Question

我有一个以Count（*）为顺序的查询，其次是偏移量获取。当我使用不同的值作为偏移量时，它总是提供相同的结果集。

我尝试按照结果集变化的顺序排列某些列，并且效果很好。有人可以帮忙吗？

select 
   Id,
   count(*) as "Count" 
from 
   some_table
group by
   "Id"
Order By
   "Count" ASC OFFSET 0 ROWS FETCH NEXT 5 ROWS ONLY;

Screenshot 1 Screenshot 2

Answer 1

这是您的查询：

select Id, count(*) as "Count" 
from some_table
group by "Id"
order by "Count" asc
OFFSET 0 ROWS FETCH NEXT 5 ROWS ONLY;

问题是您有count(*)的重复值。由于SQL表表示无序集，因此它没有默认顺序。当排序键具有相同的值时，这会带来问题。

简单的解决方案是在id中包含order by：

order by "Count" asc, id

这使得排序稳定，因此定义明确。

Answer 2

这是我的示例

创建表

create table t (id int, name varchar(100))

插入数据的

insert into t values(1,'saravnan')
insert into t values(1,'kumar')
insert into t values(1,'Ravi')
insert into t values(1,'mohan')
insert into t values(2,'Raju')
insert into t values(2,'Vikram')
insert into t values(2,'AA')
insert into t values(2,'BB')
insert into t values(2,'CC')
insert into t values(2,'DD')
insert into t values(2,'EE')
insert into t values(3,'Raju')
insert into t values(3,'Vikram')
insert into t values(3,'AA')
insert into t values(4,'BB')
insert into t values(4,'CC')
insert into t values(4,'DD')
insert into t values(4,'EE')

选择所有数据

select *from t

    id  name
1   1   saravnan
2   1   kumar
3   1   Ravi
4   1   mohan
5   2   Raju
6   2   Vikram
7   2   AA
8   2   BB
9   2   CC
10  2   DD
11  2   EE
12  3   Raju
13  3   Vikram
14  3   AA
15  4   BB
16  4   CC
17  4   DD
18  4   EE

偏移量用于从表数据的开头跳过的行数。它应该是数字。这是我的例子：

select id,count(*) from t group by id order by id offset 1 rows fetch next 3 row only

返回

下一个用于返回数字行的提取。您的示例的解决方案。

--offset 0 returns[enter image description here][1]

select Id,count(1) as "Count" from t  group by ID Order By "Count" ASC OFFSET 0 ROWS FETCH NEXT 5 ROWS ONLY

    Id  Count
1   3   3
2   4   4
3   1   4
4   2   7

-- offset 1 returns
select Id,count(1) as "Count" from t  group by ID Order By "Count" ASC OFFSET 1 ROWS FETCH NEXT 5 ROWS ONLY

    Id  Count
1   4   4
2   1   4
3   2   7

偏移不适用于Count（*），按以下顺序排序

2 个答案: