Reputation: 470
I am trying to construct a double count left join query in Slick 3.1 similar to this:
SELECT
COUNT(p.id) AS replies,
COUNT(f.id) AS images
FROM posts AS p
LEFT JOIN files AS f
ON p.id = f.post_id
WHERE thread = :thread_id
Join part of the query is quite simple and looks like this:
val joinQ = (threadId: Rep[Long]) =>
(postDAO.posts joinLeft fileRecordDAO.fileRecords on (_.id === _.postId))
.filter(_._1.thread === threadId)
Using joinQ(1L).map(_._1.id).length
generates COUNT(1)
which counts all rows - not the result I want to obtain. Using joinQ(1L).map(_._1.id).countDistinct
generates COUNT(DISTINCT p.id)
which is somewhat what I'm looking for, but trying to do two of these generates this monstrosity:
select x2.x3, x4.x5
from (select count(distinct x6.`id`) as x3
from `posts` x6
left outer join `files` x7 on x6.`id` = x7.`post_id`
where x6.`thread` = 1) x2,
(select
count(distinct (case when (x8.`id` is null) then null else 1 end)) as x5
from `posts` x9
left outer join `files` x8 on x9.`id` = x8.`post_id`
where x9.`thread` = 1) x4
Here's the double countDistinct
code:
val q = (threadId: Rep[Long]) => {
val join = joinQ(threadId)
val q1 = join.map(_._1.id).countDistinct // map arg type is
val q2 = join.map(_._2).countDistinct // (Post, Rep[Option[FileRecord]])
(q1, q2)
}
Any ideas? :)
Upvotes: 2
Views: 1110
Reputation: 586
Count is aggregation function, so you also need grouping by some field (e.x. id). Try next query:
def joinQ(threadId: Long) = (postDAO.posts joinLeft fileRecordDAO.fileRecords on (_.id === _.postId))
.filter(_._1.thread === threadId)
val q = (threadId: Long) => {
joinQ(threadId).groupBy(_._1.id).map {
case (id, qry) => (id, qry.map(_._1.id).countDistinct, qry.map(_._2.map(_.id)).countDistinct)
}
}
It generates next sql:
SELECT x2.`id`,
Count(DISTINCT x2.`id`),
Count(DISTINCT x3.`id`)
FROM `posts` x2
LEFT OUTER JOIN `files` x3
ON x2.`id` = x3.`post_id`
WHERE x2.`thread` = 10
GROUP BY x2.`id`
Upvotes: 0