The main focus of my research work lies in the area of aspect-oriented software development. I have been looking at ways of improving pointcut language of AspectJ, an AOP extension of Java, to alleviate its current limitations, namely:

• Syntactic rather than semantic nature of pointcuts
• Limitations of what can be matched by the pointcut language

In particular, I have been studying safe Datalog as an alternative pointcut language. Safe Datalog is a fully declarative logic programming language with a straightforward least-fixpoint semantics. Given an adequate relational program representation, Datalog is more expressive than the existing pointcuts in AspectJ.

As part of this work, I have co-authored a set of simple rewrite rules to translate AspectJ pointcuts to safe Datalog. These rules helped to uncover many subtleties and bugs in the existing implementations of AspectJ and are the first rigorous semantics of its static pointcuts language.

As part of my work on Datalog pointcuts, I have been also studying applications of logic programming in the context of source code querying and analysis. In particular I am involved in the CodeQuest project – an efficient and scalable implementation of a deductive database system.