hive java编写udf函数

时间:2024-05-03 11:04:26

(一)创建JAVA 代码--例子

package hiveOpt;

import org.apache.hadoop.hive.ql.exec.UDF;
import org.apache.hadoop.io.Text;

public class splitMax extends UDF {

public static int getMax(String[] arr) {
int max = Integer.valueOf(arr[0]);
for (int i = 1; i < arr.length; i++) {
if (Integer.valueOf(arr[i]) > max) {
max = Integer.valueOf(arr[i]);
}
}
return max;
}

public Text evaluate(String mdimlr)

{
Text result = new Text();
String[] a = mdimlr.split(",");

result.set(String.valueOf(getMax(a)));
return result;

}

}

(二)上传jar到hdfs

hadoop fs -put /data/soft/splitMax.jar /tmp

(三)登录hive命令行加载jar

hive> add jar hdfs:///tmp/splitMax.jar;
hive> create temporary function splitmax as 'hiveOpt.splitMax';
hive> show funcations
hive> select splitmax(id) from wobcongfu;

( 四)测试自定义函数
select phone_num,max(myCol1)
from temp_qiu_test1
LATERAL VIEW explode(split(col1,',')) myTable1 AS myCol1
group by phone_num;