Say I have this table:
id | name
-------------
1 | john
2 | steve
3 | steve
4 | john
5 | steve
I only want the rows that are unique compared to the previous row, these:
id | name
-------------
1 | john
2 | steve
4 | john
5 | steve
I can partly achieve this by using this query:
SELECT *, (
SELECT `name` FROM demotable WHERE id=t.id-1
) AS prevName FROM demotable AS t GROUP BY prevName ORDER BY id ASC
But when I am using a query with multiple UNIONs and stuff, this gets way to complicated. Is there an easy way to do this (like GROUP BY, but than more specific)?
This should work, but I don't know if it's simpler :
select demotable.*
from demotable
left join demotable as prev on prev.id = demotable.id - 1
where demotable.name != prev.name