On Tue, Oct 7, 2008 at 9:58 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> * Instead of the above, we could try to make
> ROW(some columns) = ANY (array variable)
> work. This is shorter than the above syntax and would presumably have
> a lot less overhead too. But it doesn't work right now, not even for
> named rowtypes much less anonymous ones.
By extension, would this also mean things like
select row(1,2,3)::foo = foo from foo;
or
select (1,2,3)::foo = (1,2,3)::foo;
or
select (1,2,3) = (1,2,3)::foo;
Would work (presumably as row-wise comparison does)? Also,
create index foo_idx on foo(foo);
select * from foo where (1,2,3)::foo = foo;
would be very nice.
> I'm thinking that addressing these pieces would be a generally good
> thing to do, above and beyond potential uses in recursive queries.
absolutely.
merlin