arrow
Return

Automated code-based test case reuse for software product line testing

delete2024-02-01
delete1
PRE
AI
P
Pilsu Jung
S
Seonah Lee *
U
Uicheon Lee
DOI:10.1016/j.infsof.2023.107372delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
Context: A software product line (SPL) grows in size as a new product is developed. A new product in an SPL should be tested extensively for quality assurance. For the efficient testing, previous studies suggested reusing the existing test cases of a product family. However, either their methods were not efficient because interventions from human experts, specifications, architecture and/or traceabilities for test cases were required. Objective: To address these limitations, we propose an Automated Code-based Test case reuse for SPLs (ActSPL). ActSPL automatically identifies reusable test cases for new products of a product family using source code and test cases. Method: ActSPL automatically constructs a hash-based traceability links between test cases and source code of a product family. Using the traceability links, ActSPL selects reusable test cases for a given new product from existing test cases of the product family. Results: We evaluated ActSPL in terms of the effectiveness and cost reduction of reusing test cases with five open source SPLs. The evaluation results showed that ActSPL, on average, achieved 100 % precision and 62 % recall. In addition, ActSPL, on average, saved 47.5 % of time required for testing a new product from scratch. Conclusion: Our study shows the feasibility of ActSPL reusing SPL test cases based on source code and test cases. Our results can be a basis for successive studies for automated code-based SPL testing.
Keywords:
Software product line
Product family
Software testing
Test case reuse
Traceability

Journal

Information and Software Technology cover
Information and Software Technology
IF:
4.3
Papers:
3.7K
Citations:
7.7K

Organization

G
Gyeongsang National University
Scholars:
10.0K
Papers: 8.8K
Citations: 13