Re: Make name optional in CREATE STATISTICS

From: Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org>
To: Michael Paquier <michael(at)paquier(dot)xyz>
Cc: Dean Rasheed <dean(dot)a(dot)rasheed(at)gmail(dot)com>, Matthias van de Meent <boekewurm+postgres(at)gmail(dot)com>, Simon Riggs <simon(dot)riggs(at)enterprisedb(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Make name optional in CREATE STATISTICS
Date: 2022-07-22 13:06:46
Message-ID: 20220722130646.3da5oh7tgz5qgooi@alvherre.pgsql
Views: Raw Message | Whole Thread | Download mbox | Resend email
Lists: pgsql-hackers

On 2022-Jul-22, Michael Paquier wrote:

> So this indeed has as effect to make possible the use of CONCURRENTLY
> for DATABASE and SYSTEM only within the parenthesized grammar. Seeing
> the simplifications this creates, I'd agree with dropping this part of
> the grammar.

Actually, looking at the grammar again I realized that the '('options')'
part could be refactored; and with that, keeping an extra production for
REINDEX DATABASE CONCURRENTLY is short enough. It is removed from
REINDEX SYSTEM, but that's OK because that doesn't work anyway.

I added the new test lines you proposed and amended the docs; the result
is attached.

Initially I wanted to use the "optional list of options" for all
utilities that have similar constructions, (VACUUM, ANALYZE, CLUSTER,
EXPLAIN) but it is not possible because their alternative productions
accept different keywords, so it doesn't look possible.

Álvaro Herrera 48°01'N 7°57'E —
"If you have nothing to say, maybe you need just the right tool to help you
not say it." (New York Times, about Microsoft PowerPoint)

Attachment Content-Type Size
v2-0001-Rework-grammar-for-REINDEX.patch text/x-diff 7.3 KB

In response to


Browse pgsql-hackers by date

  From Date Subject
Next Message Matthias van de Meent 2022-07-22 13:16:56 Re: Pluggable toaster
Previous Message Pavel Stehule 2022-07-22 13:04:11 proposal - enhancing plpgsql's FOREACH statement for support json type