Re: BUG #4956: Array Construct array() returning blank result

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "Niranjan Pandit" <niranjan81(at)gmail(dot)com>
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: Re: BUG #4956: Array Construct array() returning blank result
Date: 2009-07-31 14:09:41
Message-ID: 18694.1249049381@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

"Niranjan Pandit" <niranjan81(at)gmail(dot)com> writes:
> Only when the condition gets evaluated to true with at least 2 columns with
> varying values the array gets constructed, in all other cases what we get is
> a complete blank output. the label of the result is as usual as ?column? but
> there is absolutely nothing in the cell below the label.

Hmm, seems to work for me:

regression=# select array(select unique1 from tenk1 where unique2 < 2);
?column?
-------------
{8800,1891}
(1 row)

regression=# select array(select unique1 from tenk1 where unique2 < 1);
?column?
----------
{8800}
(1 row)

regression=# select array(select unique1 from tenk1 where unique2 < 0);
?column?
----------
{}
(1 row)

Pre-8.3 PG releases would return NULL instead of an empty array in the
last case, but 8,4 should not. If you're seeing differently, we'll
need a self-contained test case.

regards, tom lane

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Luke Koops 2009-07-31 14:59:56 BUG #4958: Stats collector hung on WaitForMultipleObjectsEx while attempting to recv a datagram
Previous Message Tom Lane 2009-07-31 13:59:43 Re: BUG #4957: search_path and pg_dump -T switch