Attach max from group to every row of results ? - mysql

Question

I have similar to the following database structure:
ApplicationService table :
+-----+--------------+----------+---------------+
| id | name | status | application_id
| 1 | Service 1 | 1 | 24
| 2 | Service 2 | 2 | 24
| 3 | Service 3 | 3 | 25
+-----+--------------+----------+----------------
And there is other table with status definitions:
CustomerStatus
+------------+--------------+----------+-----------+
| status_id | name | level | is_closed
| 1 | Status 1 | 2 | 1
| 2 | Status 2 | 1 | 0
| 3 | Status 3 | 3 | 1
+------------+--------------+----------+----------
The status for each row of the ApplicationServices is calculated as the max level status within the records grouped by application_id.
So to get all records from ApplicationServices with statuses would result in something like this:
+-----+--------------+----------+----------------+-------------+-----------
| id | name | status | application_id | status_name | is_closed
| 1 | Service 1 | 1 | 24 | Status 1 | 1
| 2 | Service 2 | 2 | 24 | Status 1 | 1
| 3 | Service 3 | 3 | 25 | Status 3 | 1
+-----+--------------+----------+----------------+-------------+-----------
Is there an efficient way to attach the results with max(level) grouped by application_id to every row of the result set ?

Answer 1

Try this:
SELECT A.id, A.name, A.status, A.application_id, CS.name AS status_name, CS.is_closed
FROM ApplicationService A
INNER JOIN (SELECT AA.application_id, MAX(CS.level) AS maxLevel
FROM ApplicationService AA
INNER JOIN CustomerStatus CS ON AA.status = CS.status_id
GROUP BY AA.application_id
) AS AA ON A.application_id = AA.application_id
INNER JOIN CustomerStatus CS ON AA.maxLevel = CS.level;

Answer 2

See working SQLfiddle here.
You will probably have to perform a sub-query. The subquery would simply be joining the two tables together, grouping/collapsing by application_id and then fetching the maximum is_closed value:
SELECT
t1.status AS `status`,
t1.application_id AS `app_id`,
MAX(t2.is_closed) AS `max_is_closed`
FROM applicationstatus AS t1
LEFT JOIN customerstatus AS t2 ON
t1.status = t2.status_id
GROUP BY t1.application_id
The max is_closed value should then be accessible by the max_is_closed alias when you incorporate the subquery:
SELECT
t1.id AS id,
t1.name AS name,
t1.status AS `status`,
t1.application_id AS application_id,
t2.name AS status_name,
t3.max_is_closed
FROM applicationstatus AS t1
LEFT JOIN customerstatus AS t2 ON
t1.status = t2.status_id
LEFT JOIN
(SELECT
t1.status AS `status`,
t1.application_id AS `app_id`,
MAX(t2.is_closed) AS `max_is_closed`
FROM applicationstatus AS t1
LEFT JOIN customerstatus AS t2 ON
t1.status = t2.status_id
GROUP BY t1.application_id) AS t3 ON
t1.application_id = t3.app_id
The output from the query: