Horton

Re: unsupportable composite type partition keys

Amit Langote <amitlangote09@gmail.com>

View thread View on postgr.es
From: Amit Langote <amitlangote09@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2019-12-24T09:08:48Z
Lists: pgsql-hackers

Attachments

On Mon, Dec 23, 2019 at 6:42 PM Amit Langote <amitlangote09@gmail.com> wrote:
> On Sun, Dec 22, 2019 at 6:13 AM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> > > I wonder whether we couldn't also lift
> > > the restriction against whole-row Vars in partition expressions.
> > > Doesn't seem like there is much difference between such a Var and
> > > a row(...)::table_rowtype expression.
> >
> > I didn't look into that either.  I wouldn't propose back-patching that,
> > but it'd be interesting to try to fix it in HEAD.
>
> Agreed.

I gave that a try and ended up with attached that applies on top of
your delay-loading-relcache-partition-data-2.patch.

Thanks,
Amit

Commits

  1. Fix CheckAttributeType's handling of collations for ranges.

  2. Improve comments in utils/rel.h.

  3. Allow whole-row Vars to be used in partitioning expressions.

  4. Remove equalPartitionDescs().

  5. Load relcache entries' partitioning data on-demand, not immediately.

  6. Disallow partition key expressions that return pseudo-types.

  7. Prevent a rowtype from being included in itself via a range.