讲解SQL Server2005数据项的分拆与合并

参考示例如下：

-- =============================================

-- Author: LzmTW

-- Create date: 20080102

-- Description: 连接子字符串

-- @TableName: 数据所在的表的名称

-- @KeyColName: 连接子字符串所依据的键值所在的列

-- @JoinColName: 包含要连接的子字符串所在的列

-- @Quote: 分隔子字符串

-- @Where: 选择条件，不包含Where

-- =============================================

CREATE PROCEDURE [Helper].[JoinValue]

@TableName nvarchar(100)

,@KeyColName nvarchar(20)

,@JoinColName nvarchar(20)

,@Quote nvarchar(10) = N','

,@Where nvarchar(max) = NULL

AS

BEGIN

SET NOCOUNT ON;

DECLARE

@SQL nvarchar(max)

IF @Where IS NULL

SET @SQL = N'

SELECT *

FROM

(

SELECT DISTINCT KeyCol = @KeyColName

FROM @TableName

)a

'

ELSE

SET @SQL = N'

SELECT *

FROM

(

SELECT DISTINCT KeyCol = @KeyColName

FROM @TableName

WHERE @Where

)a

'

SET @SQL = @SQL + N'

OUTER APPLY (

SELECT NewValues =

STUFF(

REPLACE(

REPLACE(

REPLACE(

(

SELECT JoinCol = @JoinColName

FROM @TableName b

WHERE @KeyColName = a.KeyCol

FOR XML RAW

)

, N'''', N'''')

, N'', N'''/>'', N'''')

, 1, LEN(N''@Quote''), N'''')

) c'

SET @SQL = REPLACE(@SQL, N'@TableName', @TableName)

SET @SQL = REPLACE(@SQL, N'@KeyColName', @KeyColName)

SET @SQL = REPLACE(@SQL, N'@JoinColName', @JoinColName)

SET @SQL = REPLACE(@SQL, N'@Quote', @Quote)

IF NOT @Where IS NULL

SET @SQL = REPLACE(@SQL, N'@Where', @Where)

--PRINT @SQL

EXEC sp_executesql @SQL

END

GO

-- =============================================

-- Author: LzmTW

-- Create date: 20080102

-- Description: 分拆字符串

-- @TableName: 数据所在的表的名称

-- @KeyColName: 分拆为子字符串所依据的键值所在的列

-- @SpliteColName: 包含要分拆的字符串所在的列

-- @Quote: 分隔子字符串

-- @Where: 选择条件，不包含Where

-- =============================================

CREATE PROCEDURE [Helper].[SpliteValues]

@TableName nvarchar(100)

,@KeyColName nvarchar(20)

,@SpliteColName nvarchar(20)

,@Quote nvarchar(10) = N','

,@Where nvarchar(max) = NULL

AS

BEGIN

SET NOCOUNT ON;

DECLARE

@SQL nvarchar(max)

IF @Where IS NULL

SET @SQL = N'

SELECT

KeyCol, NewValue

FROM

(

SELECT

KeyCol = @KeyColName

,SpliteCol = CONVERT(xml, N'''' + REPLACE(@SpliteColName, N''@Quote'', N'''') + N'''')

FROM @TableName

) a

'

ELSE

SET @SQL = N'

SELECT

KeyCol, NewValue

FROM

(

SELECT

KeyCol = @KeyColName

,SpliteCol = CONVERT(xml, N'''' + REPLACE(@SpliteColName, N''@Quote'', N'''') + N'''')

FROM @TableName

WHERE @Where

) a

'

SET @SQL = @SQL + N'

OUTER APPLY

(

SELECT NewValue = N.v.value(N''.'', ''nvarchar(max)'')

FROM SpliteCol.nodes(N''/root/v'') N(v)

) b'

SET @SQL = REPLACE(@SQL, N'@TableName', @TableName)

SET @SQL = REPLACE(@SQL, N'@KeyColName', @KeyColName)

SET @SQL = REPLACE(@SQL, N'@SpliteColName', @SpliteColName)

SET @SQL = REPLACE(@SQL, N'@Quote', @Quote)

IF NOT @Where IS NULL

SET @SQL = REPLACE(@SQL, N'@Where', @Where)

EXEC sp_executesql @Sql

END

示例：

SET NOCOUNT ON

CREATE TABLE ##Table([keyCol] varchar(3), [NewValues] varchar(max))

--原数据

SELECT

[title_id]

,[title]

FROM [pubs].[dbo].[titles]

WHERE [type] LIKE 'p%'

--以title_id的前两个字符为参考键值，合并title到一个临时表中

INSERT INTO ##Table

EXECUTE [ChineseHoliday].[Helper].[JoinValue]

@TableName = '[pubs].[dbo].[titles]'

,@KeyColName = 'LEFT([title_id], 2)'

,@JoinColName = '''《''+[title] + ''》'''

,@Quote = ','

,@Where = '[type] LIKE ''p%'''

--显示

SELECT * FROM ##Table

--对临时表NewValues的值进行分拆

EXECUTE [ChineseHoliday].[Helper].[SpliteValues]

@TableName = '##Table'

,@KeyColName = '[keyCol]'

,@SpliteColName = '[NewValues]'

,@Quote = ','

--删除临时表

DROP TABLE ##Table

结果：

title_id title

-------- --------------------------------------------------------------------------------

PC1035 But Is It User Friendly?

PC8888 Secrets of Silicon Valley

PC9999 Net Etiquette

PS1372 Computer Phobic AND Non-Phobic Individuals: Behavior Variations

PS2091 Is Anger the Enemy?

PS2106 Life Without Fear

PS3333 Prolonged Data Deprivation: Four Case Studies

PS7777 Emotional Security: A New Algorithm

keyCol NewValues

------ ------------------------------------------

PC 《But Is It User Friendly?》,《Secrets of Silicon Valley》,《Net Etiquette》

PS 《Computer Phobic AND Non-Phobic Individuals: Behavior Variations》,《Is Anger the Enemy?》,《Life Without Fear》,《Prolonged Data Deprivation: Four Case Studies》,《Emotional Security: A New Algorithm》

KeyCol NewValue

------ ------------------------------------------

PC 《But Is It User Friendly?》

PC 《Secrets of Silicon Valley》

PC 《Net Etiquette》

PS 《Computer Phobic AND Non-Phobic Individuals: Behavior Variations》

PS 《Is Anger the Enemy?》

PS 《Life Without Fear》

PS 《Prolonged Data Deprivation: Four Case Studies》

PS 《Emotional Security: A New Algorithm》

继续:字符串的分拆

-- =============================================

-- Author: LzmTW

-- Create date: 20080108

-- Description: 拆分字符串

-- =============================================

CREATE FUNCTION [Func].[Splite]

(

@Input nvarchar(max)

,@Quote nvarchar(max)

)

RETURNS

@Table TABLE

(

[ID] int identity(1,1) PRIMARY KEY

,[Value] nvarchar(max)

)

AS

BEGIN

INSERT @Table

SELECT

[Value] = NewValue

FROM

(

SELECT

SpliteCol = CONVERT(

xml

,N'' + REPLACE(

@Input

,@Quote

,N'') + N'')

) a

OUTER APPLY

(

SELECT NewValue = N.v.value(N'.', 'nvarchar(max)')

FROM SpliteCol.nodes(N'/root/v') N(v)

) b

RETURN

END

示例:

定义新行,

CREATE FUNCTION [Const].[NewLine]

(

)

RETURNS nchar(2)

AS

BEGIN

DECLARE @Result nchar(2)

SELECT @Result = char(13) + char(10)

RETURN @Result

END

DECLARE

@Input nvarchar(max)

,@Quote nvarchar(max)

SET @Input = N'90

10

20

30

40

50

60'

SET @Quote = [Const].NewLine()

SELECT * FROM [Func].[Splite] (@Input, @Quote)

结果

ID Value

----------- ------

1 90

2 10

3 20

4 30

5 40

6 50

7 60

(7 行受影响)