Forward Engineering advocates for code to be generated dynamically through model-to-text transformations that target a specific platform. In this setting, platform evolution can leave the transformation, and hence the generated code, outdated. This issue is exacerbated by the perpetual beta phenomenon in Web 2.0 platforms where continuous delta releases are a common practice. Here, manual co-evolution becomes cumbersome. This paper looks at how to automate - fully or in part - the synchronization process between the platform and the transformation. To this end, the transformation process is split in two parts: the stable part is coded as a MOFScript transformation whereas the unstable side is isolated through an adapter that is implicitly called by the transformation at generation time. In this way, platform upgrades impact the adapter but leave the transformation untouched. The work focuses on DB schema evolution, and takes MediaWiki as a vivid case study. A first case study results in the upfront cost of using the adapter paying off after three releases MediaWiki upgrades. © 2014 Springer International Publishing.
CITATION STYLE
García, J., Dìaz, O., & Cabot, J. (2014). An adapter-based approach to co-evolve generated SQL in model-to-text transformations. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8484 LNCS, pp. 518–532). Springer Verlag. https://doi.org/10.1007/978-3-319-07881-6_35
Mendeley helps you to discover research relevant for your work.