使用或逻辑的SQL列参数绑定变量

Question

I apologize if the title is misleading.

I'm trying to avoid using two different queries. With that in mind,

I have the following sample query

  SELECT COUNT (*) COUNT,
         SUM (AMT) AS DED_AMT,
         SUM (SURCOST),
         SUM (DEALSUM),
         NVL (TO_CHAR (SUM (RETAIL)), 'N/A') AS RETAIL,
         MNFCID
    FROM (SELECT B.ID, B.CD, A.*
            FROM OUTPUTS_A A JOIN OUTPUTS_B B ON A.ID = B.ID
           WHERE B.ID = :ID AND B.CD = UPPER (:CD))
GROUP BY ID;

that returns a result you can see in the first screenshot.

Notice, I'm passing two bind variables in the query, :ID, :CD. They have to go together and that's why I'm using AND operator there.

Sometimes, I have MFCID only and not :ID and :CD.

This is the logic I'm thinking about.

1. I would like to modify the query such a way that I should be able to pass MFCID as a bind variable. Let's say :mfcid is the variable I'm passing.

2. If I have the values for :ID and :CD handy, I will pass those values and pass nothing for :mfcid. (Nothing in a sense that I won't pass anything. This field CAN'T be null)

3. If I only have the value for :mfcid handy, I will pass that value and pass nothing for :ID and :CD. (Nothing in a sense that I won't pass anything. This field CAN'T be null)

Either way it should return me the same result.

I have tried putting it this way: AND B.MNFCID = NVL(:MNFCID, B.MNFCID) but it takes forever because it's always true if don't pass anything.

Answer 1

Does this do what you want?

SELECT COUNT(*) COUNT, SUM(AMT) AS DED_AMT, SUM(SURCOST),
       SUM(DEALSUM), NVL(TO_CHAR(SUM(RETAIL)), 'N/A') AS RETAIL,
       MNFCID
FROM OUTPUTS_A A JOIN
     OUTPUTS_B B 
     ON A.ID = B.ID
WHERE (B.ID = :ID OR :ID IS NULL) AND
      (B.CD = UPPER(:CD) OR :CD IS NULL)
GROUP BY B.ID;

If this doesn't work -- and it might not -- you may want to use dynamic SQL so the appropriate indexes can more readily be used.