includes/libs/filebackend/HTTPFileStreamer.php

   1 <?php
   2 /**
   3  * Functions related to the output of file content.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  */
  22
  23 /**
  24  * Functions related to the output of file content
  25  *
  26  * @since 1.28
  27  */
  28 class HTTPFileStreamer {
  29         /** @var string */
  30         protected $path;
  31         /** @var callable */
  32         protected $obResetFunc;
  33         /** @var callable */
  34         protected $streamMimeFunc;
  35
  36         // Do not send any HTTP headers unless requested by caller (e.g. body only)
  37         const STREAM_HEADLESS = 1;
  38         // Do not try to tear down any PHP output buffers
  39         const STREAM_ALLOW_OB = 2;
  40
  41         /**
  42          * @param string $path Local filesystem path to a file
  43          * @param array $params Options map, which includes:
  44          *   - obResetFunc : alternative callback to clear the output buffer
  45          *   - streamMimeFunc : alternative method to determine the content type from the path
  46          */
  47         public function __construct( $path, array $params = [] ) {
  48                 $this->path = $path;
  49                 $this->obResetFunc = isset( $params['obResetFunc'] )
  50                         ? $params['obResetFunc']
  51                         : [ __CLASS__, 'resetOutputBuffers' ];
  52                 $this->streamMimeFunc = isset( $params['streamMimeFunc'] )
  53                         ? $params['streamMimeFunc']
  54                         : [ __CLASS__, 'contentTypeFromPath' ];
  55         }
  56
  57         /**
  58          * Stream a file to the browser, adding all the headings and fun stuff.
  59          * Headers sent include: Content-type, Content-Length, Last-Modified,
  60          * and Content-Disposition.
  61          *
  62          * @param array $headers Any additional headers to send if the file exists
  63          * @param bool $sendErrors Send error messages if errors occur (like 404)
  64          * @param array $optHeaders HTTP request header map (e.g. "range") (use lowercase keys)
  65          * @param integer $flags Bitfield of STREAM_* constants
  66          * @throws MWException
  67          * @return bool Success
  68          */
  69         public function stream(
  70                 $headers = [], $sendErrors = true, $optHeaders = [], $flags = 0
  71         ) {
  72                 // Don't stream it out as text/html if there was a PHP error
  73                 if ( ( ( $flags & self::STREAM_HEADLESS ) == 0 || $headers ) && headers_sent() ) {
  74                         echo "Headers already sent, terminating.\n";
  75                         return false;
  76                 }
  77
  78                 $headerFunc = ( $flags & self::STREAM_HEADLESS )
  79                         ? function ( $header ) {
  80                                 // no-op
  81                         }
  82                         : function ( $header ) {
  83                                 is_int( $header ) ? HttpStatus::header( $header ) : header( $header );
  84                         };
  85
  86                 MediaWiki\suppressWarnings();
  87                 $info = stat( $this->path );
  88                 MediaWiki\restoreWarnings();
  89
  90                 if ( !is_array( $info ) ) {
  91                         if ( $sendErrors ) {
  92                                 self::send404Message( $this->path, $flags );
  93                         }
  94                         return false;
  95                 }
  96
  97                 // Send Last-Modified HTTP header for client-side caching
  98                 $mtimeCT = new ConvertibleTimestamp( $info['mtime'] );
  99                 $headerFunc( 'Last-Modified: ' . $mtimeCT->getTimestamp( TS_RFC2822 ) );
 100
 101                 if ( ( $flags & self::STREAM_ALLOW_OB ) == 0 ) {
 102                         call_user_func( $this->obResetFunc );
 103                 }
 104
 105                 $type = call_user_func( $this->streamMimeFunc, $this->path );
 106                 if ( $type && $type != 'unknown/unknown' ) {
 107                         $headerFunc( "Content-type: $type" );
 108                 } else {
 109                         // Send a content type which is not known to Internet Explorer, to
 110                         // avoid triggering IE's content type detection. Sending a standard
 111                         // unknown content type here essentially gives IE license to apply
 112                         // whatever content type it likes.
 113                         $headerFunc( 'Content-type: application/x-wiki' );
 114                 }
 115
 116                 // Don't send if client has up to date cache
 117                 if ( isset( $optHeaders['if-modified-since'] ) ) {
 118                         $modsince = preg_replace( '/;.*$/', '', $optHeaders['if-modified-since'] );
 119                         if ( $mtimeCT->getTimestamp( TS_UNIX ) <= strtotime( $modsince ) ) {
 120                                 ini_set( 'zlib.output_compression', 0 );
 121                                 $headerFunc( 304 );
 122                                 return true; // ok
 123                         }
 124                 }
 125
 126                 // Send additional headers
 127                 foreach ( $headers as $header ) {
 128                         header( $header ); // always use header(); specifically requested
 129                 }
 130
 131                 if ( isset( $optHeaders['range'] ) ) {
 132                         $range = self::parseRange( $optHeaders['range'], $info['size'] );
 133                         if ( is_array( $range ) ) {
 134                                 $headerFunc( 206 );
 135                                 $headerFunc( 'Content-Length: ' . $range[2] );
 136                                 $headerFunc( "Content-Range: bytes {$range[0]}-{$range[1]}/{$info['size']}" );
 137                         } elseif ( $range === 'invalid' ) {
 138                                 if ( $sendErrors ) {
 139                                         $headerFunc( 416 );
 140                                         $headerFunc( 'Cache-Control: no-cache' );
 141                                         $headerFunc( 'Content-Type: text/html; charset=utf-8' );
 142                                         $headerFunc( 'Content-Range: bytes */' . $info['size'] );
 143                                 }
 144                                 return false;
 145                         } else { // unsupported Range request (e.g. multiple ranges)
 146                                 $range = null;
 147                                 $headerFunc( 'Content-Length: ' . $info['size'] );
 148                         }
 149                 } else {
 150                         $range = null;
 151                         $headerFunc( 'Content-Length: ' . $info['size'] );
 152                 }
 153
 154                 if ( is_array( $range ) ) {
 155                         $handle = fopen( $this->path, 'rb' );
 156                         if ( $handle ) {
 157                                 $ok = true;
 158                                 fseek( $handle, $range[0] );
 159                                 $remaining = $range[2];
 160                                 while ( $remaining > 0 && $ok ) {
 161                                         $bytes = min( $remaining, 8 * 1024 );
 162                                         $data = fread( $handle, $bytes );
 163                                         $remaining -= $bytes;
 164                                         $ok = ( $data !== false );
 165                                         print $data;
 166                                 }
 167                         } else {
 168                                 return false;
 169                         }
 170                 } else {
 171                         return readfile( $this->path ) !== false; // faster
 172                 }
 173
 174                 return true;
 175         }
 176
 177         /**
 178          * Send out a standard 404 message for a file
 179          *
 180          * @param string $fname Full name and path of the file to stream
 181          * @param integer $flags Bitfield of STREAM_* constants
 182          * @since 1.24
 183          */
 184         public static function send404Message( $fname, $flags = 0 ) {
 185                 if ( ( $flags & self::STREAM_HEADLESS ) == 0 ) {
 186                         HttpStatus::header( 404 );
 187                         header( 'Cache-Control: no-cache' );
 188                         header( 'Content-Type: text/html; charset=utf-8' );
 189                 }
 190                 $encFile = htmlspecialchars( $fname );
 191                 $encScript = htmlspecialchars( $_SERVER['SCRIPT_NAME'] );
 192                 echo "<!DOCTYPE html><html><body>
 193                         <h1>File not found</h1>
 194                         <p>Although this PHP script ($encScript) exists, the file requested for output
 195                         ($encFile) does not.</p>
 196                         </body></html>
 197                         ";
 198         }
 199
 200         /**
 201          * Convert a Range header value to an absolute (start, end) range tuple
 202          *
 203          * @param string $range Range header value
 204          * @param integer $size File size
 205          * @return array|string Returns error string on failure (start, end, length)
 206          * @since 1.24
 207          */
 208         public static function parseRange( $range, $size ) {
 209                 $m = [];
 210                 if ( preg_match( '#^bytes=(\d*)-(\d*)$#', $range, $m ) ) {
 211                         list( , $start, $end ) = $m;
 212                         if ( $start === '' && $end === '' ) {
 213                                 $absRange = [ 0, $size - 1 ];
 214                         } elseif ( $start === '' ) {
 215                                 $absRange = [ $size - $end, $size - 1 ];
 216                         } elseif ( $end === '' ) {
 217                                 $absRange = [ $start, $size - 1 ];
 218                         } else {
 219                                 $absRange = [ $start, $end ];
 220                         }
 221                         if ( $absRange[0] >= 0 && $absRange[1] >= $absRange[0] ) {
 222                                 if ( $absRange[0] < $size ) {
 223                                         $absRange[1] = min( $absRange[1], $size - 1 ); // stop at EOF
 224                                         $absRange[2] = $absRange[1] - $absRange[0] + 1;
 225                                         return $absRange;
 226                                 } elseif ( $absRange[0] == 0 && $size == 0 ) {
 227                                         return 'unrecognized'; // the whole file should just be sent
 228                                 }
 229                         }
 230                         return 'invalid';
 231                 }
 232                 return 'unrecognized';
 233         }
 234
 235         protected static function resetOutputBuffers() {
 236                 while ( ob_get_status() ) {
 237                         if ( !ob_end_clean() ) {
 238                                 // Could not remove output buffer handler; abort now
 239                                 // to avoid getting in some kind of infinite loop.
 240                                 break;
 241                         }
 242                 }
 243         }
 244
 245         /**
 246          * Determine the file type of a file based on the path
 247          *
 248          * @param string $filename Storage path or file system path
 249          * @return null|string
 250          */
 251         protected static function contentTypeFromPath( $filename ) {
 252                 $ext = strrchr( $filename, '.' );
 253                 $ext = $ext === false ? '' : strtolower( substr( $ext, 1 ) );
 254
 255                 switch ( $ext ) {
 256                         case 'gif':
 257                                 return 'image/gif';
 258                         case 'png':
 259                                 return 'image/png';
 260                         case 'jpg':
 261                                 return 'image/jpeg';
 262                         case 'jpeg':
 263                                 return 'image/jpeg';
 264                 }
 265
 266                 return 'unknown/unknown';
 267         }
 268 }