Relational databases as introduced by Codd AFAIK are pretty much just storages for Prolog's fact statements: they map exactly one-to-one. The underlying data st

Relational databases as introduced by Codd AFAIK are pretty much just storages for Prolog's fact statements: they map exactly one-to-one. The underlying data structures used for implementation (B-trees etc.) may or may not be optimal for the Prolog-style queries, of course.