Hi!

I'm new to postgres and want to do something like

select substring('MTWTFSS', from (Extract (dow from current_date)), for 1)

I can't get the parser to accept anything I try.

am I way out?
/dg

danfan46 wrote on 18.06.2010 22:52:
What are you trying to do?

Thomas Kellerer wrote:
where substr(days, DAYOFWEEK_ISO(current_date),1) =
substr('MTWTFSS', DAYOFWEEK_ISO(current_date),1)
and substr(months,Month(current_date),1) = substr('JFMAMJJASOND',Month(current_date),1)

I want to find those rows that has an F in days in position 5 if current date is a friday
and a J in months position 6 if current month is June.
If not so the job is not eligible for scheduling.
This is a non normalized approach of course. :-)
/dg

danfan46 wrote on 18.06.2010 23:11:
That makes things clearer

If you check the manual you'll see that the comma is incorrect to separate the parameters


So that should be:
select substring('MTWTFSS' from extract(dow from current_date) for 1)

Unfortunately that gives the following error:
function pg_catalog.substring(unknown, double precision, integer) does not exist

Now that tells us, that the result of the extract() function seems to be a double precision, but substring expects an integer.

So all you have to do is to cast this to an int:

select substring('MTWTFSS' from extract(dow from current_date)::integer for 1)

will return what you want.

On Fri, 18 Jun 2010 23:46:05 +0200, Thomas Kellerer wrote:

Actually, that can be done with the commas, too:

mgogala=# select substring('MTWTFSS',extract(dow from
current_date)::integer,1);
substring
-----------
F
(1 row)

Time: 0.315 ms
mgogala=#



--
http://mgogala.byethost5.com

Mladen Gogala wrote:
Quick response!
I prefer this second format.
/dg