StackOverflow2013

Note that there are some explanatory texts on larger screens.

plurals

POLucene IndexReader.reopen doesn't seem to work correctly
text
Body
copied!<p>I have a problem with Lucene 2.4, the situation being as follows:</p> <p>I have to deal with the possibility that there are 2 seperate processes operating on the same Index directory and they need to have the same data. This means that when one Instance adds a Document to the Index, the other application instances shall find the added Documents on their next search. According to the Lucene Documentation, IndexReader.reopen is what I need.</p> <p>So I invented the following testcase:</p> <pre><code>package de.samedi.searcher; import static org.junit.Assert.assertEquals; import static org.junit.Assert.fail; import java.io.IOException; import org.apache.lucene.analysis.standard.StandardAnalyzer; import org.apache.lucene.document.Document; import org.apache.lucene.document.Field; import org.apache.lucene.index.CorruptIndexException; import org.apache.lucene.index.IndexReader; import org.apache.lucene.index.IndexWriter; import org.apache.lucene.queryParser.QueryParser; import org.apache.lucene.search.IndexSearcher; import org.apache.lucene.search.Query; import org.apache.lucene.search.TopDocs; import org.apache.lucene.store.FSDirectory; import org.junit.Test; public class LuceneReload { private IndexSearcher searcher1; private IndexSearcher searcher2; private FSDirectory directory1, directory2; private IndexWriter writer1, writer2; @Test public void testReload() throws Exception { String home = System.getProperty("user.home"); this.directory1 = FSDirectory.getDirectory(home + "/testIndex"); this.directory2 = FSDirectory.getDirectory(home + "/testIndex"); this.writer1 = new IndexWriter(this.directory1, new StandardAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED); this.writer2 = new IndexWriter(this.directory2, new StandardAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED); // assert that we're empty assertFound(getSearcher1(), "test", 0); assertFound(getSearcher2(), "test", 0); add(this.writer1, "test"); assertFound(getSearcher1(), "test", 1); assertFound(getSearcher2(), "test", 1); add(this.writer2, "foobar"); assertFound(getSearcher1(), "foobar", 1); assertFound(getSearcher2(), "foobar", 1); } public void assertFound(IndexSearcher searcher, String q, int expected_number) { try { QueryParser parser = new QueryParser("name", new StandardAnalyzer()); Query query = parser.parse(q); TopDocs t = searcher.search(query, null, 50); assertEquals(expected_number, t.totalHits); } catch (Exception e) { e.printStackTrace(); fail(); } } public IndexSearcher getSearcher1() throws CorruptIndexException, IOException { if (this.searcher1 == null) { this.searcher1 = new IndexSearcher(IndexReader.open(this.directory1)); } else { IndexReader new_reader, old_reader; old_reader = this.searcher1.getIndexReader(); new_reader = old_reader.reopen(); if (new_reader != old_reader) { System.err.println("index1 changed"); this.searcher1.close(); old_reader.close(); this.searcher1 = new IndexSearcher(new_reader); } } return this.searcher1; } public IndexSearcher getSearcher2() throws CorruptIndexException, IOException { if (this.searcher2 == null) { this.searcher2 = new IndexSearcher(this.directory2); } else { IndexReader new_reader, old_reader; old_reader = this.searcher2.getIndexReader(); new_reader = old_reader.reopen(); if (new_reader != old_reader) { System.err.println("index2 changed"); this.searcher2.close(); old_reader.close(); this.searcher2 = new IndexSearcher(new_reader); } } return this.searcher2; } public void add(IndexWriter writer, String name) throws CorruptIndexException, IOException { Document d = new Document(); d.add(new Field("name", name, Field.Store.YES, Field.Index.ANALYZED)); writer.addDocument(d); writer.commit(); IndexWriter.unlock(writer.getDirectory()); } } </code></pre> <p>When I instead of the reopen() calls use</p> <pre><code>new_reader = IndexReader.open(this.directory1); </code></pre> <p>the tests go green.</p> <p>Did I miss any important points</p>

Querying!

Guidance

An individual column

Larger individual text columns get their own page to allow for proper reading.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload