AutomTest 3.0: An automated test-case generation tool from User Story processing powered with LLMs
Joanna C. S. Santos, Paulo Maciel · 2024
Test-driven development (TDD) and Behavior-driven development (BDD) approaches address software development issues, proposing the creation of unit tests for each system component before implementing the code for the corresponding functionality. This approach results in an objective and compact solution. It ensures that the entire codebase is tested and enhances the overall quality of the software. However, the testing process can be time-consuming and resource-intensive. Some tools were proposed to address those issues, but they lack functionalities based on User Story in a TDD approach. In this context, we propose the AutomTest 3.0 tool, which facilitates unit test case generation from user stories prior to writing the code base related to the functionality. Built upon previous versions of AutomTest, this tool extends the other ones by including generative AI to power the creation of suggestions for methods the software under test should have. This tool creates test cases for Java using techniques such as equivalence class partitioning, natural language processing, and generative artificial intelligence. To evaluate the effectiveness of AutomTest 3.0, we conducted an exploratory study with software development professionals. Their feedback highlighted the tool’s utility in their daily work routines. AutomTest 3.0 demonstrated promising results in test case generation, scenario coverage, and speed advantages in test case creation. https://youtu.be/xoTrvhlfvu8