Re: Diff'ing RDF files

Hi all,

There is an open-source tool available from the EDM Council for converting between RDF/XML, Turtle, and JSON-LD and for consistent serialization of any of these representations of RDF and OWL. The GitHub site for it is https://github.com/edmcouncil/rdf-toolkit. It is actively maintained, freely available, and addresses a number of issues mentioned on the thread, among other things. It also allows users to turn any of its features on/off as desired. It runs on the command line, or can be invoked automatically through GitHub commit hooks, for example.

For collaborative work across development teams for large ontology projects, consistent serialization for comparison purposes was one of our first and relatively important issues. It enables visual comparison in GitHub (and likely other source code management systems), so that anyone reviewing the changes can see exactly what changed, down to the single character level. There is also an axiomatic diff tool available via the OBO Foundry that folks might find useful, available at https://robot.obolibrary.org/.  I don't know how well it works on RDF alone, mainly because I haven't attempted to use it for that, but it works well as a companion tool to the RDF Toolkit from the EDM Council as needed.

We also have a pipeline that looks for a myriad of issues in ontologies, performs regression testing using examples and reference data, and includes an html-based publication process that itself has a comparison feature, enabling comparison of any pull request or prior release with another version or with the latest version. The code for this is also open source, available from the EDM Council GitHub repository, though support is required for hosting and customization.

Best regards,

Elisa