MIME-Version: 1.0
References: <aa81xZp421DSkHiW@nathan>
 <CADkLM=dhYeJtP+8h8k4ULCu-P1pNHmFyDFBm8B0WU6A6i+NgPQ@mail.gmail.com>
 <abBIu0lYyFYAq5fx@nathan>
 <CADkLM=d-=_hbOUsMc4u_ScLmSNox6EbzFCmaMn7MTh6f_SLU_A@mail.gmail.com>
 <CADkLM=d-yUR8JsM-UmCfeRAhVj+X9MYi0kz9=zQO4EoBkd8H8A@mail.gmail.com>
 <CAA5RZ0tiL01rYm7=WaDGLUYjvz+fQRqmaXyRUv9jKRnR3_6xDQ@mail.gmail.com>
 <CADkLM=fn=RmBD0-h+yjTxhpLtbK9s_QdvT-HsGtUj+-1N=6s0g@mail.gmail.com>
 <CAA5RZ0sf7vYav+Q5GwxGikBZSYUm+4haA5bHC0N67ct570HLcg@mail.gmail.com>
 <abRugSlNTmW0-9Um@nathan>
 <CADkLM=cfpFMpgOGPEyr7U+-Cxzvukap-_s8p-ijMTyRphP=oeg@mail.gmail.com>
 <abhXqx-xffsknv3B@nathan>
In-Reply-To: <abhXqx-xffsknv3B@nathan>
From: Corey Huinker <corey.huinker@gmail.com>
Date: Mon, 16 Mar 2026 16:04:23 -0400
Message-ID: 
 <CADkLM=cuwGZU0i51vn+SUqptXB8QT8rhRRYshns+pX0KxczmAQ@mail.gmail.com>
Subject: Re: Add starelid, attnum to pg_stats and leverage this in pg_dump
To: Nathan Bossart <nathandbossart@gmail.com>
Cc: Sami Imseih <samimseih@gmail.com>, pgsql-hackers@lists.postgresql.org,
	michael@paquier.xyz
Content-Type: multipart/alternative; boundary="000000000000515c37064d29bafb"
Archived-At: 
 <https://www.postgresql.org/message-id/CADkLM%3DcuwGZU0i51vn%2BSUqptXB8QT8rhRRYshns%2BpX0KxczmAQ%40mail.gmail.com>
Precedence: bulk

--000000000000515c37064d29bafb
Content-Type: text/plain; charset="UTF-8"

>
>
> >  CREATE VIEW pg_stats_ext_exprs WITH (security_barrier) AS
> >      SELECT cn.nspname AS schemaname,
> >             c.relname AS tablename,
> > +           s.stxrelid AS tableid,
> >             sn.nspname AS statistics_schemaname,
> >             s.stxname AS statistics_name,
> > +           s.oid AS statid,
> >             pg_get_userbyid(s.stxowner) AS statistics_owner,
> > -           stat.expr,
> > +           expr.expr,
> > +           0 - expr.ordinality AS expr_attnum,
>
> I left the expr_attnum stuff out.  It seems to make this patch quite large
> and complicated, we don't plan to use it for the pg_dump patch, and I'm not
> sure about showing users a "synthetic attnum" that seems to have no other
> point of reference.  Would this information be useful in pg_dump somewhere?
> I'm curious to hear more about the intent.
>

expr_attnum was something that Michael Paquier had lamented that the view
didn't have. There is obviously no present need for it, as pg_dump isn't
being modified for extended stats at all.


> I didn't see much value in adding attnum here given the size of the changes
> to the expected output it produces.
>

Same reasons for putting that in - people had lamented that we couldn't
order the dump by attnum, and ordering by attname feels weird somehow.
Again, we don't presently need it.

--000000000000515c37064d29bafb
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_quote gmail_quote_container"><blockquo=
te class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px =
solid rgb(204,204,204);padding-left:1ex"><br>
&gt;=C2=A0 CREATE VIEW pg_stats_ext_exprs WITH (security_barrier) AS<br>
&gt;=C2=A0 =C2=A0 =C2=A0 SELECT cn.nspname AS schemaname,<br>
&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0c.relname AS tablename,=
<br>
&gt; +=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0s.stxrelid AS tableid,<br>
&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0sn.nspname AS statistic=
s_schemaname,<br>
&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0s.stxname AS statistics=
_name,<br>
&gt; +=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0s.oid AS statid,<br>
&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0pg_get_userbyid(s.stxow=
ner) AS statistics_owner,<br>
&gt; -=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0stat.expr,<br>
&gt; +=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0expr.expr,<br>
&gt; +=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A00 - expr.ordinality AS expr_=
attnum,<br>
<br>
I left the expr_attnum stuff out.=C2=A0 It seems to make this patch quite l=
arge<br>
and complicated, we don&#39;t plan to use it for the pg_dump patch, and I&#=
39;m not<br>
sure about showing users a &quot;synthetic attnum&quot; that seems to have =
no other<br>
point of reference.=C2=A0 Would this information be useful in pg_dump somew=
here?<br>
I&#39;m curious to hear more about the intent.<br></blockquote><div><br></d=
iv><div>expr_attnum was something that Michael Paquier had lamented that th=
e view didn&#39;t have. There is obviously no present need for it, as pg_du=
mp isn&#39;t being modified for extended stats at all.</div><div></div><div=
>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px =
0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">I didn&#39;t=
 see much value in adding attnum here given the size of the changes<br>
to the expected output it produces.<br></blockquote><div><br></div><div>Sam=
e reasons for putting that in - people had lamented that we couldn&#39;t or=
der the dump by attnum, and ordering by attname feels weird somehow. Again,=
 we don&#39;t presently need it.</div></div></div>

--000000000000515c37064d29bafb--