SQL Group By ignore shared fields by group - sql

Question

I have product variants that I want to get as a group and ignore any fields that are shared for the one aggregate record that is returned. It is unclear how many or what fields the various variants will share. This is an example table
CREATE TABLE IF NOT EXISTS `products` (
`id` int(6) unsigned NOT NULL,
`name` varchar(200),
`variantId` int(3),
`size` varchar(200),
`color` varchar(200),
PRIMARY KEY (`id`)
) DEFAULT CHARSET=utf8;
INSERT INTO `products` (`id`, `name`, `variantId`, `size`, `color`) VALUES
('1', 'SoccerBalls', '1', '3', 'Black-white'),
('2', 'SoccerBalls', '1', '4', 'Black-white'),
('3', 'SoccerBalls', '1', '5', 'Black-white'),
('4', 'RefereeCard', '2', null, 'yellow'),
('5', 'RefereeCard', '2', null', 'red'),
('6', 'Jersey', '3', 'L', 'Tottenham'),
('7', 'Jersey', '3', 'M', 'Chelsea'),
('8', 'Jersey', '3', 'S', 'Arsenal');
What I want as an out put is something like:
null, SoccerBalls, null, null, Black-white
null, RefereeCard, null, null, null,
null, Jersey, null, null, null

Answer 1

I think you want:
select (case when min(id) = max(id) then min(id) end) as id,
name, variantid,
(case when min(size) = max(size) then min(size) end) as size,
(case when min(color) = max(color) then min(color) end) as color
from products
group by variantid, name;
If the values can be NULL, then you need an additional condition -- if you want to treat NULL as yet-another-value:
select (case when min(id) = max(id) then min(id) end) as id,
name, variantid,
(case when min(size) = max(size) and count(size) = count(*)
then min(size)
end) as size,
(case when min(color) = max(color) and count(color) = count(*)
then min(color)
end) as color
from products
group by variantid, name;
Obviously, you don't need this check on id, because it is the primary key.