hive UDF操作

创建java类(导入对应的依赖包)：

public
classHiveHello extendsUDF {
public
Text evaluate(Text in){
Stringresult="helloWorld"+in;
return
new Text(result);
}
}
打成jar:

hive> add jar /script/hive/HelloHive.jar;

创建表：

hive> create table udf_test(name string)row format delimited fields terminated by ' ' stored as textfile;导入数据：

hive> load data local inpath'/script/hive/udf.txt' overwrite into table udf_test;

创建临时函数：

hive> create temporary function syhello as'com.test.udf.HiveHello';

使用函数：

hive> select syhello(name) from udf_test;

使用命令方式来实现上面操作：

[root@hadoop02 hive]# vi execCommend.hql

add jar /script/hive/HelloHive.jar;

drop table IF EXISTS udf_test2;

create table IF NOT EXISTS udf_test2(name string,age int) row formatdelimited fields terminated by ' ' stored as textfile;

load data local inpath'/script/hive/udf_data.txt' overwrite into table udf_test2;

create temporary function sayHello as'com.test.udf.HiveHello';

select sayHello(name),age from udf_test2;

drop temporary function sayHello;

执行上面sql:

[root@hadoop02 bin]# hive -f/script/hive/execCommend.hql ;

Logging initialized using configuration injar:file:/usr/local/hive/lib/hive-common-1.2.1.jar!/hive-log4j.properties

Added [/script/hive/HelloHive.jar] to classpath

Added resources: [/script/hive/HelloHive.jar]

OK

Time taken: 4.123 seconds

OK

Time taken: 1.415 seconds

Loading data to table default.udf_test2

Table default.udf_test2 stats: [numFiles=1,numRows=0, totalSize=29, rawDataSize=0]

OK

Time taken: 1.459 seconds

OK

Time taken: 1.187 seconds

OK

helloWorldzhangsan 45

helloWorldlisi 56

helloWorldlidan 18

Time taken: 1.024 seconds, Fetched: 3 row(s)

OK

Time taken: 0.033 seconds