Re: pgsql: Allow empty target list in SELECT. - Mailing list pgsql-committers

From Bruce Momjian
Subject Re: pgsql: Allow empty target list in SELECT.
Date
Msg-id 20150320122550.GH6317@momjian.us
Whole thread Raw
In response to Re: pgsql: Allow empty target list in SELECT.  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: pgsql: Allow empty target list in SELECT.  (Bruce Momjian <bruce@momjian.us>)
List pgsql-committers
On Wed, Oct 22, 2014 at 04:19:10PM -0400, Tom Lane wrote:
> Thom Brown <thom@linux.com> writes:
> > I can see the same output as you if I get it to ignore my .psqlrc file:
>
> > # select * from colours2;
> > --
> > (5 rows)
>
> > Turns out it's "\x auto" causing it.
>
> Ah.  Seems like a bug in the expanded-mode printout logic then.
> But, in any case, it's been like that a very long time.  You
> could do this at least as far back as 8.4:
>
> regression=# create table foo();
> CREATE TABLE
> regression=# insert into foo default values;
> INSERT 0 1
> regression=# insert into foo default values;
> INSERT 0 1
> regression=# select * from foo;
> --
> (2 rows)
>
> regression=# \x on
> Expanded display is on.
> regression=# select * from foo;
> (No rows)

I have developed a patch to fix this, e.g.:

    test=> select * from foo;
    --
    (2 rows)

    test=> \x
    Expanded display is on.
    test=> select * from foo;
    (2 rows)

I used the same footer function every other output format was using.
Patch attached.

I also found that the normal non-expanded output was forcing the use of
a pager for these tests, which seemed odd considering there were no
rows. The problem is that the variable col_count is an unsigned integer,
and there were a few places we were subtracting one from it.  When
col_count was zero, subtracting one returned a very larger positive
number, and triggered the pager.  That is fixed in this patch as well.

--
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + Everyone has their own god. +

Attachment

pgsql-committers by date:

Previous
From: Andres Freund
Date:
Subject: pgsql: Add, optional, support for 128bit integers.
Next
From: Bruce Momjian
Date:
Subject: pgsql: C comment: update lock level mention in comment