Compressing Automatically Generated Unit Test Suites Through Test Parameterization

• Published in: Fundamentals of Software Engineering Vol. 12818; pp. 215 - 221
• Main Authors: Azamnouri, Aidin, Paydar, Samad
• Format: Book Chapter Conference Proceeding
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Automated test generation; Computer Science; Maintainability; Parameterized unit tests; Randoop; Readability; Test suites; Unit test; Automated test generation;Maintainability;Readability;Unit test;Parameterized unit tests;Test suites;Randoop
• ISBN: 3030892468; 9783030892463
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-89247-0_15

Test maintenance has recently gained increasing attention from the software testing research community. When using automated unit test generation tools, the tests are typically created by random test generation or search-based algorithms. Although these tools produce a large number of tests quickly, they mostly seek to improve test coverage; overlooking other quality attributes like understandability and readability. As a result, maintaining a large and automatically generated test suite is quite challenging. In this paper, by utilizing a high level of similarity among the automatically generated tests, we propose a technique for automatically abstracting similar tests through transforming them into parameterized tests. This approach leads to the improvement of readability and understandability by reducing the size of the test suite and also by separating data and logic of the tests. We have implemented this technique as a plugin for IntelliJ IDEA and have evaluated its performance over the test suites produced by the Randoop test generation tool. The results have demonstrated that the proposed approach is able to effectively reduce the size of the test suites between 11% and 96%, with an average of 66%.