[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
RE: [aspectj-dev] not recompiling on non-changes to aspect files?


Interesting optimization.... is ClassFileReader.hasStructuralChanges() a sufficient test to avoid unneccessary whole-world-weaving? I guess we also ought to be sure that there is no around advice in the aspect too?

-- Adrian
Adrian_Colyer@xxxxxxxxxx



"Jim Hugunin" <lists@xxxxxxxxxxx>
Sent by: aspectj-dev-admin@xxxxxxxxxxx

12/03/2004 17:52
Please respond to aspectj-dev

       
        To:        <aspectj-dev@xxxxxxxxxxx>
        cc:        
        Subject:        RE: [aspectj-dev] not recompiling on non-changes to aspect files?



In 1.1 there was a design flaw in the naming conventions used for advice
that made this impossible, see this bug:
 https://bugs.eclipse.org/bugs/show_bug.cgi?id=50641

Now that Andy has fixed that problem this should be tractable.  I believe
the only issue remaining is in the analysis pass for incremental re-weaving.

BTW - Even if you change the pointcut for a piece of advice the world
doesn't need to be recompiled but only rewoven.  This can potentially be
much faster.

-Jim

> -----Original Message-----
> From: aspectj-dev-admin@xxxxxxxxxxx [mailto:aspectj-dev-admin@xxxxxxxxxxx]
> On Behalf Of Nicholas Lesiecki
> Sent: Thursday, March 11, 2004 8:29 PM
> To: aspectj-dev@xxxxxxxxxxx
> Subject: Re: [aspectj-dev] not recompiling on non-changes to aspect files?
>
> I agree this would be sweet if possible. Though I am less concerned
> about whitespace and more about method bodies...
>
> Cheers,
> Nick
> On Mar 11, 2004, at 4:27 PM, Wes Isberg wrote:
>
> > In case this can be done in the brave, new (swizzled)
> > world...
> >
> > Is there any way to not recompile the world if a file
> > containing an aspect changes only in whitespace or comments?
> > Or only in the method bodies?
> >
> > The cost of recompiling a file once you've parsed it in JDT
> > might be small enough, but recompiling the world for a
> > comment change in AJDT is a pain.
> >
> > Of course, this could mean readjusting line references into the
> > aspect, including all the "from" references to declares, if/since
> > we copy rather than reference the source locations.  Is that
> > the bigger problem?  Sigh.
> >
> > Wes
> >
> > _______________________________________________
> > aspectj-dev mailing list
> > aspectj-dev@xxxxxxxxxxx
> > http://dev.eclipse.org/mailman/listinfo/aspectj-dev
> >
>
> _______________________________________________
> aspectj-dev mailing list
> aspectj-dev@xxxxxxxxxxx
> http://dev.eclipse.org/mailman/listinfo/aspectj-dev



_______________________________________________
aspectj-dev mailing list
aspectj-dev@xxxxxxxxxxx
http://dev.eclipse.org/mailman/listinfo/aspectj-dev