Question

考虑这个问题：

 select count(*) from 
(
select pickle_id,
       pickle_code
from   pickles
where  pickle_code='DILL'
group  by pickle_id,
          pickle_code /* Returns 1000 rows */
minus 
select pickle_id,
       pickle_code
from   relish inner join pickles on relish_id=pickle_id
where  relish_code ='BURGER' /* Returns 1500 rows */
) 
/* Complete query returns 900 rows */

问题：

如何编写返回原始1000行的查询只有从减号查询中减去的100个？

结果应该总共1100行（1000 + 100减去）

Answer 1

试试这个

select pickle_id,
       pickle_code
from   pickles
where  pickle_code='DILL'
group  by pickle_id,
          pickle_code /* Returns 1000 rows */

union all
(
select pickle_id,
       pickle_code
from   pickles
where  pickle_code='DILL'
group  by pickle_id,
          pickle_code /* Returns 1000 rows */
intersect 
select pickle_id,
       pickle_code
from   relish inner join pickles on relish_id=pickle_id
where  relish_code ='BURGER' /* Returns 1500 rows */
)

Answer 2

嗯，那就是minus之前的所有内容：

select pickle_id, pickle_code
from   pickles
where  pickle_code = 'DILL'
group  by pickle_id, pickle_code;

编辑：

评论中问题的答案是：

select count(*)
from ((select pickle_id, pickle_code
      from   pickles
      where  pickle_code='DILL'
      group  by pickle_id, pickle_code /* Returns 1000 rows */
     ) intersect 
     (select pickle_id, pickle_code
      from   relish inner join pickles on relish_id=pickle_id
      where  relish_code ='BURGER' /* Returns 1500 rows */
     ) 
    ) p;

Answer 3

select count(*) from (
    select pickle_id,
           pickle_code
    from   pickles
    where  pickle_code='DILL'
    group  by pickle_id,
              pickle_code
    union all
    select pickle_id,
           pickle_code
    from   relish inner join pickles on relish_id=pickle_id
    where  relish_code ='BURGER' and pickle_code='DILL'
    group by pickle_id, pickle_code
)

如何在此查询中包含减去的行？

3 个答案: