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, null, LIBXML_NOERROR | LIBXML_NOWARNING );
  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                 $keepReading = $this->reader->read();
  71
  72                 /* Skip until first element */
  73                 while( $keepReading && $this->reader->nodeType != XmlReader::ELEMENT ) {
  74                         $keepReading = $this->reader->read();
  75                 }
  76
  77                 if ( $this->reader->name != 'svg' ) {
  78                         throw new MWException( "Expected <svg> tag, got ".
  79                                 $this->reader->name );
  80                 }
  81                 $this->debug( "<svg> tag is correct." );
  82                 $this->handleSVGAttribs();
  83
  84                 $exitDepth =  $this->reader->depth;
  85                 $keepReading = $this->reader->read();
  86                 $skip = false;
  87                 while ( $keepReading ) {
  88                         $tag = $this->reader->name;
  89                         $type = $this->reader->nodeType;
  90
  91                         $this->debug( "$tag" );
  92
  93                         if ( $tag == 'svg' && $type == XmlReader::END_ELEMENT && $this->reader->depth <= $exitDepth ) {
  94                                 break;
  95                         } elseif ( $tag == 'title' ) {
  96                                 $this->readField( $tag, 'title' );
  97                         } elseif ( $tag == 'desc' ) {
  98                                 $this->readField( $tag, 'description' );
  99                         } elseif ( $tag == 'metadata' && $type == XmlReader::ELEMENT ) {
 100                                 $this->readXml( $tag, 'metadata' );
 101                         } elseif ( $tag !== '#text' ) {
 102                                 $this->debug( "Unhandled top-level XML tag $tag" );
 103                                 $this->animateFilter( $tag );
 104                                 //$skip = true;
 105                         }
 106
 107                         if ($skip) {
 108                                 $keepReading = $this->reader->next();
 109                                 $skip = false;
 110                                 $this->debug( "Skip" );
 111                         } else {
 112                                 $keepReading = $this->reader->read();
 113                         }
 114                 }
 115
 116                 return true;
 117         }
 118
 119         /*
 120          * Read a textelement from an element
 121          *
 122          * @param String $name of the element that we are reading from
 123          * @param String $metafield that we will fill with the result
 124          */
 125         private function readField( $name, $metafield=null ) {
 126                 $this->debug ( "Read field $metafield" );
 127                 if( !$metafield || $this->reader->nodeType != XmlReader::ELEMENT ) {
 128                         return;
 129                 }
 130                 $keepReading = $this->reader->read();
 131                 while( $keepReading ) {
 132                         if( $this->reader->name == $name && $this->reader->nodeType == XmlReader::END_ELEMENT ) {
 133                                 break;
 134                         } elseif( $this->reader->nodeType == XmlReader::TEXT ){
 135                                 $this->metadata[$metafield] = $this->reader->value;
 136                         }
 137                         $keepReading = $this->reader->read();
 138                 }
 139         }
 140
 141         /*
 142          * Read an XML snippet from an element
 143          *
 144          * @param String $metafield that we will fill with the result
 145          */
 146         private function readXml( $metafield=null ) {
 147                 $this->debug ( "Read top level metadata" );
 148                 if( !$metafield || $this->reader->nodeType != XmlReader::ELEMENT ) {
 149                         return;
 150                 }
 151                 // TODO: find and store type of xml snippet. metadata['metadataType'] = "rdf"
 152                 $this->metadata[$metafield] = $this->reader->readInnerXML();
 153                 $this->reader->next();
 154         }
 155
 156         /*
 157          * Filter all children, looking for animate elements
 158          *
 159          * @param String $name of the element that we are reading from
 160          */
 161         private function animateFilter( $name ) {
 162                 $this->debug ( "animate filter" );
 163                 if( $this->reader->nodeType != XmlReader::ELEMENT ) {
 164                         return;
 165                 }
 166                 $exitDepth =  $this->reader->depth;
 167                 $keepReading = $this->reader->read();
 168                 while( $keepReading ) {
 169                         if( $this->reader->name == $name && $this->reader->depth <= $exitDepth
 170                                 && $this->reader->nodeType == XmlReader::END_ELEMENT ) {
 171                                 break;
 172                         } elseif ( $this->reader->nodeType == XmlReader::ELEMENT ) {
 173                                 switch( $this->reader->name ) {
 174                                         case 'animate':
 175                                         case 'set':
 176                                         case 'animateMotion':
 177                                         case 'animateColor':
 178                                         case 'animateTransform':
 179                                                 $this->debug( "HOUSTON WE HAVE ANIMATION" );
 180                                                 $this->metadata['animated'] = true;
 181                                                 break;
 182                                 }
 183                         }
 184                         $keepReading = $this->reader->read();
 185                 }
 186         }
 187
 188         private function throwXmlError( $err ) {
 189                 $this->debug( "FAILURE: $err" );
 190                 wfDebug( "SVGReader XML error: $err\n" );
 191         }
 192
 193         private function debug( $data ) {
 194                 if( $this->mDebug ) {
 195                         wfDebug( "SVGReader: $data\n" );
 196                 }
 197         }
 198
 199         private function warn( $data ) {
 200                 wfDebug( "SVGReader: $data\n" );
 201         }
 202
 203         private function notice( $data ) {
 204                 wfDebug( "SVGReader WARN: $data\n" );
 205         }
 206
 207         /*
 208          * Parse the attributes of an SVG element
 209          *
 210          * The parser has to be in the start element of <svg>
 211          */
 212         private function handleSVGAttribs( ) {
 213                 $defaultWidth = self::DEFAULT_WIDTH;
 214                 $defaultHeight = self::DEFAULT_HEIGHT;
 215                 $aspect = 1.0;
 216                 $width = null;
 217                 $height = null;
 218
 219                 if( $this->reader->getAttribute('viewBox') ) {
 220                         // min-x min-y width height
 221                         $viewBox = preg_split( '/\s+/', trim( $this->reader->getAttribute('viewBox') ) );
 222                         if( count( $viewBox ) == 4 ) {
 223                                 $viewWidth = $this->scaleSVGUnit( $viewBox[2] );
 224                                 $viewHeight = $this->scaleSVGUnit( $viewBox[3] );
 225                                 if( $viewWidth > 0 && $viewHeight > 0 ) {
 226                                         $aspect = $viewWidth / $viewHeight;
 227                                         $defaultHeight = $defaultWidth / $aspect;
 228                                 }
 229                         }
 230                 }
 231                 if( $this->reader->getAttribute('width') ) {
 232                         $width = $this->scaleSVGUnit( $this->reader->getAttribute('width'), $defaultWidth );
 233                 }
 234                 if( $this->reader->getAttribute('height') ) {
 235                         $height = $this->scaleSVGUnit( $this->reader->getAttribute('height'), $defaultHeight );
 236                 }
 237
 238                 if( !isset( $width ) && !isset( $height ) ) {
 239                         $width = $defaultWidth;
 240                         $height = $width / $aspect;
 241                 } elseif( isset( $width ) && !isset( $height ) ) {
 242                         $height = $width / $aspect;
 243                 } elseif( isset( $height ) && !isset( $width ) ) {
 244                         $width = $height * $aspect;
 245                 }
 246
 247                 if( $width > 0 && $height > 0 ) {
 248                         $this->metadata['width'] = intval( round( $width ) );
 249                         $this->metadata['height'] = intval( round( $height ) );
 250                 }
 251         }
 252
 253         /**
 254          * Return a rounded pixel equivalent for a labeled CSS/SVG length.
 255          * http://www.w3.org/TR/SVG11/coords.html#UnitIdentifiers
 256          *
 257          * @param $length String: CSS/SVG length.
 258          * @param $viewportSize: Float optional scale for percentage units...
 259          * @return float: length in pixels
 260          */
 261         static function scaleSVGUnit( $length, $viewportSize=512 ) {
 262                 static $unitLength = array(
 263                         'px' => 1.0,
 264                         'pt' => 1.25,
 265                         'pc' => 15.0,
 266                         'mm' => 3.543307,
 267                         'cm' => 35.43307,
 268                         'in' => 90.0,
 269                         'em' => 16.0, // fake it?
 270                         'ex' => 12.0, // fake it?
 271                         ''   => 1.0, // "User units" pixels by default
 272                         );
 273                 $matches = array();
 274                 if( preg_match( '/^\s*(\d+(?:\.\d+)?)(em|ex|px|pt|pc|cm|mm|in|%|)\s*$/', $length, $matches ) ) {
 275                         $length = floatval( $matches[1] );
 276                         $unit = $matches[2];
 277                         if( $unit == '%' ) {
 278                                 return $length * 0.01 * $viewportSize;
 279                         } else {
 280                                 return $length * $unitLength[$unit];
 281                         }
 282                 } else {
 283                         // Assume pixels
 284                         return floatval( $length );
 285                 }
 286         }
 287 }