Re: Infinite Interval - Mailing list pgsql-hackers

From Joseph Koshakow
Subject Re: Infinite Interval
Date
Msg-id CAAvxfHd3izXmKPyNNZPOROp-+8a56nbmJ3R0DiT9mR7r27qCQA@mail.gmail.com
Whole thread Raw
In response to Re: Infinite Interval  (Joseph Koshakow <koshy44@gmail.com>)
Responses Re: Infinite Interval
List pgsql-hackers
On Mon, Jan 2, 2023 at 1:21 PM Joseph Koshakow <koshy44@gmail.com> wrote:
>
> On Sat, Dec 31, 2022 at 12:09 AM jian he <jian.universality@gmail.com> wrote:
> > In float8, select float8 'inf' / float8 'inf' return NaN. Now in your patch  select interval 'infinity' / float8
'infinity';returns infinity.
 
> > I am not sure it's right. I found this related post
(https://math.stackexchange.com/questions/181304/what-is-infinity-divided-by-infinity).
>
> Good point, I agree this should return an error. We also need to
> properly handle multiplication and division of infinite intervals by
> float8 'nan'. My patch is returning an infinite interval, but it should
> be returning an error. I'll upload a new patch shortly.
>
> - Joe

Attached is the patch to handle these scenarios. Apparently dividing by
NaN is currently broken:
    postgres=# SELECT INTERVAL '1 day' / float8 'nan';
                         ?column?
    ---------------------------------------------------
     -178956970 years -8 mons -2562047788:00:54.775808
    (1 row)

This patch will fix the issue, but we may want a separate patch that
handles this specific, existing issue. Any thoughts?

- Joe

Attachment

pgsql-hackers by date:

Previous
From: Peter Geoghegan
Date:
Subject: Re: Fixing a couple of buglets in how VACUUM sets visibility map bits
Next
From: Dave Cramer
Date:
Subject: Why is char an internal-use category