Oralce一般用于较大的项目或是用于某些软件的数据库承载,通常情况下,Oralce会和java搭配。最近因项目需求,需要把员工的工作组返回给前台,但是数据库是把员工的工作组Id,都存在一个字段内了(以"逗号"分隔),而这样不符合前台的需要,他们需要一行,一行的数据。

userId, workgroup

userId, workgroup

1001 10

1001 12

1001 15

1002 2

1002 4

1002 5

一、分析思路:

大体的思路是这样的:

首先:要知道,每一员工最多有多少个组。

其次:建一个有关"数"的临时表,与上面的组数进行关联,这样就出现了"多"行

最后:多"行"有了,剩下的就是对每一行的组进行刷选。如第一行取第一个逗号左边的,第二行取第二个逗号左边的, 依此类推。

二、实现:

根据上次的思路,来实现:

第一步:

with v_usergroups as (select '1001' as userId,'10,12,15' as workgroups from dual

union

select '1002' as userId,'2,4,5' as workgroups from dual

select userid,',' || workgroups ||',' AS tempgroups,length(workgroups || ',') - nvl(length(REPLACE(workgroups, ',')), 0) AS groupcount FROM v_usergroups

PS: 这里在"workgroup" 的前后也加了逗号,是为了后面使用方面。

第二步:

select LEVEL lv from dual CONNECT BY LEVEL <= 5

PS:这里的5,我们是根据业务需要,每一员工最多分为5个组,当然也可以写其他的值,但一定要大于第一步求得的"groupcount".

到这里后,我们对这两个表进行关联,看看值怎么样:

with v_usergroups as (select '1001' as userId,'10,12,15' as workgroups from dual

union

select '1002' as userId,'2,4,5' as workgroups from dual

select * from

(select userid,',' || workgroups ||',' AS tempgroups,length(workgroups || ',') - nvl(length(REPLACE(workgroups, ',')), 0) AS groupcount FROM v_usergroups ) a,

(select LEVEL lv from dual CONNECT BY LEVEL <= 5) b where b.lv<=a.groupcount

order by userid,lv

USERID TEMPGROUPS GROUPCOUNT LV

1 1001 ,10,12,15, 3 1

2 1001 ,10,12,15, 3 2

3 1001 ,10,12,15, 3 3

4 1002 ,2,4,5, 3 1

5 1002 ,2,4,5, 3 2

6 1002 ,2,4,5, 3 3

到这里,就离我们最终的结果很近了。 只需要在外层对"tempgroups"做一下简单的处理就可以了:

第三步:

这一步的主要思路就是:截串。第一个组应该是第一逗号和第二个逗号之间的值,第二个组应该是第二个逗号与第三个逗号之间的值,那第一个,和第二个如何表示呢,其实就是利用字段lv.也就是:

substr(tempgroups,instr(tempgroups, ',', 1, lv) + 1,instr(tempgroups, ',', 1, lv + 1) - (instr(tempgroups, ',', 1, lv) + 1))

最后的SQL 如下:

with v_usergroups as (select '1001' as userId,'10,12,15' as workgroups from dual

union

select '1002' as userId,'2,4,5' as workgroups from dual

select userid,substr(tempgroups,instr(tempgroups, ',', 1, lv) + 1,instr(tempgroups, ',', 1, lv + 1) - (instr(tempgroups, ',', 1, lv) + 1)) from

(select userid,',' || workgroups ||',' AS tempgroups,length(workgroups || ',') - nvl(length(REPLACE(workgroups, ',')), 0) AS groupcount FROM v_usergroups ) a,

(select LEVEL lv from dual CONNECT BY LEVEL <= 5) b where b.lv<=a.groupcount

order by userid,lv

03-14 08:04