Exception-Driven Fault Localization for Automated Program Repair

TL;DR

This paper introduces EXCEPT, a fault localization technique for automated program repair that uses failure semantics, specifically exception types and sources, to improve fault localization accuracy over traditional spectrum-based methods.

Contribution

EXCEPT offers a novel approach focusing on exception semantics for fault localization, outperforming traditional spectrum-based techniques like Ochiai and ssFix.

Findings

EXCEPT outperforms Ochiai and ssFix on 43 exception-raising faults.

Focusing on exception semantics improves fault localization accuracy.

The technique is effective for faults involving exceptions in real benchmarks.

Abstract

Automated Program Repair (APR) techniques typically exploit spectrum-based fault localization (SBFL) to identify the program locations that should be patched, making the effectiveness of APR techniques dependent on the effectiveness of fault localization. Indeed, results show that SBFL often does not localize faults accurately, hindering the effectiveness of APR. In this paper, we propose EXCEPT, a technique that addresses the localization problem by focusing on the semantics of failures rather than on the correlation between the executed statements and the failed tests, as SBFL does. We focus on failures due to exceptions and we exploit their type and source to localize and guess the faults. Experiments with 43 exception-raising faults from the Defects4J benchmark show that EXCEPT can perform better than Ochiai and ssFix.