`TopicResponse` 쿼리 최적화

[60jong]

페이징으로 인해 일대다 관계는 fetch X
-> 한 요청 당 4개의 query가 요청됨. main / topic -> topickeyword / topickeyword -> keyword / topic -> choice)

# 1
SELECT t1_0.*,
       a1_0.*
FROM   topic t1_0
       JOIN member a1_0
         ON a1_0.id = t1_0.author_id
       JOIN topic_keyword t2_0
         ON t1_0.id = t2_0.topic_id
WHERE  t1_0.status =?
       AND NOT EXISTS(SELECT 1
                      FROM   hidden_topic h1_0
                      WHERE  h1_0.topic_id = t1_0.id
                             AND h1_0.member_id =? )
ORDER  BY t1_0.vote_count DESC
LIMIT  ?, ? 

lazy_loading -> batch 요청

# 2
select * from topic_keyword t1_0 where t1_0.topic_id in (?, ?)

# 3
select * from keyword k1_0 where k1_0.id in (?, ?)

# 4
select * from choice c1_0 where c1_0.topic_id in (?, ?)

!!! query를 topic_keyword를 기반으로 시작해도 main + 3 = 4번의 쿼리 나감

[60jong]

Keyword 3 -> 1 변경으로 수정

SELECT t1_0.*,
       a1_0.*
FROM   topic t1_0
       JOIN member a1_0
         ON a1_0.id = t1_0.author_id
       JOIN keyword k
         ON t1_0.id = k.topic_id
WHERE  t1_0.status =?
       AND k.ke
       AND NOT EXISTS(SELECT 1
                      FROM   hidden_topic h1_0
                      WHERE  h1_0.topic_id = t1_0.id
                             AND h1_0.member_id =? )
ORDER  BY t1_0.vote_count DESC
LIMIT  ?, ? 

where 절 내부의 subquery를 join으로 해결하고 픔