Supporting libraries

The concepts enables some nice streamlining for the parser and writer libraries pertaining the model. The following image showcases the idea visually.

All core graph sub-libraries are hand-coded to use relevant graphing libraries available at the implementation language (e.g. NetworkX for Python or jgrapht for Java). Note that relevant does not imply the most performant for huge graphs!

We also have a meta model for our “types” or “kinds” of vertexes, edges etc, which are injected into the soon-to-be library so that it gets realised with the graphing features and the types that a ForSyDe IO model contains.