MIME-Version: 1.0
References: <CA+FnnTwmt7BOqwZnsB5LOq47U+wzRc86wXaea9B0KEXfK2O+Kw@mail.gmail.com>
 <CAKFQuwa2ef9Bxz79Q8JtVNxmMABN829EJqiLKXxB5=5AEHyr9w@mail.gmail.com>
 <CAKFQuwa9PsXitAWuG1hg-8EPOu0AdB2TiANaBRJfgSrdjioa+Q@mail.gmail.com>
 <CA+FnnTz51Ond5HGkDgKvRB6ERi3pr5Q1wTnwoMK8rCwsm79_sA@mail.gmail.com>
 <CAKFQuwa7rK-NfVwHT7_Nob6S=GN-tKK7F1+eLQ-FQZMsW3gd8g@mail.gmail.com>
 <CA+FnnTw0bNsDBKFEQKsrjFZKqiZgRVZ27zhgXs5sqCLW=bwBVw@mail.gmail.com>
 <CA+FnnTwEWFq1BopZ4o_kEg3SuZvGE+Xhg1sur=c_HjcZ_FjTag@mail.gmail.com> <CAKFQuwbTpwDiUy=azMGk8tMo7w2ysWC1h02NmURCqGRqcAhSCw@mail.gmail.com>
In-Reply-To: <CAKFQuwbTpwDiUy=azMGk8tMo7w2ysWC1h02NmURCqGRqcAhSCw@mail.gmail.com>
From: Igor Korot <ikorot01@gmail.com>
Date: Sat, 28 Feb 2026 23:34:58 -0600
Message-ID: <CA+FnnTwaK-Z26KeqoPsAkKPJHo9OjMJdSse8tXE-KSFV9y0g_Q@mail.gmail.com>
Subject: Re: Where the info is stored
To: "David G. Johnston" <david.g.johnston@gmail.com>
Cc: "pgsql-generallists.postgresql.org" <pgsql-general@lists.postgresql.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://www.postgresql.org/message-id/CA%2BFnnTwaK-Z26KeqoPsAkKPJHo9OjMJdSse8tXE-KSFV9y0g_Q%40mail.gmail.com>
Precedence: bulk

Hi, David,

On Sat, Feb 28, 2026 at 7:42=E2=80=AFPM David G. Johnston
<david.g.johnston@gmail.com> wrote:
>
> On Sat, Feb 28, 2026 at 6:34=E2=80=AFPM Igor Korot <ikorot01@gmail.com> w=
rote:
>>
>> And why there is no WHERE populated?
>>
>> Thank you.
>>
>> On Sat, Feb 28, 2026 at 7:05=E2=80=AFPM Igor Korot <ikorot01@gmail.com> =
wrote:
>> >
>> > Hi, David,
>> >
>> > On Sat, Feb 28, 2026 at 7:02=E2=80=AFPM David G. Johnston
>> > <david.g.johnston@gmail.com> wrote:
>> > >
>> > > On Saturday, February 28, 2026, Igor Korot <ikorot01@gmail.com> wrot=
e:
>> > >>
>> > >> FROM pg_constraint co, pg_namespace n, pg_class
>> > >>
>> > >> As you can see only the constraint name and the tablespace are
>> > >> populated correctly.
>> > >
>> > >
>> > > Constraints don=E2=80=99t have included columns.  Only indexes do.  =
You need to query the index, not the constraint.
>> >
>> > I literally copied your query into my code and it didn't populated
>> > anything...
>> >
>> > Am I missing something?
>>
>
> I trimmed your query to emphasize/point-out that you were querying pg_con=
straint and that doing so to find included columns is doomed to failure (I =
suppose it could have been used to find the index, but in this case it wasn=
't.  I haven't explored that approach.).  You should step back and consider=
 why you thought the fragment I included in my reply, a bare FROM clause, w=
ould somehow be executable since it is in no way a valid query.

draft=3D#  WITH idx AS(  SELECT i.indexrelid, i.indrelid, i.indnkeyatts,
i.indkey::int2[] AS indkey FROM pg_index i, pg_class ic, pg_namespace
ns WHERE ic.oid =3D i.indexrelid AND ns.oid =3D ic.relnamespace AND
ns.nspname =3D 'public' AND ic.relname =3D 'leagues_new' ), ords AS (
SELECT idx.indexrelid, idx.indrelid, idx.indnkeyatts, s.ord,
idx.indkey[s.ord] AS attnum FROM idx CROSS JOIN LATERAL
generate_subscripts(idx.indkey, 1) AS s(ord) ) SELECT ns.nspname,
ic.relname, tc.relname, a.attname, CASE WHEN ords.ord <
ords.indnkeyatts THEN 'key' ELSE 'include' END, ords.ord + 1 AS
index_position  FROM pg_attribute a, ords, pg_class ic, pg_namespace
ns, pg_class tc WHERE a.attrelid =3D ords.indrelid AND a.attnum =3D
ords.attnum AND NOT a.attisdropped AND ic.oid =3D ords.indexrelid AND
ns.oid =3D ic.relnamespace AND tc.oid =3D ords.indrelid AND ords.ord >
ords.indnkeyatts;
 nspname | relname | relname | attname | case | index_position
---------+---------+---------+---------+------+----------------
(0 rows)

draft=3D#

This is an exact replica of your query from the first post where you put it=
.

I only removed AS statements.

As you can see 0 rows are returned.

Thank you.

>
> David J.
>