maintenance/dumpTextPass.php

   1 <?php
   2 /**
   3  * Copyright (C) 2005 Brion Vibber <brion@pobox.com>
   4  * http://www.mediawiki.org/
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License along
  17  * with this program; if not, write to the Free Software Foundation, Inc.,
  18  * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @package MediaWiki
  22  * @subpackage SpecialPage
  23  */
  24
  25 $originalDir = getcwd();
  26
  27 $optionsWithArgs = array( 'server', 'pagelist', 'start', 'end' );
  28
  29 require_once( 'commandLine.inc' );
  30 require_once( 'SpecialExport.php' );
  31 require_once( 'maintenance/backup.inc' );
  32
  33 class TextPassDumper extends BackupDumper {
  34         var $prefetch = null;
  35         var $input = "php://stdin";
  36         var $history = MW_EXPORT_FULL;
  37
  38         function dump() {
  39                 # This shouldn't happen if on console... ;)
  40                 header( 'Content-type: text/html; charset=UTF-8' );
  41
  42                 # Notice messages will foul up your XML output even if they're
  43                 # relatively harmless.
  44 //              ini_set( 'display_errors', false );
  45
  46                 $this->initProgress( $this->history );
  47
  48                 $this->db =& $this->backupDb();
  49
  50                 $this->egress = new ExportProgressFilter( $this->sink, $this );
  51
  52                 $input = fopen( $this->input, "rt" );
  53                 $result = $this->readDump( $input );
  54
  55                 if( WikiError::isError( $result ) ) {
  56                         wfDie( $result->getMessage() );
  57                 }
  58
  59                 $this->report( true );
  60         }
  61
  62         function processOption( $opt, $val, $param ) {
  63                 $url = $this->processFileOpt( $val, $param );
  64
  65                 switch( $opt ) {
  66                 case 'prefetch':
  67                         require_once 'maintenance/backupPrefetch.inc';
  68                         $this->prefetch = new BaseDump( $url );
  69                         break;
  70                 case 'stub':
  71                         $this->input = $url;
  72                         break;
  73                 case 'current':
  74                         $this->history = MW_EXPORT_CURRENT;
  75                         break;
  76                 case 'full':
  77                         $this->history = MW_EXPORT_FULL;
  78                         break;
  79                 }
  80         }
  81
  82         function processFileOpt( $val, $param ) {
  83                 switch( $val ) {
  84                 case "file":
  85                         return $param;
  86                 case "gzip":
  87                         return "compress.zlib://$param";
  88                 case "bzip2":
  89                         return "compress.bzip2://$param";
  90                 default:
  91                         return $val;
  92                 }
  93         }
  94
  95         function readDump( $input ) {
  96                 $this->buffer = "";
  97                 $this->openElement = false;
  98                 $this->atStart = true;
  99                 $this->state = "";
 100                 $this->lastName = "";
 101                 $this->thisPage = 0;
 102                 $this->thisRev = 0;
 103
 104                 $parser = xml_parser_create( "UTF-8" );
 105                 xml_parser_set_option( $parser, XML_OPTION_CASE_FOLDING, false );
 106
 107                 xml_set_element_handler( $parser, array( &$this, 'startElement' ), array( &$this, 'endElement' ) );
 108                 xml_set_character_data_handler( $parser, array( &$this, 'characterData' ) );
 109
 110                 $offset = 0; // for context extraction on error reporting
 111                 $bufferSize = 512 * 1024;
 112                 do {
 113                         $chunk = fread( $input, $bufferSize );
 114                         if( !xml_parse( $parser, $chunk, feof( $input ) ) ) {
 115                                 wfDebug( "TextDumpPass::readDump encountered XML parsing error\n" );
 116                                 return new WikiXmlError( $parser, 'XML import parse failure', $chunk, $offset );
 117                         }
 118                         $offset += strlen( $chunk );
 119                 } while( $chunk !== false && !feof( $input ) );
 120                 xml_parser_free( $parser );
 121
 122                 return true;
 123         }
 124
 125         function getText( $id ) {
 126                 if( isset( $this->prefetch ) ) {
 127                         $text = $this->prefetch->prefetch( $this->thisPage, $this->thisRev );
 128                         if( !is_null( $text ) )
 129                                 return $text;
 130                 }
 131                 $id = intval( $id );
 132                 $row = $this->db->selectRow( 'text',
 133                         array( 'old_text', 'old_flags' ),
 134                         array( 'old_id' => $id ),
 135                         'TextPassDumper::getText' );
 136                 $text = Revision::getRevisionText( $row );
 137                 $stripped = str_replace( "\r", "", $text );
 138                 $normalized = UtfNormal::cleanUp( $stripped );
 139                 return $normalized;
 140         }
 141
 142         function startElement( $parser, $name, $attribs ) {
 143                 $this->clearOpenElement( null );
 144                 $this->lastName = $name;
 145
 146                 if( $name == 'revision' ) {
 147                         $this->state = $name;
 148                         $this->egress->writeOpenPage( null, $this->buffer );
 149                         $this->buffer = "";
 150                 } elseif( $name == 'page' ) {
 151                         $this->state = $name;
 152                         if( $this->atStart ) {
 153                                 $this->egress->writeOpenStream( $this->buffer );
 154                                 $this->buffer = "";
 155                                 $this->atStart = false;
 156                         }
 157                 }
 158
 159                 if( $name == "text" && isset( $attribs['id'] ) ) {
 160                         $text = $this->getText( $attribs['id'] );
 161                         $this->openElement = array( $name, array( 'xml:space' => 'preserve' ) );
 162                         if( strlen( $text ) > 0 ) {
 163                                 $this->characterData( $parser, $text );
 164                         }
 165                 } else {
 166                         $this->openElement = array( $name, $attribs );
 167                 }
 168         }
 169
 170         function endElement( $parser, $name ) {
 171                 if( $this->openElement ) {
 172                         $this->clearOpenElement( "" );
 173                 } else {
 174                         $this->buffer .= "</$name>";
 175                 }
 176
 177                 if( $name == 'revision' ) {
 178                         $this->egress->writeRevision( null, $this->buffer );
 179                         $this->buffer = "";
 180                         $this->thisRev = "";
 181                 } elseif( $name == 'page' ) {
 182                         $this->egress->writeClosePage( $this->buffer );
 183                         $this->buffer = "";
 184                         $this->thisPage = "";
 185                 } elseif( $name == 'mediawiki' ) {
 186                         $this->egress->writeCloseStream( $this->buffer );
 187                         $this->buffer = "";
 188                 }
 189         }
 190
 191         function characterData( $parser, $data ) {
 192                 $this->clearOpenElement( null );
 193                 if( $this->lastName == "id" ) {
 194                         if( $this->state == "revision" ) {
 195                                 $this->thisRev .= $data;
 196                         } elseif( $this->state == "page" ) {
 197                                 $this->thisPage .= $data;
 198                         }
 199                 }
 200                 $this->buffer .= htmlspecialchars( $data );
 201         }
 202
 203         function clearOpenElement( $style ) {
 204                 if( $this->openElement ) {
 205                         $this->buffer .= wfElement( $this->openElement[0], $this->openElement[1], $style );
 206                         $this->openElement = false;
 207                 }
 208         }
 209 }
 210
 211
 212 $dumper = new TextPassDumper( $argv );
 213 if( isset( $options['server'] ) ) {
 214         $dumper->server = $options['server'];
 215 }
 216
 217 if( true ) {
 218         $dumper->dump();
 219 } else {
 220         $dumper->progress( <<<END
 221 This script postprocesses XML dumps from dumpBackup.php to add
 222 page text which was stubbed out (using --stub).
 223
 224 XML input is accepted on stdin.
 225 XML output is sent to stdout; progress reports are sent to stderr.
 226
 227 Usage: php dumpTextPass.php [<options>]
 228 Options:
 229   --stub=<type>:<file> To load a compressed stub dump instead of stdin
 230   --prefetch=<type>:<file> Use a prior dump file as a text source, to save
 231               pressure on the database.
 232               (Requires PHP 5.0+ and the XMLReader PECL extension)
 233   --quiet     Don't dump status reports to stderr.
 234   --report=n  Report position and speed after every n pages processed.
 235               (Default: 100)
 236   --server=h  Force reading from MySQL server h
 237   --current   Base ETA on number of pages in database instead of all revisions
 238 END
 239 );
 240 }
 241
 242 ?>