Research Statement pdfauthor=Stefan Bucur

Interpretation from Concrete Semantics. Static analysis, and type inference in particular, is helpful for finding programming errors in dynamic languages, as they commonly lack the type information available to compiled languages. However, static analysis engines are implemented by hand, hence sharing the challenges of building a symbolic execution engine. To this end, I want to take the “interpreter as executable specification” approach in the realm of abstract interpretation. Instead of writing the transition functions by hand, developers only describe the abstract domain as conversion functions between a symbolic state and an abstract state. The abstract engine performs the transitions by executing the interpreter using a symbolic state and invoking the transformation functions.