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