Towards a Common Reference Architecture for Aspect-Oriented Modeling A. Schauerhuber* W. Schwinger E. Kapsammer W. Retschitzegger M. Wimmer Women’s Postgraduate College for Internet Technologies Vienna University of Technology, Austria [email protected] Department of Telecooperation University of Linz, Austria [email protected] Information Systems Group University of Linz, Austria {ek,werner}@ifs.uni-linz.ac.at Business Informatics Group Vienna University of Technology, Austria [email protected] ABSTRACT Aspect-orientation provides a new way of modularization by clearly separating crosscutting concerns from non-crosscutting ones. Although originally emerged at the programming level, aspect-orientation meanwhile stretches also over other development phases. Not only due to the rise of model-driven engineering, some approaches already exist for dealing with aspect-orientation at the modeling level. Nevertheless, concepts from the programming level are often simply reused without proper adaptation. Consequently, such approaches fall short in considering the full spectrum of modeling concepts. This paper takes a first step towards a consolidated and more comprehensive view on aspect-orientation by discussing a common reference architecture for aspect-oriented modeling. This reference architecture identifies the basic ingredients of aspect-orientation which in turn are abstracted from specific aspect-oriented programming languages and modeling approaches. Categories and Subject Descriptors D.2.2 Design Tools and Techniques General Terms Design, Standardization, Languages, Theory Keywords reference architecture, aspect-oriented modeling 1. INTRODUCTION Aspect-oriented software development (AOSD), sometimes also called Advanced Separation of Concerns (ASoC), is a fairly young but rapidly advancing research field. AOSD aims at providing new ways of modularization in order to separate crosscutting concerns from traditional units of decomposition during software development. Today, besides Aspect-Oriented Programming (AOP) [28], different approaches initially not proposed under the term aspect-oriented, such as Adaptive Programming (AP) [29], Composition Filters (CF) [1], Subject-Oriented Programming (SOP) [21], and Multi-Dimensional Separation of Concerns (MDSoC) [33], are now called aspect-oriented, because the term is "catchier, more commonly used, and less subject to ambiguous interpretation" [13].∗ From a software development point of view, aspect-orientation originally emerged at the programming level with AspectJ [2] as one of the most prominent protagonists. Due to the rise of model-driven engineering (MDE) [8], however, the aspect-oriented paradigm is no longer restricted to the programming level but is also more and more stretching over other phases of the development life cycle such as requirements engineering (cf. aspect-oriented requirements engineering, e.g., [30], [24]) or design (cf. aspect-oriented modeling, e.g., [8], [12], [16], [25], [39]). Particularly in the field of aspect-oriented modeling (AOM) there already exist several approaches, each of them having different origins and pursuing different goals for dealing with the unique characteristics of aspect-orientation. This entails not only the problem of different terminologies but also leads to a broad variety of aspect-oriented concepts. In several cases, concepts of aspect-oriented programming languages are simply incorporated unaltered into a modeling language failing to consider the different levels of abstraction. Applying aspect-orientation at the modeling level is not just injecting code at a certain point within a program but requires the consideration of the full spectrum of modeling concepts not present in programming languages, e.g., different views on the application's structure and behavior as provided by current modeling languages such as UML [31]. This paper contributes to a consolidation of aspect-oriented modeling by taking an initial step towards a common reference architecture that identifies the basic ingredients of aspect orientation, abstracted from certain AOP languages or AOM approaches. Such a reference architecture is beneficial in three ways. First, it provides the basis for the construction of a ∗ This research has been partly funded by the Austrian Federal Ministry for Education, Science, and Culture, and the European Social Fund (ESF) under grant 31.963/46-VII/9/2002. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Conference’XX, Month X–X, XXXX, City, State, Country. Copyright XXXX ACM X-XXXXX-XXX-X/XX/XXXX…$0.00.framework of evaluation criteria, allowing for a structured and programming language independent evaluation of aspect-oriented approaches and thereby identifying their strengths and shortcomings as demonstrated in an extended version of this paper [37]. Second, concepts of different aspect-oriented approaches can be mapped onto each other via the common reference architecture, thus acting as a kind of mediator model. Third, it could act as a blueprint in terms of a metamodel for designing a new, unified aspect-oriented modeling language. The remainder of the paper is organized as follows: Section 2 discusses related work for identifying the common ingredients of aspect-orientation. On basis of this, Section 3 proposes our common reference architecture. Section 4 reflects on our proposal and identifies open problems and issues requiring further investigation. Finally, Section 5 points to future research directions. 2. RELATED WORK Although there already exist several approaches in the area of AOM, to the best of our knowledge there are only a few attempts up to now, that provide a common understanding of aspect-oriented concepts at the programming level or at the modeling level. Some of them provide a dedicated reference architecture, whereas others provide a set of evaluation criteria for surveying existing aspect-oriented approaches, only. The design of our reference architecture draws from all those sources. In van den Berg et al. [40], an attempt towards establishing a common set
View Full Document
Unlocking...