Robert Haas <robertmhaas@gmail.com> writes:
> ... But the underlying point here is that
> the only thing you really know about the function is that it's got to
> be a strategy-3 operator in some btree opclass; if that guarantees
> strictness, then so be it -- but I wasn't able to find anything in the
> code or documentation off-hand that supports that contention, so we
> might need to think a bit more about why (or if) this is guaranteed to
> be true.
FWIW, I do not think that follows. If you want to check that the
function is strict, check that explicitly.
It's very likely that in practice, all such functions are indeed strict,
but we don't have an assumption about that wired into the planner.
regards, tom lane