On the subsumption relationship in data-flow testing

• Grant number: 19/21763-9
• Support Opportunities: Scholarships abroad - Research
• Effective date (Start): March 23, 2020
• Effective date (End): March 22, 2021
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Marcos Lordello Chaim
• Grantee: Marcos Lordello Chaim
• Host Investigator: Andrew Jefferson Offutt
• Host Institution: Escola de Artes, Ciências e Humanidades (EACH). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Research place: George Mason University, United States

Abstract

Data-flow (DF) testing aims at achieving comprehensive structural evaluation of programs. It requires tests that traverse a path in which the definition of a variable and its subsequent use, i.e., a definition-use association (dua), is exercised or covered. However, DF testing has scarcely been adopted in industry because it is considered too costly. We plan to investigate the subsumption relationship defined elsewhere as a way to reduce the costs of DF testing. A dua D1 subsumes a dua D2 if every complete path that traverses D1 traverses D2 as well. This information may reduce the effort to apply DF testing. The tester would focus on those duas that subsume all others, called "unconstrained duas", and "forget'' the rest. The goal of this research proposal is to investigate several aspects underlying this premise, namely, discovery, size, scalability, and effectiveness (i.e., fault detection ability) of unconstrained duas. Moreover, we plan to investigate the role of the dua subsumption relationship in fault localization. (AU)