Question

我有下表查找给定值的行之前和之后。

例如，要显示给定列col2值125的前后两行，并且还希望在每组之后打印空白行。

表格：

CREATE TABLE PreTest
(
 col1 int,
 col2 int,
 col3 date,
 col4 time
);

样本数据：

insert into PreTest values(111,123,'2018-01-01','00:10:11'),
                          (111,124,'2018-01-01','00:10:12'),
                          (111,125,'2018-01-01','00:10:17'),
                          (111,126,'2018-01-01','00:10:16'),
                          (111,127,'2018-01-01','00:10:13');
insert into PreTest values(111,228,'2018-01-02','10:13:01'),
                            (111,229,'2018-01-02','10:13:11'),
                            (111,125,'2018-01-02','10:13:02'),
                            (111,237,'2018-01-02','10:13:30'),
                            (111,232,'2018-01-01','10:13:04');
insert into PreTest values(111,339,'2018-01-03','11:10:01'),
                            (111,338,'2018-01-03','11:10:04'),
                            (111,125,'2018-01-03','11:10:02'),
                            (111,340,'2018-01-03','00:10:11'),
                            (111,333,'2018-01-01','11:10:03');

查询：

WITH C1 AS
(
    SELECT ROW_NUMBER() OVER(order by col3,col4) rn,*
    FROM PreTest
)
SELECT * FROM 
(

    SELECT * FROM C1 WHERE rn IN (SELECT rn FROM C1 WHERE col2 = '125')
    UNION 
    SELECT * FROM C1 WHERE rn IN ( SELECT rn - 1 FROM C1 WHERE col2 = '125') 
    UNION       
    SELECT * FROM C1 WHERE rn IN ( SELECT rn - 2 FROM C1 WHERE col2 = '125') 
    UNION
    SELECT * FROM C1 WHERE rn IN ( SELECT rn + 1 FROM C1 WHERE col2 = '125') 
    UNION
    SELECT * FROM C1 WHERE rn IN ( SELECT rn + 2 FROM C1 WHERE col2 = '125')
) a

预期产量：

rn                   col1        col2        col3       col4
-------------------- ----------- ----------- ---------- ----------------
4                    111         127         2018-01-01 00:10:13.0000000
5                    111         126         2018-01-01 00:10:16.0000000
6                    111         125         2018-01-01 00:10:17.0000000
7                    111         132         2018-01-01 10:13:04.0000000
8                    111         232         2018-01-01 10:13:04.0000000

12                   111         128         2018-01-02 10:13:01.0000000
13                   111         130         2018-01-02 10:13:02.0000000
14                   111         125         2018-01-02 10:13:02.0000000
15                   111         229         2018-01-02 10:13:11.0000000
16                   111         129         2018-01-02 10:13:11.0000000

22                   111         139         2018-01-03 11:10:01.0000000
23                   111         141         2018-01-03 11:10:02.0000000
24                   111         125         2018-01-03 11:10:02.0000000
25                   111         338         2018-01-03 11:10:04.0000000
26                   111         138         2018-01-03 11:10:04.0000000

Answer 1

虽然我同意Sean的观点，但没有充分的理由愿意在SQL Server中（与表示层相反）进行此操作。

您将需要使用gaps and islands方法来确定组的更改点，然后在ORDER BY中的这些点处插入一个虚假的空白值行。

这意味着，如果要在行号4和5之间留空行，则需要将该行号CAST转换为小数，然后添加行号4.5或类似的空行。

在每组行之后打印空白行

1 个答案: