-
Home
-
European Projects
-
Property-based Testing (ProTest)
Property-based Testing
(ProTest)
Start date: May 1, 2008,
End date: Dec 31, 2011
PROJECT
FINISHED
Description
This project will develop software engineering approaches to improve reliability in future service-oriented networks.In order to improve reliability and functionality in Pervasive and Trusted Network and Service Infrastructures, PROTEST will develop software engineering approaches based on specified properties of the system which will support fault-finding and diagnosis. This will be achieved by building automated tools that will generate and run tests, observe and control execution during runtime and register occurrences for analysis. Testing in Service-oriented Networks Volumes of automated tests and logging code are written which use properties as objects for testing software. Such property-based testing will both reduce costs which will deliver economic benefits to the European software industry and improve effectiveness. It is a powerful new mechanism that will assure that a given programme is free of specific generic flaws. In case properties fail, the tools will search for simplest failing cases and analyse trace and coverage information to assist speedy diagnosis. In order to face a major challenge of concurrency, model-checking will be integrated into the tools. Today, specialists and researchers are not used to formulating general properties, which is why PROTEST will investigate possibilities of deriving them from two different sources: * Unified Modeling Language (UML) models, or UML-like models. * By refactoring existing test suites. Making an impact by combining different skills of different actorsThe combination of skills coming from universities, SMEs and larger industrial enterprises will optimise the chances of delivering results. * Academics will bring expertise in refactoring, model-checking and testing. They thereby act as real tool vendors. * Industrial expertise will be needed in telecoms in order to bring insights into what practical tools are required, what properties will need to be checked and strategies to maximise industrial impact. The initial implementation vehicle will be Erlang/OTP (Open Telecom Platform) developed by Ericsson in the early nineties. It is an open-source concurrent functional language aimed at telecoms and internet services and represents a common instrument for our research that simplifies the transfer of theory into industrial practice. Because of its good interoperability, robustness and reliability, Erlang enables PROTEST's tools to detect faults in various systems. The results of PROTEST will improve the tool vendor's products. Thus, they will be adopted by partners such as Ericsson and Lambda-stream and disseminated by ETC to their customers throughout the European telecoms sector. This three-pronged strategy will guarantee real impact.