A Methodological Framework for SPL Engineering from DSML

Abstract

For the last ten years, Software Product Line (SPL) tool developers have been facing the implementation of different variability requirements and the support of SPL engineering activities demanded by emergent domains. Despite several tools exist, few works resolve SPL process for both problem and solution space. Due to these reasons, we propose a methodological framework that overcomes the limits of existing tools and holds all the phases and activities from the requirement design till the product derivation. We start by using a Domain Specific Modelling Language (DSML) for domain description, which allows system designers working closer to the system domain as they can manipulate real concepts. Thereafter, an intermediate phase converts the DSML metamodel to a tree-structured representation similar to Feature Model (FM) notation enriched with extra-information such that cardinality, attributes, constraints, documentation, etc. The objective of this FM is to be used later as a decision tree to guide the generative process of our software factory in the following way: First, the engineer annotates the variation points with variability types such that binding time, granularity, evolution, etc, which are crucial concerns to be considered when generating the products. Second, based upon these annotations, our framework determines the possible useful variability mechanisms that could be employed to implement the product families and the engineers choose the variability programming tactic among them. Finally, the software factory produces the guidelines to implement the realization strategy and derive the related product assets through an assembly process. We provide a real industry running example giving insight into the application of the presented approach.