[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [aspectj-users] Problem compiling J2SE 5.0 source with ajc


Unfortunately, no, if the annotation has source retention then it isnt in the RuntimeVisibleAnnotation or RuntimeInvisibleAnnotation attributes in the class file - and its the class file we weave.

As a rule we don't have join points for anything that doesn't exist in the class file.  Besides this case, the other well known example is that you can't put after advice on a handler join point because the end of a catch block isn't reliably identifiable in the class file when its obvious in the source.  We have this rule so the semantics of weaving are the same regardless of whether you pass source or binary inputs to the weaver.

cheers,
Andy.

p.s. Adrian did start looking at the new compiler on schedule :)

---
Andy Clement
AspectJ Committer
clemas@xxxxxxxxxx



pope <alexandru.popescu@xxxxxxxxx>
Sent by: aspectj-users-admin@xxxxxxxxxxx

06/01/2005 19:21

Please respond to
aspectj-users@xxxxxxxxxxx

To
aspectj-users@xxxxxxxxxxx
cc
Subject
Re: [aspectj-users] Problem compiling J2SE 5.0 source with ajc





-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

[quote Adrian Colyer::on 1/5/2005 12:35 PM]
|> Do you have any sort of estimate when you'll start integrating the new
|> jdt into AspectJ?
|
| I plan on starting that tomorrow...
|
| -- Adrian
| Adrian_Colyer@xxxxxxxxxx
|
|
|
| Rafal Krzewski <Rafal.Krzewski@xxxxxxxxx>
| Sent by: aspectj-users-admin@xxxxxxxxxxx
| 05/01/2005 10:21
| Please respond to
| aspectj-users@xxxxxxxxxxx
|
|
| To
| aspectj-users@xxxxxxxxxxx
| cc
|
| Subject
| Re: [aspectj-users] Problem compiling J2SE 5.0 source with ajc
|
|
|
|
|
|
| Adrian Colyer wrote:
|
|> The Eclipse JDT
|> team released a full Java 5 compiler in their 3.1 M4 release just before
|
|> Christmas, so our next task is to pick that up and integrate it into
|> AspectJ for the M2 release (and then we *will* be able to support source
|
|> compilation).
|
| They did, 3 months ahead of original schedule, if I recall correctly.
| This means that a score of issues is to be found and fixed yet, but its
| out there to take. The work on the compiler was resumed already after
| holiday break - jdt.core v_530 was released to integration yesterday.
|
| Do you have any sort of estimate when you'll start integrating the new
| jdt into AspectJ?
|
| regards,
| Rafal
| _______________________________________________
| aspectj-users mailing list
| aspectj-users@xxxxxxxxxxx
| http://dev.eclipse.org/mailman/listinfo/aspectj-users
|
|
| _______________________________________________
| aspectj-users mailing list
| aspectj-users@xxxxxxxxxxx
| http://dev.eclipse.org/mailman/listinfo/aspectj-users
|
I see an interesting question here: can AJ work on annotations with retention type set on source? Is
this present inside the bytecode as to allow ajc in the 2nd step to take the necessary actions?

- --
:pope
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)

iD8DBQFB3Y+iTTDTje0R2dgRArfFAJ4wKeXgFb5Fg1f+uOulvhcN0/7Y1gCffaCI
ewRkwawC6psJYjSi/dtk1+o=
=S/A3
-----END PGP SIGNATURE-----
_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
http://dev.eclipse.org/mailman/listinfo/aspectj-users