package compbio.cassandra;

import java.io.IOException;
import java.util.Calendar;
import java.util.HashMap;
import java.util.List;
import java.util.ArrayList;
import java.util.Map;

import com.datastax.driver.core.Cluster;
import com.datastax.driver.core.Host;
import com.datastax.driver.core.Metadata;
import com.datastax.driver.core.Row;
import com.datastax.driver.core.Session;
import com.datastax.driver.core.ResultSet;

public class CassandraNativeConnector {
	private static Cluster cluster;
	private static Session session;
	/*
	 * connect to the cluster and look weather the dababase has any data inside
	 */
	public void Connect() {
		// local cassandra cluster
		cluster = Cluster.builder().addContactPoint("localhost").build();
		// distributed cassandra cluster
		/* cluster = Cluster.builder().addContactPoint("10.0.115.190").build(); */
		Metadata metadata = cluster.getMetadata();
		System.out.printf("Connected to cluster: %s\n", metadata.getClusterName());
		for (Host host : metadata.getAllHosts()) {
			System.out.printf("Datatacenter: %s; Host: %s; Rack: %s\n", host.getDatacenter(), host.getAddress(), host.getRack());
		}

		session = cluster.connect();
		session.execute("CREATE KEYSPACE IF NOT EXISTS ProteinKeyspace WITH replication = {'class':'SimpleStrategy', 'replication_factor':3};");
		session.execute("CREATE COLUMNFAMILY IF NOT EXISTS ProteinKeyspace.ProteinRow (Protein ascii, JobID ascii, Predictions map<ascii,ascii>, PRIMARY KEY(JobID));");
		session.execute("CREATE COLUMNFAMILY IF NOT EXISTS ProteinKeyspace.ProteinLog "
				+ "(JobID ascii, DataBegin ascii, DataEnd ascii, ip ascii, FinalStatus ascii, ExecutionStatus ascii, Protein ascii, PRIMARY KEY(JobID));");
		session.execute("CREATE COLUMNFAMILY IF NOT EXISTS ProteinKeyspace.ProteinData (jobtime bigint, JobID ascii, Protein ascii, PRIMARY KEY(JobID));");

		session.execute("CREATE INDEX IF NOT EXISTS ProteinSeq ON ProteinKeyspace.ProteinRow (protein);");
		session.execute("CREATE INDEX IF NOT EXISTS JobDateStamp ON ProteinKeyspace.ProteinData (jobtime);");

		System.out.println("Cassandra connected");
	}

	/*
	 * parsing data source and filling the database
	 */
	public void Parsing() throws IOException {
		if (true) {
			// if (source.equals("http")) {
			// get data from real Jpred production server
			System.out.println("Parsing web data source......");
			String datasrc = "http://www.compbio.dundee.ac.uk/www-jpred/results/usage-new/alljobs.dat";
			String prefix = "http://www.compbio.dundee.ac.uk/www-jpred/results";
			JpredParserHTTP parser = new JpredParserHTTP(prefix);
			parser.Parsing(datasrc, 4);
		}
		if (false) {
			// if (source.equals("file")) {
			// get irtifical data generated for the DB stress tests
			System.out.println("Parsing local file data source......");
			String datasrc = "/home/asherstnev/Projects/Java.projects/proteocache/data_stress_test/data.dat";
			String prefix = "/home/asherstnev/Projects/Java.projects/proteocache/data_stress_test/Jpreddata";
			JpredParserLocalFile parser = new JpredParserLocalFile(prefix);
			parser.Parsing(datasrc, 190);
		}
	}

	public void Closing() {
		session.shutdown();
		cluster.shutdown();
		System.out.println("Cassandra has been shut down");
	}

	/*
	 * inserting data into the db
	 */
	public void InsertData(long jobtime, String startdate, String enddate, String ip, String jobid, String statusEx, String statusFinal,
			String protein, List<FastaSequence> predictions) {

		String check1 = "SELECT * FROM ProteinKeyspace.ProteinLog WHERE JobID = '" + jobid + "';";
		ResultSet results1 = session.execute(check1);
		if (results1.isExhausted()) {
			String com1 = "INSERT INTO ProteinKeyspace.ProteinLog "
					+ "(JobID, IP, DataBegin, DataEnd, FinalStatus, ExecutionStatus, Protein)" + " VALUES ('" + jobid + "','" + ip + "','"
					+ startdate + "','" + enddate + "','" + statusFinal + "','" + statusEx + "','" + protein + "');";
			session.execute(com1);

			String com2 = "INSERT INTO ProteinKeyspace.ProteinData " + "(jobtime, JobID, Protein)" + " VALUES (" + jobtime + ",'" + jobid
					+ "','" + protein + "');";
			session.execute(com2);

			String allpredictions = "";
			for (FastaSequence pred : predictions) {
				String predictionname = pred.getId();
				String prediction = pred.getSequence().replaceAll("\n", "");
				allpredictions += "'" + predictionname + "':'" + prediction + "',";
			}
			String final_prediction = "";
			if (null != allpredictions) {
				final_prediction = allpredictions.substring(0, allpredictions.length() - 1);
			}

			String check2 = "SELECT * FROM ProteinKeyspace.ProteinRow WHERE JobID = '" + jobid + "';";
			ResultSet results2 = session.execute(check2);
			if (results2.isExhausted()) {
				String com3 = "INSERT INTO ProteinKeyspace.ProteinRow " + "(Protein, JobID, Predictions)" + " VALUES ('" 
			+ protein + "','" + jobid + "',{" + final_prediction + "});";
				session.execute(com3);
			}
		}
	}

	/*
	 * getting data from the db
	 */
	public List<Pair<String, String>> ReadProteinDataTable() {
		final long startTime = System.currentTimeMillis();
		String com = "SELECT DataBegin,DataEnd FROM ProteinKeyspace.ProteinLog;";
		System.out.println("Command: " + com);
		ResultSet results = session.execute(com);
		final long queryTime = System.currentTimeMillis();
		List<Row> rows = results.all();
		System.out.println ("Query time is " + (queryTime - startTime) + " msec");

		List<Pair<String, String>> res = new ArrayList<Pair<String, String>>();
		int c = 0;
		for (Row r : rows) {
			Pair<String, String> pair = new Pair<String, String>(r.getString("DataBegin"),r.getString("DataEnd"));
			res.add(pair);
			++c;
		}
		final long endTime = System.currentTimeMillis();
		System.out.println (c + " rows analysed, execution time is " + (endTime - startTime) + " msec");
		return res;
	}
	
	/*
	 * getting data from the db ProteinData
	 */
	public Integer ReadDateTable(long queryDate) {
		final long startTime = System.currentTimeMillis();
		String com = "SELECT jobtime, JobID FROM ProteinKeyspace.ProteinData WHERE jobtime = " + queryDate + ";";
		System.out.println("Command: " + com);
		ResultSet results = session.execute(com);
		if (results.isExhausted())
			return null;
		final long queryTime = System.currentTimeMillis();
		List<Row> rows = results.all();
		System.out.println ("Query time is " + (queryTime - startTime) + " msec");        
		return rows.size();
	}

	/*
	 * getting whole protein sequence from the db ProteinRow
	 */
	public List<StructureProteinPrediction> ReadWholeSequence(String queryProtein) {
		final long startTime = System.currentTimeMillis();
		String com = "SELECT JobID, Predictions FROM ProteinKeyspace.ProteinRow WHERE Protein = '" + queryProtein + "';";
		System.out.println("Command: " + com);
		ResultSet results = session.execute(com);
		if (results.isExhausted())
			return null;
		final long queryTime = System.currentTimeMillis();
		List<Row> rows = results.all();
		System.out.println ("Query time is " + (queryTime - startTime) + " msec");   
		System.out.println (" rows analysed,  " + rows.size());
		List<StructureProteinPrediction> res = new ArrayList<StructureProteinPrediction>();
		int c = 0;
		for (Row r : rows) {
			StructureProteinPrediction structure = new StructureProteinPrediction(queryProtein, r.getString("JobID"), r.getMap("Predictions", String.class, String.class));		
			res.add(structure);
			++c;
		}
		final long endTime = System.currentTimeMillis();
		System.out.println (c + " rows analysed, execution time is " + (endTime - startTime) + " msec");
		return res;
	}
	
	/*
	 * getting part of protein sequence from the db ProteinRow
	 */
	public List<StructureProteinPrediction>  ReadPartOfSequence(String queryProtein) {
		final long startTime = System.currentTimeMillis();
		String com = "SELECT * FROM ProteinKeyspace.ProteinRow;";
		System.out.println("Command: " + com);
		ResultSet results = session.execute(com);
		if (results.isExhausted())
			return null;
		final long queryTime = System.currentTimeMillis();
		List<Row> rows = results.all();
		System.out.println ("Query time is " + (queryTime - startTime) + " msec");   
		System.out.println (" rows analysed,  " + rows.size());
		List<StructureProteinPrediction>  res = new ArrayList<StructureProteinPrediction>();
		int c = 0;
		for (Row r : rows) {
			String prot = r.getString("Protein");
			if (prot.matches("(.*)" + queryProtein + "(.*)")) {
			//	System.out.println(prot);
				StructureProteinPrediction structure = new StructureProteinPrediction(prot, r.getString("JobID"), r.getMap("Predictions", String.class, String.class));		
				res.add(structure);
				++c;
			}
		}
		final long endTime = System.currentTimeMillis();
		System.out.println (c + " rows analysed, execution time is " + (endTime - startTime) + " msec");
		return res;
	}
	
	/*
	 * getting protein sequences by counter
	 */
	public List<Pair<String, Integer>>  ReadProteinDataByCounter(int counter) {
		final long startTime = System.currentTimeMillis();
		String com = "SELECT DISTINCT Protein FROM ProteinKeyspace.ProteinRow;";
		System.out.println("Command: " + com);
		ResultSet results = session.execute(com);
		if (results.isExhausted())
			return null;
		final long queryTime = System.currentTimeMillis();
		List<Row> rows = results.all();
		System.out.println ("Query time is " + (queryTime - startTime) + " msec");   
		System.out.println (" rows analysed,  " + rows.size());
		List<Pair<String, Integer>>  res = new ArrayList<Pair<String, Integer>>();
		int c = 0;
		for (Row r : rows) {
			String prot = r.getString("Protein");
			
		}
		final long endTime = System.currentTimeMillis();
		System.out.println (c + " rows analysed, execution time is " + (endTime - startTime) + " msec");
		return res;
	}
	
	/*
	 * getting earlest date of jobs from the db
	 */
	public long getEarliestDateInDB() {
		final long startTime = System.currentTimeMillis();
		String com = "SELECT jobtime,JobID FROM ProteinKeyspace.ProteinData;";
		System.out.println("Command: " + com);
		ResultSet results = session.execute(com);
		final long queryTime = System.currentTimeMillis();
		System.out.println ("Query time is  " + (queryTime - startTime) + " msec");

		Calendar cal = Calendar.getInstance();
		long res = cal.getTimeInMillis();
		int c = 0;
		while (!results.isExhausted()) {
			Row r = results.one();
			long d1 = r.getLong("jobtime");
			if (res > d1) {
				res = d1;
			}
			++c;
		}
		final long endTime = System.currentTimeMillis();
		System.out.println (c + " rows analysed, execution time is " + (endTime - startTime) + " msec");
		return res;
	}
	
}