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         /**
  34          * @param resource $handle
  35          */
  36         function __construct( $handle ) {
  37                 $this->mHandle = $handle;
  38         }
  39
  40         /**
  41          * @return bool
  42          */
  43         function atEnd() {
  44                 return feof( $this->mHandle );
  45         }
  46
  47         /**
  48          * @return string
  49          */
  50         function readChunk() {
  51                 return fread( $this->mHandle, 32768 );
  52         }
  53
  54         /**
  55          * @param string $filename
  56          * @return Status
  57          */
  58         static function newFromFile( $filename ) {
  59                 Wikimedia\suppressWarnings();
  60                 $file = fopen( $filename, 'rt' );
  61                 Wikimedia\restoreWarnings();
  62                 if ( !$file ) {
  63                         return Status::newFatal( "importcantopen" );
  64                 }
  65                 return Status::newGood( new ImportStreamSource( $file ) );
  66         }
  67
  68         /**
  69          * @param string $fieldname
  70          * @return Status
  71          */
  72         static function newFromUpload( $fieldname = "xmlimport" ) {
  73                 $upload =& $_FILES[$fieldname];
  74
  75                 if ( $upload === null || !$upload['name'] ) {
  76                         return Status::newFatal( 'importnofile' );
  77                 }
  78                 if ( !empty( $upload['error'] ) ) {
  79                         switch ( $upload['error'] ) {
  80                                 case UPLOAD_ERR_INI_SIZE:
  81                                         // The uploaded file exceeds the upload_max_filesize directive in php.ini.
  82                                         return Status::newFatal( 'importuploaderrorsize' );
  83                                 case UPLOAD_ERR_FORM_SIZE:
  84                                         // The uploaded file exceeds the MAX_FILE_SIZE directive that
  85                                         // was specified in the HTML form.
  86                                         // FIXME This is probably never used since that directive was removed in 8e91c520?
  87                                         return Status::newFatal( 'importuploaderrorsize' );
  88                                 case UPLOAD_ERR_PARTIAL:
  89                                         // The uploaded file was only partially uploaded
  90                                         return Status::newFatal( 'importuploaderrorpartial' );
  91                                 case UPLOAD_ERR_NO_TMP_DIR:
  92                                         // Missing a temporary folder.
  93                                         return Status::newFatal( 'importuploaderrortemp' );
  94                                 // Other error codes get the generic 'importnofile' error message below
  95                         }
  96
  97                 }
  98                 $fname = $upload['tmp_name'];
  99                 if ( is_uploaded_file( $fname ) ) {
 100                         return self::newFromFile( $fname );
 101                 } else {
 102                         return Status::newFatal( 'importnofile' );
 103                 }
 104         }
 105
 106         /**
 107          * @param string $url
 108          * @param string $method
 109          * @return Status
 110          */
 111         static function newFromURL( $url, $method = 'GET' ) {
 112                 global $wgHTTPImportTimeout;
 113                 wfDebug( __METHOD__ . ": opening $url\n" );
 114                 # Use the standard HTTP fetch function; it times out
 115                 # quicker and sorts out user-agent problems which might
 116                 # otherwise prevent importing from large sites, such
 117                 # as the Wikimedia cluster, etc.
 118                 $data = MediaWikiServices::getInstance()->getHttpRequestFactory()->request(
 119                         $method,
 120                         $url,
 121                         [
 122                                 'followRedirects' => true,
 123                                 'timeout' => $wgHTTPImportTimeout
 124                         ],
 125                         __METHOD__
 126                 );
 127                 if ( $data !== false ) {
 128                         $file = tmpfile();
 129                         fwrite( $file, $data );
 130                         fflush( $file );
 131                         fseek( $file, 0 );
 132                         return Status::newGood( new ImportStreamSource( $file ) );
 133                 } else {
 134                         return Status::newFatal( 'importcantopen' );
 135                 }
 136         }
 137
 138         /**
 139          * @param string $interwiki
 140          * @param string $page
 141          * @param bool $history
 142          * @param bool $templates
 143          * @param int $pageLinkDepth
 144          * @return Status
 145          */
 146         public static function newFromInterwiki( $interwiki, $page, $history = false,
 147                 $templates = false, $pageLinkDepth = 0
 148         ) {
 149                 if ( $page == '' ) {
 150                         return Status::newFatal( 'import-noarticle' );
 151                 }
 152
 153                 # Look up the first interwiki prefix, and let the foreign site handle
 154                 # subsequent interwiki prefixes
 155                 $firstIwPrefix = strtok( $interwiki, ':' );
 156                 $interwikiLookup = MediaWikiServices::getInstance()->getInterwikiLookup();
 157                 $firstIw = $interwikiLookup->fetch( $firstIwPrefix );
 158                 if ( !$firstIw ) {
 159                         return Status::newFatal( 'importbadinterwiki' );
 160                 }
 161
 162                 $additionalIwPrefixes = strtok( '' );
 163                 if ( $additionalIwPrefixes ) {
 164                         $additionalIwPrefixes .= ':';
 165                 }
 166                 # Have to do a DB-key replacement ourselves; otherwise spaces get
 167                 # URL-encoded to +, which is wrong in this case. Similar to logic in
 168                 # Title::getLocalURL
 169                 $link = $firstIw->getURL( strtr( "${additionalIwPrefixes}Special:Export/$page",
 170                         ' ', '_' ) );
 171
 172                 $params = [];
 173                 if ( $history ) {
 174                         $params['history'] = 1;
 175                 }
 176                 if ( $templates ) {
 177                         $params['templates'] = 1;
 178                 }
 179                 if ( $pageLinkDepth ) {
 180                         $params['pagelink-depth'] = $pageLinkDepth;
 181                 }
 182
 183                 $url = wfAppendQuery( $link, $params );
 184                 # For interwikis, use POST to avoid redirects.
 185                 return self::newFromURL( $url, "POST" );
 186         }
 187 }