Re: Date_part & cast.

Christopher Sawtell writes:

> > db=> select date_part( 'dow', date '20000421' );
> > date_part
> > ---------
> > 5
> > (1 row)
> >
> > db=> select date_part( 'dow', date ( '20000421' ) );
> > date_part
> > ---------
> > 6
> > (1 row)

This appears to work now.

> > If yes, I could not find any explanation for the second result which also
> > differ from "select date_part( 'dow', date (20000421) );"

This is the reason:

peter=# select date (20000421);
date
------------
1970-08-20
(1 row)

IMO, that's anywhere from non-obvious to violation of standard to
dangerous, but of course those who stick to the official, SQL approved,
PostgreSQL endorsed date input format

DATE '2000-04-21'

shouldn't have problems like this.

> btw, the days of the week start with Sunday = 1

No, Sunday is 0.

> template1=# select date_part( 'dow', date '19271124' );
> date_part
> -----------
> 4
> (1 row)
>
> template1=# select date_part( 'dow', date '19271125' );
> date_part
> -----------
> 5
> (1 row)
>
> template1=# select date_part( 'dow', date '19271126' );
> date_part
> -----------
> 0
> (1 row)
>
> template1=# select date_part( 'dow', date '19271127' );
> date_part
> -----------
> 1
> (1 row)
>
> There seems to be a discontinuity here doesn't there?

Hmm, these work perfectly fine for me. On some platforms you cannot trust
date calculations before 1970; perhaps that's the case here.

> If somebody could direct me to the general area in the source tree, I

Somewhere in backend/utils/adt/{datetime|timestamp}.c no doubt.

> might be able to come up with a patch & btw, to whom should I send it?

pgsql-patches(at)postgresql(dot)org

--
Peter Eisentraut Sernanders väg 10:115
peter_e(at)gmx(dot)net 75262 Uppsala
http://yi.org/peter-e/ Sweden