Reusing Single-Language Analyses for Static Analysis of Multi-language Programs

Reusing Single-Language Analyses for Static Analysis of Multi-language Programs

연구 분야: Strategies

학회: SPLASH 2023: Companion Proceedings of the 2023 ACM SIGPLAN International Conference on Systems, Programming, Languages, and Applications: Software for Humanity

State-of-the-art software is crafted in multiple programming languages. Such multi-language software challenges static analyses: Since many static analyses are focused on analyzing single-language programs, they are inherently unsound or imprecise in analyzing cross-language interaction. Existing approaches that perform multi-language analysis are not analysis- or language independent and thus lack in extensibility for new analyses and languages. We will develop an extensible, language-, framework-, and analysis-independent architecture to reuse existing static analyses for analyzing multi-language software. Our hypotheses are that, our architecture allows reusing existing single-language analyses and improves precision and soundness compared to the state of the art. We will evaluate our architecture with a points-to and data flow analysis for Java, JavaScript, and C/C++ code and compare it against the state of the art.