Combining specification-based testing, correctness proof, and inspection for program verification in practice

N/ACitations
Citations of this article
6Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Specification-based testing is limited in detecting program errors; correctness proof based on Hoare logic is difficult to perform in practice; and inspection is heavily dependent on human decisions. Each of these three is difficult to do a satisfactory job alone, but they complement each other when they come together in an appropriate manner. This paper puts forward a new method that makes good use of Hoare logic and inspection to improve the effectiveness of specification-based testing in detecting errors. The underlying principle of the method is first to use specification-based testing to discover traversed program paths and then to use Hoare logic to prove their correctness, but when proof is impossible to conduct, a special inspection is applied. During the proof or inspection process, all faults on the paths are expected to be detected. A case study is conducted to show its feasibility; an example taken from the case study is used to illustrate how the proposed method is applied; and a discussion on the important issues to be addressed in the future is presented. © 2014 Springer International Publishing Switzerland.

Cite

CITATION STYLE

APA

Liu, S., & Nakajima, S. (2014). Combining specification-based testing, correctness proof, and inspection for program verification in practice. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8332 LNCS, pp. 3–16). Springer Verlag. https://doi.org/10.1007/978-3-319-04915-1_1

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free