My Sample data:
id role_id permission_id access_granted user_id
------------------------------------------------------------------
1 1 10 1 115
------------------------------------------------------------------
2 1 5 1 115
------------------------------------------------------------------
3 2 10 0 115
------------------------------------------------------------------
4 2 8 0 115
------------------------------------------------------------------
Group by with permission_id and the result should have value true row for the repeated permission_id. As we can see the query is grouped by permission_id and the row with value true is selected for permission_id 10 and for other rows permission id is not being repeated so the value false is being picked as it is.
Expected result:
id role_id permission_id access_granted user_id
------------------------------------------------------------------
1 1 10 1 115
------------------------------------------------------------------
2 1 5 1 115
------------------------------------------------------------------
4 2 8 0 115
------------------------------------------------------------------
Below is a query that I tried to fetch the above result.
select id, role_id, permission_id,
case when max(access_granted) = true then true else false end as access_granted,
user_id as user_id
from temp_permission group by permission_id;
Need optimized and authentic query that will not fail in any case.
Create table query
CREATE TABLE temp_permission (
id int auto_increment primary key,
role_id int,
permission_id int,
access_granted tinyInt(1),
user_id int
);
Dummy data query for the table as per the screenshot
INSERT INTO `temp_permission` (`id`, `role_id`, `permission_id`, `access_granted`, `user_id`) VALUES ('1', '1', '10', '1', '115');
INSERT INTO `temp_permission` (`id`, `role_id`, `permission_id`, `access_granted`, `user_id`) VALUES ('2', '1', '5', '1', '115');
INSERT INTO `temp_permission` (`id`, `role_id`, `permission_id`, `access_granted`, `user_id`) VALUES ('3', '2', '10', '0', '115');
INSERT INTO `temp_permission` (`id`, `role_id`, `permission_id`, `access_granted`, `user_id`) VALUES ('4', '2', '8', '0', '115');
if you are unable to insert data into access_granted column replace 1 = b'1' & 0 = b'0'
You can simply order the rows based on permission_id
and access_granted
in descending order, and then select distinct rows based on permission_id
. Something like:
WITH ordered_rows AS (
SELECT id, role_id, permission_id, access_granted, user_id,
ROW_NUMBER() OVER (PARTITION BY permission_id ORDER BY access_granted DESC) as row_num
FROM temp_permission
)
SELECT id, role_id, permission_id, access_granted, user_id
FROM ordered_rows
WHERE row_num = 1;