Question

select [CLIENT_SUIT_ID]
  ,[CLIENT_CODE]
  ,[CARD_ID]
  ,[EXPIRE_DATE]
  ,ROW_NUMBER()over(partition by CARD_ID order by CARD_ID ASC)
from [KSS_TMP].[dbo].[CLIENT_SUIT]
where CARD_ID in(
SELECT CARD_ID
    from [KSS_TMP].[dbo].[CLIENT_SUIT]
    where CLIENT_CODE NOT LIKE '080%' AND card_id != '-'
    GROUP BY CARD_ID
    HAVING COUNT(CARD_ID) > 1
    )

你能告诉我如何用JOIN替换IN子句。我试图保持我的条件所以我必须使用IN。

Answer 1

如果您想将in条款替换为join：

select [CLIENT_SUIT_ID]
  ,[CLIENT_CODE]
  ,[CARD_ID]
  ,[EXPIRE_DATE]
  ,ROW_NUMBER()over(partition by CARD_ID order by CARD_ID ASC)
from [KSS_TMP].[dbo].[CLIENT_SUIT] as Tab1 JOIN 
    (SELECT CARD_ID FROM [KSS_TMP].[dbo].[CLIENT_SUIT] 
     where CLIENT_CODE NOT LIKE '080%' AND card_id != '-' 
     GROUP BY CARD_ID 
     HAVING COUNT(CARD_ID) > 1 ) as Tab2 ON Tab1.Card_Id = Tab2.Card_Id

在你的情况下，你真的不需要额外的子查询。你可以使用conditional aggregation的另一个窗口函数在1个查询中完成所有操作：

select * from (
  select *,
    row_number() over (partition by card_id order by card_id) rn,
    count(case when client_code not like '080%' and card_id != '-' 
               then 1 end) over (partition by card_id) cnt
  from client_suit
  ) t
where t.cnt > 1;

如何用JOIN替换IN子句？

1 个答案: