偶然在电脑里看到以前保存的这个函数，是将一个单独字符串切分成一组字符串，这里分隔符是英文逗号“，” 遇到其他情况只要稍加修改就好了
CREATE FUNCTION dbo.f_splitstr(
@str varchar(8000)
)RETURNS @r TABLE(id int IDENTITY(1, 1), value varchar(5000))
AS
BEGIN
/* Function body */
DECLARE @pos int
SET @pos = CHARINDEX(',', @str)
WHILE @pos > 0
BEGIN
INSERT @r(value) VALUES(LEFT(@str, @pos - 1))
SELECT
@str = STUFF(@str, 1, @pos, ''),
@pos = CHARINDEX(',', @str)
END
IF @str > ''
INSERT @r(value) VALUES(@str)
RETURN
END
截图如下
2011-11-15 10:15:28
今天在园子里看到另外一个实现方法，不妨借鉴一下：
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
/*by kudychen 2011-9-28 */
CREATE function [dbo].[SplitString]
(
@Input nvarchar(max), --input string to be separated
@Separator nvarchar(max)=',', --a string that delimit the substrings in the input string
@RemoveEmptyEntries bit=1 --the return value does not include array elements that contain an empty string
)
returns @TABLE table
(
[Id] int identity(1,1),
[Value] nvarchar(max))
as
begin
declare @Index int, @Entry nvarchar(max)
set @Index = charindex(@Separator,@Input)
while (@Index>0)
begin
set @Entry=ltrim(rtrim(substring(@Input, 1, @Index-1)))
if (@RemoveEmptyEntries=0) or (@RemoveEmptyEntries=1 and @Entry<>'')
begin
insert into @TABLE([Value]) Values(@Entry)
end
set @Input = substring(@Input, @Index+datalength(@Separator)/2, len(@Input))
set @Index = charindex(@Separator, @Input)
end
set @Entry=ltrim(rtrim(@Input))
if (@RemoveEmptyEntries=0) or (@RemoveEmptyEntries=1 and @Entry<>'')
begin
insert into @TABLE([Value]) Values(@Entry)
end
return
end
使用方法
declare @str1 varchar(max), @str2 varchar(max), @str3 varchar(max)
set @str1 = '1,2,3'
set @str2 = '1###2###3'
set @str3 = '1###2###3###'
select [Value] from [dbo].[SplitString](@str1, ',', 1)
select [Value] from [dbo].[SplitString](@str2, '###', 1)
select [Value] from [dbo].[SplitString](@str3, '###', 0)
里面还有个自增的[Id]字段哦，在某些情况下有可能会用上的，例如根据Id来保存排序等等。
例如根据某表的ID保存排序：
update a set a.[Order]=t.[Id]
from [dbo].[表] as a join [dbo].SplitString('1,2,3', ',', 1) as t on a.[Id]=t.[Value]
具体的应用请根据自己的情况来吧