[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
RE: [aspectj-dev] Re: after() and around() advice does not work on handler join poi nts.
|
Sorry for the latest of the email, just wanted to put my $0.02 in...
> I have a similar thought. Could something like following be
> implemented:
> 1. When ajc is supplied the source code, it annotates the byte code
> generated such that weaving phase will know where exaclty
> to instrucment the advice.
> 2. If weaver sees the annotations, it performs weaving as per
> language semantics. However, if it does not find annotation
> (because the code was compiled with, say, javac), it issues
> compiler-limiation warning (as it does right now.).
>
> BTW, thanks Jim for your detailed reply.
>
> -Ramnivas
It think if ajc treated bytecodes differently from source code then that might
just encourage people to run decompilers in order to get around the problem. So
what about if `after' was always allowed and it would have the same semantics
for every source/bytecode given (ie. the `after' affects what can be inferred as
the largest possible scope of the finally).
-Macneil
