Hive+UDTF简单示例

         UDTF(User-Defined Table-Generating Functions)  用来解决输入一行输出多行(On-to-many maping) 的需求。

         我们来看一个简单的例子。

         现在有一个名为studentScore.txt的文本,里面的内容如下：

                        A       90
                        A       80
                        A       70
                        A       50
                        B       60
                        B       90
                        B       95
                        B       80

         我们 要统计 A和B的总分。

          1）将studentScore.txt上传至hdfs.

                 bin/hadoop fs -put /home/wangzhun/tmp/studentScore.txt wz

          2)  在hive里面，创建studentScore表,并导入 数据

                 create table studentScore(name string,score int) ROW FORMAT DELIMITED Fields TERMINATED BY '\t';

                 load data inpath '/user/root/wz/studentScore.txt' overwrite into table studentScore;               

          3）编写UDTF。

                

package com.wz.udf;
import java.util.ArrayList;

import org.apache.hadoop.hive.ql.udf.generic.GenericUDTF;
import org.apache.hadoop.hive.ql.exec.UDFArgumentException;
import org.apache.hadoop.hive.ql.exec.UDFArgumentLengthException;
import org.apache.hadoop.hive.ql.metadata.HiveException;
import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory;
import org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory;

public class staticScore extends GenericUDTF{
  Integer nTotalScore = Integer.valueOf(0);       //总分
  Object forwardObj[] = new Object[1];
  String strStudent;       //学生姓名
  @Override
  public void close() throws HiveException {  
        //输出最后一个学生的总分
	forwardObj[0]=(strStudent+":"+String.valueOf(nTotalScore));
	forward(forwardObj);
    }

  
  @Override
  public StructObjectInspector initialize(ObjectInspector[] args)throws UDFArgumentException {
          strStudent="";
	  ArrayList<String> fieldNames = new ArrayList<String>();
          ArrayList<ObjectInspector> fieldOIs = new ArrayList<ObjectInspector>();
          fieldNames.add("col1");
          fieldOIs.add(PrimitiveObjectInspectorFactory.javaStringObjectInspector);
          return ObjectInspectorFactory.getStandardStructObjectInspector(fieldNames,fieldOIs);
       }


  @Override
  public void process(Object[] args) throws HiveException {
	if(!strStudent.isEmpty() && !strStudent.equals(args[0].toString()))
	{
           //当学生名字变化时，输出该学生的总分
	   String[] newRes = new String[1];
	   newRes[0]=(strStudent+":"+String.valueOf(nTotalScore));
	   forward(newRes);
	   nTotalScore=0;
	}	
	strStudent=args[0].toString();
	nTotalScore+=Integer.parseInt(args[1].toString());
       }

   }

          4）打包编译成jar包.

                 javac -classpath /home/wangzhun/hadoop/hadoop-0.20.2/hadoop-0.20.2-core.jar:/home/wangzhun/hive/hive-0.8.1/lib/hive-exec-0.8.1.jar staticScore.java

                 jar -cvf staticScore.jar com/wz/udf/staticScore.class

           5) 在hive里面添加jar包，创建临时函数，并执行得到结果。

                add jar /home/wangzhun/hive/hive-0.8.1/lib/staticScore.jar;
                create temporary function statics as 'com.wz.udf.staticScore';
               

hive> select statics(studentScore.name,studentScore.score) as col1 from studentSselect statics(studentScore.name,studentScore.score) mytable as col1 from studencreate temporary function statics as 'com.wz.udf.staticScore';       
OK
Time taken: 0.213 seconds
hive> select statics(studentScore.name,studentScore.score) as col1 from studentScore;
Total MapReduce jobs = 1
Launching Job 1 out of 1
Number of reduce tasks is set to 0 since there's no reduce operator
Starting Job = job_201311282251_0001, Tracking URL = http://localhost:50030/jobdetails.jsp?jobid=job_201311282251_0001
Kill Command = /home/wangzhun/hadoop/hadoop-0.20.2/bin/../bin/hadoop job  -Dmapred.job.tracker=localhost:9001 -kill job_201311282251_0001
Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 0
2013-11-28 23:02:24,612 Stage-1 map = 0%,  reduce = 0%
2013-11-28 23:02:30,689 Stage-1 map = 100%,  reduce = 0%
2013-11-28 23:02:33,732 Stage-1 map = 100%,  reduce = 100%
Ended Job = job_201311282251_0001
MapReduce Jobs Launched: 
Job 0: Map: 1   HDFS Read: 40 HDFS Write: 12 SUCESS
Total MapReduce CPU Time Spent: 0 msec
OK
A:290
B:325
Time taken: 34.356 seconds