source: trunk/gsdl3/src/java/org/greenstone/gsdl3/gs3build/extractor/GMLExtractor.java@ 5800

package org.greenstone.gsdl3.gs3build.extractor;

import java.io.FileReader;

import org.xml.sax.XMLReader;
import org.xml.sax.InputSource;
import org.xml.sax.SAXException;
import org.xml.sax.Attributes;
import org.xml.sax.helpers.XMLReaderFactory;
import org.xml.sax.helpers.DefaultHandler;

import org.greenstone.gsdl3.gs3build.doctypes.DocumentID;
import org.greenstone.gsdl3.gs3build.doctypes.DocumentInterface;
import org.greenstone.gsdl3.gs3build.doctypes.GMLDocument;

public class GMLExtractor implements ExtractorInterface
{
    /**
     *  An inner class to handle GML files
     */
    class GMLHandler extends DefaultHandler
    { String file;
      String label;
      StringBuffer value;
      boolean inElement;

      GMLHandler()
      { super();

        this.file = null;
    this.label = null;
    this.value = null;
      }

        public void startElement(String URI, String localName, String qName, Attributes attributes)
        {   if (localName.equals("Filename"))
            {   this.value = new StringBuffer();
            }
            else if (localName.equals("Metadata"))
            {   this.label = attributes.getValue("name");
                this.value = new StringBuffer();
            }
        }

        public void endElement(String URI, String localName, String qName)
        {   if (localName.equals("Filename"))
            {   this.file = this.value.toString();
                this.value = null;
            }
            else if (localName.equals("Metadata"))
            {   GMLExtractor.postMetadata(this.file, this.label, this.value.toString());
                this.value = null;
                this.label = null;
            }
        }

        public void characters(char c[], int start, int length)
        {   if (this.label != null)
            { String string = new String(c, start, length);
                this.value.append(string);
            }
        }
    }

    /**
     *  Construct of extractor
     */
    public GMLExtractor()
    { // Intentionally left blank
    }

    /**
     *  This extractor doesn't need to do any preparation/completion work,
     *  so this member function is empty.
     */
    public void configure(String outputDir)
    { // Intentionally left blank
    }

    /**
     *  This extractor doesn't need to do any preparation/completion work,
     *  so this member function is empty.
     */
    public void startPass(int passNo)
    { // Intentionally left blank
    }

    /**
     *  Process the document - for a GML document, this results in the
     *  decoration of other files, for other documents, it does nothing.
     */
    public void extractDocument(DocumentID docID, DocumentInterface document)
    {   
      if (document.getDocumentType().equals(GMLDocument.GML_DOCUMENT_TYPE))
      { // Extract the content from the GML file
    try {
      XMLReader reader = XMLReaderFactory.createXMLReader();
      GMLHandler handler = new GMLHandler();
      reader.setContentHandler(handler);
      reader.setErrorHandler(handler);

      // A GML document consists of one file only - get it.
      FileReader fileReader = new FileReader(document.getDocumentFiles().get(0).toString());
      reader.parse(new InputSource(fileReader));
    }
    catch (SAXException saxException)
    { // TODO: log error
    }
    catch (java.io.FileNotFoundException fileException)
    {
    }
    catch (java.io.IOException ioException)
    {
    }
    // for each document post it to the corresponding document
      }
    }

    protected static void postMetadata(String file, String value, String label)
    {
    }

    /**
     *  This extractor doesn't need to do any preparation/completion work,
     *  so this member function is empty.
     */
    public void endPass(int passNo)
    { // Intentionally left blank
    }

    /**
     *  This extractor is a simple, single-pass extractor
     *
     *  @see: org.greenstone.gsdl3.gs3build.extractor.ExtractorInterface:getNumberOfPasses
     */
    public int getNumberOfPasses()
    {   return 1;
    }
}