *
   5  * Copyright © 2005 Brion Vibber <brion@pobox.com>
   6  * http://www.mediawiki.org/
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with this program; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  21  * http://www.gnu.org/copyleft/gpl.html
  22  *
  23  * @file
  24  * @ingroup Maintenance
  25  */
  26
  27 /**
  28  * Readahead helper for making large MediaWiki data dumps;
  29  * reads in a previous XML dump to sequentially prefetch text
  30  * records already normalized and decompressed.
  31  *
  32  * This can save load on the external database servers, hopefully.
  33  *
  34  * Assumes that dumps will be recorded in the canonical order:
  35  * - ascending by page_id
  36  * - ascending by rev_id within each page
  37  * - text contents are immutable and should not change once
  38  *   recorded, so the previous dump is a reliable source
  39  *
  40  * @ingroup Maintenance
  41  */
  42 class BaseDump {
  43         var $reader = null;
  44         var $atEnd = false;
  45         var $atPageEnd = false;
  46         var $lastPage = 0;
  47         var $lastRev = 0;
  48
  49         function BaseDump( $infile ) {
  50                 $this->reader = new XMLReader();
  51                 $this->reader->open( $infile );
  52         }
  53
  54         /**
  55          * Attempts to fetch the text of a particular page revision
  56          * from the dump stream. May return null if the page is
  57          * unavailable.
  58          *
  59          * @param $page Integer: ID number of page to read
  60          * @param $rev Integer: ID number of revision to read
  61          * @return string or null
  62          */
  63         function prefetch( $page, $rev ) {
  64                 $page = intval( $page );
  65                 $rev = intval( $rev );
  66                 while ( $this->lastPage < $page && !$this->atEnd ) {
  67                         $this->debug( "BaseDump::prefetch at page $this->lastPage, looking for $page" );
  68                         $this->nextPage();
  69                 }
  70                 if ( $this->lastPage > $page || $this->atEnd ) {
  71                         $this->debug( "BaseDump::prefetch already past page $page looking for rev $rev  [$this->lastPage, $this->lastRev]" );
  72                         return null;
  73                 }
  74                 while ( $this->lastRev < $rev && !$this->atEnd && !$this->atPageEnd ) {
  75                         $this->debug( "BaseDump::prefetch at page $this->lastPage, rev $this->lastRev, looking for $page, $rev" );
  76                         $this->nextRev();
  77                 }
  78                 if ( $this->lastRev == $rev && !$this->atEnd ) {
  79                         $this->debug( "BaseDump::prefetch hit on $page, $rev [$this->lastPage, $this->lastRev]" );
  80                         return $this->nextText();
  81                 } else {
  82                         $this->debug( "BaseDump::prefetch already past rev $rev on page $page  [$this->lastPage, $this->lastRev]" );
  83                         return null;
  84                 }
  85         }
  86
  87         function debug( $str ) {
  88                 wfDebug( $str . "\n" );
  89                 // global $dumper;
  90                 // $dumper->progress( $str );
  91         }
  92
  93         /**
  94          * @access private
  95          */
  96         function nextPage() {
  97                 if ( $this->skipTo( 'page', 'mediawiki' ) ) {
  98                         if ( $this->skipTo( 'id' ) ) {
  99                                 $this->lastPage = intval( $this->nodeContents() );
 100                                 $this->lastRev = 0;
 101                                 $this->atPageEnd = false;
 102                         }
 103                 } else {
 104                         $this->atEnd = true;
 105                 }
 106         }
 107
 108         /**
 109          * @access private
 110          */
 111         function nextRev() {
 112                 if ( $this->skipTo( 'revision' ) ) {
 113                         if ( $this->skipTo( 'id' ) ) {
 114                                 $this->lastRev = intval( $this->nodeContents() );
 115                         }
 116                 } else {
 117                         $this->atPageEnd = true;
 118                 }
 119         }
 120
 121         /**
 122          * @access private
 123          */
 124         function nextText() {
 125                 $this->skipTo( 'text' );
 126                 return strval( $this->nodeContents() );
 127         }
 128
 129         /**
 130          * @access private
 131          */
 132         function skipTo( $name, $parent = 'page' ) {
 133                 if ( $this->atEnd ) {
 134                         return false;
 135                 }
 136                 while ( $this->reader->read() ) {
 137                         if ( $this->reader->nodeType == XMLReader::ELEMENT &&
 138                                 $this->reader->name == $name ) {
 139                                 return true;
 140                         }
 141                         if ( $this->reader->nodeType == XMLReader::END_ELEMENT &&
 142                                 $this->reader->name == $parent ) {
 143                                 $this->debug( "BaseDump::skipTo found </$parent> searching for <$name>" );
 144                                 return false;
 145                         }
 146                 }
 147                 return $this->close();
 148         }
 149
 150         /**
 151          * Shouldn't something like this be built-in to XMLReader?
 152          * Fetches text contents of the current element, assuming
 153          * no sub-elements or such scary things.
 154          *
 155          * @return String
 156          * @access private
 157          */
 158         function nodeContents() {
 159                 if ( $this->atEnd ) {
 160                         return null;
 161                 }
 162                 if ( $this->reader->isEmptyElement ) {
 163                         return "";
 164                 }
 165                 $buffer = "";
 166                 while ( $this->reader->read() ) {
 167                         switch( $this->reader->nodeType ) {
 168                         case XMLReader::TEXT:
 169 //                      case XMLReader::WHITESPACE:
 170                         case XMLReader::SIGNIFICANT_WHITESPACE:
 171                                 $buffer .= $this->reader->value;
 172                                 break;
 173                         case XMLReader::END_ELEMENT:
 174                                 return $buffer;
 175                         }
 176                 }
 177                 return $this->close();
 178         }
 179
 180         /**
 181          * @access private
 182          */
 183         function close() {
 184                 $this->reader->close();
 185                 $this->atEnd = true;
 186                 return null;
 187         }
 188 }