SQL合并行

--1. 创建表，添加测试数据
CREATETABLE tb(id int, [value]varchar(10))
INSERT tb SELECT1, 'aa'
UNIONALLSELECT1, 'bb'
UNIONALLSELECT2, 'aaa'
UNIONALLSELECT2, 'bbb'
UNIONALLSELECT2, 'ccc'

--SELECT * FROM tb
/**//*
id value
----------- ----------
1           aa
1           bb
2           aaa
2           bbb
2           ccc

(5 row(s) affected)
*/


--2 在SQL2000只能用自定义函数实现
----2.1 创建合并函数fn_strSum，根据id合并value值
GO
CREATEFUNCTION dbo.fn_strSum(@idint)
RETURNSvarchar(8000)
AS
BEGIN
    DECLARE@valuesvarchar(8000)
    SET@values=''
    SELECT@values=@values+','+ value FROM tb WHERE id=@id
    RETURNSTUFF(@values, 1, 1, '')
END
GO

-- 调用函数
SELECT id, VALUE = dbo.fn_strSum(id) FROM tb GROUPBY id
DROPFUNCTION dbo.fn_strSum

----2.2 创建合并函数fn_strSum2，根据id合并value值
GO
CREATEFUNCTION dbo.fn_strSum2(@idint)
RETURNSvarchar(8000)
AS
BEGIN
    DECLARE@valuesvarchar(8000)   
    SELECT@values=isnull(@values+',', '') + value FROM tb WHERE id=@id
    RETURN@values
END
GO

-- 调用函数
SELECT id, VALUE = dbo.fn_strSum2(id) FROM tb GROUPBY id
DROPFUNCTION dbo.fn_strSum2


--3 在SQL2005中的新解法
----3.1 使用OUTER APPLY
SELECT*
FROM (SELECTDISTINCT id FROM tb) A OUTER APPLY(
        SELECT[values]=STUFF(REPLACE(REPLACE(
            (
                SELECT value FROM tb N
                WHERE id = A.id
                FOR XML AUTO
            ), '<N value="', ','), '"/>', ''), 1, 1, '')
)N

----3.2 使用XML
SELECT id, [values]=STUFF((SELECT','+[value]FROM tb t WHERE id=tb.id FOR XML PATH('')), 1, 1, '')
FROM tb
GROUPBY id

--4 删除测试表tb
droptable tb

/**//*
id          values
----------- --------------------
1           aa,bb
2           aaa,bbb,ccc

(2 row(s) affected)