Question

I’m trying to select two distinct numbers id1 and id2 from the following table:

tb_table1(
bigint id1
bigint id2
bigint userid)

if I do

select distinct id1, id2 from tb_table1 

I’ll get, for example, two rows, (111, 222) and (222,111).

I only want one of those rows since I don’t care which column, id1, or id2 that the result gets returned in. Basically, I want distinct pairs where order doesn’t matter.

Thoughts? Thanks in advance.

Answer 1

It would be remiss of me to not point out that this suggests your table is not quite as normalized as it should be – what will you do when users acquire a third id? But anyway.

Using the fact that UNION (as opposed to UNION ALL) will automatically de-duplicate, you could do

SELECT id1, id2 FROM tb_table1 WHERE id1 < id2
UNION
SELECT id2, id1 FROM tb_table1 WHERE NOT id1 < id2