I wrote:> ... So I think this patch is missing a bet by not> accepting equal() expressions.
I've committed this with that logic, a comment explaining exactly whythis is the way to do it, and some other cosmetic improvements.
regards, tom lane