mapreduce实现结构化查询(求最大值,求最小值&&求和)

求最大值

package com.panther.max;

import com.panther.util.HadoopLogger;
import com.panther.util.Util;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.util.Tool;
import org.slf4j.Logger;

import java.io.IOException;

/**
 * Created by panther on 15-6-7.
 */
public class MAX extends Configured implements Tool {
    private static final Logger LOG = HadoopLogger.getLog(MAX.class);
    private static final String JOB_NAME = "max job";
    private static final String OUTPUT_NAME = "max = ";

    private static int index = Util.getIndex();

    private static MAX max;

    public static MAX getMax() {
        if (max == null) {
            max = new MAX();
        }
        return max;
    }

    public static class MaxMapper extends Mapper<LongWritable, Text, Text, Text> {
        @Override protected void map(LongWritable key, Text value, Context context) {
            Util.getUtil().getResult(value, index, context);
        }
    }

    public static class MaxReducer extends Reducer<Text, Text, Text, Text> {
        private float max = 0.0f;

        @Override protected void reduce(Text key, Iterable<Text> values, Context context)
                throws IOException, InterruptedException {
            for (Text value : values) {
                LOG.info(value.toString());
                float tmp = Float.parseFloat(value.toString());
                if (max <= tmp) {
                    max = tmp;
                }
            }
            context.write(new Text(OUTPUT_NAME + max), new Text(""));
        }

    }

    @Override public int run(String[] args) throws Exception {
        Configuration conf = getConf(); //获得配置文件对象
        Job job = Util.getUtil().getJob(conf, JOB_NAME, MAX.class, args, MaxMapper.class, MaxReducer.class);
        job.waitForCompletion(true);
        return job.isSuccessful() ? 0 : 1;
    }
}

求最小值
package com.panther.min;

import com.panther.util.HadoopLogger;
import com.panther.util.Util;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.util.Tool;
import org.slf4j.Logger;

import java.io.IOException;

/**
 * Created by panther on 15-6-7.
 */
public class MIN extends Configured implements Tool {
    private static final Logger LOG = HadoopLogger.getLog(MIN.class);
    private static final String JOB_NAME = "min job";
    private static final String OUTPUT_NAME = "min = ";

    private static int index = Util.getIndex();

    private static MIN min;

    public static MIN getMin() {
        if (min == null) {
            min = new MIN();
        }
        return min;
    }

    public static class MinMapper extends Mapper<LongWritable, Text, Text, Text> {
        private Text out = new Text("");

        @Override protected void map(LongWritable key, Text value, Context context) {
            Util.getUtil().getResult(value, index, context);
        }
    }

    public static class MinReducer extends Reducer<Text, Text, Text, Text> {
        private float min = Float.MAX_VALUE;

        @Override protected void reduce(Text key, Iterable<Text> values, Context context)
                throws IOException, InterruptedException {
            for (Text value : values) {
                LOG.info(value.toString());
                float tmp = Float.parseFloat(value.toString());
                if (min >= tmp) {
                    min = tmp;
                }

            }
            context.write(new Text(OUTPUT_NAME + min), new Text(""));
        }

    }

    @Override public int run(String[] args) throws Exception {
        Configuration conf = getConf(); //获得配置文件对象
        Job job = Util.getUtil().getJob(conf, JOB_NAME, MIN.class, args, MinMapper.class, MinReducer.class);
        job.waitForCompletion(true);
        return job.isSuccessful() ? 0 : 1;
    }
}

求和:
package com.panther.sum;

import com.panther.util.HadoopLogger;
import com.panther.util.Util;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.util.Tool;
import org.slf4j.Logger;

import java.io.IOException;

/**
 * Created by panther on 15-6-5.
 */
public class SUM extends Configured implements Tool {
    private static final Logger LOG = HadoopLogger.getLog(SUM.class);

    private static final String JOB_NAME = "sum job";
    private static final String OUTPUT_NAME = "sum = ";

    private static int index = Util.getIndex();

    private static SUM sum;

    public static SUM getSum() {
        if (sum == null) {
            sum = new SUM();
        }
        return sum;
    }

    public static void setIndex(int index) {
        SUM.index = index;
    }

    public static class SumMapper extends Mapper<LongWritable, Text, Text, Text> {
        private Text out = new Text("");

        @Override protected void map(LongWritable key, Text value, Context context) {
            Util.getUtil().getResult(value, index, context);
        }
    }

    public static class SumReducer extends Reducer<Text, Text, Text, Text> {
        private float sum = 0.0f;

        @Override protected void reduce(Text key, Iterable<Text> values, Context context)
                throws IOException, InterruptedException {
            for (Text text : values) {
                LOG.info(text.toString());
                sum += Float.parseFloat(text.toString());
            }
            context.write(new Text(OUTPUT_NAME + sum), new Text(""));
        }

    }

    @Override public int run(String[] args) throws Exception {
        Configuration conf = getConf(); //获得配置文件对象
        Job job = Util.getUtil().getJob(conf, JOB_NAME, SUM.class, args, SumMapper.class, SumReducer.class);
        job.waitForCompletion(true);
        return job.isSuccessful() ? 0 : 1;
    }

}

通用日志操作

package com.panther.util;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;

/**
 * Created by panther on 15-6-7.
 */
public class HadoopLogger {
    private static Map<Class<?>, Logger> loggerMap = new ConcurrentHashMap<Class<?>, Logger>();

    public static Logger getLog(Class<?> cls) {
        Logger logger = loggerMap.get(cls);
        if (logger == null) {
            logger = LoggerFactory.getLogger(cls);
            loggerMap.put(cls, logger);
        }
        return logger;
    }
}

公用代码提取

package com.panther.util;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.text.SimpleDateFormat;
import java.util.Date;

/**
 * Created by panther on 15-6-5.
 */
public class Util {
    private static final Logger LOG = LoggerFactory.getLogger(Util.class);
    private static final String DATEFORMAT = "yyyy-MM-dd-HH-mm-ss";
    private static int index = 3;

    private static Util util = null;

    public static Util getUtil() {
        if (util == null) {
            util = new Util();
        }
        return util;
    }

    public static int getIndex() {
        return index;
    }

    /**
     * 分割字符串
     *
     * @param value
     * @return
     */
    public static String[] getSplit(Text value) {
        String regex = "[\\|,]{1}";
        return value.toString().split(regex);
    }

    /**
     * 过滤出每列
     *
     * @param value
     * @param index
     * @param context
     */
    public void getResult(Text value, int index, Mapper.Context context) {
        LOG.info("value = {}", value);
        String[] result = Util.getSplit(value);
        if (result.length > index + 1) {
            LOG.info(result[index]);
            try {
                context.write(new Text(""), new Text(result[index]));
            } catch (Exception e) {
                LOG.error("sum aggregation failure exception is : {}", e);
            } finally {
                LOG.info("index = {},result = {}", index, result[index]);
            }
        }
    }

    /**
     * 初始化job参数
     *
     * @param conf
     * @param jobName
     * @param jarCls
     * @param args
     * @param mapper
     * @param reducer
     * @return
     * @throws Exception
     */
    public Job getJob(Configuration conf, String jobName, Class<?> jarCls, String[] args,
            Class<? extends Mapper> mapper, Class<? extends Reducer> reducer) throws Exception {
        Job job = null;
        job = new Job(conf, jobName);
        job.setJarByClass(jarCls);
        FileInputFormat.addInputPath(job, new Path(args[0])); //设置map输入文件路径
        FileOutputFormat
                .setOutputPath(job, new Path(args[1] + jobName + Util.getUtil().getCurrentDate())); //设置reduce输出文件路径
        job.setMapperClass(mapper);
        job.setReducerClass(reducer);

        job.setInputFormatClass(TextInputFormat.class); //设置文件输入格式
        job.setOutputFormatClass(TextOutputFormat.class);//使用默认的output格格式
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(Text.class);

        //设置reduce的输出key和value类型
        job.setOutputKeyClass(Text.class);
        job.setOutputValueClass(Text.class);
        return job;
    }

    /**
     * 得到当前时间
     *
     * @return
     */
    public String getCurrentDate() {
        Date mData = new Date();
        SimpleDateFormat sdf = new SimpleDateFormat(DATEFORMAT);
        return sdf.format(mData).toString();
    }

    public void getColumn(String string) {
        int column = -1;
        for (int i = 0; i < string.length(); ++i) {
            if (string.charAt(i) > '0' && string.charAt(i) <= '9') {
                column = Integer.parseInt("" + string.charAt(i)) - 1;
            }
        }
        if (column != -1) {
            index = column;
        }
    }
}

结果输出：

package com.panther.result;

import com.panther.max.MAX;
import com.panther.min.MIN;
import com.panther.sum.SUM;
import com.panther.util.Util;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

import java.util.Scanner;

/**
 * Created by panther on 15-6-7.
 */
public class Result {
    private static final String VALUE_MAX = "MAX";
    private static final String VALUE_MIN = "MIN";
    private static final String VALUE_SUM = "SUM";
    private static final String MESSAGE_INFO = "please input command:";

    public static void main(String[] args) throws Exception {
        int res = 0;
        while (true) {
            System.out.println(MESSAGE_INFO);
            Scanner in = new Scanner(System.in);        //使用Scanner类定义对象
            String s = in.nextLine();
            System.out.println(s);
            Tool tool = null;
            Util.getUtil().getColumn(s.toUpperCase());
            if (s.toUpperCase().contains(VALUE_MAX)) {
                tool = MAX.getMax();
            } else if (s.toUpperCase().contains(VALUE_MIN)) {
                tool = MIN.getMin();
            } else if (s.toUpperCase().contains(VALUE_SUM)) {
                tool = SUM.getSum();
            }
            if (tool != null) {
                res = ToolRunner.run(new Configuration(), tool, args);
            }
        }
    }
}