Automated Test Generation Based on an Applicational Model

• Main Author: Cabeda, Mariana Araújo
• Publication Date: 2018
• Format: Master thesis
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10362/62913
• topic: Software testing; Software test automation; Software test coverage; OutSystems; Visual Programming Language; OutSystems applicational model; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

Context: As testing is an extremely costly and time-consuming process, tools to automatically generate test cases have been proposed throughout the literature. OutSystems provides a software development environment where with the aid of the visual OutSystems language, developers can create their applications in an agile form, thus improving their productivity. Problem: As OutSystems aims at accelerating software development, automating the test case generation activity would bring great value to their clients. Objectives: The main objectives of this work are to: develop an algorithm that generates, automatically, test cases for OutSystems applications and evaluates the coverage they provide to the code, according to a set of criteria. Methods: The OutSystems language is represented as a graph to which developers can then add pieces of code by dragging nodes to the screen and connecting them to the graph. The methodology applied in this work consists in traversing these graphs with depth and breadth-first search algorithms, employing a boundary-value analysis to identify the test inputs and a cause-effect graphing to reduce the number of redundant inputs generated. To evaluate these test inputs, coverage criteria regarding the control flow of data are analysed according to node, branch, condition, modified condition-decision and multiple condition coverage. Results: This tool is able to generate test inputs that cover 100% of reachable code and the methodologies employed help greatly in reducing the inputs generated, as well as displaying a minimum set of test inputs with which the developer is already able to cover all traversable code. Usability tests also yield very optimistic feedback from users. Conclusions: This work’s objectives were fully met, seen as we have a running tool able to act upon a subset of the OutSystems applicational model. This work provides crucial information for assessing the quality of OutSystems applications, with value for OutSystems developers, in the form of efficiency and visibility.