includes/import/ImportStreamSource.php

   1 <?php
   2 /**
   3  * MediaWiki page data importer.
   4  *
   5  * Copyright © 2003,2005 Brion Vibber <brion@pobox.com>
   6  * https://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 SpecialPage
  25  */
  26 use MediaWiki\MediaWikiServices;
  27
  28 /**
  29  * Imports a XML dump from a file (either from file upload, files on disk, or HTTP)
  30  * @ingroup SpecialPage
  31  */
  32 class ImportStreamSource implements ImportSource {
  33         function __construct( $handle ) {
  34                 $this->mHandle = $handle;
  35         }
  36
  37         /**
  38          * @return bool
  39          */
  40         function atEnd() {
  41                 return feof( $this->mHandle );
  42         }
  43
  44         /**
  45          * @return string
  46          */
  47         function readChunk() {
  48                 return fread( $this->mHandle, 32768 );
  49         }
  50
  51         /**
  52          * @param string $filename
  53          * @return Status
  54          */
  55         static function newFromFile( $filename ) {
  56                 MediaWiki\suppressWarnings();
  57                 $file = fopen( $filename, 'rt' );
  58                 MediaWiki\restoreWarnings();
  59                 if ( !$file ) {
  60                         return Status::newFatal( "importcantopen" );
  61                 }
  62                 return Status::newGood( new ImportStreamSource( $file ) );
  63         }
  64
  65         /**
  66          * @param string $fieldname
  67          * @return Status
  68          */
  69         static function newFromUpload( $fieldname = "xmlimport" ) {
  70                 $upload =& $_FILES[$fieldname];
  71
  72                 if ( $upload === null || !$upload['name'] ) {
  73                         return Status::newFatal( 'importnofile' );
  74                 }
  75                 if ( !empty( $upload['error'] ) ) {
  76                         switch ( $upload['error'] ) {
  77                                 case 1:
  78                                         # The uploaded file exceeds the upload_max_filesize directive in php.ini.
  79                                         return Status::newFatal( 'importuploaderrorsize' );
  80                                 case 2:
  81                                         # The uploaded file exceeds the MAX_FILE_SIZE directive that
  82                                         # was specified in the HTML form.
  83                                         return Status::newFatal( 'importuploaderrorsize' );
  84                                 case 3:
  85                                         # The uploaded file was only partially uploaded
  86                                         return Status::newFatal( 'importuploaderrorpartial' );
  87                                 case 6:
  88                                         # Missing a temporary folder.
  89                                         return Status::newFatal( 'importuploaderrortemp' );
  90                                 # case else: # Currently impossible
  91                         }
  92
  93                 }
  94                 $fname = $upload['tmp_name'];
  95                 if ( is_uploaded_file( $fname ) ) {
  96                         return ImportStreamSource::newFromFile( $fname );
  97                 } else {
  98                         return Status::newFatal( 'importnofile' );
  99                 }
 100         }
 101
 102         /**
 103          * @param string $url
 104          * @param string $method
 105          * @return Status
 106          */
 107         static function newFromURL( $url, $method = 'GET' ) {
 108                 global $wgHTTPImportTimeout;
 109                 wfDebug( __METHOD__ . ": opening $url\n" );
 110                 # Use the standard HTTP fetch function; it times out
 111                 # quicker and sorts out user-agent problems which might
 112                 # otherwise prevent importing from large sites, such
 113                 # as the Wikimedia cluster, etc.
 114                 $data = Http::request(
 115                         $method,
 116                         $url,
 117                         [
 118                                 'followRedirects' => true,
 119                                 'timeout' => $wgHTTPImportTimeout
 120                         ],
 121                         __METHOD__
 122                 );
 123                 if ( $data !== false ) {
 124                         $file = tmpfile();
 125                         fwrite( $file, $data );
 126                         fflush( $file );
 127                         fseek( $file, 0 );
 128                         return Status::newGood( new ImportStreamSource( $file ) );
 129                 } else {
 130                         return Status::newFatal( 'importcantopen' );
 131                 }
 132         }
 133
 134         /**
 135          * @param string $interwiki
 136          * @param string $page
 137          * @param bool $history
 138          * @param bool $templates
 139          * @param int $pageLinkDepth
 140          * @return Status
 141          */
 142         public static function newFromInterwiki( $interwiki, $page, $history = false,
 143                 $templates = false, $pageLinkDepth = 0
 144         ) {
 145                 if ( $page == '' ) {
 146                         return Status::newFatal( 'import-noarticle' );
 147                 }
 148
 149                 # Look up the first interwiki prefix, and let the foreign site handle
 150                 # subsequent interwiki prefixes
 151                 $firstIwPrefix = strtok( $interwiki, ':' );
 152                 $interwikiLookup = MediaWikiServices::getInstance()->getInterwikiLookup();
 153                 $firstIw = $interwikiLookup->fetch( $firstIwPrefix );
 154                 if ( !$firstIw ) {
 155                         return Status::newFatal( 'importbadinterwiki' );
 156                 }
 157
 158                 $additionalIwPrefixes = strtok( '' );
 159                 if ( $additionalIwPrefixes ) {
 160                         $additionalIwPrefixes .= ':';
 161                 }
 162                 # Have to do a DB-key replacement ourselves; otherwise spaces get
 163                 # URL-encoded to +, which is wrong in this case. Similar to logic in
 164                 # Title::getLocalURL
 165                 $link = $firstIw->getURL( strtr( "${additionalIwPrefixes}Special:Export/$page",
 166                         ' ', '_' ) );
 167
 168                 $params = [];
 169                 if ( $history ) {
 170                         $params['history'] = 1;
 171                 }
 172                 if ( $templates ) {
 173                         $params['templates'] = 1;
 174                 }
 175                 if ( $pageLinkDepth ) {
 176                         $params['pagelink-depth'] = $pageLinkDepth;
 177                 }
 178
 179                 $url = wfAppendQuery( $link, $params );
 180                 # For interwikis, use POST to avoid redirects.
 181                 return ImportStreamSource::newFromURL( $url, "POST" );
 182         }
 183 }