Detecting source code changes to maintain the consistence of behavioral model

Abstract

It is well-known that as software system evolves, the source code tends to deviate from its design model so that maintaining their consistence is challenging. Our objective is to detect code changes that influence designed program behaviour which are referred as design level changes and update the behavioural model timely and automatically to maintain consistence. We propose an approach that filters out low-level source code changes that do not influence program behaviour, abstracts code changes into updating operations for behavioral model, and automates the integration and update of activity diagrams to maintain consistence. We've recognised that it is not uncommon for developers to introduce quick and dirty implementation that unnecessarily increases program complexity or introduces suboptimal behaviour changes. So while merging code changes into behaviour model, our approach also calculates cyclometric complexity variation before and after the process so that developers can be alerted of significant and/or detrimental changes. Our tool allows the user to approve the change in code before merging and updating the model.