FASTEN: an open extensible framework to experiment with formal specification approaches: using language engineering to develop a multi-paradigm specification environment for NuSMV

Proceedings of the 7th International Workshop on Formal Methods in Software Engineering(2019)

引用 6|浏览21
暂无评分
摘要
Formal specification approaches have been successfully used to specify and verify complex systems. Verification engineers so far either directly use formal specification languages which can be consumed by verification tools (e.g. SMV, Promela) or main stream modeling languages which are then translated into formal languages and verified (e.g. SysML, AADL). The first approach is expressive and effective but difficult to use by non-experts. The second approach lowers the entry barrier for novices but users are limited to the constructs of the chosen modeling languages and thereby end up abusing the language to encode behaviors of interest. In this paper, we introduce a third approach that we call FASTEN, in which modular and extensible Domain Specific Languages (DSLs) are used to raise the abstraction level of specification languages towards the domain of interest. The approach aims to help novice users to use formal specification, enable experts to use multi-paradigm modeling, and provide tools for the developers of verification technologies to easily experiment with various types of specification approaches. To show the feasibility of the approach, we release an open-source tool based on Jetbrains' MPS language workbench that provides an extensible stack of more than ten DSLs, situated at different levels of abstraction, built on top of the SMV language. We use the NuSMV model checker to perform verification, to simulate the models and lift the traces at the abstraction level of the DSLs. We detail on the experience with designing and developing the DSLs stack and briefly report on using the DSLs in practice for the study of a communication protocol of a safety critical system.
更多
查看译文
关键词
domain specific languages, formal methods
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要