在我的应用程序中，我使用了商店产品描述值，如下所示：
ID BILLNO CUS_NAME DATE TOT_BAL S1 S2 S3 S4 D1 D2 D3 D4 Q1 Q2 Q3 Q4 U1 U2 U3 U4 T1 T2 T3 T4 TOTAL CUSCODE
-------------------------------------------------------------------------------------------------------------------------------------------------------------------
29 1 XXX Apr-03-2017 1932 1 2 3 NULL AAA BBB CCC NULL 6 30 6 NULL 80 35 67 NULL 480 1050 402 0 1932 DF
40 2 YYYY Apr-04-2017 6454 1 2 3 NULL AAA DDD FFF NULL 30 24 50 NULL 80 96 35 NULL 2400 2304 1750 0 6454 MQ
-------------------------------------------------------------------------------------------------------------------------------------------------------------------
我正在将我的应用程序迁移到最新技术，因此我正在重新构建mt表以避免复杂化。所以我希望结果将是如下调整我的新表：
BILLNO CUS_NAME DATE TOT_BAL SNO Pdt QTY Unit Tot CUSCODE
1 XXX Apr-03-2017 1932 1 AAA 6 80 480 DF
1 XXX Apr-03-2017 1932 2 BBB 30 35 1050 DF
1 XXX Apr-03-2017 1932 3 CCC 6 67 402 DF
2 YYY Apr-04-2017 6454 1 AAA 30 80 2400 MQ
2 YYY Apr-04-2017 6454 2 DDD 24 96 2304 MQ
2 YYY Apr-04-2017 6454 3 FFF 50 35 1750 MQ
我尝试了以下查询，但结果错误：
我不想要Null值。我只需要有效数据。对于示例S4，D4，Q4是Null值，因此Output应忽略该值。
SELECT BILLNO, CUS_NAME,SNOA,SNO,PDTA,PDT
FROM [tmpFormat]
UNPIVOT
(
SNO
FOR SNOA IN (S1,S2,S3,S4,S5,S6,S7,S8,S9,S10,S11,S12,S13,S14,S15,S16,S17)
) UnPIV
UNPIVOT
(
PDT
FRO PDTA IN (D1,D2,D3,D4,D5,D6,D7,D8,D9,D10,D11,D12,D13,D14,D15,D16,D17)
) UnPiv1;
答案 0 :(得分：2)
尝试交叉申请
SELECT BILLNO, CUS_NAME, UnP.SNOA, UnP.PDTA --,..
FROM [tmpFormat]
CROSS APPLY
( SELECT S1 AS SNOA, D1 as PDTA --,..
UNION ALL
SELECT S2 AS SNOA, D2 as PDTA --,..
-- ..
) UnP
答案 1 :(得分：0)
您可以手动取消它：
select BILLNO, CUS_NAME, DATE, TOT_BAL, S1 as SNO, D1 as Pdt, Q1 as QTY, U1 as Unit, T1 as Tot, CUSCODE
from [tmpFormat]
union all
select BILLNO, CUS_NAME, DATE, TOT_BAL, S2, D2, Q2, U2, T2, CUSCODE
from [tmpFormat]
uniona all
select BILLNO, CUS_NAME, DATE, TOT_BAL, S3, D3, Q3, U3, T3, CUSCODE
from [tmpFormat]
答案 2 :(得分：0)
Dynamic Query UnPivoting：
使用Sys.Columns和Stuff（动态查询）连接|将多列（需要追加）值分组到一列下。
示例S1，S2，S3 ......作为SNO
使用Unpivot处理稍后的动态查询，以将多列数据转换为单行。 示例D1，D2，D3 .....在PDT列下，ID用静态数据映射。
动态查询结果存储在Table变量中以供以后处理。
创建的表变量与CTE一起用于将Unpivoted数据与静态数据连接以获得最终结果。
If OBJECT_ID('UnpivotData','u') is not null
Drop table UnPivotData
go
Create table UnPivotData
(
Id int,
BILLNO int,
Cus_Name varchar(10),
[Date] date,
Tot_Bal Int,
S1 smallint,
S2 smallint,
S3 smallint,
S4 smallint,
D1 varchar(3),
D2 varchar(3),
D3 varchar(3),
D4 varchar(3),
Q1 smallint,
Q2 smallint,
Q3 smallint,
Q4 smallint,
U1 smallint,
U2 smallint,
U3 smallint,
U4 smallint,
T1 smallint,
T2 smallint,
T3 smallint,
T4 smallint,
Total int,
CusCode varchar(2)
)
insert into UnPivotData
Select 29, 1, 'XXX', '03-04-2017', 1932, 1, 2, 3, NULL, 'AAA', 'BBB', 'CCC', NULL, 6, 30, 6, NULL, 80, 35, 67, NULL, 480, 1050, 402, 0, 1932, 'DF'
union
Select 40, 2,'YYY', '04-04-2017', 6454, 1, 2, 3, NULL, 'AAA', 'DDD', 'FFF', NULL, 30, 24, 50, NULL, 80, 96, 35, NULL, 2400, 2304, 1750, 0, 6454, 'MQ'
DECLARE @SNO AS NVARCHAR(MAX)='',@PDT AS NVARCHAR(MAX)='',@QTY AS NVARCHAR(MAX)='', @UNIT AS NVARCHAR(MAX)='', @TOT AS NVARCHAR(MAX)=''
SET @SNO = STUFF((SELECT distinct ',' + QuoteName(C.name)
from sys.tables T join sys.columns C on T.object_id=C.object_id and T.object_id=OBJECT_ID('UnPivotData','U') and C.Name like'S'+'[0-9]%'
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)'),1,1,'')
SET @PDT = STUFF((SELECT distinct ',' + QuoteName(C.name)
from sys.tables T join sys.columns C on T.object_id=C.object_id and T.object_id=OBJECT_ID('UnPivotData','U') and C.Name like'D'+'[0-9]%'
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)'),1,1,'')
SET @QTY = STUFF((SELECT distinct ',' + QuoteName(C.name)
from sys.tables T join sys.columns C on T.object_id=C.object_id and T.object_id=OBJECT_ID('UnPivotData','U') and C.Name like'Q'+'[0-9]%'
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)'),1,1,'')
SET @UNIT = STUFF((SELECT distinct ',' + QuoteName(C.name)
from sys.tables T join sys.columns C on T.object_id=C.object_id and T.object_id=OBJECT_ID('UnPivotData','U') and C.Name like'U'+'[0-9]%'
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)'),1,1,'')
SET @TOT = STUFF((SELECT distinct ',' + QuoteName(C.name)
from sys.tables T join sys.columns C on T.object_id=C.object_id and T.object_id=OBJECT_ID('UnPivotData','U') and C.Name like'T'+'[0-9]%'
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)'),1,1,'')
Declare @Query1 Nvarchar(MAX),@Query2 Nvarchar(MAX),@Query3 Nvarchar(MAX),@Query4 Nvarchar(MAX),@Query5 Nvarchar(MAX)
Set @Query1='select ID,SNO from
(
Select ID,'+@SNO+' from UnPivotData
)L
unpivot
(
SNo for col in('+ @SNO +')
) u'
Set @Query2='select ID,PDT from
(
Select ID,'+@PDT+' from UnPivotData
)L
unpivot
(
PDT for col in('+ @PDT +')
) u'
Set @Query3='select ID,QTY from
(
Select ID,'+@QTY+' from UnPivotData
)L
unpivot
(
QTY for col in('+ @QTY +')
) u'
Set @Query4='select ID,UNIT from
(
Select ID,'+@UNIT+' from UnPivotData
)L
unpivot
(
UNIT for col in('+ @UNIT +')
) u'
Set @Query5='select ID, TOT from
(
Select ID,'+@TOT+' from UnPivotData
)L
unpivot
(
TOT for col in('+ @TOT +')
) u'
Declare @t1 table
(
id int,
SNO Smallint
)
Declare @t2 table
(
id int,
PDT Varchar(3)
)
Declare @t3 table
(
id int,
QTY Smallint
)
Declare @t4 table
(
id int,
UNIT Smallint
)
Declare @t5 table
(
id int,
TOT Smallint
)
Insert into @T1 execute ( @Query1 )
Insert into @T2 execute ( @Query2 )
Insert into @T3 execute ( @Query3 )
Insert into @T4 execute ( @Query4 )
Insert into @T5 execute ( @Query5 )
;With Cte
as
(
Select UD.Id,BILLNO,Cus_Name,[Date],Tot_Bal,SNO,Row_Number() over(partition by T1.ID Order by T1.ID)Row_NO--,PDT,QTY,UNIT,TOT
,Total,CusCode
from UnPivotData UD join @t1 T1 on UD.Id=T1.id
)
,
Cte1
as
(
Select Cte.*,PDT from Cte Join
(Select *,Row_Number() over(partition by ID Order by ID)Row_NO from @t2) T2 on Cte.ID=T2.ID and Cte.Row_NO=T2.Row_NO
)
,
Cte2
as
(
Select Cte1.*,QTY from Cte1 Join
(Select *,Row_Number() over(partition by ID Order by ID)Row_NO from @t3) T3 on Cte1.ID=T3.ID and Cte1.Row_NO=T3.Row_NO
)
,
Cte3
as
(
Select Cte2.*,UNIT from Cte2 Join
(Select *,Row_Number() over(partition by ID Order by ID)Row_NO from @t4) T4 on Cte2.ID=T4.ID and Cte2.Row_NO=T4.Row_NO
)
,
Cte4
as
(
Select Cte3.*,TOT from Cte3 Join
(Select *,Row_Number() over(partition by ID Order by ID)Row_NO from @t5) T5 on Cte3.ID=T5.ID and Cte3.Row_NO=T5.Row_NO
)
Select Id,BILLNO,Cus_Name,[Date],Tot_Bal,SNO,PDT,QTY,UNIT,TOT
,Cuscode
from Cte4