MapReduce（4）

一、处理多个文件
求每个同学每科成绩的总分
chinese.txt

english.txt

math.txt

1.新建 path : score , 上传源文件

Mapper.java

package com.study.score.day01;

import java.io.IOException;

import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileSplit;

/**
 * 统计每个人三科的每科各月的总成绩
 * key : 姓名
 * value : student
 * Map : 映射数据
 * 
 * Mapper 数量 = 切片的数量 
 */
public class ScoreMapper extends Mapper<LongWritable, Text, Text, Student> {

	@Override
	protected void map(LongWritable key, Text value, Mapper<LongWritable, Text, Text, Student>.Context context)
			throws IOException, InterruptedException {

		// 文件名称，根据文件名称判断成绩是哪个科目的
		FileSplit split = (FileSplit) context.getInputSplit();
		String textName = split.getPath().getName();
		String [] textNames = textName.split("\\.");
		textName = textNames[0];
		// 每行的内容
		// 1 zhang 89  月份 姓名 成绩
		String lineContent = value.toString() ;
		// 
		String [] datas = lineContent.split(" ");
		String name = datas[1];
		String score = datas[2];
		
		Student student = new Student();
		student.setName(name);
		if("chinese".equals(textName)){
			student.setChinese(Integer.valueOf(score));
			student.setEnglish(0);
			student.setMath(0);
		}else if("english".equals(textName)){
			student.setEnglish(Integer.valueOf(score));
			student.setMath(0);
			student.setChinese(0);
		}else if("math".equals(textName)){
			student.setMath(Integer.valueOf(score));
			student.setChinese(0);
			student.setEnglish(0);
		}
		context.write(new Text(name), student);
	}
}

2.Reducer.java

package com.study.score.day01;

import java.io.IOException;

import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Reducer;

public class ScoreReducer extends Reducer<Text, Student, Text, Student> {

	@Override
	protected void reduce(Text key, Iterable<Student> values, Reducer<Text, Student, Text, Student>.Context context)
			throws IOException, InterruptedException {

		Student student = new Student();
		student.setName(key.toString());
		
		Integer chinese = 0;
		Integer english = 0 ;
		Integer math = 0 ;
		for(Student stu : values){
			chinese = chinese + stu.getChinese();
			english = english + stu.getEnglish();
			math = math + stu.getMath();
		}
		student.setChinese(chinese);
		student.setEnglish(english);
		student.setMath(math);
		
		context.write(key, student);
	}
}

3.Student.java

package com.study.score.day01;

import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;

import org.apache.hadoop.io.Writable;

public class Student implements Writable{

	private String name ;
	private Integer chinese ;
	private Integer english ;
	private Integer math ;
	
	
	
	public String getName() {
		return name;
	}
	public void setName(String name) {
		this.name = name;
	}
	public Integer getChinese() {
		return chinese;
	}
	public void setChinese(Integer chinese) {
		this.chinese = chinese;
	}
	public Integer getEnglish() {
		return english;
	}
	public void setEnglish(Integer english) {
		this.english = english;
	}
	public Integer getMath() {
		return math;
	}
	public void setMath(Integer math) {
		this.math = math;
	}
	@Override
	public void readFields(DataInput input) throws IOException {
		this.name = input.readUTF();
		this.chinese = input.readInt();
		this.english = input.readInt();
		this.math = input.readInt();
	}
	@Override
	public void write(DataOutput output) throws IOException {
		output.writeUTF(name);
		output.writeInt(chinese);
		output.writeInt(english);
		output.writeInt(math);
	}
	@Override
	public String toString() {
		return "Student [name=" + name + ", chinese=" + chinese + ", english=" + english + ", math=" + math + "]";
	}
	
	
}

4.Driver.java

public class ScoreDriver {

	public static void main(String[] args) throws ClassNotFoundException, IOException, InterruptedException {

		Configuration conf = new Configuration();
		Job job = Job.getInstance(conf);
		
		job.setJarByClass(ScoreDriver.class);
		job.setMapperClass(ScoreMapper.class);
		job.setReducerClass(ScoreReducer.class);
		
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(Student.class);
		
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(Student.class);
	// 读取路径下的所有文件，此时 result 文件夹不存在	
		FileInputFormat.setInputPaths(job, new Path("hdfs://192.168.76.131:9000/score"));
		FileOutputFormat.setOutputPath(job, new Path("hdfs://192.168.76.131:9000/score/result"));
	
		job.waitForCompletion(true);
	}

}

总结：
利用 content.getInputFile().getPath().getName() 文件名称来区分成绩的归属，即是哪个科目的成绩

二、排序

备注：若使用插件上传文件后乱码，检查 Eclipse 工作空间的编码格式，设置为UTF-8与文件的编码格式相同

根据电影热度对电影排序
惊天破 72
机械师2 83
奇异博士 67
但丁密码 79
比利林恩的中场战事 84
侠探杰克:永不回头 68
龙珠Z:复活的弗利萨 79
长城 56


1.Po

package com.study.sort.day01;

import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;

import org.apache.hadoop.io.WritableComparable;

public class MovieBean implements WritableComparable<MovieBean>{

	private String name ;
	private Integer hotNum ;
	public String getName() {
		return name;
	}
	public void setName(String name) {
		this.name = name;
	}
	public Integer getHotNum() {
		return hotNum;
	}
	public void setHotNum(Integer hotNum) {
		this.hotNum = hotNum;
	}
	@Override
	public void readFields(DataInput input) throws IOException {
		this.name = input.readUTF();
		this.hotNum = input.readInt();
	}
	@Override
	public void write(DataOutput output) throws IOException {
		output.writeUTF(this.name);
		output.writeInt(this.hotNum);
	}
	@Override
	public String toString() {
		return "MovieBean [name=" + name + ", hotNum=" + hotNum + "]";
	}
// 降序排序：旧对象 - 当前对象
	@Override
	public int compareTo(MovieBean o) {
		return o.getHotNum() - this.getHotNum();
	}

}

2.Mapper

public class SortMapper extends Mapper<LongWritable, Text, MovieBean, NullWritable> {

	@Override
	protected void map(LongWritable key, Text value, Mapper<LongWritable, Text, MovieBean, NullWritable>.Context context)
			throws IOException, InterruptedException {
		String line = value.toString() ;
		String [] datas = line.split(" ");
		MovieBean movieBean = new MovieBean();
		movieBean.setName(datas[0]);
		movieBean.setHotNum(Integer.valueOf(datas[1]));
		
		context.write(movieBean, NullWritable.get());
	}
}

3.Driver

public class SortDriver {

	public static void main(String[] args) throws IllegalArgumentException, IOException, ClassNotFoundException, InterruptedException {

		Configuration conf = new Configuration();
		Job job = Job.getInstance(conf);
		
		job.setJarByClass(SortDriver.class);
		job.setMapperClass(SortMapper.class);
		
		job.setMapOutputKeyClass(MovieBean.class);
		job.setMapOutputValueClass(NullWritable.class);
		
		FileInputFormat.setInputPaths(job, new Path("hdfs://192.168.76.131:9000/sort"));
		FileOutputFormat.setOutputPath(job, new Path("hdfs://192.168.76.131:9000/sort/result"));
		
		job.waitForCompletion(true);
		
	}

}

重点：
PO 实现 比较接口，定义比较结果
Mapper 排序是根据KEY值进行排序的，所以 PO类作为KEY值

三、多层MR处理

在第一层MR处理基础上
添加第二个JOB处理第一个JOB的运行结果
例子：
计算每人3个月的总收入并排序

第一个MR：计算每人的总收入
第二个MR：按照收入进行排序Mapper

四、combine 减轻 reducer压力

统计单词数量
若多个mapper 读取一份文件，每个mapper 对文件处理的结果为
hello 1
hello 1
...

最后发送到reducer 上进行合并处理，若文件数量很大，则reducer的压力很大

所以，在 mapper的处理过程中进行预处理，先进行合并
再发送到 reducer 上进行二次合并，此时只是对两个数据进行合并

1.Mapper
读取数据，分割，发送

2.Combine extends Reducer

3.Reducer
接收 combine 处理后的数据

4.Driver
job.serCombineClass();

不需对 combine 的输出 key value 进行设置