I have a working query using INNER JOIN and a subquery but was wondering if there is a more effient way of writing it.
with prl
as
(
SELECT `number`, creator, notes
FROM ratings
INNER JOIN
projects on ratings.project_id = projects.project_id
WHERE ratings.rating = 5 AND projects.active = 1
)
SELECT prl.`number`, creator, notes
FROM prl
INNER JOIN(
SELECT `number`
HAVING COUNT(creator) > 1
)temp ON prl.`number` = temp.`number`
ORDER BY temp.`number`
projects table
project_id| number | creator | active |
| 1 | 3 | bob | 1 |
| 2 | 4 | mary | 1 |
| 3 | 5 | asi | 1 |
rating table
project_id| notes | rating |
| 1 | note1 | 5 |
| 1 | note2 | 5 |
| 3 | note3 | 5 |
| 1 | note4 | 1 |
| 2 | note5 | 5 |
| 3 | note6 | 2 |
result
| number | creator | notes |
| 3 | bob | note1 |
| 3 | bob | note2 |
CodePudding user response:
It seems like you're using MySQL version that support window function. If so, then try this:
SELECT number, creator, notes
FROM
(SELECT p.number, p.creator, r.notes,
COUNT(creator) OVER (PARTITION BY creator) AS cnt
FROM project p
JOIN rating r ON p.project_id=r.project_id
WHERE r.rating=5
AND p.active = 1) v
WHERE cnt=2;
As far as whether this is more efficient, I'm not really sure because it depends in your table indexes but for a small dataset, I assume this will do well.