日期:2014-05-18  浏览次数:20740 次

求一SQL语句,比较难 (要求最新2次采购价格不同的记录,即什么时候有降过价)
求最新2次采购价格,物料代码及单价相同的只取一个,根据时间降序排列,这样可以知道最近的降价是什么时候,并把这2次采购单的数据输出。
数据为:
A表: B表:
pono datex pono code price  
0809010001 2008-09-01 0809010001 a123 1.5
0901010001 2009-01-01 0809010001 b123 5
0902010001 2009-02-01 0809010001 c123 4
  0901010001 a123 1.5
  0901010001 b123 3 
  0901010001 c123 3.5 
  0902010001 a123 1.2
  0902010001 b123 2.5
  0902010001 c123 3.5 
   

输出结果为:
pono datex code price
0902010001 2009-02-01 a123 1.2
0901010001 2009-01-01 a123 1.5
0902010001 2009-02-01 b123 2.5
0901010001 2009-01-01 b123 3
0902010001 2009-02-01 c123 3.5
0809010001 2008-09-01 c123 4

------解决方案--------------------
SQL code

---------------------------------
--  Author: htl258(Tony)
--  Date  : 2009-07-10 15:55:44
---------------------------------
--> 生成测试数据表:a

If not object_id('[a]') is null
    Drop table [a]
Go
Create table [a]([pono] nvarchar(10),[datex] Datetime)
Insert a
Select '0809010001','2008-09-01' union all
Select '0901010001','2009-01-01' union all
Select '0902010001','2009-02-01'
Go
--Select * from a

--> 生成测试数据表:b

If not object_id('[b]') is null
    Drop table [b]
Go
Create table [b]([pono] nvarchar(10),[code] nvarchar(4),[price] decimal(18,1))
Insert b
Select '0809010001','a123',1.5 union all
Select '0809010001','b123',5 union all
Select '0809010001','c123',4 union all
Select '0901010001','a123',1.5 union all
Select '0901010001','b123',3 union all
Select '0901010001','c123',3.5 union all
Select '0902010001','a123',1.2 union all
Select '0902010001','b123',2.5 union all
Select '0902010001','c123',3.5
Go
--Select * from b

-->SQL查询如下:
select pono,datex,code,price
from (
    select a.pono,a.datex,b.code,b.price,
        rn=row_number() over(partition by code order by datex desc)
    from a
      join (select * from b t where not exists(select 1 from b where code=t.code and price=t.price and [pono]>t.[pono])) b
        on a.pono=b.pono
) as t
where rn<=2
/*
pono       datex                   code price
---------- ----------------------- ---- ---------------------------------------
0902010001 2009-02-01 00:00:00.000 a123 1.2
0901010001 2009-01-01 00:00:00.000 a123 1.5
0902010001 2009-02-01 00:00:00.000 b123 2.5
0901010001 2009-01-01 00:00:00.000 b123 3.0
0902010001 2009-02-01 00:00:00.000 c123 3.5
0809010001 2008-09-01 00:00:00.000 c123 4.0

(6 行受影响)
*/

------解决方案--------------------
SQL code

declare @a table (pono nvarchar(10),datex datetime)
insert into @a select '0809010001','2008-09-01'
      union all select '0901010001','2009-01-01'
      union all select '0902010001','2009-02-01'
declare @b table (pono nvarchar(10),code nvarchar(10),price money)
insert into @b select '0809010001','a123',1.5
           union all select '0809010001','b123',5
           union all select '0809010001','c123',4
           union all select '0901010001','a123',1