On Fri, Sep 18, 2009 at 3:06 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> On Fri, Sep 18, 2009 at 1:58 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>> Robert Haas <robertmhaas@gmail.com> writes:
>>>> Mmm, I like that. Putting that bunch of hairy logic in a subroutine
>>>> instead of repeating it in several places definitely seems better. I
>>>> don't really like the name "clause_matches_join", though.
>
>>> It was the first thing that came to mind ... got a better idea?
>
>> clause_has_well_defined_sides()?
>
> Nah ... they're "well defined" in any case, they might just not be what
> we need for the current join. As an example,
>
> (a.f1 + b.f2) = c.f3
>
> would be usable if joining {A B} to {C}, but not when joining
> {A} to {B C}.
The clauses are well-defined, but they don't have well-defined sides.
I see now what you're going for with clause_matches_join, but
"matches" is a pretty broad term, IMO.
...Robert