package org.sf.mustru.docs;
import java.io.IOException;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.index.IndexWriter;
import org.sf.mustru.crawl.ClassifyDoc;
//import com.aliasi.tokenizer.NGramTokenizerFactory;
import com.sleepycat.bind.tuple.TupleBinding;
import com.sleepycat.bind.tuple.TupleInput;
import com.sleepycat.bind.tuple.TupleOutput;
/**
* Media class for image, audio, animation, and video objects
*
*/
public class MediaDoc extends IndexableDoc
{
String textType = "";
/**
* A generic media document class
* String File name containing the image
*/
public MediaDoc ()
{ super(); setBdbBinding( new MediaDocDb() ); }
public MediaDoc (String ifile)
{ super(ifile); setBdbBinding( new MediaDocDb() ); }
public void loadSpecific(ClassifyDoc cdoc)
{
setTextType(cdoc.classifyTextContents(this));
//*-- file type is set in CrawlThread
}
//*-- create the Lucene Index
public void loadIndex(IndexWriter ramIW, boolean storeTermVector) throws IOException
{
Document doc = new Document();
doc.add( new Field("key", getFileName(), Field.Store.YES, Field.Index.NO) );
doc.add(new Field("contents", getContents().toString(), Field.Store.NO, Field.Index.TOKENIZED) );
doc.add( new Field("type", getFileType(), Field.Store.YES, Field.Index.NO) );
ramIW.addDocument(doc);
}
public TupleBinding getBdbBinding()
{ return bdbBinding; }
public void setBdbBinding(TupleBinding bdbBinding)
{ this.bdbBinding = bdbBinding; }
public String toString()
{
StringBuffer sb = new StringBuffer();
sb.append(super.toString());
//*-- add media specific data
return sb.toString();
}
public String getTextType()
{ return textType; }
public void setTextType(String textType)
{ this.textType = textType; }
}
final class MediaDocDb extends TupleBinding
{
static IndexableDocBinding idb = new IndexableDocBinding();
IndexableDoc idoc;
public Object entryToObject(TupleInput ti)
{
idoc = (IndexableDoc) idb.entryToObject(ti);
MediaDoc o = new MediaDoc();
o.loadGeneric(idoc); idoc = null;
//*-- write any media specific information to o
return o;
}
public void objectToEntry(Object o, TupleOutput to)
{
idb.objectToEntry(o, to);
//MediaDoc id = (MediaDoc) o;
//*-- write any media specific information to tuple output
}
}