org.apache.hadoop.mapred.lib.db.DBOutputFormat.java Source code

Java tutorial

Introduction

Here is the source code for org.apache.hadoop.mapred.lib.db.DBOutputFormat.java

Source

/**
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package org.apache.hadoop.mapred.lib.db;

import java.io.IOException;
import java.sql.Connection;
import java.sql.PreparedStatement;
import java.sql.SQLException;

import org.apache.hadoop.classification.InterfaceAudience;
import org.apache.hadoop.classification.InterfaceStability;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.mapred.JobConf;
import org.apache.hadoop.mapred.OutputFormat;
import org.apache.hadoop.mapred.RecordWriter;
import org.apache.hadoop.mapred.Reporter;
import org.apache.hadoop.mapreduce.MRJobConfig;
import org.apache.hadoop.mapreduce.TaskAttemptID;
import org.apache.hadoop.mapreduce.task.TaskAttemptContextImpl;
import org.apache.hadoop.util.Progressable;

@InterfaceAudience.Public
@InterfaceStability.Stable
public class DBOutputFormat<K extends DBWritable, V> extends org.apache.hadoop.mapreduce.lib.db.DBOutputFormat<K, V>
        implements OutputFormat<K, V> {

    /**
     * A RecordWriter that writes the reduce output to a SQL table
     */
    protected class DBRecordWriter extends org.apache.hadoop.mapreduce.lib.db.DBOutputFormat<K, V>.DBRecordWriter
            implements RecordWriter<K, V> {

        protected DBRecordWriter(Connection connection, PreparedStatement statement) throws SQLException {
            super(connection, statement);
        }

        /** {@inheritDoc} */
        public void close(Reporter reporter) throws IOException {
            super.close(null);
        }
    }

    /** {@inheritDoc} */
    public void checkOutputSpecs(FileSystem filesystem, JobConf job) throws IOException {
    }

    /** {@inheritDoc} */
    public RecordWriter<K, V> getRecordWriter(FileSystem filesystem, JobConf job, String name,
            Progressable progress) throws IOException {
        org.apache.hadoop.mapreduce.RecordWriter<K, V> w = super.getRecordWriter(
                new TaskAttemptContextImpl(job, TaskAttemptID.forName(job.get(MRJobConfig.TASK_ATTEMPT_ID))));
        org.apache.hadoop.mapreduce.lib.db.DBOutputFormat.DBRecordWriter writer = (org.apache.hadoop.mapreduce.lib.db.DBOutputFormat.DBRecordWriter) w;
        try {
            return new DBRecordWriter(writer.getConnection(), writer.getStatement());
        } catch (SQLException se) {
            throw new IOException(se);
        }
    }

    /**
     * Initializes the reduce-part of the job with the appropriate output settings
     * 
     * @param job The job
     * @param tableName The table to insert data into
     * @param fieldNames The field names in the table.
     */
    public static void setOutput(JobConf job, String tableName, String... fieldNames) {
        if (fieldNames.length > 0 && fieldNames[0] != null) {
            DBConfiguration dbConf = setOutput(job, tableName);
            dbConf.setOutputFieldNames(fieldNames);
        } else {
            if (fieldNames.length > 0)
                setOutput(job, tableName, fieldNames.length);
            else
                throw new IllegalArgumentException("Field names must be greater than 0");
        }
    }

    /**
     * Initializes the reduce-part of the job with the appropriate output settings
     * 
     * @param job The job
     * @param tableName The table to insert data into
     * @param fieldCount the number of fields in the table.
     */
    public static void setOutput(JobConf job, String tableName, int fieldCount) {
        DBConfiguration dbConf = setOutput(job, tableName);
        dbConf.setOutputFieldCount(fieldCount);
    }

    private static DBConfiguration setOutput(JobConf job, String tableName) {
        job.setOutputFormat(DBOutputFormat.class);
        job.setReduceSpeculativeExecution(false);

        DBConfiguration dbConf = new DBConfiguration(job);

        dbConf.setOutputTableName(tableName);
        return dbConf;
    }

}