Question

如何排除某些行？例如，我有下表：

+------+------+------+
| Col1 | Col2 | Col3 |
+------+------+------+
|    1 | 1    |    R |
|    1 | 2    |    D |
|    2 | 3    |    R |
|    2 | 4    |    R |
|    3 | 5    |    R |
|    4 | 6    |    D |
+------+------+------+

我只需要选择：

|    2 | 3    |    R |  
|    2 | 4    |    R |  
|    3 | 5    |    R |

我的选择无法正常运行：

with t (c1,c2,c3) as(
select 1 , 1 , 'R' from dual union all
select 1 , 2 , 'D' from dual union all
select 2 , 3 , 'R' from dual union all
select 2 , 4 , 'R' from dual union all
select 3 , 5 , 'R' from dual union all
select 4 , 6 , 'D' from dual), 
tt as (select t.*,count(*) over (partition by c1) cc from t ) select * from tt where cc=1 and c3='R';

谢谢！

Answer 1

使用row_number()窗口功能

    with t (c1,c2,c3) as(
select 1 , 1 , 'R' from dual union all
select 1 , 2 , 'D' from dual union all
select 2 , 3 , 'R' from dual union all
select 2 , 4 , 'R' from dual union all
select 3 , 5 , 'R' from dual union all
select 4 , 6 , 'D' from dual
    ),
  t1 as
  (
      select c1,c2,c3,row_number() over(order by c2) rn from t 
  ) select * from t1  where t1.rn>=3 and t1.rn<=5

demo link

Answer 2

这将起作用：

select * from (select * from table_name) where rownum<=4
minus 
select * from ( select * from table_name) where rownum<=2

Answer 3

您可以尝试使用相关子查询

select * from tablename a
from 
where exists (select 1 tablename b where a.col1=b.col1 having count(*)>1)

Answer 4

select * from table where col2 = 'R'

或者如果您只想排除D值的行

select * from table where col2 != 'D'

Answer 5

这取决于您的要求，但是您可以采用以下方式：

if(Build.VERSION.SDK_INT > Build.VERSION_CODES.LOLLIPOP_MR1) {}

如果您想排除，只需像WHERE col1！= 1

您也可以使用IN子句，例如

SELECT * FROM `table` WHERE col1 = 2 AND col3 = "R"

此语法适用于MySql，但您可以根据自己的要求或所使用的数据库对其进行修改。

Answer 6

我的猜测是，您希望col1的所有行都没有col1 = D的行，而col1 = R至少有1行。@可能存在[not]

DROP TABLE T;
CREATE TABLE T
(Col1 NUMBER, Col2 NUMBER, Col3 VARCHAR(1));
INSERT INTO T VALUES (    1 , 1    ,    'R');
INSERT INTO T VALUES (       1 , 2    ,    'D'); 
INSERT INTO T VALUES (       2 , 3    ,    'R'); 
INSERT INTO T VALUES (       2 , 4    ,    'R'); 
INSERT INTO T VALUES (       3 , 5    ,    'R');
INSERT INTO T VALUES (       3 , 6    ,    'D');
INSERT INTO T VALUES (       4 , 5    ,    'X');
INSERT INTO T VALUES (       4 , 6    ,    'Y');
INSERT INTO T VALUES (       5 , 6    ,    'X');
INSERT INTO T VALUES (       5 , 5    ,    'R');
INSERT INTO T VALUES (       5 , 6    ,    'Y');

SELECT * 
FROM T
WHERE NOT EXISTS(SELECT 1 FROM T T1 WHERE T1.COL1 = T.COL1 AND COL3 = 'D') AND
      EXISTS(SELECT 1 FROM T T1 WHERE T1.COL1 = T.COL1 AND COL3 = 'R');

结果

      COL1       COL2 COL3
---------- ---------- ----
         5          6 X    
         5          5 R    
         5          6 Y    
         2          3 R    
         2          4 R

Answer 7

根据您提供的信息，我只能推测，唯一的要求是COL1等于2或3，在这种情况下，您要做的就是（假设您实际上有表）；

SELECT * FROM <table_name>
WHERE col1 IN (2,3);

这将为问题中提供的特定示例提供所需的输出。如果有一个选择要求超出了检索数据的范围（第1列是2或3），则可以提供更具体或更精确的答案。

如何从SQL选择中排除某些行

7 个答案: