如何在 SQL 中创建特殊情况的 Intersect 子句？

Question

我需要一组值，只有当它们在两个不同的选择语句中可用时，所以我使用了“相交”

(select statement 1) intersect (select statement 2)

但是当另一个 select 语句返回没有值时，我想要一个 select 语句中的所有值。

示例 1：

Statement 1 returns(1,2,3,4,5) and Statement 2 returns(2,5,8,9);
Expected result 1: (2,5)

示例 2：

Statement 1 returns(NULL) and Statement 2 returns(2,5,8,9);
Expected result 1: (2,5,8,9)

---编辑：为清楚起见添加第三个示例---

示例 3：

Statement 1 returns(2,5,8,9) and Statement 2 returns(NULL);
Expected result 1: (2,5,8,9)

---编辑 2--- 我在

中使用此选择

Answer 1

对于第一种情况，您可以使用 INNER JOIN

select  col 
from  (
  select 1 col union all 
  select 2 union all 
  select 3 union all 
  select 4 union all 
  select 5
) t1  
inner join  (
  select 2 col union all 
  select 5 union all 
  select 8  union all 
  select 9
) t2  ON t1.col = t2.col

第二个你可以尝试使用右连接检查空值

select  col 
from  (
  select 1 col union all 
  select 2 union all 
  select 3 union all 
  select 4 union all 
  select 5
) t1  
right join  (
  select 2 col union all 
  select 5 union all 
  select 8  union all 
  select 9
) t2  ON t1.col = t2.col
where t1.col is null 

Answer 2

有几种方法可以做你想做的事，但你需要根据你期望空语句的位置正确放置表格：

DECLARE @t1 TABLE
(
ID INTEGER
);

DECLARE @t2 TABLE
(
ID INTEGER
);

DECLARE @t3 TABLE
(
ID INTEGER
);

INSERT INTO @t1 VALUES (1),(2),(3),(4),(5);

INSERT INTO @t2 VALUES (2),(5),(8),(9);

SELECT *
FROM @t1 t1
INNER JOIN @t2 t2 ON t1.ID = t2.ID;

--Example 2, you can use either a FULL JOIN
SELECT *
FROM @t1 t1
FULL JOIN @t3 t3 ON t1.ID = t3.ID;

--Example 2, or you can use a LEFT JOIN. But you need to put the table that is likely to have data as the 
--main table in a LEFT JOIN
SELECT *
FROM @t1 t1
LEFT JOIN @t3 t3 ON t1.ID = t3.ID;

Answer 3

伪代码

bool flag := exists (select 1 from Statement_2);
select *
from Statement_1 t1
where exists (select 1 from Statement_2 t2 where t1.key = t2.key) or not flag;

Sql server 版本，将key替换成真实key比较。

declare @flag bit = case when exists (select 1 from Statement_2)
                   then 1 else 0 end;
select *
from Statement_1 t1
where exists (select 1 from Statement_2 t2 where t1.key = t2.key) or @flag = 0;

Answer 4

这将是另一种方式：

--Setup
create table #table1(id int);
insert into #table1(id) values (1), (2), (3), (4), (5);

create table #table2(id int);
insert into #table2(id) values (2), (5), (8), (9);

--Query
select 
    ISNULL(a.id, b.id) id
from
    #table1 a full outer join #table2 b on a.id = b.id
where 
    not exists(select null from #table1) or 
    not exists(select null from #table2) or
    (a.id is not null and b.id is not null)

Answer 5

像这样使用UNION ALL：

(
  select statement 1
  union all
  select statement 2  
  where not exists (
    select statement 1
  )
)
intersect
(
  select statement 2
  union all
  select statement 1  
  where not exists (
    select statement 2
  )
) 

如果select statement X 已经包含WHERE 子句，请在末尾使用AND 来应用条件not exists (...)。
请参阅demo。
对于这些表：

create table table1(id int);
insert into table1(id) values (1), (2), (3), (4), (5);

create table table2(id int);
insert into table2(id) values (2), (5), (8), (9);

案例：

(
  select * from table1
  union all
  select * from table2  
  where not exists (
    select * from table1
  )
)
intersect
(
  select * from table2
  union all
  select * from table1  
  where not exists (
    select * from table2
  )
)

结果是：

> | id |
> | -: |
> |  2 |
> |  5 |

案例：

(
  select * from table1 where id > 9
  union all
  select * from table2  
  where not exists (
    select * from table1 where id > 9
  )
)
intersect
(
  select * from table2
  union all
  select * from table1 where id > 9 
  and not exists (
    select * from table2
  )
)

结果是：

> | id |
> | -: |
> |  2 |
> |  5 |
> |  8 |
> |  9 |

对于案例：

(
  select * from table1
  union all
  select * from table2 where id > 9 
  and not exists (
    select * from table1
  )
)
intersect
(
  select * from table2 where id > 9
  union all
  select * from table1
  where not exists (
    select * from table2 where id > 9
  )
)

结果是：

> | id |
> | -: |
> |  1 |
> |  2 |
> |  3 |
> |  4 |
> |  5 |

Answer 6

您可能想试试这个，看看它是否更快：

with
    s1 as (select count(*) over () c1, <query1> ...),
    s2 as (select count(*) over () c2, <query2> ...),
)
select coalesce(t1.id, t2.id) as id
from table1 t1 full outer join table2 t2
    on t2.id = t1.id and c1 > 0 and c2 > 0
);

Answer 7

我想我会这样做：

select t1.id
from t1
where exists (select 1 from t2 where t2.id = t1.id)
union all
select t1.id
from t1
where not exists (select 1 from t2)
union all
select t2.id
from t2
where not exists (select 1 from t1);

如果您的语句是表，那么优化器应该能够产生高效的执行计划。如果语句是复杂的表达式，那么您可能需要不同的方法。如果是这样，了解该声明的内容会很有帮助。可能有一种最佳方式来表达更多信息的逻辑。