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  */
  29
  30 class DjVuImage {
  31         function __construct( $filename ) {
  32                 $this->mFilename = $filename;
  33         }
  34
  35         /**
  36          * Check if the given file is indeed a valid DjVu image file
  37          * @return bool
  38          */
  39         public function isValid() {
  40                 $info = $this->getInfo();
  41                 return $info !== false;
  42         }
  43
  44
  45         /**
  46          * Return data in the style of getimagesize()
  47          * @return array or false on failure
  48          */
  49         public function getImageSize() {
  50                 $data = $this->getInfo();
  51
  52                 if( $data !== false ) {
  53                         $width  = $data['width'];
  54                         $height = $data['height'];
  55
  56                         return array( $width, $height, 'DjVu',
  57                                 "width=\"$width\" height=\"$height\"" );
  58                 }
  59                 return false;
  60         }
  61
  62         // ---------
  63
  64         /**
  65          * For debugging; dump the IFF chunk structure
  66          */
  67         function dump() {
  68                 $file = fopen( $this->mFilename, 'rb' );
  69                 $header = fread( $file, 12 );
  70                 // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
  71                 extract( unpack( 'a4magic/a4chunk/NchunkLength', $header ) );
  72                 echo "$chunk $chunkLength\n";
  73                 $this->dumpForm( $file, $chunkLength, 1 );
  74                 fclose( $file );
  75         }
  76
  77         private function dumpForm( $file, $length, $indent ) {
  78                 $start = ftell( $file );
  79                 $secondary = fread( $file, 4 );
  80                 echo str_repeat( ' ', $indent * 4 ) . "($secondary)\n";
  81                 while( ftell( $file ) - $start < $length ) {
  82                         $chunkHeader = fread( $file, 8 );
  83                         if( $chunkHeader == '' ) {
  84                                 break;
  85                         }
  86                         // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
  87                         extract( unpack( 'a4chunk/NchunkLength', $chunkHeader ) );
  88                         echo str_repeat( ' ', $indent * 4 ) . "$chunk $chunkLength\n";
  89
  90                         if( $chunk == 'FORM' ) {
  91                                 $this->dumpForm( $file, $chunkLength, $indent + 1 );
  92                         } else {
  93                                 fseek( $file, $chunkLength, SEEK_CUR );
  94                                 if( $chunkLength & 1 == 1 ) {
  95                                         // Padding byte between chunks
  96                                         fseek( $file, 1, SEEK_CUR );
  97                                 }
  98                         }
  99                 }
 100         }
 101
 102         function getInfo() {
 103                 $file = fopen( $this->mFilename, 'rb' );
 104                 if( $file === false ) {
 105                         wfDebug( __METHOD__ . ": missing or failed file read\n" );
 106                         return false;
 107                 }
 108
 109                 $header = fread( $file, 16 );
 110                 $info = false;
 111
 112                 if( strlen( $header ) < 16 ) {
 113                         wfDebug( __METHOD__ . ": too short file header\n" );
 114                 } else {
 115                         // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
 116                         extract( unpack( 'a4magic/a4form/NformLength/a4subtype', $header ) );
 117
 118                         if( $magic != 'AT&T' ) {
 119                                 wfDebug( __METHOD__ . ": not a DjVu file\n" );
 120                         } elseif( $subtype == 'DJVU' ) {
 121                                 // Single-page document
 122                                 $info = $this->getPageInfo( $file, $formLength );
 123                         } elseif( $subtype == 'DJVM' ) {
 124                                 // Multi-page document
 125                                 $info = $this->getMultiPageInfo( $file, $formLength );
 126                         } else  {
 127                                 wfDebug( __METHOD__ . ": unrecognized DJVU file type '$formType'\n" );
 128                         }
 129                 }
 130                 fclose( $file );
 131                 return $info;
 132         }
 133
 134         private function readChunk( $file ) {
 135                 $header = fread( $file, 8 );
 136                 if( strlen( $header ) < 8 ) {
 137                         return array( false, 0 );
 138                 } else {
 139                         // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
 140                         extract( unpack( 'a4chunk/Nlength', $header ) );
 141                         return array( $chunk, $length );
 142                 }
 143         }
 144
 145         private function skipChunk( $file, $chunkLength ) {
 146                 fseek( $file, $chunkLength, SEEK_CUR );
 147
 148                 if( $chunkLength & 0x01 == 1 && !feof( $file ) ) {
 149                         // padding byte
 150                         fseek( $file, 1, SEEK_CUR );
 151                 }
 152         }
 153
 154         private function getMultiPageInfo( $file, $formLength ) {
 155                 // For now, we'll just look for the first page in the file
 156                 // and report its information, hoping others are the same size.
 157                 $start = ftell( $file );
 158                 do {
 159                         list( $chunk, $length ) = $this->readChunk( $file );
 160                         if( !$chunk ) {
 161                                 break;
 162                         }
 163
 164                         if( $chunk == 'FORM' ) {
 165                                 $subtype = fread( $file, 4 );
 166                                 if( $subtype == 'DJVU' ) {
 167                                         wfDebug( __METHOD__ . ": found first subpage\n" );
 168                                         return $this->getPageInfo( $file, $length );
 169                                 }
 170                                 $this->skipChunk( $file, $length - 4 );
 171                         } else {
 172                                 wfDebug( __METHOD__ . ": skipping '$chunk' chunk\n" );
 173                                 $this->skipChunk( $file, $length );
 174                         }
 175                 } while( $length != 0 && !feof( $file ) && ftell( $file ) - $start < $formLength );
 176
 177                 wfDebug( __METHOD__ . ": multi-page DJVU file contained no pages\n" );
 178                 return false;
 179         }
 180
 181         private function getPageInfo( $file, $formLength ) {
 182                 list( $chunk, $length ) = $this->readChunk( $file );
 183                 if( $chunk != 'INFO' ) {
 184                         wfDebug( __METHOD__ . ": expected INFO chunk, got '$chunk'\n" );
 185                         return false;
 186                 }
 187
 188                 if( $length < 9 ) {
 189                         wfDebug( __METHOD__ . ": INFO should be 9 or 10 bytes, found $length\n" );
 190                         return false;
 191                 }
 192                 $data = fread( $file, $length );
 193                 if( strlen( $data ) < $length ) {
 194                         wfDebug( __METHOD__ . ": INFO chunk cut off\n" );
 195                         return false;
 196                 }
 197
 198                 // FIXME: Would be good to replace this extract() call with something that explicitly initializes local variables.
 199                 extract( unpack(
 200                         'nwidth/' .
 201                         'nheight/' .
 202                         'Cminor/' .
 203                         'Cmajor/' .
 204                         'vresolution/' .
 205                         'Cgamma', $data ) );
 206                 # Newer files have rotation info in byte 10, but we don't use it yet.
 207
 208                 return array(
 209                         'width' => $width,
 210                         'height' => $height,
 211                         'version' => "$major.$minor",
 212                         'resolution' => $resolution,
 213                         'gamma' => $gamma / 10.0 );
 214         }
 215
 216         /**
 217          * Return an XML string describing the DjVu image
 218          * @return string
 219          */
 220         function retrieveMetaData() {
 221                 global $wgDjvuToXML;
 222                 if ( isset( $wgDjvuToXML ) ) {
 223                         $cmd = $wgDjvuToXML . ' --without-anno --without-text ' .
 224                                 wfEscapeShellArg( $this->mFilename );
 225                         $xml = wfShellExec( $cmd );
 226                 } else {
 227                         $xml = null;
 228                 }
 229                 return $xml;
 230         }
 231
 232 }
 233
 234
 235 ?>