mysql中的group_concat分组连接功能相当强大,可以先分组再连接成字符串,还可以进行排序连接。但是hive中并没有这个函数,那么hive中怎么实现这个功能呢?
这里要用到:concat_ws函数和collect_list、collect_set 函数。
1. 建立测试表(无分区表):
create table if not exists db_name.test_tb(id string,content string,comment string) row format delimited fields terminated by '\1' stored as textfile
2. 插入几条数据:
insert into db_name.test_tb values('1','Tom','测试1') insert into db_name.test_tb values('1','Bob','测试2') insert into db_name.test_tb values('1','Wendy','测试3') insert into db_name.test_tb values('2','Bob','测试22') insert into db_name.test_tb values('2','Tom','测试11')
3. concat_ws + collect_set + group by:
select id, concat_ws(',',collect_set(content)) as con_con, concat_ws(',',collect_set(comment)) as con_com from db_name.test_tb group by id
结果:无序且不对应(con_con与con_com的位置) —— 但是注意 collect_set会将重复的数据删除,因为集合的性质。
每次运行连接的结果顺序都可能不一样。
4. concat_ws + collect_list + group by:
select id, concat_ws(',',collect_list(content)) as con_con, concat_ws(',',collect_list(comment)) as con_com from db_name.test_tb group by id
结果:对应(con_con与con_com的位置)但无序。
5. concat_ws + collect_list + group by + row_number():
select id, concat_ws(',',collect_list(content)) as con_con, concat_ws(',',collect_list(comment)) as con_com, concat_ws(',',collect_list(cast(rn as string))) as con_rn from db_name.test_tb ( select id, content, comment, row_number() over(partition by id order by content asc) as rn from db_name.test_tb ) group by id
结果:对应(con_con与con_com的位置)且有序。