Selecting Effective Test Sets

Published

May 26, 2006

Author(s)

Leonard E. Gebase, Robert D. Snelick, Roch Bertucat

Abstract

One of the first problems that must be addressed when attempting to test the reliability of any software implementation is determining what is to be tested. The number of paths an implementation can follow is generally much too large to be able to test all possibilities. Pragmatic decisions must be made that allow selective testing, but exhaustive testing covering all possible paths is not possible. In this paper we investigate an approach for finding sets of messages that provide effective coverage for testing HL7 applications. HL7 messages are constrained by an XML document that determines the form and content of the messages. We describe a method for determining one metric for measuring the set of messages that encompass all paths allowed by the XML document. In general this set of messages is also very large. Nevertheless, the metric provides us with a concrete assessment of the scope of the testing problem. Furthermore, we are able to formulate techniques for filtering this large message set and substantially reduce its size while maintaining the essential coverage capabilities inherent in the larger message set.

Conference Dates

June 26-29, 2006

Conference Location

Las Vegas, NV, USA

Conference Title

2006 International Conference on Software Engineering Research and Practice (SERP'06)

Pub Type

Conferences

Download Paper

Local Download

Keywords

automated methods, combinations, conformance testing, filtering, messaging systems, test messages

Citation

Gebase, L. , Snelick, R. and Bertucat, R. (2006), Selecting Effective Test Sets, 2006 International Conference on Software Engineering Research and Practice (SERP'06), Las Vegas, NV, USA, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=50830 (Accessed October 11, 2024)