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  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @package MediaWiki
  22  * @subpackage SpecialPage
  23  */
  24
  25 $originalDir = getcwd();
  26
  27 require_once( 'commandLine.inc' );
  28 require_once( 'SpecialExport.php' );
  29 require_once( 'maintenance/backup.inc' );
  30
  31 class TextPassDumper extends BackupDumper {
  32         var $prefetch = null;
  33         var $input = "php://stdin";
  34         var $history = MW_EXPORT_FULL;
  35
  36         function dump() {
  37                 # This shouldn't happen if on console... ;)
  38                 header( 'Content-type: text/html; charset=UTF-8' );
  39
  40                 # Notice messages will foul up your XML output even if they're
  41                 # relatively harmless.
  42 //              ini_set( 'display_errors', false );
  43
  44                 $this->initProgress( $this->history );
  45
  46                 $this->db =& $this->backupDb();
  47
  48                 $this->egress = new ExportProgressFilter( $this->sink, $this );
  49
  50                 $input = fopen( $this->input, "rt" );
  51                 $result = $this->readDump( $input );
  52
  53                 if( WikiError::isError( $result ) ) {
  54                         wfDie( $result->getMessage() );
  55                 }
  56
  57                 $this->report( true );
  58         }
  59
  60         function processOption( $opt, $val, $param ) {
  61                 $url = $this->processFileOpt( $val, $param );
  62
  63                 switch( $opt ) {
  64                 case 'prefetch':
  65                         require_once 'maintenance/backupPrefetch.inc';
  66                         $this->prefetch = new BaseDump( $url );
  67                         break;
  68                 case 'stub':
  69                         $this->input = $url;
  70                         break;
  71                 case 'current':
  72                         $this->history = MW_EXPORT_CURRENT;
  73                         break;
  74                 case 'full':
  75                         $this->history = MW_EXPORT_FULL;
  76                         break;
  77                 }
  78         }
  79
  80         function processFileOpt( $val, $param ) {
  81                 switch( $val ) {
  82                 case "file":
  83                         return $param;
  84                 case "gzip":
  85                         return "compress.zlib://$param";
  86                 case "bzip2":
  87                         return "compress.bzip2://$param";
  88                 default:
  89                         return $val;
  90                 }
  91         }
  92
  93         function readDump( $input ) {
  94                 $this->buffer = "";
  95                 $this->openElement = false;
  96                 $this->atStart = true;
  97                 $this->state = "";
  98                 $this->lastName = "";
  99                 $this->thisPage = 0;
 100                 $this->thisRev = 0;
 101
 102                 $parser = xml_parser_create( "UTF-8" );
 103                 xml_parser_set_option( $parser, XML_OPTION_CASE_FOLDING, false );
 104
 105                 xml_set_element_handler( $parser, array( &$this, 'startElement' ), array( &$this, 'endElement' ) );
 106                 xml_set_character_data_handler( $parser, array( &$this, 'characterData' ) );
 107
 108                 $offset = 0; // for context extraction on error reporting
 109                 $bufferSize = 512 * 1024;
 110                 do {
 111                         $chunk = fread( $input, $bufferSize );
 112                         if( !xml_parse( $parser, $chunk, feof( $input ) ) ) {
 113                                 wfDebug( "TextDumpPass::readDump encountered XML parsing error\n" );
 114                                 return new WikiXmlError( $parser, 'XML import parse failure', $chunk, $offset );
 115                         }
 116                         $offset += strlen( $chunk );
 117                 } while( $chunk !== false && !feof( $input ) );
 118                 xml_parser_free( $parser );
 119
 120                 return true;
 121         }
 122
 123         function getText( $id ) {
 124                 if( isset( $this->prefetch ) ) {
 125                         $text = $this->prefetch->prefetch( $this->thisPage, $this->thisRev );
 126                         if( $text === null ) {
 127                                 // Entry missing from prefetch dump
 128                         } elseif( $text === "" ) {
 129                                 // Blank entries may indicate that the prior dump was broken.
 130                                 // To be safe, reload it.
 131                         } else {
 132                                 return $text;
 133                         }
 134                 }
 135                 $id = intval( $id );
 136                 $row = $this->db->selectRow( 'text',
 137                         array( 'old_text', 'old_flags' ),
 138                         array( 'old_id' => $id ),
 139                         'TextPassDumper::getText' );
 140                 $text = Revision::getRevisionText( $row );
 141                 $stripped = str_replace( "\r", "", $text );
 142                 $normalized = UtfNormal::cleanUp( $stripped );
 143                 return $normalized;
 144         }
 145
 146         function startElement( $parser, $name, $attribs ) {
 147                 $this->clearOpenElement( null );
 148                 $this->lastName = $name;
 149
 150                 if( $name == 'revision' ) {
 151                         $this->state = $name;
 152                         $this->egress->writeOpenPage( null, $this->buffer );
 153                         $this->buffer = "";
 154                 } elseif( $name == 'page' ) {
 155                         $this->state = $name;
 156                         if( $this->atStart ) {
 157                                 $this->egress->writeOpenStream( $this->buffer );
 158                                 $this->buffer = "";
 159                                 $this->atStart = false;
 160                         }
 161                 }
 162
 163                 if( $name == "text" && isset( $attribs['id'] ) ) {
 164                         $text = $this->getText( $attribs['id'] );
 165                         $this->openElement = array( $name, array( 'xml:space' => 'preserve' ) );
 166                         if( strlen( $text ) > 0 ) {
 167                                 $this->characterData( $parser, $text );
 168                         }
 169                 } else {
 170                         $this->openElement = array( $name, $attribs );
 171                 }
 172         }
 173
 174         function endElement( $parser, $name ) {
 175                 if( $this->openElement ) {
 176                         $this->clearOpenElement( "" );
 177                 } else {
 178                         $this->buffer .= "</$name>";
 179                 }
 180
 181                 if( $name == 'revision' ) {
 182                         $this->egress->writeRevision( null, $this->buffer );
 183                         $this->buffer = "";
 184                         $this->thisRev = "";
 185                 } elseif( $name == 'page' ) {
 186                         $this->egress->writeClosePage( $this->buffer );
 187                         $this->buffer = "";
 188                         $this->thisPage = "";
 189                 } elseif( $name == 'mediawiki' ) {
 190                         $this->egress->writeCloseStream( $this->buffer );
 191                         $this->buffer = "";
 192                 }
 193         }
 194
 195         function characterData( $parser, $data ) {
 196                 $this->clearOpenElement( null );
 197                 if( $this->lastName == "id" ) {
 198                         if( $this->state == "revision" ) {
 199                                 $this->thisRev .= $data;
 200                         } elseif( $this->state == "page" ) {
 201                                 $this->thisPage .= $data;
 202                         }
 203                 }
 204                 $this->buffer .= htmlspecialchars( $data );
 205         }
 206
 207         function clearOpenElement( $style ) {
 208                 if( $this->openElement ) {
 209                         $this->buffer .= wfElement( $this->openElement[0], $this->openElement[1], $style );
 210                         $this->openElement = false;
 211                 }
 212         }
 213 }
 214
 215
 216 $dumper = new TextPassDumper( $argv );
 217
 218 if( true ) {
 219         $dumper->dump();
 220 } else {
 221         $dumper->progress( <<<END
 222 This script postprocesses XML dumps from dumpBackup.php to add
 223 page text which was stubbed out (using --stub).
 224
 225 XML input is accepted on stdin.
 226 XML output is sent to stdout; progress reports are sent to stderr.
 227
 228 Usage: php dumpTextPass.php [<options>]
 229 Options:
 230   --stub=<type>:<file> To load a compressed stub dump instead of stdin
 231   --prefetch=<type>:<file> Use a prior dump file as a text source, to save
 232               pressure on the database.
 233               (Requires PHP 5.0+ and the XMLReader PECL extension)
 234   --quiet     Don't dump status reports to stderr.
 235   --report=n  Report position and speed after every n pages processed.
 236               (Default: 100)
 237   --server=h  Force reading from MySQL server h
 238   --current   Base ETA on number of pages in database instead of all revisions
 239 END
 240 );
 241 }
 242
 243 ?>