A Survey on the Automated Analyses of Feature Models.

This paper surveys the state of the art on the automated analyses of feature models outlining the main advances made until now and concludes that the proposal based on constraint programming is one of the most promising ones since it supports all the operations identified in this paper over feature models.

Abstract

Palabras clave: Software Product Lines, Domain Engineering, Feature ModelsResumen. Feature models are one of the most important assets in software productline engineering when capturing variability. Although the analyses of feature models wasstated as challenging problem since they were fist proposed, it has not been fully achievedyet. In this paper we survey the state of the art on the automated analyses of featuremodels outlining the main advances made until now. We conclude that our proposal basedon constraint programming is one of the most promising ones since it supports all theoperations identified in this paper over feature models.1 IntroductionThe industrialization of the software engineering discipline is not achieved yet. We un-derstand the industrialization of software as the capability of producing software systemsthat fulfill budget and quality in a prescribed way. In other industries this is achieved byproducing product lines. Indeed, Boing, Dell or even Mc Donalds builds product lines ofaircraft, computers or hamburgers. Why not Software Product Lines (SPLs)?A key technical question that confronts software engineers is how to specify a particularproduct in an SPL. One of the possible solutions is to do it in terms of features, where afeature is an increment in product functionality. Designing a family of software systemsin terms of features is more natural than doing it in terms of objects or classes becausethey are understood by all stakeholders [19]. While single software systems are specifiedin terms of features, software product lines are specified using feature models.Feature models are recognized in the literature to be one of the most important contri-butions to SPL engineering [3]. A key task is to capture commonalities and variabilitiesamong products and feature models are used to this end.Although there has been advances in SPL engineering regarding feature models, a keytechnical aspect remains open: their automated analyses. In this paper we survey thestate of the art on the automated analyses of feature models. First, we review the different1