[aosd-discuss] AOP, Design Patterns and Metadata - master thesis

[Günter Kniesel]

andrei oprisan schrieb:
>              Hello all,
>    My name is Andrei Oprisan and I am a student at the  Department of 
> Computer Science, University of Joensuu, Finland. I am currently writing 
> my master thesis. The thesis is centered around using AOP and Metadata 
> to implement the GOF's Design Patterns. I am using AspectJ and 
> annotations for that, but I am kind of stuck now because things are not 
> so clear for me where should I go. I am aware of the previous 
> implementation of Design Patterns using AspectJ done by Jan Hannemann. I 
> would like to kindly ask you for opinions or advices on the the subject 
> of the my thesis. If required, I can provide more information on the 
> work that I had already done.
> Thank you very much for your time.
> 
> Kind Regards,
>     Andrei.

Hi Andrei,

if you got stuck with AspectJ you might want to try LogicAJ 
(http://roots.iai.uni-bonn.de/research/logicaj). A discussion
of how LAJ can express design patterns that are hard to model
properly in AspectJ (e.g. different variants of Decorators) is
contained in

   Günter Kniesel, Tobias Rho, Stefan Hanenberg:
   "Reusable Pattern Implementations Need Generic Aspects"
   Proceedings of ECOOP 2004 Workshop on Reflection, AOP and
   Meta-Data for Software Evolution (RAM-SE), 2004.
   http://www.cs.uni-bonn.de/~gk/papers/knieselRhoHanenberg-RAMSE04.pdf

For background information on LogicAJ see 	

   Tobias Rho, Günter Kniesel
   "Uniform Genericity for Aspect Languages"
   Technical Report IAI-TR-2004-4, Computer Science Department III,
   University of Bonn. ISSN 0944-8535. December 2004.
http://www.cs.uni-bonn.de/~gk/papers/IAI-TR-2004-4-RhoKniesel-UniformAspectGenericity.pdf

For a survey of aspect genericity including other languages see

     Günter Kniesel, Tobias Rho:
     "A Definition, Overview and Taxonomy of Generic Aspect Languages"
     In Lionel Senturier (Ed.): Aspect-oriented Software Development,
     Special issue of L'Objet, vol. 12, no. 2-3 (Sept. 2006), p. 9-39,
     Hermes Science Publishing, London, 2006.
     http://www.cs.uni-bonn.de/~gk/papers/kniesel-lobjet2006.pdf

By the way, LAJ can also use annotations, although the above papers 
don't mention it. However, for most design patterns that we have 
analysed, it does not depend on annotations, since its pointcut language 
is powerful enough to identify the relevant join points without hints
from the user.

Hope this helps. "Spor la treaba!" :)

Günter