Section: New Results

Reverse Engineering & Evolution

Model Driven Reverse Engineering (MDRE), with its applications on software modernization or tool evolution for example, is a discipline in which model-based principles and techniques are used to treat various kinds of (sometimes very large) existing systems. In the continuity the work started several years ago, AtlanMod has been working actively on this research area this year again. The main contributions are the following:

• In the context of the ARTIST FP7 project, the work has been continued on reusing (and extending accordingly) MoDisco and several of its components to provide the Reverse Engineering support required within the project (and more particularly in the context of the use cases provided by our industrial partners). This has been an important year for the team in this project since it successfully ended in November 2015 after final review at the European Commission. At conceptual-level, the proposed overall approach (as a main result of the ARTIST project) and the main lessons we have learned from its application to concrete industrial scenarios have been published and promoted to a large and high-level audience [11] . The ARTIST project in itself, the various research aspects it addresses and the offered technical solutions have also been presented to the Modeling community [22] . At tooling-level, several (MoDisco-based) model discovery components from Java and SQL have been enhanced while made available as part of a second version of the official ARTIST OS Release (http://www.artist-project.eu/tools-of-toolbox/193 ). A promising work has also been started on studying deeper the automated discovery of behavioral aspects of software applications, notably by working on a pragmatic mapping between a programming language (Java) and a modeling language (the OMG fUML standard) that focuses on executable aspects.

• To facilitate the understanding of existing software applications via the different models describing them, a significant work has been performed related to providing a generic support for dealing with viewpoints and views expressed on set of possibly heterogeneous and large models. To this intent, and directly capitalizing on the work performed in the TEAP FUI project that ended by the end of 2014, the EMF Views prototype has been significantly refined and enhanced with a ViewPoint Definition Language (the VPDL domain-specific language having a SQL-like syntax) notably [18] . Based on this same model viewpoints/views approach, and more particularly on its underlying (meta)model virtualization support, the general problem of lightweight (meta)model extension has been studied more deeply in the context of our work within the MoNoGe FUI project (national). This has already resulted in a corresponding prototype and a DSL for expressing metamodel extensions [17] . Within the coming year, the plan is to continue further this global work on model viewpoints/views in a software understanding and evolution context.

• Software development projects are notoriously complex and difficult to deal with. Several support tools have been introduced in the past decades to ease the development activities such as issue tracking, code review and Source Control Management (SCM) systems. While such tools efficiently track the evolution of a given aspect of the project (e.g., issues or code), they provide just a partial view of the software project and they often lack of querying mechanisms beyond basic support (e.g., command line, simple gui). This is particularly true for projects that rely on Git, one of the most popular SCM systems. Nowadays many tools are built on top of it, however, they do not complement Git with query functionalities and currently none of them proposes a mechanism that unifies the project information scattered in such different tools. In [28] , we propose a conceptual schema for Git and an approach that, given a Git repository, exports its data to a relational database in order to (1) promote data integration with other existing Git-based tools relying on databases and (2) provide query functionalities expressed through standard SQL syntax. To ensure efficiency, our approach comes with an incremental propagation mechanism that refreshes the database content with the latest modifications.