Reputation: 11
I've got the following JPQL :
SELECT a.b.id, a.b.name, a.c.id,a.c.name
left join a.b left join a.c
group by a.b.id,a.b.name,a.c.id,a.c.name
now b and c are both referencing the same table. the generated SQL is doing the left join I asked, and another join for a.b.name and a.c.name (which is unnecessary because the left join includes the name, and it retrieves more results than expected) how do I make the SQL generated not include the unnecessary join?
1 solution came up is not select the names and retrieve them individually by a different query.. but it's not the most elegant way I suppose..
(btw I tried selecting a.b,a.c and group by a.b,a.c but it throws ORA not a group by expression because the generated sql retrieves all rows but group by is only by ID) and the left join is necessary since I want to allow null values.
Thanks a lot.
Upvotes: 0
Views: 289
Reputation: 692081
SELECT a.b.id, a.b.name, a.c.id,a.c.name
The above implicitly creates an inner join between a abd b,a nd another inner join between a and c. The query should be
select b.id, b.name, c.id, c.name
from A a
left join a.b b
left join a.c c
The group by
clause doesn't make any sense, since you have no aggregate in your select clause. group by
would be useful if you had, for example
select b.id, b.name, c.id, c.name, count(c.foo)
from A a
left join a.b b
left join a.c c
group by b.id, b.name, c.id, c.name
Upvotes: 0