Обсуждение: BUG in date_part

Hello,
    I've discovered a bug in the date_part function. It returns 25 when
doing select date_part('day','2000-03-26'::date);
    Anyway, this isn't happening if casting the date to timestamp
    select date_part('day','2000-03-26'::timestamp);

                Alexander

[ Charset ISO-8859-1 unsupported, converting... ]
>     Hello,
>     I've discovered a bug in the date_part function. It returns 25 when
> doing select date_part('day','2000-03-26'::date);
>     Anyway, this isn't happening if casting the date to timestamp
>     select date_part('day','2000-03-26'::timestamp);

It is fixed in current sources:

    test=>  select date_part('day','2000-03-26'::date);
     date_part
    -----------
            26
    (1 row)

--
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 853-3000
  +  If your life is a hard drive,     |  830 Blythe Avenue
  +  Christ can be your backup.        |  Drexel Hill, Pennsylvania 19026

"Alexandru COSTIN" <acostin@NOSPAMrds.ro> writes:
>     I've discovered a bug in the date_part function. It returns 25 when
> doing select date_part('day','2000-03-26'::date);

Is 2000-03-26 a daylight savings changeover day where you live?
If so, this is a known bug that's fixed for 7.1.  date_part is not
the problem, it's being handed a wrong timestamp input value ---
there's a DST boundary bug in the date-to-timestamp converter.
In my timezone 2000-04-02 was a changeover day, and 7.0.2 gives:

select '2000-04-02'::date::timestamp;
        ?column?
------------------------
 2000-04-01 23:00:00-05
(1 row)

            regards, tom lane