Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: pass (asf.osuosl.org: local policy)
Message-ID: <002401c66eec$16420340$ad471bac@intraet.cesar.org.br>
From: "Rodrigo Marcuschi" <rodrigo.marcuschi@cesar.org.br>
To: <java-user@lucene.apache.org>
Subject: Problem with lucene 1.9.1
Date: Wed, 3 May 2006 16:59:28 -0300
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0021_01C66ED2.F09A2820"

------=_NextPart_000_0021_01C66ED2.F09A2820
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hello,=20

I am having problems using lucene 1.9.1. I was using 1.4.3 successfully, =
and tried upgrading to 1.9.1. As I changed my source code so I would no =
longer be invocating deprecated methods, the application no longer =
worked properly, i.e. the IndexModifier was no longer storing documents. =
Here is my source code:

-------------------------------------------------------------------------=
------------------------------
package org.cesar.trulog.search.api;

import java.io.File;
import java.io.IOException;

import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.index.IndexModifier;

public class UserIndexer {

 public static final String ID =3D "userId";
 public static final String KEYWORDS =3D "searchKeywords";
 public static final String USERNAME =3D "searchUsername";

 private final File index;
 private final Analyzer analyzer;
 private IndexModifier modifier;=20

 public UserIndexer(final Analyzer analyzer, final String dir) throws =
IOException {
      index =3D new File(dir);
 =20
      this.analyzer =3D analyzer;
      synchronized(UserIndexer.class) {
           if (!index.exists()) {
                index.mkdirs();
           }
      }
 =20
      this.modifier =3D new IndexModifier(this.index, this.analyzer, =
true);
 }

 public UserIndexer(final String dir) throws IOException {
      this(new StandardAnalyzer(), dir);
 }
=20
 private Document getDocument(final SearchItem searchItem) {
      Document doc =3D new Document();=20
      doc.add(new Field(ID, ((Integer) =
searchItem.getValue(SearchItemField.ID)).toString(), Field.Store.YES, =
Field.Index.UN_TOKENIZED));
      doc.add(new Field(KEYWORDS, (String) =
searchItem.getValue(SearchItemField.KEYWORD), Field.Store.YES, =
Field.Index.TOKENIZED));
      doc.add(new Field(USERNAME, (String) =
searchItem.getValue(SearchItemField.NAME), Field.Store.YES, =
Field.Index.TOKENIZED));
 =20
      return doc;
 }
=20
 public synchronized void add(final SearchItem searchItem) throws =
IOException {=20
     modifier.deleteDocuments(new Term(ID, ((Integer) =
searchItem.getValue(SearchItemField.ID)).toString())); =20
     Document doc =3D this.getDocument(searchItem);
     modifier.addDocument(doc);
     modifier.flush();
     System.out.println(modifier.docCount() + " docs in index");
     modifier.close();  =20
 }=20
}

-------------------------------------------------------------------------=
------------------------------

SearchItem is a custom class that has the data to be used by lucene. The =
relevant method in there is=20

void add(final SearchItem searchItem)=20

which calls lucene's 'addDocument' method. It used to work back when I =
was using Field.Keyword, Field.Text and modifier.delete --- but all =
these got deprecated, so.... Here's what happens: no matter how many =
documents I add, all I get is '1 docs in index'. If I DON'T close the =
modifier, I get the right results 2, 3, 4 docs in index etc --- but if =
the application is restarted, the docs are lost, i.e. a search that =
should return 2 or 3 results returns no result at all.

Any ideas where to start? What to fix? If I should go back using the =
deprecated methods? I tried using FSDirectory instead of File, didn't =
work. I tried not deleting the documents before adding them, it didn't =
work. I tried using a custom analyzer, it didn't work. I'm running out =
of ideas.

Thank you very much for your attention,
    Rodrigo
------=_NextPart_000_0021_01C66ED2.F09A2820--