0

在 SQL Server 2000 中使用视图

表格1：

id  z1      z2      z3      z4   type
--------------------------------------
01A 300     400     300     400  01
2B  300     400     300     500  02
3C  700     600     400     300  01
04A 500     400     800     900  01
05B 400     300     400     300  02
06  150     200     200     150  03
....

表2：

type  value1         value2
------------------------------------    
01    0      300
01    301    500          
02    0      200
02    201    400
03
.....

我想根据 table2 范围选择 table1 行：

表 1 的组合-max(Z1, Z2) and max(Z3, Z4)

如果 Max(z1,z2) range 小于或等于 table2 max(value2) where table1.type = table2.type If Max(z1, z2) range 小于或等于 table2 max(value2) where table1.type = table2.type

如果 z1、z2 范围小于或等于 table2，那么我想显示 z1、z2 行，否则为 null 如果 z3、z4 范围小于或等于 table2，那么我想显示 z3、z4 行，否则为 null

预期产出

表格1：

 id  z1      z2      z3      z4   type
    --------------------------------------

01A 300     400     300     400  01 ' `Both (z1, z2), (z3, z4) rows are matching with table2 for type 01`
2B  300     400     null     null  02 ' (z1, z2) are matching, (z3, z4) rows are not matching with table2 for type 02
3C  null    null     400     300  01 ' (z1, z2) rows are not matching, (z3, z4) rows are matching with table2 for type 01
04A 500     400     null     null  01 ' (z1, z2) rows are matching, (z3, z4) rows are not matching with table2 for type 01
05B 400     300     400     300  02 ' Both (z1, z2), (z3, z4) rows are matching with table2 for type 02

....

目前我正在使用一个视图，我不想更改为存储过程，因为大多数报告都在使用这个视图。

如何在 sql.. 中做到这一点？

4

1 回答 1

1

这似乎给出了预期的结果。（我更喜欢使用 CTE，但您指定了 2000）。

样本数据：

declare @T1 table (Id varchar(3) not null,z1 int not null,z2 int not null,z3 int not null,z4 int not null,type char(2) not null)
insert into @T1 (id,z1,z2,z3,z4,type) values
('01A',300,400,300,400,'01'),
('2B',300,400,300,500,'02'),
('3C',700,600,400,300,'01'),
('04A',500,400,800,900,'01'),
('05B',400,300,400,300,'02'),
('06',150,200,200,150,'03')

declare @T2 table (type char(2) not null,value1 int not null,value2 int not null)
insert into @T2 (type,value1,value2) values
('01',0,300),
('01',301,500),
('02',0,200),
('02',201,400)

询问：

select
    t1.Id,
    CASE WHEN t2.type is not null then z1 END as z1,
    CASE WHEN t2.type is not null then z2 END as z2,
    CASE WHEN t3.type is not null then z3 END as z3,
    CASE WHEN t3.type is not null then z4 END as z4,
    t1.type
from
    @T1 t1
        left join
    (select type,MAX(value2) as val2 from @T2 group by type) t2
        on
            t1.type = t2.type and
            t1.z1 <= t2.val2 and
            t1.z2 <= t2.val2
        left join
    (select type,MAX(value2) as val2 from @T2 group by type) t3
        on
            t1.type = t3.type and
            t1.z3 <= t3.val2 and
            t1.z4 <= t3.val2
where
    t2.type is not null or
    t3.type is not null

结果：

Id   z1          z2          z3          z4          type
---- ----------- ----------- ----------- ----------- ----
01A  300         400         300         400         01
2B   300         400         NULL        NULL        02
3C   NULL        NULL        400         300         01
04A  500         400         NULL        NULL        01
05B  400         300         400         300         02

我不确定你的问题标题的相关性是什么，中间的措辞很混乱，我对最后一个WHERE子句进行了猜测，因为该06行似乎消失了。

于 2012-11-07T07:39:46.187 回答