On Tue, Feb 15, 2011 at 1:40 PM, Heikki Linnakangas
<heikki.linnakangas@enterprisedb.com> wrote:
> I'm actually surprised we don't need to distinguish them in more places, but
> nevertheless it feels like we should have that info available more
> conveniently, and without requiring a catalog lookup like get_rel_relkind()
> does. At first I thought we should add a field to RelOptInfo, but that
> doesn't help transformLockingClause. Adding the field to RangeTblEntry seems
> quite invasive, and it doesn't feel like the right place to cache that kind
> of information anyway. Thoughts on that?
Maybe it should be a new RTEKind.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company