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         var $infiles = null;
  49
  50         function BaseDump( $infile ) {
  51                 $this->infiles = explode( ';', $infile );
  52                 $this->reader = new XMLReader();
  53                 $infile = array_shift( $this->infiles );
  54                 if ( defined( 'LIBXML_PARSEHUGE' ) ) {
  55                         $this->reader->open( $infile, null, LIBXML_PARSEHUGE );
  56                 }
  57                 else {
  58                         $this->reader->open( $infile );
  59                 }
  60         }
  61
  62         /**
  63          * Attempts to fetch the text of a particular page revision
  64          * from the dump stream. May return null if the page is
  65          * unavailable.
  66          *
  67          * @param $page Integer: ID number of page to read
  68          * @param $rev Integer: ID number of revision to read
  69          * @return string or null
  70          */
  71         function prefetch( $page, $rev ) {
  72                 $page = intval( $page );
  73                 $rev = intval( $rev );
  74                 while ( $this->lastPage < $page && !$this->atEnd ) {
  75                         $this->debug( "BaseDump::prefetch at page $this->lastPage, looking for $page" );
  76                         $this->nextPage();
  77                 }
  78                 if ( $this->lastPage > $page || $this->atEnd ) {
  79                         $this->debug( "BaseDump::prefetch already past page $page looking for rev $rev  [$this->lastPage, $this->lastRev]" );
  80                         return null;
  81                 }
  82                 while ( $this->lastRev < $rev && !$this->atEnd && !$this->atPageEnd ) {
  83                         $this->debug( "BaseDump::prefetch at page $this->lastPage, rev $this->lastRev, looking for $page, $rev" );
  84                         $this->nextRev();
  85                 }
  86                 if ( $this->lastRev == $rev && !$this->atEnd ) {
  87                         $this->debug( "BaseDump::prefetch hit on $page, $rev [$this->lastPage, $this->lastRev]" );
  88                         return $this->nextText();
  89                 } else {
  90                         $this->debug( "BaseDump::prefetch already past rev $rev on page $page  [$this->lastPage, $this->lastRev]" );
  91                         return null;
  92                 }
  93         }
  94
  95         function debug( $str ) {
  96                 wfDebug( $str . "\n" );
  97                 // global $dumper;
  98                 // $dumper->progress( $str );
  99         }
 100
 101         /**
 102          * @access private
 103          */
 104         function nextPage() {
 105                 if ( $this->skipTo( 'page', 'mediawiki' ) ) {
 106                         if ( $this->skipTo( 'id' ) ) {
 107                                 $this->lastPage = intval( $this->nodeContents() );
 108                                 $this->lastRev = 0;
 109                                 $this->atPageEnd = false;
 110                         }
 111                 } else {
 112                         $this->close();
 113                         if ( count( $this->infiles ) ) {
 114                                 $infile = array_shift( $this->infiles );
 115                                 $this->reader->open( $infile );
 116                                 $this->atEnd = false;
 117                         }
 118                 }
 119         }
 120
 121         /**
 122          * @access private
 123          */
 124         function nextRev() {
 125                 if ( $this->skipTo( 'revision' ) ) {
 126                         if ( $this->skipTo( 'id' ) ) {
 127                                 $this->lastRev = intval( $this->nodeContents() );
 128                         }
 129                 } else {
 130                         $this->atPageEnd = true;
 131                 }
 132         }
 133
 134         /**
 135          * @access private
 136          * @return string
 137          */
 138         function nextText() {
 139                 $this->skipTo( 'text' );
 140                 return strval( $this->nodeContents() );
 141         }
 142
 143         /**
 144          * @access private
 145          * @param $name string
 146          * @param $parent string
 147          * @return bool|null
 148          */
 149         function skipTo( $name, $parent = 'page' ) {
 150                 if ( $this->atEnd ) {
 151                         return false;
 152                 }
 153                 while ( $this->reader->read() ) {
 154                         if ( $this->reader->nodeType == XMLReader::ELEMENT &&
 155                                 $this->reader->name == $name ) {
 156                                 return true;
 157                         }
 158                         if ( $this->reader->nodeType == XMLReader::END_ELEMENT &&
 159                                 $this->reader->name == $parent ) {
 160                                 $this->debug( "BaseDump::skipTo found </$parent> searching for <$name>" );
 161                                 return false;
 162                         }
 163                 }
 164                 return $this->close();
 165         }
 166
 167         /**
 168          * Shouldn't something like this be built-in to XMLReader?
 169          * Fetches text contents of the current element, assuming
 170          * no sub-elements or such scary things.
 171          *
 172          * @return String
 173          * @access private
 174          */
 175         function nodeContents() {
 176                 if ( $this->atEnd ) {
 177                         return null;
 178                 }
 179                 if ( $this->reader->isEmptyElement ) {
 180                         return "";
 181                 }
 182                 $buffer = "";
 183                 while ( $this->reader->read() ) {
 184                         switch ( $this->reader->nodeType ) {
 185                         case XMLReader::TEXT:
 186 //                      case XMLReader::WHITESPACE:
 187                         case XMLReader::SIGNIFICANT_WHITESPACE:
 188                                 $buffer .= $this->reader->value;
 189                                 break;
 190                         case XMLReader::END_ELEMENT:
 191                                 return $buffer;
 192                         }
 193                 }
 194                 return $this->close();
 195         }
 196
 197         /**
 198          * @access private
 199          * @return null
 200          */
 201         function close() {
 202                 $this->reader->close();
 203                 $this->atEnd = true;
 204                 return null;
 205         }
 206 }