From: Igor Shevchenko <igor@carcass.ath.cx>
To: Tomasz Myrta <jasiek@klaster.net>
Subject: Re: LIBPQ Question
Date: Mon, 27 Oct 2003 09:25:25 +0200
User-Agent: KMail/1.5.2
References: <ILECJKPFOLBAGINIFFFBMECECAAA.creid@netbcg.com>
	<3F9CC61C.8010103@klaster.net>
In-Reply-To: <3F9CC61C.8010103@klaster.net>
Cc: pgsql-interfaces@postgresql.org
MIME-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 8bit
Content-Disposition: inline
Message-Id: <200310270925.26056.igor@carcass.ath.cx>
X-HOSTING.UKR.NET-ID: c0a013d97ea9b9470e2e2466d3332ab4
X-Virus-Scanned: by amavisd-new at postgresql.org
X-Archive-Number: 200310/53
X-Sequence-Number: 4542

On Monday 27 October 2003 09:15, Tomasz Myrta wrote:
> Dnia 2003-10-27 00:08, Użytkownik creid napisał:
> > Problem:  Assigning a COUNT(*) result into an integer variable in my C
> > program consistently fails except when I assign the same result to a char
> > variable.  I can only assume that the internal data type the COUNT
> > function uses is integer.
> >
> > Can anyone help put me in the proper mindset so I may deal with this,
> > seemingly simple issue, to resolution.
> >
> > I need the integer result to to help me satisfy a dynamic memory
> > requirement... COUNT(*) result will tell me how many rows of data I need
> > to malloc and I cannot perform a math operation on a char variable.
>
> All libpq results are strings.
> some_int_value=atoi(PQgetvalue(...))

Not true anymore with protocol v3, which added the binary format. Text format 
is still the default.

> Anyway why do you need count(*) ? When you retrieve your rows, you can
> always check how many are them using PQntuples(...) and then malloc your
> memory tables.
>
> Regards,
> Tomasz Myrta
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: the planner will ignore your desire to choose an index scan if your
>       joining column's datatypes do not match

-- 
Best regards,
Igor Shevchenko