ant-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Les Hughes <>
Subject RE: Identify which classes were created from a Java task?
Date Thu, 18 Oct 2001 19:19:59 GMT

True, of course there isnt a 1:1 mapping of source to class - which is why I
mentioned <depends> as well. But Java really is a pain when it comes to
trying to work out this kind of stuff esp. since javac does its own thing
when compiling. I do wish it would only compile those files I tell it
to....Hmm, jikes?

Anyway here's a thought. What about having v1.0.jar of your system built and
on the classpath for the compiler. Blow away your source tree and then sync
out of your SCM only those files that changed post v1.0. Next compile this
source tree and jar up anything it produces. Then stick this jar
(v1.0patch1.jar) in front of the old one on the classpath and perhaps Bob's
your uncle. Maybe? Or am I talking  rubbish again and have I been doing too
much powerpointing recently :-)


> -----Original Message-----
> From: Diane Holt []
> Sent: 18 October 2001 18:00
> To:
> Subject: RE: Identify which classes were created from a Java task?
> --- Les Hughes <> wrote:
> > Perforce supports labels and "what changed" kinds of 
> queries - others
> > may be able to help out depending on your SCM.
> The problem with that approach is that Java can end up with more class
> files than source files, so just having a list of which source files
> changed (or even which files were handed off to the compiler) wouldn't
> necessarily give you a list of all the build-output files 
> you'd need to
> pick up. I think the "find -newer" approach is probably your 
> best bet (the
> only other approaches I could think of were far more 
> inelegant than that
> one, which is why I didn't offer any of them :)
> Diane
> =====
> (
> __________________________________________________
> Do You Yahoo!?
> Make a great connection at Yahoo! Personals.

View raw message