/*
 * IndexTask.java
 */
package org.ngbw.utils;


import java.io.File;
import java.io.FileInputStream;
import java.io.InputStream;
import java.io.IOException;
import java.util.concurrent.Future;
import java.util.zip.GZIPInputStream;

import org.apache.lucene.analysis.KeywordAnalyzer;
import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.index.IndexWriter;


/**
 *
 * @author Paul Hoover
 *
 */
abstract class IndexTask implements Task {

	private final IndexWriter m_fieldWriter;
	private final IndexWriter m_sourceWriter;

	protected final String m_dataset;
	protected final String m_dataDir;
	protected final String m_indexDir;
	protected final String m_fileName;


	protected IndexTask(String dataset, String dataDir, String indexDir, String fileName) throws IOException
	{
		m_dataset = dataset;
		m_dataDir = dataDir;
		m_indexDir = indexDir;
		m_fileName = fileName;

		String separator = System.getProperty("file.separator");
		String fieldDir = m_indexDir + separator + m_dataset + separator + "FIELD_INDEX";
		String sourceDir = m_indexDir + separator + m_dataset + separator + "SOURCE_INDEX";

		m_fieldWriter = new IndexWriter(fieldDir, new StandardAnalyzer(), true);
		m_sourceWriter = new IndexWriter(sourceDir, new KeywordAnalyzer(), true);
	}


	public void run()
	{
		try {
			String separator = System.getProperty("file.separator");
			int offset = m_fileName.lastIndexOf(separator);
			String baseDir;
			String fileName;

			if (offset != -1) {
				baseDir = m_dataDir + separator + m_fileName.substring(0, offset);
				fileName = m_fileName.substring(offset + 1);
			}
			else {
				baseDir = m_dataDir;
				fileName = m_fileName;
			}

			String[] names = (new File(baseDir)).list(new WildcardFilter(fileName));

			for (int i = 0 ; i < names.length ; i += 1) {
				String name = baseDir + separator + names[i];
				InputStream inStream;

				try {
					inStream = new GZIPInputStream(new FileInputStream(name));
				}
				catch (IOException ioErr) {
					inStream = new FileInputStream(name);
				}

				try {
					addRecords(inStream);
				}
				finally {
					inStream.close();
				}
			}

			optimizeIndex();
			closeIndex();
		}
		catch (Exception err) {
			err.printStackTrace(System.err);
		}
	}

	public void runInChildProcess()
	{
		try {
			closeIndex();

			String separator = System.getProperty("file.separator");
			String jvmPathName = System.getProperty("java.home") + separator + "bin" + separator + "java";
			String javaClassPath = System.getProperty("java.class.path");
			String className = this.getClass().getName();
			String[] commandArray = new String[10];

			commandArray[0] = jvmPathName;
			commandArray[1] = "-Xms1024m";
			commandArray[2] = "-Xmx8192m";
			commandArray[3] = "-cp";
			commandArray[4] = javaClassPath;
			commandArray[5] = className;
			commandArray[6] = m_dataset;
			commandArray[7] = m_dataDir;
			commandArray[8] = m_indexDir;
			commandArray[9] = m_fileName;

			Process indexJob = Runtime.getRuntime().exec(commandArray);
			Future<String> stdOut = InputStreamCollector.readInputStream(indexJob.getInputStream());
			Future<String> stdErr = InputStreamCollector.readInputStream(indexJob.getErrorStream());
			int exitCode = indexJob.waitFor();

			System.out.print(stdOut.get());

			if (exitCode != 0)
				System.err.println("Indexing failed for dataset " + m_dataset);

			System.err.print(stdErr.get());
		}
		catch (Exception err) {
			err.printStackTrace(System.err);
		}
	}


	protected abstract void addRecords(InputStream inStream) throws IOException;

	protected void addSequence(SequenceRecord record) throws IOException
	{
		Document newDoc = new Document();

		newDoc.add(new Field("RECORD_ID", record.primaryId, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("PRIMARY_ID", record.primaryId, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("ALTERNATIVE_ID", record.alternativeId, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("NAME", record.name, Field.Store.YES, Field.Index.TOKENIZED));
		newDoc.add(new Field("ORGANISM", record.organism, Field.Store.YES, Field.Index.TOKENIZED));
		newDoc.add(new Field("VERSION", record.version, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("SIMPLE_SEARCH", record.filteredSource, Field.Store.NO, Field.Index.TOKENIZED));
		newDoc.add(new Field("DATASET", m_dataset, Field.Store.YES, Field.Index.UN_TOKENIZED));

		m_fieldWriter.addDocument(newDoc);

		addSource(record);
	}

	protected void addStructure(StructureRecord record) throws IOException
	{
		Document newDoc = new Document();

		newDoc.add(new Field("RECORD_ID", record.primaryId, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("PRIMARY_ID", record.primaryId, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("TYPE", record.type, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("RESOLUTION", record.resolution, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("NAME", record.name, Field.Store.YES, Field.Index.TOKENIZED));
		newDoc.add(new Field("AUTHOR", record.author, Field.Store.YES, Field.Index.TOKENIZED));
		newDoc.add(new Field("MOLECULE", record.molecule, Field.Store.YES, Field.Index.TOKENIZED));
		newDoc.add(new Field("ORGANISM", record.organism, Field.Store.YES, Field.Index.TOKENIZED));
		newDoc.add(new Field("DEPOSITION_DATE", record.depositionDate, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("MODIFICATION_DATE", record.modificationDate, Field.Store.YES, Field.Index.TOKENIZED));
		newDoc.add(new Field("SIMPLE_SEARCH", record.filteredSource, Field.Store.NO, Field.Index.TOKENIZED));
		newDoc.add(new Field("DATASET", m_dataset, Field.Store.YES, Field.Index.UN_TOKENIZED));

		m_fieldWriter.addDocument(newDoc);

		addSource(record);
	}


	private void optimizeIndex() throws IOException
	{
		m_fieldWriter.optimize();
		m_sourceWriter.optimize();
	}

	private void closeIndex() throws IOException
	{
		m_fieldWriter.close();
		m_sourceWriter.close();
	}

	private void addSource(Record record) throws IOException
	{
		Document newDoc = new Document();

		newDoc.add(new Field("RECORD_ID", record.primaryId, Field.Store.YES, Field.Index.UN_TOKENIZED));
		newDoc.add(new Field("RECORD_SOURCE", record.completeSource, Field.Store.COMPRESS, Field.Index.NO));

		m_sourceWriter.addDocument(newDoc);
	}
}