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                 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                         extract( unpack( 'a4chunk/NchunkLength', $chunkHeader ) );
  87                         echo str_repeat( ' ', $indent * 4 ) . "$chunk $chunkLength\n";
  88
  89                         if( $chunk == 'FORM' ) {
  90                                 $this->dumpForm( $file, $chunkLength, $indent + 1 );
  91                         } else {
  92                                 fseek( $file, $chunkLength, SEEK_CUR );
  93                                 if( $chunkLength & 1 == 1 ) {
  94                                         // Padding byte between chunks
  95                                         fseek( $file, 1, SEEK_CUR );
  96                                 }
  97                         }
  98                 }
  99         }
 100
 101         function getInfo() {
 102                 $file = fopen( $this->mFilename, 'rb' );
 103                 if( $file === false ) {
 104                         wfDebug( __METHOD__ . ": missing or failed file read\n" );
 105                         return false;
 106                 }
 107
 108                 $header = fread( $file, 16 );
 109                 $info = false;
 110
 111                 if( strlen( $header ) < 16 ) {
 112                         wfDebug( __METHOD__ . ": too short file header\n" );
 113                 } else {
 114                         extract( unpack( 'a4magic/a4form/NformLength/a4subtype', $header ) );
 115
 116                         if( $magic != 'AT&T' ) {
 117                                 wfDebug( __METHOD__ . ": not a DjVu file\n" );
 118                         } elseif( $subtype == 'DJVU' ) {
 119                                 // Single-page document
 120                                 $info = $this->getPageInfo( $file, $formLength );
 121                         } elseif( $subtype == 'DJVM' ) {
 122                                 // Multi-page document
 123                                 $info = $this->getMultiPageInfo( $file, $formLength );
 124                         } else  {
 125                                 wfDebug( __METHOD__ . ": unrecognized DJVU file type '$formType'\n" );
 126                         }
 127                 }
 128                 fclose( $file );
 129                 return $info;
 130         }
 131
 132         private function readChunk( $file ) {
 133                 $header = fread( $file, 8 );
 134                 if( strlen( $header ) < 8 ) {
 135                         return array( false, 0 );
 136                 } else {
 137                         extract( unpack( 'a4chunk/Nlength', $header ) );
 138                         return array( $chunk, $length );
 139                 }
 140         }
 141
 142         private function skipChunk( $file, $chunkLength ) {
 143                 fseek( $file, $chunkLength, SEEK_CUR );
 144
 145                 if( $chunkLength & 0x01 == 1 && !feof( $file ) ) {
 146                         // padding byte
 147                         fseek( $file, 1, SEEK_CUR );
 148                 }
 149         }
 150
 151         private function getMultiPageInfo( $file, $formLength ) {
 152                 // For now, we'll just look for the first page in the file
 153                 // and report its information, hoping others are the same size.
 154                 $start = ftell( $file );
 155                 do {
 156                         list( $chunk, $length ) = $this->readChunk( $file );
 157                         if( !$chunk ) {
 158                                 break;
 159                         }
 160
 161                         if( $chunk == 'FORM' ) {
 162                                 $subtype = fread( $file, 4 );
 163                                 if( $subtype == 'DJVU' ) {
 164                                         wfDebug( __METHOD__ . ": found first subpage\n" );
 165                                         return $this->getPageInfo( $file, $length );
 166                                 }
 167                                 $this->skipChunk( $file, $length - 4 );
 168                         } else {
 169                                 wfDebug( __METHOD__ . ": skipping '$chunk' chunk\n" );
 170                                 $this->skipChunk( $file, $length );
 171                         }
 172                 } while( $length != 0 && !feof( $file ) && ftell( $file ) - $start < $formLength );
 173
 174                 wfDebug( __METHOD__ . ": multi-page DJVU file contained no pages\n" );
 175                 return false;
 176         }
 177
 178         private function getPageInfo( $file, $formLength ) {
 179                 list( $chunk, $length ) = $this->readChunk( $file );
 180                 if( $chunk != 'INFO' ) {
 181                         wfDebug( __METHOD__ . ": expected INFO chunk, got '$chunk'\n" );
 182                         return false;
 183                 }
 184
 185                 if( $length < 9 ) {
 186                         wfDebug( __METHOD__ . ": INFO should be 9 or 10 bytes, found $length\n" );
 187                         return false;
 188                 }
 189                 $data = fread( $file, $length );
 190                 if( strlen( $data ) < $length ) {
 191                         wfDebug( __METHOD__ . ": INFO chunk cut off\n" );
 192                         return false;
 193                 }
 194
 195                 extract( unpack(
 196                         'nwidth/' .
 197                         'nheight/' .
 198                         'Cminor/' .
 199                         'Cmajor/' .
 200                         'vresolution/' .
 201                         'Cgamma', $data ) );
 202                 # Newer files have rotation info in byte 10, but we don't use it yet.
 203
 204                 return array(
 205                         'width' => $width,
 206                         'height' => $height,
 207                         'version' => "$major.$minor",
 208                         'resolution' => $resolution,
 209                         'gamma' => $gamma / 10.0 );
 210         }
 211
 212         /**
 213          * Return an XML string describing the DjVu image
 214          * @return string
 215          */
 216         function retrieveMetaData() {
 217                 global $wgDjvuToXML;
 218                 if ( isset( $wgDjvuToXML ) ) {
 219                         $cmd = $wgDjvuToXML . ' --without-anno --without-text ' . $this->mFilename;
 220                         $xml = wfShellExec( $cmd, $retval );
 221                 } else {
 222                         $xml = null;
 223                 }
 224                 return $xml;
 225         }
 226
 227 }
 228
 229
 230 ?>