Scheduling Refactoring Opportunities Using Computational Search

code smell;refactoring;optimization;bi-level;SBSE;scheduling;Software engineering;Computer Science, College of Engineering and Computer Science

Tama, NivinMaxim, Bruce ;

University of Michigan

关键词: code smell;    refactoring;    optimization;    bi-level;    SBSE;    scheduling;    Software engineering;    Computer Science, College of Engineering and Computer Science;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/136063/Scheduling%20Refactoring%20Opportunities%20Using%20Computational%20Search.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Maintaining a high-level code quality can be extremely expensive since time and monetarypressures force programmers to neglect improving the quality of their source code. Refactoring is an extremely important solution to reduce and manage the growing complexity of software systems. Developers often need to make trade-offs between code quality, available resources and delivering a product on time, and such management support is beyond the scope and capability of existing refactoring engines.The problem of finding the optimal sequence in which the refactoring opportunities, such as bad smells, should be ordered is rarely studied. Due to the large number of possible scheduling solutions to explore, software engineers cannot manually find an optimal sequence of refactoring opportunities that may reduce the effort and time required to efficiently improve the quality of software systems. In this paper, we use bi-level multi-objective optimization to the refactoring opportunities management problem. The upper level generates a population of solutions where each solution is defined as an ordered list of code smells to fix which maximize the benefits in terms of quality improvements and minimize the cost in terms of number of refactorings to apply. The lower level finds the best sequence of refactorings that fixes the maximum number of code smells with a minimum number of refactorings for each solution (code smells sequence) in the upper level. The statistical analysis of our experiments over 30 runs on 6 open source systems and 1 industrial project shows a significant reduction in effort and better improvements of quality when compared to state-of-art bad smells prioritization techniques. The manual evaluation performed by software engineers also confirms the relevance of our refactoring opportunities scheduling solutions.