• Model transformation development is a complex task. Therefore, having mechanisms for transformation testing and understanding becomes a matter of utmost importance. Understanding, among others, implies being able to trace back bugs to their causes. In model transformations, causes can be related with either the input model or
    the transformation code. This work describes HandyMOF, a tool that first eases the transition between the effect (i.e. generated code file) and the causes (i.e. input model and transformations) and then provides the
    means to check the transformation coverage obtained by a test suite. The challenges are twofold. First, the obtainment of input model suites which yield to a quantifiable transformation coverage. Second, providing
    fine-grained traces that permit to trace back code not just to the transformation rule but to the inner 'print' statements. A transformation that generates Google Web Toolkit (GWT) code is used as the running example.

  • Model transformations are precious and eortful outcomes
    of Model-Driven Engineering. As any other artifact, transformations are
    also subject to evolution forces. Not only are they aected by changes to
    transformation requirements, but also by the changes to the associated
    metamodels. Manual co-evolution of transformations after these metamodel
    changes is cumbersome and error-prone. In this setting, this paper
    introduces a semi-automatic process for the co-evolution of transformations
    after metamodel evolution. The process is divided in two main
    stages: at the detection stage, the changes to the metamodel are detected
    and classied, while the required actions for each type of change
    are performed at the co-evolution stage. The contributions of this paper
    include the automatic co-evolution of breaking and resolvable changes
    and the assistance to the transformation developer to aid in the coevolution
    of breaking and unresolvable changes. The presented process
    is implemented for ATL in the CO-URE prototype.

  • Form-intensive Web applications are common among institutions that collect bulks of data in a piecemeal fashion. European projects, income tax return and the like, illustrate this situation. In some cases, the very same form might need to be fed with different data (e.g. presenting different projects to the very same institution), while in other cases different forms might need to be fed with the same data (e.g. presenting a single project to different institutions). In these scenarios, manual filling is not only time-consuming but it is also terribly repetitive.

ONEKIN, UNIVERSITY OF THE BASQUE COUNTRY

University of the basque country