# 建表语句，注意！注意！注意！本题测试环境为Hive3.1.0
-- 部门表(dept) dno dname
create table dept(
dno int,
dname string
)
row format delimited
fields terminated by ‘,‘
lines terminated by ‘\n‘
stored as textfile;

-- 人员表(emp) ename dno1 dno2 dno3
create table emp(
ename string,
dno1 int,
dno2 int,
dno3 int
)
row format delimited
fields terminated by ‘,‘
lines terminated by ‘\n‘
stored as textfile;

# dept表
99,人力资源部
100,技术部门
111,大数据
2345,移动端开发
321,IOS开发
5578,数据仓库
334,ETL
7890,Android研发

# emp表
小王,100,111,5578
小李,100,2345,321
小张,100,111,334

load data local inpath 
‘/opt/module/data/dept.txt‘ into table default.dept;

load data local inpath 
‘/opt/module/data/emp.txt‘ into table default.emp;

-- 原始表初始化
select e.ename,d.dname
from emp e
join dept d
on e.dno1 = d.dno or e.dno2 = d.dno or e.dno3 = d.dno;

-- 失败1：拼接。hive中不能查询非group by的字段
select 
    ename,
    dname
from (
select e.ename,d.dname
from emp e
join dept d
on e.dno1 = d.dno or e.dno2 = d.dno or e.dno3 = d.dno
)tmp
group by ename;


-- 失败2：hive中不能查询非group by的字段
with tmp as (
select e.ename,d.dname
from emp e
join dept d
on e.dno1 = d.dno or e.dno2 = d.dno or e.dno3 = d.dno
)
select ename,concat_ws(‘>‘, dname) from tmp group by ename;

-- 成功，注意，本环境为3.1.0；如果是Hive2.x的版本，不能使用or，需要使用where条件，然后union 起来
with tmp as (
    select 
        e.ename,
        d.dname
    from emp e
    join dept d
    on e.dno1 = d.dno 
    or e.dno2 = d.dno 
    or e.dno3 = d.dno
)
select 
    ename,
    concat_ws(‘>‘, collect_set(dname))
from tmp
    group by ename;