Semantic-based weaving of scenarios

Abstract

The notion of aspect looks promising for handling cross-cutting concerns earlier in the software life-cycle, up from programming to design, analysis and even requirements. Support for aspects is thus now raising interest also at the modeling level, including with behavioral modeling languages such as scenarios. With this kind of modeling languages, even if aspect weaving can be performed at the abstract syntax level, a weaving at the semantics level seems a far more appealing and powerful mechanism. In this paper we present a semantic-based aspect weaving algorithm for Hierarchical Message Sequence Charts (HMSCs). The algorithm proposed uses a set of transformations that take into account the compositional semantics of HMSCs to weave an initial HMSC and a behavioral aspect expressed with scenarios.