Property-based testing for functional programs
Licentiate thesis, 2011

This thesis advances the view that property-based testing is a powerful way of testing functional programs, that has advantages not shared by traditional unit testing. It does this by showing two new applications of property-based testing to functional programming as well as a study of the effectiveness of property-based testing. First, we present a tool, QuickSpec, which attempts to infer an equational specification from a functional program with the help of testing. The resulting specifications can be used to improve your understanding of the code or as properties in a test suite. The tool is applicable to quite a wide variety of situations. Second, we describe a system that helps to find race conditions in Erlang programs. It consists of two parts: a randomised scheduler to provoke unusual behaviour in the program under test and allow replayability of test cases, and a module that tests that all of the functions of an API behave atomically with respect to each other. Finally, we present an experiment we carried out to compare property-based testing against test-driven development. The results were inconclusive, but in the process we developed a black-box algorithm for automatically grading student programs by testing, by inferring for each program a set of bugs that the program contains.

testing

Functional programming

Room EC, EDIT building, Hörsalsvägen 11, Göteborg
Opponent: Colin Runciman, Department of Computer Science, University of York

Author

Nicholas Smallbone

Chalmers, Computer Science and Engineering (Chalmers), Software Technology (Chalmers)

Finding Race Conditions in Erlang with Quick Check and PULSE

ICFP'09: Proceedings of the ACM SIGPLAN International Conference on Functional Programming,; (2009)p. 149-160

Paper in proceedings

Ranking programs using black box testing

Proceedings - International Conference on Software Engineering,; (2010)p. 103-110

Paper in proceedings

QuickSpec: Guessing Formal Specifications using Testing

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics),; Vol. 6143(2010)p. 6-21

Paper in proceedings

Areas of Advance

Information and Communication Technology

Subject Categories

Software Engineering

Publisher

Chalmers University of Technology

Room EC, EDIT building, Hörsalsvägen 11, Göteborg

Opponent: Colin Runciman, Department of Computer Science, University of York

More information

Latest update

1/18/2019