Automated Generation of Similar Paths for Localizing Program Faults
Junmin Ye, Kai Zhu, Cong Jin, Quan Zheng · 2009
Localizing a program fault accurately in debugging is complex and time-consuming. In the process of fault diagnosing, identifying or generating the successful test paths as similar as possible to the failed test is core for the effectiveness of faults localization. A method for calculating the similarity between two test paths based on analyzing difference between program control-flow is defined, and a novel algorithm based on DD-graph for generating similar path set directly from a failed test is proposed. It is experimentally proved that the proposed algorithm can generate a similar path set for a failed path and can help to localize the program faults.