English Español français rss
> Accueil > Programme > Colloque ATL

Coll. ATL : Short paper session

Intervenant(s) :Manuel Wimmer, Martina Seidl, Skander Turki, Eric Senn, Dominique Blouin, Saadia Dhouib, Jean-Philippe Diguet, Johan Laurent
Type d'événement :Conférence
Niveau :Confirmé
Date :Mercredi 8 juillet 2009
Horaire :11h30
Durée :60 minutes
Langue :English
Lieu :Ecole des Mines de Nantes

On Using UML Profiles in ATL Transformations by Manuel Wimmer and Martina Seidl

For defining modeling languages, metamodels and UML profiles are the proposed options. While metamodels are supported by several dedicated model transformation approaches, currently no transformation language exists which support UML profiles as first class language definitions. Instead, the usage of UML profiles in transformations is implicit by using calls to external UML APIs. In this paper, we first discuss the state-of-the-art of using UML profiles in ATL. Subsequently, three approaches for supporting profiles as first class language definitions within ATL transformations are introduced and discussed. In particular, these approaches aim at using stereotypes and tagged values within declarative rules without using external API calls. The benefits are : first, the enhanced static checking of ATL transformations, second, the more explicit representation of transformations logic enhances the application of higher-order transformations, and third, enhanced tool support such as code completion may be provided.

Checking syntactic constraints on models using ATL model transformations by Skander Turki, Eric Senn, Dominique Blouin, Saadia Dhouib, Jean-Philippe Diguet and Johan Laurent.

This paper presents a use case on using ATL model transformations to check syntactic correctness constraints on a UML/MARTE model. This model transformation is a part of an interoperability model transformation from UML/MARTE to AADL in which UML/MARTE is used as a UML profile for AADL. As input, the constraints checking model transformation takes a UML/MARTE model and produces as output a diagnosis model that conforms to a diagnosis meta-model (VERIF) which is also presented in this paper.