Semantic Fault Localization for Mutation-based Program Repair

Abstract

In this paper, we combine the semantics-based fault localization and the mutation-based program repair techniques to design an efficient algorithm for automatic repair of programs. Given an erroneous program and an assertion, we first use the fault localization to automatically identify statements relevant for the error, and then repeatedly mutate the expressions in those relevant statements until a correct program that satisfies the assertion is obtained. This way, we significantly reduce the space of all mutated programs without dropping any potentially correct program, and so locate a repaired program much faster than a program repair with no fault localization.A prototype tool has been developed for repairing C programs by combining fault localization and code mutations. The experimental results on C benchmarks demonstrate that this approach is significantly faster than the previous program repair approach with no fault localization.