On Mon, Jul 09, 2001 at 08:22:43PM -0400, Tom Lane wrote:
> AFAIR, cnfify doesn't modify its inputs. But watch out for the
> difference between explicit and implicit ANDing.
OK, I'm pretty sure I got it right now.
> In practice, I seem to recall that VACUUM is broken for partial indexes
> anyway, specifically because it does not pay attention to partial-ness:
OK, I don't feel too confident playing with the vacuum code, seems to be a
real quick way to destroy your database.
To actually be able to use ExecInsertIndexTuples, you need to create an
EState, a ResultRelInfo and put each tuple in a TupleSlot for a while. Does
it sound like I'm on the right track? That's quite a few changes.
Isn't someone else playing with the vacuum code for 7.2 anyway (for
background vacuums)? We'd better make sure we don't clash.
Back to other issues. pg_dump now works for partial indicies, as long as the
pg_get_expr function is defined. To make that an internal function I have
add it to pg_proc.h and initdb again, right?
http://svana.org/kleptog/pgsql/partial-indicies.patch
http://svana.org/kleptog/pgsql/expr.c
--
Martijn van Oosterhout <kleptog@svana.org>
http://svana.org/kleptog/
> It would be nice if someone came up with a certification system that
> actually separated those who can barely regurgitate what they crammed over
> the last few weeks from those who command secret ninja networking powers.
