3

我正在尝试将这两个表连接在一起以显示基数中的记录和历史中的空值。我遇到的问题是左外连接。当 BO.val2 = 0 时：我希望连接仅使用 BO.VAL5 = TR.VAL5 但是当 BO.val2 = TR.val2 我希望它使用它来连接。下面的代码只是一个想法。如果您对如何执行此操作有任何建议，那就太好了！

 SELECT DISTINCT      
   BO.RUN_DATE,
   BO.val2,
   BO.val3,
   BO.val4,
   BO.VAL5
   TR.DTDATE,
   TR.val2,
   TR.val3,
   TR.val4,
   TR.val5
 FROM BASE BO                    
 LEFT OUTER JOIN HISTORY AS TR ON CASE 
   WHEN BO.val2 = 0 and  BO.VAL5 = TR.VAL5  THEN ????
   WHEN BO.val2 = TR.val2 and BO.VAL5 = TR.VAL5 then ???
   ELSE ??
   END 
 WHERE TRUNC(BO.POST_DATE)= TRUNC(SYSDATE)
 ORDER BY BO.VAL4  ;
4

2 回答 2

1

您的两个变体有BO.VAL5 = TR.VAL5共同点。

逻辑的第二部分是BO.val2 = TR.val2or BO.val2 = 0。

所以，最终条件应该是：

SELECT DISTINCT
    BO.RUN_DATE,
    BO.val2,
    BO.val3,
    BO.val4,
    BO.VAL5,
    TR.DTDATE,
    TR.val2,
    TR.val3,
    TR.val4,
    TR.val5
FROM
    BASE AS BO
    LEFT JOIN HISTORY AS TR
        ON BO.VAL5 = TR.VAL5 AND (BO.val2 = 0 OR BO.val2 = TR.val2)
WHERE
    TRUNC(BO.POST_DATE) = TRUNC(SYSDATE)
ORDER BY BO.VAL4;
于 2017-02-16T05:28:53.617 回答
0

您需要做的是两次加入历史，使用您的不同条件。然后您可以使用 COALESCE() 返回数据

  SELECT DISTINCT 
       BO.RUN_DATE,
       BO.val2,
       BO.val3,
       BO.val4,
       BO.VAL5
       COALESCE(TR.DTDATE, TR2.DTDATE) AS DTDATE,
       COALESCE(TR.val2, TR2.val2) AS val2,
       COALESCE(TR.val3, TR2.val3) AS val3,
       COALESCE(TR.val4, TR2.val4) AS val4,
       COALESCE(TR.val5, TR2.val5) AS val5
  FROM BASE BO                    
  LEFT OUTER JOIN History AS TR ON TR.VAL5 = BO.VAL5
  LEFT OUTER JOIN History AS TR2 ON TR2.VAL2 = BO.VAL2
  WHERE TRUNC(BO.POST_DATE)= TRUNC(SYSDATE)
  ORDER BY BO.VAL4 ;
于 2017-02-15T23:29:45.667 回答