On 2020-05-20 01:47, Tom Lane wrote:
I wrote:
However, we do have to have a benefit to show those people whose
queries we break. Hence my insistence on having a working AS fix
(or some other benefit) before not after.
I experimented with this a bit more, and came up with the attached.
It's not a working patch, just a set of grammar changes that Bison
is happy with. (Getting to a working patch would require fixing the
various build infrastructure that knows about the keyword classification,
which seems straightforward but tedious.)
What I was hoping to get out of this was to resolve some of the weird
precedence hacks that were blamed on postfix operators. But building on
your patch, the best I could achieve was
-%nonassoc IDENT GENERATED NULL_P PARTITION RANGE ROWS GROUPS PRECEDING
FOLLOWING CUBE ROLLUP
+%nonassoc IDENT PARTITION RANGE ROWS GROUPS PRECEDING FOLLOWING CUBE
ROLLUP
which is a pretty poor yield.
Maybe this isn't worth it after all.
--
Peter Eisentraut http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
