复制代码 代码如下: –创建 test 表 ，插入数据 CREATE TABLE test(code varchar(50), [values] varchar(10),[count] int) INSERT test SELECT ‘001’, ‘aa’,1 UNION ALL SELECT ‘001’, ‘bb’,2 UNION ALL SELECT ‘002’, ‘aaa’,4 UNION ALL SEL
复制代码 代码如下:–创建 test 表 ，插入数据CREATE TABLE test(code varchar(50), [values] varchar(10),[count] int)INSERT test SELECT ‘001’, ‘aa’,1UNION ALL SELECT ‘001’, ‘bb’,2UNION ALL SELECT ‘002’, ‘aaa’,4UNION ALL SELECT ‘002’, ‘bbb’,5UNION ALL SELECT ‘002’, ‘ccc’,3; –方法一–将多行合并成一行，并做分组统计SELECT code, [values] = stuff(b.[values].value(‘/R[1]’, ‘nvarchar(max)’),,, ”),[count] FROM (SELECT code,sum([count]) as [count] FROM test GROUP BY code) a CROSS apply ( SELECT [values] =( SELECT N’,’ + [values] FROM test WHERE code = a.code FOR XML PATH(”), ROOT(‘R’), TYPE )) b; –方法二—SQL2005中的新解法 使用XMLSELECT code, data=STUFF((SELECT ‘,’+[values] FROM test t WHERE code=t1.code FOR XML PATH(”)), 1, 1, ”),sum([count]) as [count]FROM test t1GROUP BY code –查询结果–001 aa,bb 3–002 aaa,bbb,ccc 12 drop table test
