JDBC

Search Lucene Index created in database using JdbcDirectory

Posted on Updated on

In our last excercise we created Lucene index in database using JdbcDirectory which comes with Compass in this post we will search against index created. With this much of text let’s get our hand dirty and write some code.

Please note following files will be used from create lucene index post

  1. pom.xml
  2. MyJDBCDirectory.java
  3. JDBCBatchInsert.java
  4. JDBCIndexer.java (without creating index we cannot search against it ;), so in our test case below we will create an index before searching)
  5. JDBCDatabaseUtil.java
  6. Database schema

Now with some code from our backyard with let’s finish up search quickly. Below is JDBCSearcher.java which has a very simple search method which takes the name of the index field and value which we want to search. It returns true if valid search is performed or else false. Please note with this search we are only interested in search result with highest hit, others we don’t care.

package com.mumz.test.lucene.jdbc;

import java.io.IOException;

import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.SimpleAnalyzer;
import org.apache.lucene.index.CorruptIndexException;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.queryParser.ParseException;
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.Directory;
import org.apache.lucene.util.Version;

/**
 * The Class InMemoryDirectorySearcher.
 * 
 * @author prabhat.jha
 */
public class JDBCSearcher {

	/** The directory. */
	private Directory	directory	= null;

	/**
	 * The Constructor.
	 * 
	 * @param directory
	 *            the directory
	 */
	public JDBCSearcher(Directory directory) {
		this.directory = directory;
	}

	/**
	 * Search.
	 * 
	 * @param fileName
	 *            the file name
	 * @return the string
	 */
	public boolean search(String columnName, String value) {
		IndexSearcher indexSearcher = null;
		try {
			/**
			 * Specify the version
			 */
			Analyzer analyzer = new SimpleAnalyzer(Version.LUCENE_36);
			/**
			 * Create query columnname (index name passed), we built out index for name, author and publisher so
			 * we have to search against the same.
			 */
			Query query = new QueryParser(Version.LUCENE_36, columnName, analyzer).parse(value);
			IndexReader indexReader = IndexReader.open(directory);
			indexSearcher = new IndexSearcher(indexReader);
			/**
			 * This will hold all the results which results from the search
			 * operation
			 */
			TopDocs topDocs = indexSearcher.search(query, 1);
			if (topDocs.scoreDocs.length > 0) {
				System.out.println("Found :  Book with id = " + indexSearcher.doc(topDocs.scoreDocs[0].doc).get("BOOKID") + " , Name = "
						+ indexSearcher.doc(topDocs.scoreDocs[0].doc).get("name") + " ,Author = "
						+ indexSearcher.doc(topDocs.scoreDocs[0].doc).get("author") + " ,Publisher = "
						+ indexSearcher.doc(topDocs.scoreDocs[0].doc).get("publisher") + " with hits : " + topDocs.scoreDocs[0].doc);
				return true;
			} else {
				System.out.println("No Record found");
				return false;
			}
		} catch (ParseException e) {
			e.printStackTrace();
		} catch (CorruptIndexException e) {
			e.printStackTrace();
		} catch (IOException e) {
			e.printStackTrace();
		} finally {
			if (indexSearcher != null) {
				try {
					indexSearcher.close();
				} catch (IOException e) {
					e.printStackTrace();
				}
			}
			indexSearcher = null;
		}
		return false;
	}
}

In you want to see all the hits you can use below code snippet.

TopScoreDocCollector results = TopScoreDocCollector.create(100, true);
TopDocs topDocs = indexSearcher.search(query, 1);
indexSearcher.search(query, results);
ScoreDoc[] scores = results.topDocs().scoreDocs;
for (ScoreDoc scoreDoc : scores) {
	System.out.println("Found :  Book with id = " + indexSearcher.doc(scoreDoc.doc).get("id") + " ,        
    Name = " + indexSearcher.doc(scoreDoc.doc).get("name") + " ,Author = " + 
    indexSearcher.doc(scoreDoc.doc).get("author") + " ,Publisher = " +   
    indexSearcher.doc(scoreDoc.doc).get("publisher") + " with hits : " + scoreDoc.score);

}

And finally here is our JUnit test case, please note this test case assumes that you also have indexer code with you.

package com.mumz.test.lucene.jdbc;

import org.apache.lucene.store.Directory;
import org.apache.lucene.store.jdbc.dialect.MySQLDialect;

import junit.framework.TestCase;

/**
 * The Class LuceneJDBCTest.
 * @author prabhat.jha
 */
public class LuceneJDBCTest extends TestCase {

	/** The directory. */
	private Directory	directory	= null;

       /** (non-Javadoc)
	 * @see junit.framework.TestCase#setUp()
	 */
	protected void setUp() throws Exception {
		directory = new MyJDBCDirectory(JDBCDatabaseUtil.getDataSource(), new MySQLDialect(), "LUCENE_INDEX_TABLE");
		super.setUp();
	}

	/**
	 * Test insert record.
	 */
	public void testInsertRecord() {		
		new JDBCBatchInsert().insertRecords();
	}

	/**
	 * Test build index.
	 */
	public void testBuildIndex() {
		new JDBCIndexer(directory).buildIndex();
	}

	/**
	 * Test search record on name.
	 */
	public void testSearchRecordOnName() {
		boolean found = new JDBCSearcher(directory).search("name", "Spring In Action");
		assertEquals(found, true);
	}

	/**
	 * Test search record fail on name.
	 */
	public void testSearchRecordFailOnName() {
		boolean found = new JDBCSearcher(directory).search("name", "No Such BookName");
		assertEquals(found, false);
	}

	/**
	 * Test search record on author.
	 */
	public void testSearchRecordOnAuthor() {
		boolean found = new JDBCSearcher(directory).search("author", "Test Author Hibernate In Action10");
		assertEquals(found, true);
	}

	/**
	 * Test search record fail on author.
	 */
	public void testSearchRecordFailOnAuthor() {
		boolean found = new JDBCSearcher(directory).search("name", "No Such Author");
		assertEquals(found, false);
	}

	/**
	 * Test search record on publisher.
	 */
	public void testSearchRecordOnPublisher() {
		boolean found = new JDBCSearcher(directory).search("publisher", "Test Publisher Spring Bible7");
		assertEquals(found, true);
	}

	/**
	 * Test search record fail on publisher.
	 */
	public void testSearchRecordFailOnPublisher() {
		boolean found = new JDBCSearcher(directory).search("name", "No Such Publisher");
		assertEquals(found, false);
	}
	
	/* (non-Javadoc)
	 * @see junit.framework.TestCase#tearDown()
	 */
	protected void tearDown() throws Exception {
		if(directory != null) {
			directory.close();
		}
		super.tearDown();
	}
}