Cassandraにデータをいれ、ARFFで書き出し、Wekaで分析する－その２ソース

１月２３日、「ビッグデータ管理入門」の後半を、ＮＩＩで聞いてきた
その内容（＝表題のことをする）をメモメモのつづき

今回は、Cassandraアクセス。
hectorを使ってアクセスしている。
今回は、いただいたソースコードを貼ってみる。
（説明は次回）

import java.util.Iterator;

import me.prettyprint.cassandra.serializers.IntegerSerializer;
import me.prettyprint.cassandra.serializers.StringSerializer;
import me.prettyprint.cassandra.service.ColumnSliceIterator;
import me.prettyprint.cassandra.service.KeyIterator;
import me.prettyprint.cassandra.service.ThriftKsDef;
import me.prettyprint.cassandra.service.template.ColumnFamilyResult;
import me.prettyprint.cassandra.service.template.ColumnFamilyTemplate;
import me.prettyprint.cassandra.service.template.ColumnFamilyUpdater;
import me.prettyprint.cassandra.service.template.ThriftColumnFamilyTemplate;
import me.prettyprint.hector.api.Cluster;
import me.prettyprint.hector.api.Keyspace;
import me.prettyprint.hector.api.beans.HColumn;
import me.prettyprint.hector.api.ddl.ColumnFamilyDefinition;
import me.prettyprint.hector.api.ddl.ComparatorType;
import me.prettyprint.hector.api.ddl.KeyspaceDefinition;
import me.prettyprint.hector.api.factory.HFactory;
import me.prettyprint.hector.api.query.SliceQuery;

/**
 * Hello world!
 *
 */
public class Sample {
    public static String CLUSTER_NAME = "Test Cluster";
    public static String DB_SERVER = "127.0.0.1";
    public static String KEYSPACE_NAME = "MyKeyspace";
    public static String COLUMN_FAMILY_NAME = "AccessLog";

    public static String LOG_FILE = "access.log";

    private Cluster cluster;
    private KeyspaceDefinition keyspaceDef;
    private Keyspace keyspace;
    private ColumnFamilyTemplate<Integer, String> template;

    public static void main(String[] args) {
        Sample app = new Sample();

//        app.dropKeyspace();
        app.prepare();

        app.insertSample();
        app.findAllSample();
        app.updateSample();
        app.deleteSample();
        app.findAllKeys();
    }
    
    public Sample() {
        cluster = HFactory.getOrCreateCluster(CLUSTER_NAME, DB_SERVER + ":9160");
    }

    private void prepare() {
        keyspaceDef = cluster.describeKeyspace(KEYSPACE_NAME);
        if (keyspaceDef == null) {
            KeyspaceDefinition newKeyspace = HFactory.createKeyspaceDefinition(KEYSPACE_NAME, ThriftKsDef.DEF_STRATEGY_CLASS, 1, null);
            cluster.addKeyspace(newKeyspace, true);

            ColumnFamilyDefinition cfDef = HFactory.createColumnFamilyDefinition(KEYSPACE_NAME, COLUMN_FAMILY_NAME, ComparatorType.BYTESTYPE);
            cluster.addColumnFamily(cfDef);
        }
        keyspace = HFactory.createKeyspace(KEYSPACE_NAME, cluster);

        template = new ThriftColumnFamilyTemplate<Integer, String>(keyspace, COLUMN_FAMILY_NAME, IntegerSerializer.get(), StringSerializer.get());
    }

    private void dropKeyspace() {
        cluster.dropKeyspace(KEYSPACE_NAME);
    }

    private void insertSample() {
        System.out.println("----- insertSample -----");
        for (int i = 0; i < 10; ++i) {
            ColumnFamilyUpdater<Integer, String> updater = template.createUpdater(i);
            updater.setString("c1", "Hello" + i);
            updater.setString("c2", "World" + i * 2);
            updater.setString("c3", "!!" + i * i);
            template.update(updater);
            System.out.println("Inserted : " + i);
        }
    }

    private void updateSample() {
        System.out.println("----- updateSample -----");
        for (int i = 0; i < 5; ++i) {
            int id = (int) (Math.random() * 10);
            ColumnFamilyUpdater<Integer, String> updater = template.createUpdater(id);
            updater.setString("c3", "!!!" + id * id * id);
            template.update(updater);
            System.out.println("Updated : " + id);
        }
    }

    private void deleteSample() {
        System.out.println("----- deleteSample -----");

	//	ランダムに３つc3消す
        for (int i = 0; i < 3; ++i) {
            int id = (int) (Math.random() * 10);
            template.deleteColumn(id, "c3");
            System.out.println("Deleted : " + id + ".c3");
        }

	//	ランダムに１つ消す
        int id = (int) (Math.random() * 10);
        template.deleteRow(id);
        System.out.println("Deleted : " + id);
    }

    private void findAllSample() {
        System.out.println("----- findAllSample1 -----");
        for (int i = 0; i < 10; ++i) {
            int id = i;
            ColumnFamilyResult<Integer, String> res = template.queryColumns(id);
            String value = "" + res.getKey() + " : " + "c1=" + res.getString("c1") + " " + "c3=" + res.getString("c3");
            System.out.println(value);
        }
    }

    private void findAllKeys() {
        System.out.println("----- findAllKeys -----");
        @SuppressWarnings("deprecation")
        KeyIterator<Integer> keyIterator = new KeyIterator<Integer>(keyspace, COLUMN_FAMILY_NAME, IntegerSerializer.get());
        Iterator<Integer> itr = keyIterator.iterator();
        while (itr.hasNext()) {
            Integer key = itr.next();
            System.out.println(key);
        }
    }
}

とりあえず、今日はここまで。
あしたから、説明する。