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         /** @var resource */
  34         private $mHandle;
  35
  36         /**
  37          * @param resource $handle
  38          */
  39         function __construct( $handle ) {
  40                 $this->mHandle = $handle;
  41         }
  42
  43         /**
  44          * @return bool
  45          */
  46         function atEnd() {
  47                 return feof( $this->mHandle );
  48         }
  49
  50         /**
  51          * @return string
  52          */
  53         function readChunk() {
  54                 return fread( $this->mHandle, 32768 );
  55         }
  56
  57         /**
  58          * @param string $filename
  59          * @return Status
  60          */
  61         static function newFromFile( $filename ) {
  62                 Wikimedia\suppressWarnings();
  63                 $file = fopen( $filename, 'rt' );
  64                 Wikimedia\restoreWarnings();
  65                 if ( !$file ) {
  66                         return Status::newFatal( "importcantopen" );
  67                 }
  68                 return Status::newGood( new ImportStreamSource( $file ) );
  69         }
  70
  71         /**
  72          * @param string $fieldname
  73          * @return Status
  74          */
  75         static function newFromUpload( $fieldname = "xmlimport" ) {
  76                 $upload =& $_FILES[$fieldname];
  77
  78                 if ( $upload === null || !$upload['name'] ) {
  79                         return Status::newFatal( 'importnofile' );
  80                 }
  81                 if ( !empty( $upload['error'] ) ) {
  82                         switch ( $upload['error'] ) {
  83                                 case UPLOAD_ERR_INI_SIZE:
  84                                         // The uploaded file exceeds the upload_max_filesize directive in php.ini.
  85                                         return Status::newFatal( 'importuploaderrorsize' );
  86                                 case UPLOAD_ERR_FORM_SIZE:
  87                                         // The uploaded file exceeds the MAX_FILE_SIZE directive that
  88                                         // was specified in the HTML form.
  89                                         // FIXME This is probably never used since that directive was removed in 8e91c520?
  90                                         return Status::newFatal( 'importuploaderrorsize' );
  91                                 case UPLOAD_ERR_PARTIAL:
  92                                         // The uploaded file was only partially uploaded
  93                                         return Status::newFatal( 'importuploaderrorpartial' );
  94                                 case UPLOAD_ERR_NO_TMP_DIR:
  95                                         // Missing a temporary folder.
  96                                         return Status::newFatal( 'importuploaderrortemp' );
  97                                 // Other error codes get the generic 'importnofile' error message below
  98                         }
  99
 100                 }
 101                 $fname = $upload['tmp_name'];
 102                 if ( is_uploaded_file( $fname ) ) {
 103                         return self::newFromFile( $fname );
 104                 } else {
 105                         return Status::newFatal( 'importnofile' );
 106                 }
 107         }
 108
 109         /**
 110          * @param string $url
 111          * @param string $method
 112          * @return Status
 113          */
 114         static function newFromURL( $url, $method = 'GET' ) {
 115                 global $wgHTTPImportTimeout;
 116                 wfDebug( __METHOD__ . ": opening $url\n" );
 117                 # Use the standard HTTP fetch function; it times out
 118                 # quicker and sorts out user-agent problems which might
 119                 # otherwise prevent importing from large sites, such
 120                 # as the Wikimedia cluster, etc.
 121                 $data = MediaWikiServices::getInstance()->getHttpRequestFactory()->request(
 122                         $method,
 123                         $url,
 124                         [
 125                                 'followRedirects' => true,
 126                                 'timeout' => $wgHTTPImportTimeout
 127                         ],
 128                         __METHOD__
 129                 );
 130                 if ( $data !== false ) {
 131                         $file = tmpfile();
 132                         fwrite( $file, $data );
 133                         fflush( $file );
 134                         fseek( $file, 0 );
 135                         return Status::newGood( new ImportStreamSource( $file ) );
 136                 } else {
 137                         return Status::newFatal( 'importcantopen' );
 138                 }
 139         }
 140
 141         /**
 142          * @param string $interwiki
 143          * @param string $page
 144          * @param bool $history
 145          * @param bool $templates
 146          * @param int $pageLinkDepth
 147          * @return Status
 148          */
 149         public static function newFromInterwiki( $interwiki, $page, $history = false,
 150                 $templates = false, $pageLinkDepth = 0
 151         ) {
 152                 if ( $page == '' ) {
 153                         return Status::newFatal( 'import-noarticle' );
 154                 }
 155
 156                 # Look up the first interwiki prefix, and let the foreign site handle
 157                 # subsequent interwiki prefixes
 158                 $firstIwPrefix = strtok( $interwiki, ':' );
 159                 $interwikiLookup = MediaWikiServices::getInstance()->getInterwikiLookup();
 160                 $firstIw = $interwikiLookup->fetch( $firstIwPrefix );
 161                 if ( !$firstIw ) {
 162                         return Status::newFatal( 'importbadinterwiki' );
 163                 }
 164
 165                 $additionalIwPrefixes = strtok( '' );
 166                 if ( $additionalIwPrefixes ) {
 167                         $additionalIwPrefixes .= ':';
 168                 }
 169                 # Have to do a DB-key replacement ourselves; otherwise spaces get
 170                 # URL-encoded to +, which is wrong in this case. Similar to logic in
 171                 # Title::getLocalURL
 172                 $link = $firstIw->getURL( strtr( "${additionalIwPrefixes}Special:Export/$page",
 173                         ' ', '_' ) );
 174
 175                 $params = [];
 176                 if ( $history ) {
 177                         $params['history'] = 1;
 178                 }
 179                 if ( $templates ) {
 180                         $params['templates'] = 1;
 181                 }
 182                 if ( $pageLinkDepth ) {
 183                         $params['pagelink-depth'] = $pageLinkDepth;
 184                 }
 185
 186                 $url = wfAppendQuery( $link, $params );
 187                 # For interwikis, use POST to avoid redirects.
 188                 return self::newFromURL( $url, "POST" );
 189         }
 190 }