Manuscript Title:

FAULT LOCALIZATION MODEL FOR EVALUATING THE CAPABILITIES OF AUTOMATIC PROGRAM REPAIR WITH CHECKPOINTS

Author:

T. MAMATHA, B. RAMA SUBBA REDDY, C SHOBA BINDU

DOI Number:

DOI:10.17605/OSF.IO/F3ZCE

Published : 2022-10-10

About the author(s)

1. T. MAMATHA - Research Scholar, Dept. of CSE, JNTUA University, Anantapuramu, Andhra Pradesh, India.
2. B. RAMA SUBBA REDDY - Professor, Dept. of CSE, Anantha Lakshmi Institute of Technology and Sciences, Anantapuramu, AP, India.
3. C SHOBA BINDU - Professor, Dept. of CSE, JNTUA University, Anantapuramu, Andhra Pradesh India.

Full Text : PDF

Abstract

Fault localization is a key phase in the automatic repair of programs because correct identification of program areas that are most closely linked with a fault significantly influences the efficiency of the patching. Most Automated Program Repair (APR) tools use common fault location methods, which are not integrated tightly into the overall program repair process and hence provide only a poor efficiency. Each year, software businesses spend several hours debugging and correcting faults for developers. Automated program repair can decrease debugging expenses. Existing automated repair solutions, such as Genprog, TSP Repair, and Sketch Fix, are highly promising but cannot repair all defects. We analyzed fault localization on automatic repair techniques with fault reduction techniques. The main challenges are to discover code semantically comparable to defective code and integrate it into the faulty program. By extending the time needed to discover a possible remedy, APR performance will be degraded. Furthermore, the correctness of the program repair will be compromised since APR will update fault-free declarations that have a higher repair priority than a real incorrect declaration. In this research, an Effective Model for Fault Localization using Checkpoints for Automatic Program Repair (EMFLC-APR) is proposed that improves the performance of the system and the time for spending faults can also be reduced. The proposed technique provides a feedback loop between the operations for identifying the sources of the issue and the work to generate and evaluate possible remedies. The feedback loop allows partial assessment findings of potential remedies to be used to more correctly discover errors and ultimately lead to processes with enhanced efficiency and effectiveness. The proposed model exhibits better efficiency in fault localization for automatic program repair.


Keywords

Fault Localization, Automatic Program Repair, Debug, Error Correction Time, Faulty Code.