Systematic Component-Oriented Language Reuse

Modern-day software is increasingly developed by domain experts. Domain-specific languages (DSLs) help them to bridge the conceptual gap between the problem and the solution domain. Engineering these languages is complex, as it requires comprehensive white-box knowledge of constituents of DSLs, i.e., syntax and semantics, and their composition and realization in respective technological spaces of language workbenches. For efficient language engineering reusing existing languages by composition is essential. For this purpose, various language composition and variability mechanisms have been developed. However, in the literature, there exists no approach enabling systematic language reuse holistically, i.e., including syntax and semantics, in a black-box fashion. We propose an approach for the systematic component-oriented reuse of DSLs (SCOLaR) that supports the integrated reuse of syntax, well-formedness rules, and semantics via black-box language components that are composable along their interface in a language family. We implemented a prototypical version of our solution with the MontiCore language workbench, that supports embedding and aggregation of grammar-based languages with translational semantics realized with code-generators. In the future, we plan to improve our implementation with the support of other language workbenches and to provide a comprehensive evaluation with a series of case studies. With our work, we aim to advance the field of software language engineering by providing a novel method for black-box language reuse capturing syntax and semantics definitions for textual external, translational DSLs. This paper outlines our approach by stating the problem, motivating our research question, and proposing our solution, our plan for evaluation, and presenting expected contributions.