Automating Integration-Level Test Case Generation for Object-Oriented .Net Applications

Sammanfattning: In spite of introducing many techniques and tools, nowadays still most of software testing is done manually. This means spending more cost and time and increasing the possibility of bugs. In addition, by the emergence of new distributed development environments and agile methodologies in recent years, the process of software development has considerably speeded-up, and as a consequence, the concept of DevOps including continuous integration and continuous delivery (CI/CD) has become important more and more. In this context, automatically generation and execution of test cases, specially at integration level has been getting software specialists’ attention more than before; so as to improve the scalability of test process. While many tools have been created for automating Unit Testing in industry, the Integration Testing automation, because of its complexity, has always been a challenge in software engineering and no automation tool has been used in industry for testing at this level. In this thesis, an automated solution for integration-level test-case generation for .Net application was provided. Its test-case generation is technically based on a combination of data-flow analysis and object-oriented concepts such as coupling, and it was implemented by the newly presented .Net compiler named Roslyn. At the next step, the quality of generated test cases was evaluated by examining the solution on a couple of C# projects as benchmarks and confronting the results to 10 mostly used Integration-Level Mutation Operators (, which are specific to Object-Oriented applications). Despite some limitations such as not taking all object-oriented aspects of tested code (e.g. inheritance) into account in the implementation, by considering the reflection of most interface mutations on the generated test paths, and the average execution times, the proposed algorithm showed promising potentials of acceptable coverage on integration-specific parts of code with a reasonable performance. At the end, based on Coupling-Based Analysis, and applicable Roslyn features, a comprehensive automated integration testing (both generation and execution) are proposed as future works.