MontiCore: Modular Development of Textual Domain Specific Languages

Abstract

Reuse is a key technique for a more efficient development and ensures the quality of the results. In object technology explicit encapsulation, interfaces, and inheritance are well-known principles for independent development that enable combination and reuse of developed artifacts. In this paper we apply modularity concepts for domain specific languages (DSLs) and discuss how they help to design new languages by extending existing ones and composing fragments to new DSLs. We use an extended grammar format with appropriate tool support that avoids redefinition of existing functionalities by introducing language inheritance and embedding as first class artifacts in a DSL definition. Language embedding and inheritance is not only assisted by the parser, but also by the editor, and algorithms based on tree traversal like context checkers, pretty printers, and code generators. We demonstrate that compositional engineering of new languages becomes a useful concept when starting to define project-individual DSLs using appropriate tool support.

Extending Architecture Description Languages With Exchangeable Component Behavior Languages

“…MontiArcAutomaton relies on the MontiCore [9,5] language workbench. MontiCore languages are extended context-free grammars (CFG) with well-formedness rules implemented in Java.…”

Bridging the Chasm between Executable Metamodeling and Models of Computation

“…With a similar support for object-orientation and static introductions, Kermeta and its aspect paradigm can be seen as an analogue of JastAdd in the DSML world. Rebernak et al [20] and Krahn et al [21] contributed to the field in the context of model-driven DSLs. While they also advocate modularity of DSL compilers and interpreters, we go further: we take advantage of modularity mechanisms for integrating the body of knowledge on models of computation, and allow their reuse and variability.…”

Modeling Language Variability

“…Astereotype<<prio:outer>> attached to a hierarchical Statechart model would override the default priority rule that the innermost enabled transition is taken if there are multiple transitions with the same trigger enabled in the same step, see also [Rum04]. Another form of syntactic variability is given by so called language parameters, also termed language embedding in [KRV08]. Consider again, for example, the language of Statecharts in which transitions may be guarded by a precondition.…”