We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Automatic Program Repair of Java Single Bugs using Two-level Mutation Operators.
- Authors
Oo, Cherry; Hnin Min Oo
- Abstract
Automatic program repair (APR) is one of the necessary software maintenance tasks because most software systems have errors that need to be fixed. APR techniques are considered as a search problem where the search space includes all potential repair candidates, with the aim of identifying the correct repair code in space. This paper proposes a repair approach that finds the correct repair code for object-oriented program bugs such as Java bugs in the minimized search space using the type of buggy statement and mutation system, MuJava. This approach consists of four main phases. First, program bugs are localized by prioritizing the statements based on their suspiciousness of containing bugs. Second, the mutation system is employed to mutate the program using two-level operators of the mutation system. Third, we extract the fixed candidates that are similar to the buggy statement type within the mutants and after receiving the ordered list of candidate patches, the last phase validates their correctness one by one using the test suite until a correct patch found. The experiment demonstrates that our strategy can effectively fix 19 out of 21 bugs from four real-world projects and achieves 90.48% accuracy.
- Subjects
SOFTWARE maintenance; OBJECT-oriented programming; SYSTEMS software
- Publication
IAENG International Journal of Computer Science, 2020, Vol 47, Issue 2, p223
- ISSN
1819-656X
- Publication type
Article