includes/MacBinary.php

   1 <?php
   2 /**
   3  * MacBinary signature checker and data fork extractor, for files
   4  * uploaded from Internet Explorer for Mac.
   5  *
   6  * Copyright (C) 2005 Brion Vibber <brion@pobox.com>
   7  * Portions based on Convert::BinHex by Eryq et al
   8  * http://www.mediawiki.org/
   9  *
  10  * This program is free software; you can redistribute it and/or modify
  11  * it under the terms of the GNU General Public License as published by
  12  * the Free Software Foundation; either version 2 of the License, or
  13  * (at your option) any later version.
  14  *
  15  * This program is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  18  * GNU General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU General Public License along
  21  * with this program; if not, write to the Free Software Foundation, Inc.,
  22  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  23  * http://www.gnu.org/copyleft/gpl.html
  24  *
  25  * @package MediaWiki
  26  * @subpackage SpecialPage
  27  */
  28
  29 class MacBinary {
  30         function MacBinary( $filename ) {
  31                 $this->open( $filename );
  32                 $this->loadHeader();
  33         }
  34
  35         /**
  36          * The file must be seekable, such as local filesystem.
  37          * Remote URLs probably won't work.
  38          *
  39          * @param string $filename
  40          */
  41         function open( $filename ) {
  42                 $this->valid = false;
  43                 $this->version = 0;
  44                 $this->filename = '';
  45                 $this->dataLength = 0;
  46                 $this->resourceLength = 0;
  47                 $this->handle = fopen( $filename, 'rb' );
  48         }
  49
  50         /**
  51          * Does this appear to be a valid MacBinary archive?
  52          * @return bool
  53          */
  54         function isValid() {
  55                 return $this->valid;
  56         }
  57
  58         /**
  59          * Get length of data fork
  60          * @return int
  61          */
  62         function dataForkLength() {
  63                 return $this->dataLength;
  64         }
  65
  66         /**
  67          * Copy the data fork to an external file or resource.
  68          * @param resource $destination
  69          * @return bool
  70          */
  71         function extractData( $destination ) {
  72                 if( !$this->isValid() ) {
  73                         return false;
  74                 }
  75
  76                 // Data fork appears immediately after header
  77                 fseek( $this->handle, 128 );
  78                 return $this->copyBytesTo( $destination, $this->dataLength );
  79         }
  80
  81         /**
  82          *
  83          */
  84         function close() {
  85                 fclose( $this->handle );
  86         }
  87
  88         // --------------------------------------------------------------
  89
  90         /**
  91          * Check if the given file appears to be MacBinary-encoded,
  92          * as Internet Explorer on Mac OS may provide for unknown types.
  93          * http://www.lazerware.com/formats/macbinary/macbinary_iii.html
  94          * If ok, load header data.
  95          *
  96          * @return bool
  97          * @access private
  98          */
  99         function loadHeader() {
 100                 $fname = 'MacBinary::loadHeader';
 101
 102                 fseek( $this->handle, 0 );
 103                 $head = fread( $this->handle, 128 );
 104                 $this->hexdump( $head );
 105
 106                 if( strlen( $head ) < 128 ) {
 107                         wfDebug( "$fname: couldn't read full MacBinary header\n" );
 108                         return false;
 109                 }
 110
 111                 if( $head{0} != "\x00" || $head{74} != "\x00" ) {
 112                         wfDebug( "$fname: header bytes 0 and 74 not null\n" );
 113                         return false;
 114                 }
 115
 116                 $signature = substr( $head, 102, 4 );
 117                 $a = unpack( "ncrc", substr( $head, 124, 2 ) );
 118                 $storedCRC = $a['crc'];
 119                 $calculatedCRC = $this->calcCRC( substr( $head, 0, 124 ) );
 120                 if( $storedCRC == $calculatedCRC ) {
 121                         if( $signature == 'mBIN' ) {
 122                                 $this->version = 3;
 123                         } else {
 124                                 $this->version = 2;
 125                         }
 126                 } else {
 127                         $crc = sprintf( "%x != %x", $storedCRC, $calculatedCRC );
 128                         if( $storedCRC == 0 && $head{82} == "\x00" &&
 129                                 substr( $head, 101, 24 ) == str_repeat( "\x00", 24 ) ) {
 130                                 wfDebug( "$fname: no CRC, looks like MacBinary I\n" );
 131                                 $this->version = 1;
 132                         } elseif( $signature == 'mBIN' && $storedCRC == 0x185 ) {
 133                                 // Mac IE 5.0 seems to insert this value in the CRC field.
 134                                 // 5.2.3 works correctly; don't know about other versions.
 135                                 wfDebug( "$fname: CRC doesn't match ($crc), looks like Mac IE 5.0\n" );
 136                                 $this->version = 3;
 137                         } else {
 138                                 wfDebug( "$fname: CRC doesn't match ($crc) and not MacBinary I\n" );
 139                                 return false;
 140                         }
 141                 }
 142
 143                 $nameLength = ord( $head{1} );
 144                 if( $nameLength < 1 || $nameLength > 63 ) {
 145                         wfDebug( "$fname: invalid filename size $nameLength\n" );
 146                         return false;
 147                 }
 148                 $this->filename = substr( $head, 2, $nameLength );
 149
 150                 $forks = unpack( "Ndata/Nresource", substr( $head, 83, 8 ) );
 151                 $this->dataLength = $forks['data'];
 152                 $this->resourceLength = $forks['resource'];
 153                 $maxForkLength = 0x7fffff;
 154
 155                 if( $this->dataLength < 0 || $this->dataLength > $maxForkLength ) {
 156                         wfDebug( "$fname: invalid data fork length $this->dataLength\n" );
 157                         return false;
 158                 }
 159
 160                 if( $this->resourceLength < 0 || $this->resourceLength > $maxForkLength ) {
 161                         wfDebug( "$fname: invalid resource fork size $this->resourceLength\n" );
 162                         return false;
 163                 }
 164
 165                 wfDebug( "$fname: appears to be MacBinary $this->version, data length $this->dataLength\n" );
 166                 $this->valid = true;
 167                 return true;
 168         }
 169
 170         /**
 171          * Calculate a 16-bit CRC value as for MacBinary headers.
 172          * Adapted from perl5 Convert::BinHex by Eryq,
 173          * based on the mcvert utility (Doug Moore, April '87),
 174          * with magic array thingy by Jim Van Verth.
 175          * http://search.cpan.org/~eryq/Convert-BinHex-1.119/lib/Convert/BinHex.pm
 176          *
 177          * @param string $data
 178          * @param int $seed
 179          * @return int
 180          * @access private
 181          */
 182         function calcCRC( $data, $seed = 0 ) {
 183                 # An array useful for CRC calculations that use 0x1021 as the "seed":
 184                 $MAGIC = array(
 185                         0x0000, 0x1021, 0x2042, 0x3063, 0x4084, 0x50a5, 0x60c6, 0x70e7,
 186                         0x8108, 0x9129, 0xa14a, 0xb16b, 0xc18c, 0xd1ad, 0xe1ce, 0xf1ef,
 187                         0x1231, 0x0210, 0x3273, 0x2252, 0x52b5, 0x4294, 0x72f7, 0x62d6,
 188                         0x9339, 0x8318, 0xb37b, 0xa35a, 0xd3bd, 0xc39c, 0xf3ff, 0xe3de,
 189                         0x2462, 0x3443, 0x0420, 0x1401, 0x64e6, 0x74c7, 0x44a4, 0x5485,
 190                         0xa56a, 0xb54b, 0x8528, 0x9509, 0xe5ee, 0xf5cf, 0xc5ac, 0xd58d,
 191                         0x3653, 0x2672, 0x1611, 0x0630, 0x76d7, 0x66f6, 0x5695, 0x46b4,
 192                         0xb75b, 0xa77a, 0x9719, 0x8738, 0xf7df, 0xe7fe, 0xd79d, 0xc7bc,
 193                         0x48c4, 0x58e5, 0x6886, 0x78a7, 0x0840, 0x1861, 0x2802, 0x3823,
 194                         0xc9cc, 0xd9ed, 0xe98e, 0xf9af, 0x8948, 0x9969, 0xa90a, 0xb92b,
 195                         0x5af5, 0x4ad4, 0x7ab7, 0x6a96, 0x1a71, 0x0a50, 0x3a33, 0x2a12,
 196                         0xdbfd, 0xcbdc, 0xfbbf, 0xeb9e, 0x9b79, 0x8b58, 0xbb3b, 0xab1a,
 197                         0x6ca6, 0x7c87, 0x4ce4, 0x5cc5, 0x2c22, 0x3c03, 0x0c60, 0x1c41,
 198                         0xedae, 0xfd8f, 0xcdec, 0xddcd, 0xad2a, 0xbd0b, 0x8d68, 0x9d49,
 199                         0x7e97, 0x6eb6, 0x5ed5, 0x4ef4, 0x3e13, 0x2e32, 0x1e51, 0x0e70,
 200                         0xff9f, 0xefbe, 0xdfdd, 0xcffc, 0xbf1b, 0xaf3a, 0x9f59, 0x8f78,
 201                         0x9188, 0x81a9, 0xb1ca, 0xa1eb, 0xd10c, 0xc12d, 0xf14e, 0xe16f,
 202                         0x1080, 0x00a1, 0x30c2, 0x20e3, 0x5004, 0x4025, 0x7046, 0x6067,
 203                         0x83b9, 0x9398, 0xa3fb, 0xb3da, 0xc33d, 0xd31c, 0xe37f, 0xf35e,
 204                         0x02b1, 0x1290, 0x22f3, 0x32d2, 0x4235, 0x5214, 0x6277, 0x7256,
 205                         0xb5ea, 0xa5cb, 0x95a8, 0x8589, 0xf56e, 0xe54f, 0xd52c, 0xc50d,
 206                         0x34e2, 0x24c3, 0x14a0, 0x0481, 0x7466, 0x6447, 0x5424, 0x4405,
 207                         0xa7db, 0xb7fa, 0x8799, 0x97b8, 0xe75f, 0xf77e, 0xc71d, 0xd73c,
 208                         0x26d3, 0x36f2, 0x0691, 0x16b0, 0x6657, 0x7676, 0x4615, 0x5634,
 209                         0xd94c, 0xc96d, 0xf90e, 0xe92f, 0x99c8, 0x89e9, 0xb98a, 0xa9ab,
 210                         0x5844, 0x4865, 0x7806, 0x6827, 0x18c0, 0x08e1, 0x3882, 0x28a3,
 211                         0xcb7d, 0xdb5c, 0xeb3f, 0xfb1e, 0x8bf9, 0x9bd8, 0xabbb, 0xbb9a,
 212                         0x4a75, 0x5a54, 0x6a37, 0x7a16, 0x0af1, 0x1ad0, 0x2ab3, 0x3a92,
 213                         0xfd2e, 0xed0f, 0xdd6c, 0xcd4d, 0xbdaa, 0xad8b, 0x9de8, 0x8dc9,
 214                         0x7c26, 0x6c07, 0x5c64, 0x4c45, 0x3ca2, 0x2c83, 0x1ce0, 0x0cc1,
 215                         0xef1f, 0xff3e, 0xcf5d, 0xdf7c, 0xaf9b, 0xbfba, 0x8fd9, 0x9ff8,
 216                         0x6e17, 0x7e36, 0x4e55, 0x5e74, 0x2e93, 0x3eb2, 0x0ed1, 0x1ef0
 217                 );
 218                 $len = strlen( $data );
 219                 $crc = $seed;
 220                 for( $i = 0; $i < $len; $i++ ) {
 221                         $crc ^= ord( $data{$i} ) << 8;
 222                         $crc &= 0xFFFF;
 223                         $crc = ($crc << 8) ^ $MAGIC[$crc >> 8];
 224                         $crc &= 0xFFFF;
 225                 }
 226                 return $crc;
 227         }
 228
 229         /**
 230          * @param resource $destination
 231          * @param int $bytesToCopy
 232          * @return bool
 233          * @access private
 234          */
 235         function copyBytesTo( $destination, $bytesToCopy ) {
 236                 $bufferSize = 65536;
 237                 for( $remaining = $bytesToCopy; $remaining > 0; $remaining -= $bufferSize ) {
 238                         $thisChunkSize = min( $remaining, $bufferSize );
 239                         $buffer = fread( $this->handle, $thisChunkSize );
 240                         fwrite( $destination, $buffer );
 241                 }
 242         }
 243
 244         /**
 245          * Hex dump of the header for debugging
 246          * @access private
 247          */
 248         function hexdump( $data ) {
 249                 global $wgDebugLogFile;
 250                 if( !$wgDebugLogFile ) return;
 251
 252                 $width = 16;
 253                 $at = 0;
 254                 for( $remaining = strlen( $data ); $remaining > 0; $remaining -= $width ) {
 255                         $line = sprintf( "%04x:", $at );
 256                         $printable = '';
 257                         for( $i = 0; $i < $width && $remaining - $i > 0; $i++ ) {
 258                                 $byte = ord( $data{$at++} );
 259                                 $line .= sprintf( " %02x", $byte );
 260                                 $printable .= ($byte >= 32 && $byte <= 126 )
 261                                         ? chr( $byte )
 262                                         : '.';
 263                         }
 264                         if( $i < $width ) {
 265                                 $line .= str_repeat( '   ', $width - $i );
 266                         }
 267                         wfDebug( "MacBinary: $line $printable\n" );
 268                 }
 269         }
 270 }
 271
 272 ?>