| From: | Rodrigo De León <rdeleonp(at)gmail(dot)com> |
|---|---|
| To: | pgsql-sql(at)postgresql(dot)org |
| Cc: | Pablo Barrón <amentoraz(at)gmail(dot)com> |
| Subject: | Re: ordering by multiple columns |
| Date: | 2007-03-14 16:14:49 |
| Message-ID: | a55915760703140914q5883f01fud0572153e575285d@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-sql |
On 3/14/07, Pablo Barrón <amentoraz(at)gmail(dot)com> wrote:
> Hi!
>
> I'm trying to order a list in which the first parameter to order is a
> specific field, and the second should vary depending on a specific
> condition. More explicit, I would like to do something like:
>
> - ORDER BY a.column1, [b.column2 if c.value != 19 || c.column2 if
> c.value==19]
>
> That is to say, use a column if a value on each row isn't 19, OR use a
> different column if it isn't. How could I solve this? I've got to have a
> global order over my query, so I cannot separate it into two queries, etc
> :-?
>
> Thank you!
SELECT (...)
ORDER BY a.column1, CASE
WHEN c.VALUE <> 19
THEN b.column2
ELSE c.column2
END;
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Frank Bax | 2007-03-14 17:58:01 | Re: Joins on many-to-many relations. |
| Previous Message | Andreas Kretschmer | 2007-03-14 16:12:35 | Re: ordering by multiple columns |