Re: [HACKERS] Interval aggregate regression failure

Michael Glaesemann wrote:
> > Yea, just an optimization, but I was worried that the computations
> > might
> > throw problems for certain numbers, so I figured I would only
> > trigger it
> > when necessary.
>
> Thanks for the explanation. Helps me know I might actually be
> learning this.
>
> > Patch attached. It also fixes a regression test output too.
>
> Thanks for the patch. I'll look at it more closely tonight.

OK, here is a much nicer patch. The fix is to do no rounding, but to
find the number of days before applying the factor adjustment. It
passes all your tests here:

test=> select '41 mon 10:00:00'::interval / 10 as "pos";
pos
------------------------
4 mons 3 days 01:00:00
(1 row)

test=> select interval '41 mon 12 days 360:00' * 0.3 as product_a
, interval '41 mon -12 days -360:00' * 0.3 as product_b
, interval '-41 mon 12 days 360:00' * 0.3 as product_c
, interval '-41 mon -12 days -360:00' * 0.3 as product_d;
product_a | product_b |
product_c | product_d
--------------------------+--------------------------+-----------------------------+------------------------------

1 year 12 days 122:24:00 | 1 year 6 days -122:24:00 | -1 years -6 days
+122:24:00 | -1 years -12 days -122:24:00
(1 row)

test=> select interval '41 mon 12 days 360:00' / 10 as quotient_a
, interval '41 mon -12 days -360:00' / 10 as quotient_b
, interval '-41 mon 12 days 360:00' / 10 as quotient_c
, interval '-41 mon -12 days -360:00' / 10 as quotient_d;
quotient_a | quotient_b | quotient_c
| quotient_d
------------------------+-------------------------+---------------------------+---------------------------

4 mons 4 days 40:48:00 | 4 mons 2 days -40:48:00 | -4 mons -2 days
+40:48:00 | -4 mons -4 days -40:48:00
(1 row)

What do you see on your platform?

--
Bruce Momjian bruce(at)momjian(dot)us
EnterpriseDB http://www.enterprisedb.com

+ If your life is a hard drive, Christ can be your backup. +