> > If so, maybe the docstring for `functionp' should be changed (after
> > all, it was what I modeled this one off of):
> >
> >     Non-nil if OBJECT is a function.
>
> Yes, I think so.

The patch is attached.