Detecting duplicate examples in behaviour driven development specifications

In Behaviour-Driven Development (BDD), the behaviour of the software to be built is specified as a set of example interactions with the system, expressed using a “Given-When-Then” structure. The examples are written using customer language, and are readable by end-users. They are also executable, and act as tests that determine whether the implementation matches the desired behaviour or not. This approach can be effective in building a common understanding of the requirements, but it can also face problems. When the suites of examples grow large, they can be difficult and expensive to change. Duplication can creep in, and can be challenging to detect manually. Current tools for detecting duplication in code are also not effective for BDD examples. Moreover, human concerns of readability and clarity can rise. We present an approach for detecting duplication in BDD suites that is based around dynamic tracing, and describe an evaluation based on three open source systems.