摘要:基于互联网的软件开发要求产品快速迭代，同时保持产品的质量，其关键的环节就是持续集成.持续集成通过自动化测试来保证集成到主干的代码质量.持续集成时的测试用例选择是一个非常有挑战性的问题.如果运行所有的测试用例，需要消耗大量的计算资源，并造成测试反馈周期过长.如果选择的测试用例集不合适，又不足以覆盖必要的测试代码以保证待集成代码的质量.持续集成测试用例集优化的目的是平衡测试资源和测试质量，在尽可能不影响测试质量的情况下，减少持续集成的资源需求.对近年来国内外****在该领域的研究工作进行了系统的分析、提炼和总结.为此，首先从研究主题、影响因子、研究方法、研究对象以及性能评价等方面提出5个研究问题；然后从电子文献数据库搜索最近10年的研究工作，经过仔细审查和筛选后选择39篇文献作为研究对象；最后，从选择文献中收集数据，通过定量分析和可视化展示来回答提出的研究问题.总的来说，回顾了持续集成测试用例集优化的研究进展，为该领域的研究者提供了一些有用的发现，并总结了面临的问题和挑战.



Abstract:Currently, many software applications are rapidly developed and deployed. Rapid iterations are used to implement new requirements while keeping high quality. Continuous integration (CI) is an important activity to make it work. CI tests the code automatically before it will be integrated into the master branch to ensure its quality. The big challenge is how to select an appropriate test case set for continuous integration. Running all test cases will consume a large amount of resource and get feedback in a long cycle, but selecting an improper test case subset will lose some required test, and therefore increase quality risk. The goal of test set optimization is to select an appropriate test case subset which can satisfy the test requirement for each unit or iteration test while reducing the test resource as much as possible. In this paper, a systematic literature review is performed to reveal the current status of CI-test set optimization based on existing literature. More specifically, the related studies published between 2007 and 2017 from four databases of electronic literature are searched to finally select 39 important studies for careful review. Five research questions are presented in order to achieve the above goal, and data from the selected studies are extracted to analyze and answer the questions. The analysis results contribute some useful findings and new challenges for researches and practitioners.



PDF全文下载地址:

http://jos.org.cn/jos/article/pdf/5613