A Document Builder Generic for TIka and Lucene -> Yankees to offer Girardi a new deal

/**

* example usage of the configurator pattern. E Lucene nao eh banco de dados, eh banco de indice.

* os dados ficam no banco ou nao, pode ser o sistema de arquivos. mas nao ficam no lucene.

* Lucene eh Bussieness Inteligence e serviço de macaco eh ficar pulando feito um idiota.

* Mouse Bussiness eh coisa seria.. ;-)

**/

package org.nanotek.lucene.tika;

import org.nanotek.Kong;

import org.apache.tika.Tika;

import org.nanotek.util.Configurator;

import org.apache.tika.io.TikaInputStream;

import org.apache.lucene.document.Document;

import org.nanotek.lucene.LuceneDocumentBuilder;

import org.nanotek.lucene.TikaDocumentTransformer;

public abstract class  TikaBaseDocumentBuilder<S extends TikaInputStream, P extends Kong<String>> implements LuceneDocumentBuilder<S> , TikaDocumentTransformer<S,P,Document>{

private Tika tika;

private String textFieldName = "TextData";

protected Configurator<S,P,Document> configurator;

public TikaBaseDocumentBuilder() {

initTika();

}

@Override

public Document buildDocument(S source) {

return transform(source);

}

private void initTika() {

tika = new Tika();

}

public Tika getTika() {

return tika;

}

public void setTika(Tika tika) {

this.tika = tika;

}

public String getTextFieldName() {

return textFieldName;

}

public void setTextFieldName(String textFieldName) {

this.textFieldName = textFieldName;

}

@Override

public abstract Document transform(S source,P parameters);

//      {

//              Document document = null;

//              String txtExtracted;

//              if (tika == null)

//                      initTika();

//              try {

//                      txtExtracted = tika.parseToString(source);

//                      document = new Document();

//                      Field fileLocationField = new TextField("file_location", source.toString() , Field.Store.YES); //$NON-NLS-1$ //$NON-NLS-2$

//                      document.add(fileLocationField);

//                      Field wordField = new TextField(textFieldName, txtExtracted, Field.Store.YES); //$NON-NLS-1$ //$NON-NLS-2$

//                      document.add(wordField);

//                      Field modifiedField = new LongField("modified" , new Date().getTime() , Field.Store.YES); //$NON-NLS-1$

//                      document.add(modifiedField);

//              } catch (IOException | TikaException e) {

//                      throw new TransformerException(e);

//              }

//              return document;

//      }

}