Question

Hi I have a coupon system that uses one table (called “coupons”) to store info about each type of available coupon that one could generate, and another table (generatedcoupons) to store info about each coupon generated. I’m having a hard time comparing info in each table.

The schemas:

table: coupons
+----+------+--------------------+---------------+
|  id| owner|     expiration_date| limit_per_user|
|  15|    34| 2011-09-18 00:00:00|              2|
+----+------+--------------------+---------------+

table: generatedcoupons
+----+----------+------+--------------------+------+--------+
|  id| coupon_id| owner|                date|  used| user_id|
|   1|        15|    34| 2011-09-17 00:00:00| false|     233|
+----+----------+------+--------------------+------+--------+

I’m trying to run queries to display coupon from the point of view of a user (i.e. all queries will have where user_id='$userid'. I can’t figure out how to display all coupons where the limit_per_user has not been met… here’s what I’ve got that doesn’t work:

select * 
from coupons 
where owner=34 
and (count(SELECT * from generatedcoupons where user_id=233 and coupon_id=coupons.id)<limit_per_user)

Answer 1

For MySQL, joining is usually faster and more reliable than subqueries, but makes you think a bit harder. In this case, you need to limit based on the number of joined rows, which requires a post-grouping calculation; fortunately, that’s precisely what the HAVING clause is for:

select coupons.*
from coupons
left join generatedcoupons on user_id = 233 and coupon_id = coupons.id
where coupons.owner = 34
group by coupons.id
having count(generatedcoupons.id) < limit_per_user