includes/media/SVGMetadataExtractor.php

   1 <?php
   2 /**
   3  * SVGMetadataExtractor.php
   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  * @author Derk-Jan Hartman <hartman _at_ videolan d0t org>
  23  * @author Brion Vibber
  24  * @copyright Copyright © 2010-2010 Brion Vibber, Derk-Jan Hartman
  25  * @license http://www.gnu.org/copyleft/gpl.html GNU General Public License
  26  */
  27
  28 class SVGMetadataExtractor {
  29         static function getMetadata( $filename ) {
  30                 $svg = new SVGReader( $filename );
  31                 return $svg->getMetadata();
  32         }
  33 }
  34
  35 class SVGReader {
  36         const DEFAULT_WIDTH = 512;
  37         const DEFAULT_HEIGHT = 512;
  38
  39         private $reader = null;
  40         private $mDebug = false;
  41         private $metadata = Array();
  42
  43         /**
  44          * Constructor
  45          *
  46          * Creates an SVGReader drawing from the source provided
  47          * @param $source String: URI from which to read
  48          */
  49         function __construct( $source ) {
  50                 $this->reader = new XMLReader();
  51                 $this->reader->open( $source );
  52
  53                 $this->metadata['width'] = self::DEFAULT_WIDTH;
  54                 $this->metadata['height'] = self::DEFAULT_HEIGHT;
  55
  56                 $this->read();
  57         }
  58
  59         /*
  60          * @return Array with the known metadata
  61          */
  62         public function getMetadata() {
  63                 return $this->metadata;
  64         }
  65
  66         /*
  67          * Read the SVG
  68          */
  69         public function read() {
  70                 $this->reader->read();
  71
  72                 if ( $this->reader->name != 'svg' ) {
  73                         throw new MWException( "Expected <svg> tag, got ".
  74                                 $this->reader->name );
  75                 }
  76                 $this->debug( "<svg> tag is correct." );
  77
  78                 $this->debug( "Starting primary dump processing loop." );
  79                 $this->handleSVGAttribs();
  80                 $exitDepth =  $this->reader->depth;
  81
  82                 $keepReading = $this->reader->read();
  83                 $skip = false;
  84                 while ( $keepReading ) {
  85                         $tag = $this->reader->name;
  86                         $type = $this->reader->nodeType;
  87
  88                         $this->debug( "$tag" );
  89
  90                         if ( $tag == 'svg' && $type == XmlReader::END_ELEMENT && $this->reader->depth <= $exitDepth ) {
  91                                 break;
  92                         } elseif ( $tag == 'title' ) {
  93                                 $this->readField( $tag, 'title' );
  94                         } elseif ( $tag == 'desc' ) {
  95                                 $this->readField( $tag, 'description' );
  96                         } elseif ( $tag == 'metadata' && $type == XmlReader::ELEMENT ) {
  97                                 $this->readXml( $tag, 'metadata' );
  98                         } elseif ( $tag !== '#text' ) {
  99                                 $this->debug( "Unhandled top-level XML tag $tag" );
 100                                 $this->animateFilter( $tag );
 101                                 //$skip = true;
 102                         }
 103
 104                         if ($skip) {
 105                                 $keepReading = $this->reader->next();
 106                                 $skip = false;
 107                                 $this->debug( "Skip" );
 108                         } else {
 109                                 $keepReading = $this->reader->read();
 110                         }
 111                 }
 112
 113                 return true;
 114         }
 115
 116         /*
 117          * Read a textelement from an element
 118          *
 119          * @param String $name of the element that we are reading from
 120          * @param String $metafield that we will fill with the result
 121          */
 122         private function readField( $name, $metafield=null ) {
 123                 $this->debug ( "Read field $metafield" );
 124                 if( !$metafield || $this->reader->nodeType != XmlReader::ELEMENT ) {
 125                         return;
 126                 }
 127                 $keepReading = $this->reader->read();
 128                 while( $keepReading ) {
 129                         if( $this->reader->name == $name && $this->reader->nodeType == XmlReader::END_ELEMENT ) {
 130                                 $keepReading = false;
 131                                 break;
 132                         } elseif( $this->reader->nodeType == XmlReader::TEXT ){
 133                                 $this->metadata[$metafield] = $this->reader->value;
 134                         }
 135                         $keepReading = $this->reader->read();
 136                 }
 137         }
 138
 139         /*
 140          * Read an XML snippet from an element
 141          *
 142          * @param String $metafield that we will fill with the result
 143          */
 144         private function readXml( $metafield=null ) {
 145                 $this->debug ( "Read top level metadata" );
 146                 if( !$metafield || $this->reader->nodeType != XmlReader::ELEMENT ) {
 147                         return;
 148                 }
 149                 // TODO: find and store type of xml snippet. metadata['metadataType'] = "rdf"
 150                 $this->metadata[$metafield] = $this->reader->readInnerXML();
 151                 $this->reader->next();
 152         }
 153
 154         /*
 155          * Filter all children, looking for animate elements
 156          *
 157          * @param String $name of the element that we are reading from
 158          */
 159         private function animateFilter( $name ) {
 160                 $this->debug ( "animate filter" );
 161                 if( $this->reader->nodeType != XmlReader::ELEMENT ) {
 162                         return;
 163                 }
 164                 $exitDepth =  $this->reader->depth;
 165                 $keepReading = $this->reader->read();
 166                 while( $keepReading ) {
 167                         if( $this->reader->name == $name && $this->reader->depth <= $exitDepth
 168                                 && $this->reader->nodeType == XmlReader::END_ELEMENT ) {
 169                                 $keepReading = false;
 170                                 break;
 171                         } elseif( $this->reader->nodeType == XmlReader::ELEMENT ){
 172                                 switch( $this->reader->name ) {
 173                                         case 'animate':
 174                                         case 'set':
 175                                         case 'animateMotion':
 176                                         case 'animateColor':
 177                                         case 'animateTransform':
 178                                                 $this->debug( "HOUSTON WE HAVE ANIMATION" );
 179                                                 $this->metadata['animated'] = true;
 180                                                 break;
 181                                 }
 182                         }
 183                         $keepReading = $this->reader->read();
 184                 }
 185         }
 186
 187         private function throwXmlError( $err ) {
 188                 $this->debug( "FAILURE: $err" );
 189                 wfDebug( "SVGReader XML error: $err\n" );
 190         }
 191
 192         private function debug( $data ) {
 193                 if( $this->mDebug ) {
 194                         wfDebug( "SVGReader: $data\n" );
 195                 }
 196         }
 197
 198         private function warn( $data ) {
 199                 wfDebug( "SVGReader: $data\n" );
 200         }
 201
 202         private function notice( $data ) {
 203                 wfDebug( "SVGReader WARN: $data\n" );
 204         }
 205
 206         /*
 207          * Parse the attributes of an SVG element
 208          *
 209          * The parser has to be in the start element of <svg>
 210          */
 211         private function handleSVGAttribs( ) {
 212                 $defaultWidth = self::DEFAULT_WIDTH;
 213                 $defaultHeight = self::DEFAULT_HEIGHT;
 214                 $aspect = 1.0;
 215                 $width = null;
 216                 $height = null;
 217
 218                 if( $this->reader->getAttribute('viewBox') ) {
 219                         // min-x min-y width height
 220                         $viewBox = preg_split( '/\s+/', trim( $this->reader->getAttribute('viewBox') ) );
 221                         if( count( $viewBox ) == 4 ) {
 222                                 $viewWidth = $this->scaleSVGUnit( $viewBox[2] );
 223                                 $viewHeight = $this->scaleSVGUnit( $viewBox[3] );
 224                                 if( $viewWidth > 0 && $viewHeight > 0 ) {
 225                                         $aspect = $viewWidth / $viewHeight;
 226                                         $defaultHeight = $defaultWidth / $aspect;
 227                                 }
 228                         }
 229                 }
 230                 if( $this->reader->getAttribute('width') ) {
 231                         $width = $this->scaleSVGUnit( $this->reader->getAttribute('width'), $defaultWidth );
 232                 }
 233                 if( $this->reader->getAttribute('height') ) {
 234                         $height = $this->scaleSVGUnit( $this->reader->getAttribute('height'), $defaultHeight );
 235                 }
 236
 237                 if( !isset( $width ) && !isset( $height ) ) {
 238                         $width = $defaultWidth;
 239                         $height = $width / $aspect;
 240                 } elseif( isset( $width ) && !isset( $height ) ) {
 241                         $height = $width / $aspect;
 242                 } elseif( isset( $height ) && !isset( $width ) ) {
 243                         $width = $height * $aspect;
 244                 }
 245
 246                 if( $width > 0 && $height > 0 ) {
 247                         $this->metadata['width'] = intval( round( $width ) );
 248                         $this->metadata['height'] = intval( round( $height ) );
 249                 }
 250         }
 251
 252         /**
 253          * Return a rounded pixel equivalent for a labeled CSS/SVG length.
 254          * http://www.w3.org/TR/SVG11/coords.html#UnitIdentifiers
 255          *
 256          * @param $length String: CSS/SVG length.
 257          * @param $viewportSize: Float optional scale for percentage units...
 258          * @return float: length in pixels
 259          */
 260         static function scaleSVGUnit( $length, $viewportSize=512 ) {
 261                 static $unitLength = array(
 262                         'px' => 1.0,
 263                         'pt' => 1.25,
 264                         'pc' => 15.0,
 265                         'mm' => 3.543307,
 266                         'cm' => 35.43307,
 267                         'in' => 90.0,
 268                         'em' => 16.0, // fake it?
 269                         'ex' => 12.0, // fake it?
 270                         ''   => 1.0, // "User units" pixels by default
 271                         );
 272                 $matches = array();
 273                 if( preg_match( '/^\s*(\d+(?:\.\d+)?)(em|ex|px|pt|pc|cm|mm|in|%|)\s*$/', $length, $matches ) ) {
 274                         $length = floatval( $matches[1] );
 275                         $unit = $matches[2];
 276                         if( $unit == '%' ) {
 277                                 return $length * 0.01 * $viewportSize;
 278                         } else {
 279                                 return $length * $unitLength[$unit];
 280                         }
 281                 } else {
 282                         // Assume pixels
 283                         return floatval( $length );
 284                 }
 285         }
 286 }