includes/DjVuImage.php

   1 <?php
   2 /**
   3  * Support for detecting/validating DjVu image files and getting
   4  * some basic file metadata (resolution etc)
   5  *
   6  * File format docs are available in source package for DjVuLibre:
   7  * http://djvulibre.djvuzone.org/
   8  *
   9  *
  10  * Copyright (C) 2006 Brion Vibber <brion@pobox.com>
  11  * http://www.mediawiki.org/
  12  *
  13  * This program is free software; you can redistribute it and/or modify
  14  * it under the terms of the GNU General Public License as published by
  15  * the Free Software Foundation; either version 2 of the License, or
  16  * (at your option) any later version.
  17  *
  18  * This program is distributed in the hope that it will be useful,
  19  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  20  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  21  * GNU General Public License for more details.
  22  *
  23  * You should have received a copy of the GNU General Public License along
  24  * with this program; if not, write to the Free Software Foundation, Inc.,
  25  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  26  * http://www.gnu.org/copyleft/gpl.html
  27  *
  28  * @package MediaWiki
  29  */
  30
  31 class DjVuImage {
  32         function __construct( $filename ) {
  33                 $this->mFilename = $filename;
  34         }
  35
  36         /**
  37          * Check if the given file is indeed a valid DjVu image file
  38          * @return bool
  39          */
  40         public function isValid() {
  41                 $info = $this->getInfo();
  42                 return $info !== false;
  43         }
  44
  45
  46         /**
  47          * Return data in the style of getimagesize()
  48          * @return array or false on failure
  49          */
  50         public function getImageSize() {
  51                 $data = $this->getInfo();
  52
  53                 if( $data !== false ) {
  54                         $width  = $data['width'];
  55                         $height = $data['height'];
  56
  57                         return array( $width, $height, 'DjVu',
  58                                 "width=\"$width\" height=\"$height\"" );
  59                 }
  60                 return false;
  61         }
  62
  63         // ---------
  64
  65         /**
  66          * For debugging; dump the IFF chunk structure
  67          */
  68         function dump() {
  69                 $file = fopen( $this->mFilename, 'rb' );
  70                 $header = fread( $file, 12 );
  71                 // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
  72                 extract( unpack( 'a4magic/a4chunk/NchunkLength', $header ) );
  73                 echo "$chunk $chunkLength\n";
  74                 $this->dumpForm( $file, $chunkLength, 1 );
  75                 fclose( $file );
  76         }
  77
  78         private function dumpForm( $file, $length, $indent ) {
  79                 $start = ftell( $file );
  80                 $secondary = fread( $file, 4 );
  81                 echo str_repeat( ' ', $indent * 4 ) . "($secondary)\n";
  82                 while( ftell( $file ) - $start < $length ) {
  83                         $chunkHeader = fread( $file, 8 );
  84                         if( $chunkHeader == '' ) {
  85                                 break;
  86                         }
  87                         // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
  88                         extract( unpack( 'a4chunk/NchunkLength', $chunkHeader ) );
  89                         echo str_repeat( ' ', $indent * 4 ) . "$chunk $chunkLength\n";
  90
  91                         if( $chunk == 'FORM' ) {
  92                                 $this->dumpForm( $file, $chunkLength, $indent + 1 );
  93                         } else {
  94                                 fseek( $file, $chunkLength, SEEK_CUR );
  95                                 if( $chunkLength & 1 == 1 ) {
  96                                         // Padding byte between chunks
  97                                         fseek( $file, 1, SEEK_CUR );
  98                                 }
  99                         }
 100                 }
 101         }
 102
 103         function getInfo() {
 104                 $file = fopen( $this->mFilename, 'rb' );
 105                 if( $file === false ) {
 106                         wfDebug( __METHOD__ . ": missing or failed file read\n" );
 107                         return false;
 108                 }
 109
 110                 $header = fread( $file, 16 );
 111                 $info = false;
 112
 113                 if( strlen( $header ) < 16 ) {
 114                         wfDebug( __METHOD__ . ": too short file header\n" );
 115                 } else {
 116                         // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
 117                         extract( unpack( 'a4magic/a4form/NformLength/a4subtype', $header ) );
 118
 119                         if( $magic != 'AT&T' ) {
 120                                 wfDebug( __METHOD__ . ": not a DjVu file\n" );
 121                         } elseif( $subtype == 'DJVU' ) {
 122                                 // Single-page document
 123                                 $info = $this->getPageInfo( $file, $formLength );
 124                         } elseif( $subtype == 'DJVM' ) {
 125                                 // Multi-page document
 126                                 $info = $this->getMultiPageInfo( $file, $formLength );
 127                         } else  {
 128                                 wfDebug( __METHOD__ . ": unrecognized DJVU file type '$formType'\n" );
 129                         }
 130                 }
 131                 fclose( $file );
 132                 return $info;
 133         }
 134
 135         private function readChunk( $file ) {
 136                 $header = fread( $file, 8 );
 137                 if( strlen( $header ) < 8 ) {
 138                         return array( false, 0 );
 139                 } else {
 140                         // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
 141                         extract( unpack( 'a4chunk/Nlength', $header ) );
 142                         return array( $chunk, $length );
 143                 }
 144         }
 145
 146         private function skipChunk( $file, $chunkLength ) {
 147                 fseek( $file, $chunkLength, SEEK_CUR );
 148
 149                 if( $chunkLength & 0x01 == 1 && !feof( $file ) ) {
 150                         // padding byte
 151                         fseek( $file, 1, SEEK_CUR );
 152                 }
 153         }
 154
 155         private function getMultiPageInfo( $file, $formLength ) {
 156                 // For now, we'll just look for the first page in the file
 157                 // and report its information, hoping others are the same size.
 158                 $start = ftell( $file );
 159                 do {
 160                         list( $chunk, $length ) = $this->readChunk( $file );
 161                         if( !$chunk ) {
 162                                 break;
 163                         }
 164
 165                         if( $chunk == 'FORM' ) {
 166                                 $subtype = fread( $file, 4 );
 167                                 if( $subtype == 'DJVU' ) {
 168                                         wfDebug( __METHOD__ . ": found first subpage\n" );
 169                                         return $this->getPageInfo( $file, $length );
 170                                 }
 171                                 $this->skipChunk( $file, $length - 4 );
 172                         } else {
 173                                 wfDebug( __METHOD__ . ": skipping '$chunk' chunk\n" );
 174                                 $this->skipChunk( $file, $length );
 175                         }
 176                 } while( $length != 0 && !feof( $file ) && ftell( $file ) - $start < $formLength );
 177
 178                 wfDebug( __METHOD__ . ": multi-page DJVU file contained no pages\n" );
 179                 return false;
 180         }
 181
 182         private function getPageInfo( $file, $formLength ) {
 183                 list( $chunk, $length ) = $this->readChunk( $file );
 184                 if( $chunk != 'INFO' ) {
 185                         wfDebug( __METHOD__ . ": expected INFO chunk, got '$chunk'\n" );
 186                         return false;
 187                 }
 188
 189                 if( $length < 9 ) {
 190                         wfDebug( __METHOD__ . ": INFO should be 9 or 10 bytes, found $length\n" );
 191                         return false;
 192                 }
 193                 $data = fread( $file, $length );
 194                 if( strlen( $data ) < $length ) {
 195                         wfDebug( __METHOD__ . ": INFO chunk cut off\n" );
 196                         return false;
 197                 }
 198
 199                 // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
 200                 extract( unpack(
 201                         'nwidth/' .
 202                         'nheight/' .
 203                         'Cminor/' .
 204                         'Cmajor/' .
 205                         'vresolution/' .
 206                         'Cgamma', $data ) );
 207                 # Newer files have rotation info in byte 10, but we don't use it yet.
 208
 209                 return array(
 210                         'width' => $width,
 211                         'height' => $height,
 212                         'version' => "$major.$minor",
 213                         'resolution' => $resolution,
 214                         'gamma' => $gamma / 10.0 );
 215         }
 216
 217         /**
 218          * Return an XML string describing the DjVu image
 219          * @return string
 220          */
 221         function retrieveMetaData() {
 222                 global $wgDjvuToXML;
 223                 if ( isset( $wgDjvuToXML ) ) {
 224                         $cmd = $wgDjvuToXML . ' --without-anno --without-text ' .
 225                                 wfEscapeShellArg( $this->mFilename );
 226                         $xml = wfShellExec( $cmd );
 227                 } else {
 228                         $xml = null;
 229                 }
 230                 return $xml;
 231         }
 232
 233 }
 234
 235
 236 ?>