Return-Path: <JFernandez@viquity.com>
Mailing-List: contact ant-dev-help@jakarta.apache.org; run by ezmlm
Delivered-To: mailing list ant-dev@jakarta.apache.org
Received: (qmail 15450 invoked from network); 20 Sep 2000 20:01:46 -0000
Received: from postbox.viquity.com (HELO dcsrv0.ecom2ecom.com)
 (63.198.126.137)
  by locus.apache.org with SMTP; 20 Sep 2000 20:01:46 -0000
Received: by dcsrv0 with Internet Mail Service (5.5.2650.21)
	id <SGY31432>; Wed, 20 Sep 2000 13:01:12 -0700
Message-ID: <635802DA64D4D31190D500508B9B04104E0F0A@dcsrv0>
From: Jose  Alberto Fernandez <JFernandez@viquity.com>
To: "'ant-dev@jakarta.apache.org'" <ant-dev@jakarta.apache.org>
Subject: RE: Additional behavior for Javac
Date: Wed, 20 Sep 2000 13:01:12 -0700
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Rating: locus.apache.org 1.6.2 0/1000/N

> From: Conor MacNeill [mailto:conor@ebinteractive.com.au]
> 
> > -----Original Message-----
> > From: Stefan Bodewig [mailto:bodewig@bost.de]
> > Sent: Wednesday, 20 September 2000 22:34
> > To: ant-dev@jakarta.apache.org
> > Subject: Re: Additional behavior for Javac
> >
> >
> > >>>>> "JAF" == Jose Alberto Fernandez 
> <JFernandez@viquity.com> writes:
> >
> >  JAF> If I understand correctly, the current Javac's srcdir 
> and <src>
> >  JAF> elements define a kind of search path for in which <include>
> >  JAF> <exclude> options are applied.
> >
> > This is the "is <src> a Path or a FileSet" question we had 
> some weeks
> > ago. Think we just forgot to come to a conclusion.
> >
> 
> It is a bit like asking whether light is a wave or a particle 
> :-). I think
> the src attribute is closer to a path at present.
> 
> I think what Jose Alberto is trying to achieve is very risky. 
> Even if he can
> convince ant to select the right files, the compiler will 
> still be using the
> sourcepath and may pick up other files from the wrong path.
> 

The problem for me is that <src>s are not quite a path. In a path, you know
that you will be getting only one version of each file (the first one in the
path), but here you may be finish passing to the compiler two versions, one
from each <src> because the includes apply to both.

So eventhough I could work around the compiler limitations, I cannot work
around <javac>'s limitations. It would be nice if the behaviour of both
where the same. Since this may be kind of an expensive operation one may
want to have some sort of "checkduplicates" flag for this situation.

Jose Alberto