I-Corps: Translation Potential of Optimizing Regression Testing in Software Development

这个I-Corps项目的更广泛的影响是技术的发展，它将使公司能够大大减少他们必须为任何代码更改运行的软件测试的数量，从而降低他们的软件测试成本并提高开发人员的生产力。这项技术有可能在大量的代码更改中平均减少超过50%的测试时间。将该技术集成到已经存在的工作流程中，将引入最小的中断，并且对大多数软件开发人员来说仍然是隐藏的，从而增加了采用该技术的机会。这项研究将带来：(a)减少测试时间并更快地向开发人员提供反馈，(b)减少运行测试所需的资源，这将降低总体成本和维护成本，以及(c)与绿色计算努力和碳中和保持一致。这个I-Corps项目利用体验式学习和对行业生态系统的第一手调查来评估该技术的翻译潜力。该解决方案基于回归测试选择工具（RTS）的开发，该工具通过跳过不受代码更改影响的测试来降低测试成本。该技术有可能集成到大多数常见的开发工作流中。该工具可以在细粒度代码级别（例如，类、方法和函数）上捕获每个测试（或一组类似的测试）的依赖关系。一旦用户更改了其中一个代码元素，测试工具将自动识别一组要运行的测试，以及那些不受更改影响而要跳过的测试。该工具由特定事件触发，并且需要能够访问各种存储服务以检索和存储测试依赖项。该工具的早期评估显示测试预算减少了50%以上。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact of this I-Corps project is the development of technology that will enable companies to substantially reduce the number of software tests they have to run for any code change, thus reducing their software testing cost and improving the developer’s productivity. This technology has the potential to reduce testing time by over 50% on average, across a large number of code changes. The integration of the technology into already existing workflows will introduce minimal disruptions and remain hidden from most software developers, increasing the chance for adoption. This research will bring: (a) a reduction in testing time and faster feedback to developers, (b) a reduction in required resources to run tests which will reduce overall cost and maintenance cost, and (c) an alignment with green computing efforts and carbon neutrality.This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on the development of a regression test selection tool (RTS) that reduces testing cost by skipping tests that are not impacted by code changes. This technology has the potential to be integrated in most common development workflows. The tool can capture dependencies for each test (or a group of similar tests) on fine-grained code level (e.g., classes, methods, and functions). Once a user changes one of the code elements, the test tool will automatically identify a set of tests to run, and those to skip as they are not impacted by changes. The tool is triggered by specific events and needs to be able to access various storage services to retrieve and store test dependencies. Early evaluations of the tool have shown a reduction of the testing budget by over 50%.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.