JAVA maven 编写UDF适用于hive和impala

hive 内置函数很少，我们可以通过自定义的方式添加新的UDF上去，来增强hive的处理能力。

比如hive没有字符串包含的UDF.

我们通过Java+maven的方式来编写一个字符串包含的UDF

1、新建maven工程

2、修改pom.xml

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
	xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
	<modelVersion>4.0.0</modelVersion>
	<groupId>com.lr.udf</groupId>
	<artifactId>common.udf</artifactId>
	<version>0.0.1-SNAPSHOT</version>
	<properties>
		<java.source.version>1.8</java.source.version>
		<java.target.version>1.8</java.target.version>
		<maven.compiler.plugin.version>3.3</maven.compiler.plugin.version>
		<file.encoding>UTF-8</file.encoding>
		<hadoop.version>2.6.0</hadoop.version>
	</properties>


	<dependencies>
		<dependency>
			<groupId>org.apache.hadoop</groupId>
			<artifactId>hadoop-client</artifactId>
			<version>2.7.3</version>
		</dependency>
		<dependency>
			<groupId>org.apache.hive</groupId>
			<artifactId>hive-exec</artifactId>
			<version>1.2.1</version>
			<exclusions>
				<exclusion>
					<groupId>org.pentaho</groupId>
					<artifactId>pentaho-aggdesigner-algorithm</artifactId>
				</exclusion>
			</exclusions>
		</dependency>
	</dependencies>
	
	<build>
		<finalName>${project.artifactId}</finalName>
	</build>
</project>

3、编写UDF类

package common.udf;


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


/**
 * 需要继承UDF
 * @author liurui
 *
 */
public class StringUDF extends UDF{


	/**
	 * 重写evaluate方法，参数可以任意
	 * @param orgStr
	 * @param dest
	 * @return
	 */
	public Integer evaluate(String orgStr,String dest){
		if(orgStr.contains(dest)){
			return 1;
		}
		return 0;
	}
	
	public static void main(String[] args) {
		StringUDF udf = new StringUDF();
		System.out.println(udf.evaluate("aaabc", "b"));
	}
}

4、打包

mvn clean package

5、上传到hdfs

hdfs dfs -put common.udf.jar /tmp/test/common.udf.jar

[root@bigdata-cdh001 home]# ll
total 4
drwx------. 2 centos-user centos-user   62 Aug 25  2017 centos-user
-rw-r--r--  1 root        root        2518 May 22 18:56 common.udf.jar

[root@bigdata-cdh001 home]# hadoop dfs -ls /tmp/test
DEPRECATED: Use of this script to execute hdfs command is deprecated.
Instead use the hdfs command for it.
Found 1 items
drwxr-xr-x   - hdfs supergroup          0 2018-04-10 15:11 /tmp/test/8a18f7a6b63a4b8180b242e448cea705

[root@bigdata-cdh001 home]# hdfs dfs -put common.udf.jar /tmp/test/common.udf.jar
[root@bigdata-cdh001 home]# hadoop dfs -ls /tmp/test
DEPRECATED: Use of this script to execute hdfs command is deprecated.
Instead use the hdfs command for it.
Found 2 items
drwxr-xr-x   - hdfs supergroup          0 2018-04-10 15:11 /tmp/test/8a18f7a6b63a4b8180b242e448cea705
-rw-r--r--   3 root supergroup       2819 2018-05-21 22:34 /tmp/test/csot-hive-udf.jar

6、将UDF添加到hive

#将jar添加到hive

add jar hdfs:///tmp/test/common.udf.jar

#创建临时的UDF

create temporary function my_contains as 'common.udf.StringUDF';

#创建永久的只需要去掉temporary

[root@bigdata-cdh001 home]# hive
Java HotSpot(TM) 64-Bit Server VM warning: ignoring option MaxPermSize=512M; support was removed in 8.0
Java HotSpot(TM) 64-Bit Server VM warning: ignoring option MaxPermSize=512M; support was removed in 8.0

Logging initialized using configuration in jar:file:/opt/cloudera/parcels/CDH-5.10.1-1.cdh5.10.1.p0.10/jars/hive-common-1.1.0-cdh5.10.1.jar!/hive-log4j.properties
WARNING: Hive CLI is deprecated and migration to Beeline is recommended.
hive> add jar hdfs:///tmp/test/common.udf.jar;
converting to local hdfs:///tmp/test/common.udf.jar
Added [/tmp/6eaaaad7-1446-4bed-9543-0a02d901e8bc_resources/common.udf.jar] to class path
Added resources: [hdfs:///tmp/test/common.udf.jar]
hive> create temporary function my_contains as 'common.udf.StringUDF';
OK
Time taken: 1.692 seconds
hive> 
    > SELECT * FROM default.i_f_task_status where my_contains(task_id,'c')=0 LIMIT 5;
OK
1	100
1	100
1	100
1	100
1	100
Time taken: 0.109 seconds, Fetched: 5 row(s)
hive> 

7、查询task_id不包含字符c的 

my_contains(task_id,'c')=0 不包含  my_contains(task_id,'c')=1 包含

SELECT * FROM default.i_f_task_status where my_contains(task_id,'c')=0 LIMIT 5;

在hive添加的udf 在impala是不支持的，所以需要另外再加

如果是impala，创建函数的时候需要指定参数和返回值类型，如下：

则在impala-shell或者hue的impala上执行

create function csot_contains(string,string) returns int location 'hdfs:///tmp/test/common.udf.jar' symbol='cn.com.cloudstar.udf.StringUDF';

[bigdata-cdh001:21000] > create function my_contains(string,string) returns int location 'hdfs:///tmp/test/common.udf.jar' symbol='common.udf.StringUDF';
Query: create function my_contains(string,string) returns int location 'hdfs:///tmp/test/common.udf.jar' symbol='common.udf.StringUDF'

Fetched 0 row(s) in 0.07s
[bigdata-cdh001:21000] > 

[bigdata-cdh001:21000] > SELECT * FROM default.i_f_task_status where my_contains(task_id,'c')=0 LIMIT 5;
Query: select * FROM default.i_f_task_status where my_contains(task_id,'c')=0 LIMIT 5
Query submitted at: 2018-05-22 19:19:09 (Coordinator: http://bigdata-cdh001:25000)
Query progress can be monitored at: http://bigdata-cdh001:25000/query_plan?query_id=34f236d41e4812c:9f1e208500000000
+---------+---------+
| process | task_id |
+---------+---------+
| 1       | 100     |
| 1       | 100     |
| 1       | 100     |
| 1       | 100     |
| 1       | 100     |
+---------+---------+
Fetched 5 row(s) in 4.55s
[bigdata-cdh001:21000] > 

源码：https://gitee.com/qianqianjie/common.udf.git