includes/import/ImportStreamSource.php

   1 <?php
   2 /**
   3  * MediaWiki page data importer.
   4  *
   5  * Copyright © 2003,2005 Brooke Vibber <bvibber@wikimedia.org>
   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
  27 use MediaWiki\MainConfigNames;
  28 use MediaWiki\MediaWikiServices;
  29 use MediaWiki\Status\Status;
  30 use Wikimedia\AtEase\AtEase;
  31
  32 /**
  33  * Imports a XML dump from a file (either from file upload, files on disk, or HTTP)
  34  * @ingroup SpecialPage
  35  */
  36 class ImportStreamSource implements ImportSource {
  37         /** @var resource */
  38         private $mHandle;
  39
  40         /**
  41          * @param resource $handle
  42          */
  43         public function __construct( $handle ) {
  44                 $this->mHandle = $handle;
  45         }
  46
  47         /**
  48          * @return bool
  49          */
  50         public function atEnd() {
  51                 return feof( $this->mHandle );
  52         }
  53
  54         /**
  55          * @return string
  56          */
  57         public function readChunk() {
  58                 return fread( $this->mHandle, 32768 );
  59         }
  60
  61         /**
  62          * @return bool
  63          */
  64         public function isSeekable() {
  65                 return stream_get_meta_data( $this->mHandle )['seekable'] ?? false;
  66         }
  67
  68         /**
  69          * @param int $offset
  70          * @return int
  71          */
  72         public function seek( int $offset ) {
  73                 return fseek( $this->mHandle, $offset );
  74         }
  75
  76         /**
  77          * @param string $filename
  78          * @return Status
  79          */
  80         public static function newFromFile( $filename ) {
  81                 AtEase::suppressWarnings();
  82                 $file = fopen( $filename, 'rt' );
  83                 AtEase::restoreWarnings();
  84                 if ( !$file ) {
  85                         return Status::newFatal( "importcantopen" );
  86                 }
  87                 return Status::newGood( new ImportStreamSource( $file ) );
  88         }
  89
  90         /**
  91          * @param string $fieldname
  92          * @return Status
  93          */
  94         public static function newFromUpload( $fieldname = "xmlimport" ) {
  95                 // phpcs:ignore MediaWiki.Usage.SuperGlobalsUsage.SuperGlobals
  96                 $upload =& $_FILES[$fieldname];
  97
  98                 if ( $upload === null || !$upload['name'] ) {
  99                         return Status::newFatal( 'importnofile' );
 100                 }
 101                 if ( !empty( $upload['error'] ) ) {
 102                         switch ( $upload['error'] ) {
 103                                 case UPLOAD_ERR_INI_SIZE:
 104                                         // The uploaded file exceeds the upload_max_filesize directive in php.ini.
 105                                         return Status::newFatal( 'importuploaderrorsize' );
 106                                 case UPLOAD_ERR_FORM_SIZE:
 107                                         // The uploaded file exceeds the MAX_FILE_SIZE directive that
 108                                         // was specified in the HTML form.
 109                                         // FIXME This is probably never used since that directive was removed in 8e91c520?
 110                                         return Status::newFatal( 'importuploaderrorsize' );
 111                                 case UPLOAD_ERR_PARTIAL:
 112                                         // The uploaded file was only partially uploaded
 113                                         return Status::newFatal( 'importuploaderrorpartial' );
 114                                 case UPLOAD_ERR_NO_TMP_DIR:
 115                                         // Missing a temporary folder.
 116                                         return Status::newFatal( 'importuploaderrortemp' );
 117                                 // Other error codes get the generic 'importnofile' error message below
 118                         }
 119
 120                 }
 121                 $fname = $upload['tmp_name'];
 122                 if ( is_uploaded_file( $fname ) ) {
 123                         return self::newFromFile( $fname );
 124                 } else {
 125                         return Status::newFatal( 'importnofile' );
 126                 }
 127         }
 128
 129         /**
 130          * @param string $url
 131          * @param string $method
 132          * @return Status
 133          */
 134         public static function newFromURL( $url, $method = 'GET' ) {
 135                 $httpImportTimeout = MediaWikiServices::getInstance()->getMainConfig()->get(
 136                         MainConfigNames::HTTPImportTimeout );
 137                 wfDebug( __METHOD__ . ": opening $url" );
 138                 # Use the standard HTTP fetch function; it times out
 139                 # quicker and sorts out user-agent problems which might
 140                 # otherwise prevent importing from large sites, such
 141                 # as the Wikimedia cluster, etc.
 142                 $data = MediaWikiServices::getInstance()->getHttpRequestFactory()->request(
 143                         $method,
 144                         $url,
 145                         [
 146                                 'followRedirects' => true,
 147                                 'timeout' => $httpImportTimeout
 148                         ],
 149                         __METHOD__
 150                 );
 151                 if ( $data !== null ) {
 152                         $file = tmpfile();
 153                         fwrite( $file, $data );
 154                         fflush( $file );
 155                         fseek( $file, 0 );
 156                         return Status::newGood( new ImportStreamSource( $file ) );
 157                 } else {
 158                         return Status::newFatal( 'importcantopen' );
 159                 }
 160         }
 161
 162         /**
 163          * @param string $interwiki
 164          * @param string $page
 165          * @param bool $history
 166          * @param bool $templates
 167          * @param int $pageLinkDepth
 168          * @return Status
 169          */
 170         public static function newFromInterwiki( $interwiki, $page, $history = false,
 171                 $templates = false, $pageLinkDepth = 0
 172         ) {
 173                 if ( $page == '' ) {
 174                         return Status::newFatal( 'import-noarticle' );
 175                 }
 176
 177                 # Look up the first interwiki prefix, and let the foreign site handle
 178                 # subsequent interwiki prefixes
 179                 $firstIwPrefix = strtok( $interwiki, ':' );
 180                 $interwikiLookup = MediaWikiServices::getInstance()->getInterwikiLookup();
 181                 $firstIw = $interwikiLookup->fetch( $firstIwPrefix );
 182                 if ( !$firstIw ) {
 183                         return Status::newFatal( 'importbadinterwiki' );
 184                 }
 185
 186                 $additionalIwPrefixes = strtok( '' );
 187                 if ( $additionalIwPrefixes ) {
 188                         $additionalIwPrefixes .= ':';
 189                 }
 190                 # Have to do a DB-key replacement ourselves; otherwise spaces get
 191                 # URL-encoded to +, which is wrong in this case. Similar to logic in
 192                 # Title::getLocalURL
 193                 $link = $firstIw->getURL( strtr( "{$additionalIwPrefixes}Special:Export/$page",
 194                         ' ', '_' ) );
 195
 196                 $params = [];
 197                 if ( $history ) {
 198                         $params['history'] = 1;
 199                 }
 200                 if ( $templates ) {
 201                         $params['templates'] = 1;
 202                 }
 203                 if ( $pageLinkDepth ) {
 204                         $params['pagelink-depth'] = $pageLinkDepth;
 205                 }
 206
 207                 $url = wfAppendQuery( $link, $params );
 208                 # For interwikis, use POST to avoid redirects.
 209                 return self::newFromURL( $url, "POST" );
 210         }
 211 }