From mboxrd@z Thu Jan  1 00:00:00 1970
From: Zdenek Kabelac <zkabelac@redhat.com>
Date: Thu, 08 Oct 2009 13:02:05 +0200
Subject: [PATCH 01/13] Replicator: export out_ function for modules
In-Reply-To: <20091007150503.GC3054@agk-dp.fab.redhat.com>
References: <1254751240-17431-1-git-send-email-zkabelac@redhat.com>
	<20091007150503.GC3054@agk-dp.fab.redhat.com>
Message-ID: <4ACDC6AD.2090207@redhat.com>
List-Id: <lvm-devel.redhat.com>
To: lvm-devel@redhat.com
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

Dne 7.10.2009 17:05, Alasdair G Kergon napsal(a):
> On Mon, Oct 05, 2009 at 04:00:28PM +0200, Zdenek Kabelac wrote:
>> +/* insert new line */
>> +int out_newline(struct formatter *f)
>> +{
>> +	outnl(f);
>> +	return 0;
> 
> Should return 1 here, and the caller has to check that.

true

> Instead, I think the outnl macro should be changed to work like the outf one.
> (make the 'f->nl' part internal to an export.c function instead of the header file)
> The idea was that the callers would be using this so frequently, it was
> clearer just writing 'outf(...)'  than 'if (!out_text(...)) return_0;' everywhere.
> 
> Alasdair
> 

So how about this:

+/* insert new line */
+int out_newline(struct formatter *f)
+{
+	return f->nl(f);
+}
+

I think this will be the easiest way - nl() is virtual function thus it can't
be easily 'macro-ised'.

Zdenek