- simplified

java/org/slub/rosetta/dps/repository/plugin/SLUBTechnicalMetadataExtractorMediaConchPlugin.java

@@ -20,12 +20,10 @@ package org.slub.rosetta.dps.repository.plugin;
 import com.exlibris.core.sdk.strings.StringUtils;
 import com.exlibris.dps.sdk.techmd.MDExtractorPlugin;
 import org.xml.sax.InputSource;
-import org.xml.sax.XMLReader;
 
-import javax.xml.parsers.SAXParserFactory;
+import javax.xml.XMLConstants;
 import javax.xml.transform.Transformer;
 import javax.xml.transform.TransformerFactory;
-import javax.xml.transform.sax.SAXSource;
 import javax.xml.transform.stream.StreamResult;
 import javax.xml.transform.stream.StreamSource;
 import java.io.BufferedReader;
@@ -122,6 +120,7 @@ public class SLUBTechnicalMetadataExtractorMediaConchPlugin implements MDExtract
         }
 
 
+
         /* ffprobe output of metadata
            supports different outputs. we are using the flat-model, see WRITERS section in ffprobe manual
            the streams will be mapped as: streams.stream.0.$property
@@ -137,43 +136,37 @@ public class SLUBTechnicalMetadataExtractorMediaConchPlugin implements MDExtract
             String line=reader.readLine();
             StringBuilder mediainfo_output= new StringBuilder();
             while (line != null) {
-                System.out.println(line);
-                mediainfo_output.append(line);
+                    /* we should patched out, because not allowed to download xsd */
+                String regex = " https://mediaarea\\.net/mediainfo/mediainfo.*\\.xsd";
+                String line_patched = line.replaceAll( regex, "");
+                mediainfo_output.append(line_patched);
                 line = reader.readLine();
             }
             /* xslt transform */
             InputStream stylestream = getClass().getResourceAsStream("resources/transformer.xsl");
             StreamSource stylesource = new StreamSource( stylestream);
+            System.out.println("stylesource, systemID=" + stylesource.getSystemId());
+            System.out.println("stylesource, publicID=" + stylesource.getPublicId());
+
             // Use a Transformer for output
-            SAXParserFactory spf = SAXParserFactory.newInstance();
-            spf.setNamespaceAware(true);
-            /* disabled external DTD loading, which does not work with class ressource */
-            spf.setValidating(false);
-            spf.setNamespaceAware(true);
-            spf.setFeature("http://apache.org/xml/features/nonvalidating/load-external-dtd", false);
-            spf.setFeature("http://xml.org/sax/features/use-entity-resolver2", false);
-            spf.setFeature("http://xml.org/sax/features/validation", false);
-            spf.setFeature("http://apache.org/xml/features/nonvalidating/load-dtd-grammar", false);
-            spf.setFeature("http://xml.org/sax/features/allow-dtd-events-after-endDTD", false);
-            XMLReader r = spf.newSAXParser().getXMLReader();
-//            StreamSource mediainfo_source = new StreamSource(mediainfo_output);
-            SAXSource mediainfo_sax_source = new SAXSource(r, new InputSource(mediainfo_output.toString()));
+            StreamSource mediainfo_source = new StreamSource(String.valueOf(new InputSource(mediainfo_output.toString())));
+            System.out.println("mediainfo_source, systemID=" + mediainfo_source.getSystemId());
+            System.out.println("mediainfo_source, publicID=" + mediainfo_source.getPublicId());
+
             TransformerFactory tFactory = TransformerFactory.newInstance();
-            //tFactory.setAttribute(XMLConstants.ACCESS_EXTERNAL_DTD, "");
-            //tFactory.setAttribute(XMLConstants.ACCESS_EXTERNAL_STYLESHEET, "");
-//            tFactory.setAttribute(XMLConstants.ACCESS_EXTERNAL_SCHEMA, "");
-//            tFactory.setAttribute(XMLConstants.USE_CATALOG, "");
+            tFactory.setAttribute(XMLConstants.ACCESS_EXTERNAL_DTD, "");
+            tFactory.setAttribute(XMLConstants.ACCESS_EXTERNAL_STYLESHEET, "");
             Transformer transformer = tFactory.newTransformer(stylesource);
 
 /* ok, mediainfo is loaded correctly, and xslt loaded too */
             OutputStream debugFile = new FileOutputStream("DEBUG.xml");
             StreamResult result = new StreamResult( debugFile);
 //            StreamResult result = new StreamResult(System.out); /* FIXME , use StringOutputStream */
-            transformer.transform(mediainfo_sax_source, result);
+            transformer.transform(mediainfo_source, result);
             attributes.put("key", "value");
         } catch (InterruptedException e) {
+            Thread.currentThread().interrupt();
             e.printStackTrace();
-
         }
     }
 
@@ -209,6 +202,7 @@ public class SLUBTechnicalMetadataExtractorMediaConchPlugin implements MDExtract
         } catch (IOException e) {
             //log.error("exception creation socket, clamd not available at host=" + host + "port=" + port, e);
         } catch (InterruptedException e) {
+            Thread.currentThread().interrupt();
             e.printStackTrace();
         }
         return response.toString().trim();
@@ -224,7 +218,8 @@ public class SLUBTechnicalMetadataExtractorMediaConchPlugin implements MDExtract
 
     @Override
     public List<String> getExtractionErrors() {
-      return this.extractionErrors;
+        List<String> extractionErrors = this.extractionErrors;
+        return extractionErrors;
     }
 
     /* following list is build using: