An Automated Approach to Generate Test Cases From Use Case Description Model

Abstract

Test complexity and test adequacy are frequently raised by software developers and testing agents. However, there is little research works at this aspect on specification-based testing at the use case description level. Thus, this research proposes an automatic test cases generator approach to reduce the test complexity and to enhance the percentage of test coverage. First, to support the infrastructure for performing automatic, this proposed approach refines the use cases using use case describing template and save it in the text file. Then, the saved file is input to the Algorithm of Control Flow Diagram (ACFD) to convert use case details to a control flow diagram. After that, the Proposed Tool of Generating Test Paths (PTGTP) is used to generate test cases from the control flow diagram. Finally, the genetic algorithm associated with transition coverage is adapted to optimize and evaluate the adequacy of such test cases. A money withdrawal use case in the ATM system is used as an ongoing case study. Preliminary results show that the generated test cases achieve high coverage with an optimal test case. This automatic test case generation approach is effective and efficient. Therefore, it could promote to use other test case coverage criteria.

A Review of Automatic Generation of Test Cases from Use Case Specification

“…According to [18] modeling speci cations can reduce the time and effort spent in test case generation, thus the use of models in the approach [13] also decreases the time and effort required by software engineers to add missing speci cations into new or existing test cases. Likewise, the approach in [14], uses a template [15] to create a complete description of use cases by including all the relevant details to eliminate manual intervention to ensure the coverage of essential requirements by test cases.…”

“…Manual intervention is required by the software engineer between nearly every automated step. On the contrary, the approaches proposed in [12,14] are fully automatic and do not require manual work. Nonetheless, UMTG [13] can generate test cases parallelly in a multicore platform, decreasing the performance time which would be increased due to manual intervention.…”

Mutated Specification-Based Test Data Generation with a Genetic Algorithm

“…SBT is characterized by test data being generated from the specification without concerning the structure of the corresponding program and test results being analyzed based on the specification [4][5][6][7][8]. Formal specifications may allow the test data generation and test result analysis to be done rigorously, systematically, and even automatically in many cases [9][10][11][12]. In our work, we mainly use formal specifications in pre-and postconditions, such as Vienna Development Method (VDM) [13], a formal method that has been developed over past years [14,15], and Structured-Object-Oriented Formal Language (SOFL) [16], which has the potential of practical use in industry and serves as a solid foundation to develop a method of functional scenario-based test data generation [17].…”

A Prioritization Approach for Regression Test Cases Based on a Revised Genetic Algorithm

“…Many research works had employed the heuristic search algorithms to propose an approach to generate or to prioritize the regression test cases [2,8,14,24,18,19].…”

“…Alrawashed et al [2] developed an automated approach for test suite optimization. In their approach, they generated test cases from the use case description model.…”