与MysQL JOIN的区别总和

Question

I have two tables tableA and tableB.

tableA
-------
ID  sid     amount

1   swerdd  20    
2   swerdd  30    
3   fdff    40

tableB
------

new_id  token    date

10      swerdd  2014 -10-12    
11      swerdd  2014 -10-13    
12      swerdd  2014 -10-13

Now when I use the query

SELECT * FROM tableA JOIN tableB on tableA.sid = tableB.token 

is giving multiple rows and so my other query

SELECT SUM(amount) FROM tableA JOIN tableB on tableA.sid = tableB.token 

is giving wrong results. How can I select the DISTINCT SUM ?

EDIT : The thing is that I have a many to many relation here as the sid and token is repeating. So in a nutshell the question is how can I calculate distinct sum in a many to many join

Answer 1

A couple of approaches that don't involve a join operation...

An EXISTS predicate...

SELECT SUM(a.reward_amount) 
  FROM tableA a
 WHERE EXISTS (SELECT 1 FROM tableB b WHERE b.token = a.sid) 

An IN (subquery) predicate...

SELECT SUM(a.reward_amount) 
  FROM tableA a
 WHERE a.sid IN (SELECT b.token FROM tableB b)

For a join operation, one option is to use an inline view to get a distinct list of token...

SELECT SUM(a.reward_amount)
  FROM tableA a
  JOIN ( SELECT b.token
           FROM tableB b
          GROUP BY b.token
       ) c
    ON c.token = a.sid

Answer 2

To get distinct sum you can use below query

SELECT reward_amount
FROM tableA 
     JOIN tableB on tableA.sid = tableB.token
GROUP BY reward_amount;