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