How do you test software stability and quality while reusing multiple FLOSS components from diverse sources? How do you ensure all components will behave as expected? The DECODER central database (PKM) stores code-related knowledge: artifacts and bindings between them, obtained by generating formal specifications from informal data (like documentations) or semi-formal models out of source code. DECODER provides general-purpose languages and methods applicable to any application domain, some considering source code as natural language and applying NLP (Natural Language Processing) algorithms to it. Innovative tools are being experimented on several Java/C/C++ use cases such as IoT, AI-based image recognition, enterprise computing, Cloud computing, Big Data and middleware.
DECODER is developing the first Software Project Intelligence platform. It aims to assist developers in the task of understanding the software project in which they are involved.
This is achieved by providing different views of the software project and mechanisms to navigate between the different software assets. DECODER also brings automatic generation of annotations, documents and resources to help understand and improve software code. With the DECODER open source Integrated Development Environment (IDE), developers will write high quality code that is more secure and better aligned with requirements and maintainers will immediately know what has been done, how and with what tools.
DECODER is reducing the software project learning curve while increasing the developer’s productivity. It provides a new interface to developers with several views at different abstraction levels, and with a focus on file dependencies, modularization, and data structures being used. DECODER stands for DEveloper COmpanion for Documented and annotatEd code Reference. This collaborative project is driven by CEA (Commissariat à l’Énergie Atomique et aux Énergies Alternatives) as technicalleader and coordinated by Technikon. Project partners include Universitat Politècnica de València, Capgemini Group, Sysgo, Tree Technology, and OW2. It has received funding from the European Union’s H2020 research and innovation programme under the grant agreement 824231.