On Sat, 2021-03-20 at 15:51 +0000, Ron Clarke wrote: > In SQL Server this is easy, we insert the records into a temporary table with separate Unique > indexes on the id for set a and the ids for set b and put the 'ignore_dup_key' on which tells > SQL Server to ignore duplicate rows and carry on. > > The nearest to this at first appears to be the ON CONFLICT ON CONSTRAINT IGNORE in Postgres. > But this only works with a single constraint, at a time i.e. we can't set the ON CONFLICT ON > CONSTRAINT IGNORE to work with multiple UNIQUE indexes.
