连接两个表并显示不同的记录

Question

我在 MS Access 2010 中有两个表：

部分登出表：

PART_ID DRAWN_DATE LOCATION_ID
 C0001  07/29/2013     501
 C0002  07/29/2013     604
 C0003  08/01/2013     703
 C0004  08/01/2013     807
 C0005  08/02/2013     505
 C0006  08/02/2013     602
 C0007  08/02/2013     707
 C0008  08/03/2013     802
 C0009  08/03/2013     803
 C0001  10/01/2013     605
 C0002  10/02/2013     704
 C0004  10/05/2013     806

零件退货表：

PART_ID RETURN_DATE LOCATION_ID
 C0001  09/04/2013     STORE
 C0002  09/05/2013     STORE
 C0004  09/10/2013     STORE
 C0007  09/12/2013     STORE
 C0008  09/13/2013     STORE
 C0002  10/03/2013     STORE

这是我想要的结果：

PART_ID DRAWN_DATE LOG-OUT LOCATION RETURN_DATE RETURN LOCATION
 C0001  07/29/2013       501        09/04/2013       STORE
 C0001  10/01/2013       605
 C0002  07/29/2013       604        09/05/2013       STORE
 C0002  10/02/2013       704        10/03/2013       STORE
 C0003  08/01/2013       703
 C0004  08/01/2013       807        09/10/2013       STORE
 C0004  10/05/2013       806        
 C0005  08/02/2013       505
 C0006  08/02/2013       602
 C0007  08/02/2013       707        09/13/2013       STORE
 C0008  08/03/2013       802        10/03/2013       STORE
 C0009  08/03/2013       803

但我只能得到这个：

PART_ID DRAWN_DATE LOG-OUT LOCATION RETURN_DATE RETURN LOCATION
 C0001  07/29/2013       501        09/04/2013       STORE
 C0001  10/01/2013       605        `09/04/2013       STORE`
 C0002  07/29/2013       604        09/05/2013       STORE
 `C0002  07/29/2013       604`        10/03/2013       STORE
 C0002  10/02/2013       704        `09/05/2013       STORE`
 `C0002  10/02/2013       704        10/03/2013       STORE`
 C0003  08/01/2013       703
 C0004  08/01/2013       807        09/10/2013       STORE
 C0004  10/05/2013       806        `09/10/2013       STORE`
 C0005  08/02/2013       505
 C0006  08/02/2013       602
 C0007  08/02/2013       707        09/13/2013       STORE
 C0008  08/03/2013       802        10/03/2013       STORE
 C0009  08/03/2013       803

在我写完之后：

SELECT L.PART_ID, L.DRAWN_DATE, L.LOCATION_ID AS [LOG-OUT LOCATION], R.RETURN_DATE, R.LOCATION_ID AS RETURN_LOCATION FROM (SELECT * FROM [PART LOG-OUT] ORDER BY PART_ID) AS L LEFT JOIN (SELECT * FROM [PART RETURN] ORDER BY PART_ID) AS R ON L.PART_ID = R.PART_ID ORDER BY L.PART_ID, L.DRAWN_DATE, R.RETURN_DATE;

有人可以纠正我吗？谢谢！

Answer 1

您的要求有点模糊，所以我可能会离开，但您似乎想将退货与仅退货之前的时间配对，像这样？

SELECT L.PART_ID, L.DRAWN_DATE, L.LOCATION_ID AS [LOG-OUT LOCATION], 
       MIN(R.RETURN_DATE), MIN(R.LOCATION_ID) AS RETURN_LOCATION 
FROM (SELECT * FROM [LOG_OUT LOCATION]) AS L 
LEFT JOIN (SELECT * FROM [PART_RETURN]) AS R 
  ON L.PART_ID = R.PART_ID AND L.DRAWN_DATE < R.RETURN_DATE
GROUP BY L.LOCATION_ID,L.PART_ID,L.DRAWN_DATE
ORDER BY L.PART_ID, L.DRAWN_DATE, MIN(R.RETURN_DATE)

An SQLfiddle to test with.

请注意，由于没有将单次购买与单次退货配对（并且您的问题中没有此类示例），因此配对它们的逻辑非常基本。

Answer 2

当您加入PART_ID 列时，您得到的结果就是您期望得到的结果。

例如，LOG-OUT 表中有两个C0001，它们将连接到PART RETURN 表中的同一个C0001，因为没有任何东西可以区分返回表中的行：

PART LOG-OUT TABLE         -> PART RETURN TABLE
C0001  07/29/2013     501  -> C0001  09/04/2013     STORE
C0001  10/01/2013     605  -> C0001  09/04/2013     STORE

您需要加入另一个标准，或者您需要加入一些更独特的标识符才能获得您正在寻找的结果。

Answer 3

SELECT L.PART_ID, L.DRAWN_DATE, L.LOCATION_ID AS [LOG-OUT LOCATION], R.RETURN_DATE, R.LOCATION_ID AS RETURN_LOCATION FROM (SELECT * FROM [PART LOG-OUT] ORDER BY PART_ID) AS L LEFT OUTER JOIN (SELECT * FROM [PART RETURN] ORDER BY PART_ID) AS R ON L.PART_ID = R.PART_ID ORDER BY L.PART_ID, L.DRAWN_DATE, R.RETURN_DATE;