Reputation: 85
i have the following SQL query and i was wondering if i can make the SUM() part of it dynamic so that i don't have to put category_id (2 and 3)
SELECT
a.project_id,
COUNT(a.id) AS Total,
SUM(CASE WHEN a.category_id = 2 AND a.`status` < 80 THEN 1 ELSE 0 END) AS 'Bugs En cours',
SUM(CASE WHEN a.category_id = 2 AND a.`status` >= 80 THEN 1 ELSE 0 END) AS 'Bugs Resolu',
SUM(CASE WHEN a.category_id = 3 AND a.`status` < 80 THEN 1 ELSE 0 END) AS 'Ameliorations En cours',
SUM(CASE WHEN a.category_id = 3 AND a.`status` >= 80 THEN 1 ELSE 0 END) AS 'Ameliorations Resolu'
FROM bugs a
GROUP BY a.project_id
HAVING COUNT(a.id) > 0
The goal is to list the project id
and the count
of different Anomalies depending on the category_id
and on the status
('En cours' OR 'Resolu')
The issue in this query is that if we added another category, i will have to manually edit this query which is not ideal.
Upvotes: 1
Views: 64
Reputation: 29647
The dynamic sql below is building the conditional sums from on a reference table with the categories.
SET @sums := ( SELECT GROUP_CONCAT( CONCAT( ', SUM(CASE WHEN category_id = ', category_id, ' AND `status` < 80 THEN 1 ELSE 0 END)', ' AS `', category_name, ' En cours`', CHAR(10), ', SUM(CASE WHEN category_id = ', category_id, ' AND `status` >= 80 THEN 1 ELSE 0 END)', ' AS `', category_name, ' Resolu`', CHAR(10) ) ORDER BY category_id SEPARATOR '') as sums FROM categories ); SET @dynsql := CONCAT('SELECT', CHAR(10), ' bug.project_id', CHAR(10), ', COUNT(bug.id) AS Total', CHAR(10), @sums, 'FROM bugs bug', CHAR(10), 'GROUP BY bug.project_id', CHAR(10), 'HAVING COUNT(bug.id) > 0;'); -- SELECT @dynsql AS dynsql; PREPARE dyn_stmt FROM @dynsql; EXECUTE dyn_stmt; DEALLOCATE PREPARE dyn_stmt;
project_id | Total | Conneries En cours | Conneries Resolu | Bugs En cours | Bugs Resolu | Ameliorations En cours | Ameliorations Resolu |
---|---|---|---|---|---|---|---|
0 | 5 | 1 | 0 | 1 | 1 | 1 | 1 |
Demo on db<>fiddle here
Upvotes: 1