SHF: Small:Automatically Synthesizing System and Integration Tests

点击翻译按钮获取中文摘要

System and integration testing is a set of key activities in software development and maintenance to determine if applications and their constituent components behave correctly. System/integration testing is vitally important for ensuring software quality, since many serious software defects occur in interactions among different components. Unfortunately, creating effective system/integration tests requires significant time and effort, since it is not feasible to test the enormous set of all combinations of the components from a nontrivial software application. It is crucial to reduce the cost and increase the quality of software with system/integration testing that is effective in fault finding and efficient in resource consumption and execution time. In sum, there are unique factors in system/integration testing, not previously considered, that if not addressed will continue to limit the quality of software applications. In this project, a novel framework will be created for automatically creating effective system/integration tests that find bugs efficiently. A three-pronged research program will be developed for Automatic Synthesis of System and Integration Software Tests (ASSIST). New algorithms and techniques will be created and evaluated as part of the work on this project for automatically obtaining models that describe interacting components, thus reducing the number of synthesized system and integration tests and increasing their fault-finding power. Also, a novel way will be created in which static and dynamic analyses and machine learning are used to obtain test input data as well as oracles for the synthesized system and integration tests. As a result of the work on this project, system and integration tests will be automatically synthesized using ASSIST that have a comparable bug-finding power with manually created integration tests. A suite of new ASSIST tools that will be developed, evaluated, and applied to different open-source and commercial applications, and made available to the broader community.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.