includes/media/DjVu.php

   1 <?php
   2 /**
   3  * Handler for DjVu images.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Media
  22  */
  23
  24 /**
  25  * Handler for DjVu images
  26  *
  27  * @ingroup Media
  28  */
  29 class DjVuHandler extends ImageHandler {
  30
  31         /**
  32          * @return bool
  33          */
  34         function isEnabled() {
  35                 global $wgDjvuRenderer, $wgDjvuDump, $wgDjvuToXML;
  36                 if ( !$wgDjvuRenderer || ( !$wgDjvuDump && !$wgDjvuToXML ) ) {
  37                         wfDebug( "DjVu is disabled, please set \$wgDjvuRenderer and \$wgDjvuDump\n" );
  38                         return false;
  39                 } else {
  40                         return true;
  41                 }
  42         }
  43
  44         /**
  45          * @param $file
  46          * @return bool
  47          */
  48         function mustRender( $file ) {
  49                 return true;
  50         }
  51
  52         /**
  53          * @param $file
  54          * @return bool
  55          */
  56         function isMultiPage( $file ) {
  57                 return true;
  58         }
  59
  60         /**
  61          * @return array
  62          */
  63         function getParamMap() {
  64                 return array(
  65                         'img_width' => 'width',
  66                         'img_page' => 'page',
  67                 );
  68         }
  69
  70         /**
  71          * @param $name
  72          * @param $value
  73          * @return bool
  74          */
  75         function validateParam( $name, $value ) {
  76                 if ( in_array( $name, array( 'width', 'height', 'page' ) ) ) {
  77                         if ( $value <= 0 ) {
  78                                 return false;
  79                         } else {
  80                                 return true;
  81                         }
  82                 } else {
  83                         return false;
  84                 }
  85         }
  86
  87         /**
  88          * @param $params
  89          * @return bool|string
  90          */
  91         function makeParamString( $params ) {
  92                 $page = isset( $params['page'] ) ? $params['page'] : 1;
  93                 if ( !isset( $params['width'] ) ) {
  94                         return false;
  95                 }
  96                 return "page{$page}-{$params['width']}px";
  97         }
  98
  99         /**
 100          * @param $str
 101          * @return array|bool
 102          */
 103         function parseParamString( $str ) {
 104                 $m = false;
 105                 if ( preg_match( '/^page(\d+)-(\d+)px$/', $str, $m ) ) {
 106                         return array( 'width' => $m[2], 'page' => $m[1] );
 107                 } else {
 108                         return false;
 109                 }
 110         }
 111
 112         /**
 113          * @param $params
 114          * @return array
 115          */
 116         function getScriptParams( $params ) {
 117                 return array(
 118                         'width' => $params['width'],
 119                         'page' => $params['page'],
 120                 );
 121         }
 122
 123         /**
 124          * @param $image File
 125          * @param  $dstPath
 126          * @param  $dstUrl
 127          * @param  $params
 128          * @param int $flags
 129          * @return MediaTransformError|ThumbnailImage|TransformParameterError
 130          */
 131         function doTransform( $image, $dstPath, $dstUrl, $params, $flags = 0 ) {
 132                 global $wgDjvuRenderer, $wgDjvuPostProcessor;
 133
 134                 // Fetch XML and check it, to give a more informative error message than the one which
 135                 // normaliseParams will inevitably give.
 136                 $xml = $image->getMetadata();
 137                 if ( !$xml ) {
 138                         $width = isset( $params['width'] ) ? $params['width'] : 0;
 139                         $height = isset( $params['height'] ) ? $params['height'] : 0;
 140                         return new MediaTransformError( 'thumbnail_error', $width, $height,
 141                                 wfMessage( 'djvu_no_xml' )->text() );
 142                 }
 143
 144                 if ( !$this->normaliseParams( $image, $params ) ) {
 145                         return new TransformParameterError( $params );
 146                 }
 147                 $width = $params['width'];
 148                 $height = $params['height'];
 149                 $page = $params['page'];
 150                 if ( $page > $this->pageCount( $image ) ) {
 151                         return new MediaTransformError(
 152                                 'thumbnail_error',
 153                                 $width,
 154                                 $height,
 155                                 wfMessage( 'djvu_page_error' )->text()
 156                         );
 157                 }
 158
 159                 if ( $flags & self::TRANSFORM_LATER ) {
 160                         return new ThumbnailImage( $image, $dstUrl, $width, $height, $dstPath, $page );
 161                 }
 162
 163                 if ( !wfMkdirParents( dirname( $dstPath ), null, __METHOD__ ) ) {
 164                         return new MediaTransformError(
 165                                 'thumbnail_error',
 166                                 $width,
 167                                 $height,
 168                                 wfMessage( 'thumbnail_dest_directory' )->text()
 169                         );
 170                 }
 171
 172                 $srcPath = $image->getLocalRefPath();
 173                 # Use a subshell (brackets) to aggregate stderr from both pipeline commands
 174                 # before redirecting it to the overall stdout. This works in both Linux and Windows XP.
 175                 $cmd = '(' . wfEscapeShellArg( $wgDjvuRenderer ) . " -format=ppm -page={$page}" .
 176                         " -size={$params['physicalWidth']}x{$params['physicalHeight']} " .
 177                         wfEscapeShellArg( $srcPath );
 178                 if ( $wgDjvuPostProcessor ) {
 179                         $cmd .= " | {$wgDjvuPostProcessor}";
 180                 }
 181                 $cmd .= ' > ' . wfEscapeShellArg($dstPath) . ') 2>&1';
 182                 wfProfileIn( 'ddjvu' );
 183                 wfDebug( __METHOD__.": $cmd\n" );
 184                 $retval = '';
 185                 $err = wfShellExec( $cmd, $retval );
 186                 wfProfileOut( 'ddjvu' );
 187
 188                 $removed = $this->removeBadFile( $dstPath, $retval );
 189                 if ( $retval != 0 || $removed ) {
 190                         wfDebugLog( 'thumbnail',
 191                                 sprintf( 'thumbnail failed on %s: error %d "%s" from "%s"',
 192                                         wfHostname(), $retval, trim($err), $cmd ) );
 193                         return new MediaTransformError( 'thumbnail_error', $width, $height, $err );
 194                 } else {
 195                         return new ThumbnailImage( $image, $dstUrl, $width, $height, $dstPath, $page );
 196                 }
 197         }
 198
 199         /**
 200          * Cache an instance of DjVuImage in an Image object, return that instance
 201          *
 202          * @return DjVuImage
 203          */
 204         function getDjVuImage( $image, $path ) {
 205                 if ( !$image ) {
 206                         $deja = new DjVuImage( $path );
 207                 } elseif ( !isset( $image->dejaImage ) ) {
 208                         $deja = $image->dejaImage = new DjVuImage( $path );
 209                 } else {
 210                         $deja = $image->dejaImage;
 211                 }
 212                 return $deja;
 213         }
 214
 215         /**
 216          * Cache a document tree for the DjVu XML metadata
 217          * @param $image File
 218          * @param $gettext Boolean: DOCUMENT (Default: false)
 219          * @return bool
 220          */
 221         function getMetaTree( $image , $gettext = false ) {
 222                 if ( isset( $image->dejaMetaTree ) ) {
 223                         return $image->dejaMetaTree;
 224                 }
 225
 226                 $metadata = $image->getMetadata();
 227                 if ( !$this->isMetadataValid( $image, $metadata ) ) {
 228                         wfDebug( "DjVu XML metadata is invalid or missing, should have been fixed in upgradeRow\n" );
 229                         return false;
 230                 }
 231                 wfProfileIn( __METHOD__ );
 232
 233                 wfSuppressWarnings();
 234                 try {
 235                         // Set to false rather than null to avoid further attempts
 236                         $image->dejaMetaTree = false;
 237                         $image->djvuTextTree = false;
 238                         $tree = new SimpleXMLElement( $metadata );
 239                         if( $tree->getName() == 'mw-djvu' ) {
 240                                 foreach($tree->children() as $b){
 241                                         if( $b->getName() == 'DjVuTxt' ) {
 242                                                 $image->djvuTextTree = $b;
 243                                         }
 244                                         elseif ( $b->getName() == 'DjVuXML' ) {
 245                                                 $image->dejaMetaTree = $b;
 246                                         }
 247                                 }
 248                         } else {
 249                                 $image->dejaMetaTree = $tree;
 250                         }
 251                 } catch( Exception $e ) {
 252                         wfDebug( "Bogus multipage XML metadata on '{$image->getName()}'\n" );
 253                 }
 254                 wfRestoreWarnings();
 255                 wfProfileOut( __METHOD__ );
 256                 if( $gettext ) {
 257                         return $image->djvuTextTree;
 258                 } else {
 259                         return $image->dejaMetaTree;
 260                 }
 261         }
 262
 263         function getImageSize( $image, $path ) {
 264                 return $this->getDjVuImage( $image, $path )->getImageSize();
 265         }
 266
 267         function getThumbType( $ext, $mime, $params = null ) {
 268                 global $wgDjvuOutputExtension;
 269                 static $mime;
 270                 if ( !isset( $mime ) ) {
 271                         $magic = MimeMagic::singleton();
 272                         $mime = $magic->guessTypesForExtension( $wgDjvuOutputExtension );
 273                 }
 274                 return array( $wgDjvuOutputExtension, $mime );
 275         }
 276
 277         function getMetadata( $image, $path ) {
 278                 wfDebug( "Getting DjVu metadata for $path\n" );
 279                 return $this->getDjVuImage( $image, $path )->retrieveMetaData();
 280         }
 281
 282         function getMetadataType( $image ) {
 283                 return 'djvuxml';
 284         }
 285
 286         function isMetadataValid( $image, $metadata ) {
 287                 return !empty( $metadata ) && $metadata != serialize(array());
 288         }
 289
 290         function pageCount( $image ) {
 291                 $tree = $this->getMetaTree( $image );
 292                 if ( !$tree ) {
 293                         return false;
 294                 }
 295                 return count( $tree->xpath( '//OBJECT' ) );
 296         }
 297
 298         function getPageDimensions( $image, $page ) {
 299                 $tree = $this->getMetaTree( $image );
 300                 if ( !$tree ) {
 301                         return false;
 302                 }
 303
 304                 $o = $tree->BODY[0]->OBJECT[$page-1];
 305                 if ( $o ) {
 306                         return array(
 307                                 'width' => intval( $o['width'] ),
 308                                 'height' => intval( $o['height'] )
 309                         );
 310                 } else {
 311                         return false;
 312                 }
 313         }
 314
 315         function getPageText( $image, $page ){
 316                 $tree = $this->getMetaTree( $image, true );
 317                 if ( !$tree ) {
 318                         return false;
 319                 }
 320
 321                 $o = $tree->BODY[0]->PAGE[$page-1];
 322                 if ( $o ) {
 323                         $txt = $o['value'];
 324                         return $txt;
 325                 } else {
 326                         return false;
 327                 }
 328
 329         }
 330
 331 }