Feedback-ID: i0fe9450f:Fastmail
Date: Fri, 14 Nov 2025 15:25:27 +0900
From: Michael Paquier <michael@paquier.xyz>
To: Corey Huinker <corey.huinker@gmail.com>
Cc: jian he <jian.universality@gmail.com>, Tomas Vondra <tomas@vondra.me>,
	pgsql-hackers@lists.postgresql.org, tgl@sss.pgh.pa.us
Subject: Re: Extended Statistics set/restore/clear functions.
Message-ID: <aRbLV7Tru3ulO5ru@paquier.xyz>
References: <aQ55MDBjbO8_0fnv@paquier.xyz>
 <CADkLM=eGwDZX9DQ_GouzqjMG7x_kCedBS6br2SciJ3wjwivfZg@mail.gmail.com>
 <aRLumVDIQHUTKQYG@paquier.xyz>
 <CACJufxHZzUeu_fyem2oA_FSrjk0ky5vn_h1Dw68=UVLekXPjXA@mail.gmail.com>
 <CACJufxHP+wZ9Th2U9JW27cCq-me222e76pv7vcjqt+oCL03m5Q@mail.gmail.com>
 <CADkLM=c8-U4GLMw5VdeDdfp1ae6BW=PCfEQqAky04iZMbckCFw@mail.gmail.com>
 <CACJufxFo5b0FvmV4CmOb74NFrXm_znZHh7bLiiVkAzVzTnJiVw@mail.gmail.com>
 <CADkLM=fky861BrifcuWc0CbWfo3yMbRB+m1jG4qqXG_3hnqdtA@mail.gmail.com>
 <CACJufxFni1oB-VFuLbVs9VSgVx0czvoZyoW0z_tJ2kq4E3SXYw@mail.gmail.com>
 <CADkLM=d_mmTn3+ChL-z_ZomwOKT-BssQU+=6_a8WDv6A4b7uAA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha512;
	protocol="application/pgp-signature"; boundary="VCBbQdVohHX4k75v"
Content-Disposition: inline
In-Reply-To: 
 <CADkLM=d_mmTn3+ChL-z_ZomwOKT-BssQU+=6_a8WDv6A4b7uAA@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/aRbLV7Tru3ulO5ru%40paquier.xyz>
Precedence: bulk


--VCBbQdVohHX4k75v
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Fri, Nov 14, 2025 at 12:49:23AM -0500, Corey Huinker wrote:
> Negative numbers represent the Nth expression defined in the extended
> statistics object. So if you have extended statistics on a, b, length(a),
> length(b) then you can legally have -1 and -2 in the attributes, but
> nothing lower than that.
>
> See functions pg_ndistinct_validate_items() and
> pg_depdendencies_validate_deps() as these check the attributes in the value
> against the definition of the extended stats object.

Exactly.  Extended stats on system columns don't work because they
don't really make sense as we want to track correlations between the
attributes defined, and these reflect internal states:
create table poo (a int, b int);
create statistics poos (ndistinct ) ON cmax, a from poo;
ERROR:  0A000: statistics creation on system columns is not supported

Note that the expressions are also stored in pg_stats_ext_exprs.

> I'm trying to implement those test cases, but I may have missed some.

I've found a lot of them with coverage-html during a previous lookup.
I'd like to think that we should aim for something close to 100%
coverage for the two input functions.

> Implemented many, but not all of these suggestions.

Thanks for the new versions, I'll also look at all these across the
next couple of days.  Probably not at 0005~ for now.
--
Michael

--VCBbQdVohHX4k75v
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----
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=8wUv
-----END PGP SIGNATURE-----

--VCBbQdVohHX4k75v--