Quality of Information and Communications Technology - 13th International Conference, QUATIC 2020, Faro, Portugal, September 9-11, 2020, Proceedings
Quality of Information and Communications Technology - 13th International Conference, QUATIC 2020, Faro, Portugal, September 9-11, 2020, Proceedings
复制标题
信息和通信技术质量 - 第 13 届国际会议，QUATIC 2020，葡萄牙法鲁，2020 年 9 月 9-11 日，会议记录
DOI：
10.1007/978-3-030-58793-2_7
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Counsell S
中科院分区：
文献类型：
--
作者：
Counsell S
在这篇简短的文章中，我们探讨了一个简单的，但尚未探索的问题，重构和bug之间的关系。重构后立即重新测试代码的复杂性是一些重构有缺陷的原因吗？为了方便我们的分析，我们使用了一组从三个开源系统中挖掘出来的超过4000个重构，并将其分解为货车Deursen和Moonen的四个测试类别。初步结果表明，与非bug类相比，bug类已经受到更多的重构，需要大量的重新测试的承诺;因此，重新测试的程度可能是一个重要的因素，在确定重构是否会产生错误。我们的发现支持了Bavota等人的观点，即在某些重构实践之后进行更多更好的测试可以减少重构造成的危害。
In this short paper, we explore one simple, yet unexplored question about the relationship between refactoring and bugs. Is the complexity of re-testing code immediately after refactoring a reason why some refactorings are buggy? To facilitate our analysis, we use a set of over four thousand refactorings mined from three open-source systems and decomposed into the four test categories of van Deursen and Moonen. Preliminary results showed that, compared with non-buggy classes, buggy classes had been subjected to more refactorings where a large re-test commitment was required; extent of re-test may therefore be a significant factor in determining whether refactoring creates bugs. Our finding supports that of Bavota et al. - that more and better testing after certain refactoring practices could reduce the harm that refactorings cause.