Re: [RFC] Extend namespace of valid guc names

From: Andres Freund <andres(at)2ndquadrant(dot)com>
To: pgsql-hackers(at)postgresql(dot)org
Subject: Re: [RFC] Extend namespace of valid guc names
Date: 2013-09-14 21:21:05
Message-ID: 20130914212105.GB4071@awork2.anarazel.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 2013-02-25 22:15:33 +0100, Andres Freund wrote:
> Currently guc-file.c allows the following as guc names:
>
> ID {LETTER}{LETTER_OR_DIGIT}*
> QUALIFIED_ID {ID}"."{ID}
>
> That is either one token starting with a letter followed by numbers or
> letters or exactly two of those separated by a dot.
> Those restrictions are existing for neither SET/set_config() via SQL nor
> for postgres -c styles of setting options.
>
> I propose loosening those restrictions to
> a) allow repeatedly qualified names like a.b.c
> b) allow variables to start with a digit from the second level onwards.

The attached patch does a) but not b) as Tom argued it's not allowed in
a plain fashion in SQL either. There you need to quote the variable name.

> Additionally, should we perhaps enforce the same rules for -c and
> set_config()/SET?

The discussion seems to have concluded that this isn't neccessary, so I
haven't included this.

The patch still is pretty trivial - I've looked around and I really
didn't see anything else that requires changes. I haven't even found
documentation to adapt.

Greetings,

Andres Freund

--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

Attachment Content-Type Size
0001-Allow-custom-GUCs-to-be-nested-more-than-one-level-i.patch text/x-patch 1.1 KB

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Pavel Stehule 2013-09-14 21:28:44 Re: PL/pgSQL, RAISE and error context
Previous Message Pavel Stehule 2013-09-14 21:05:23 Re: Assertions in PL/PgSQL