Lack of automated feedback / validation

[dbooth-boston]

"One of the greatest things that happened in learning HTML (1994, in my case) was the existence of validators to provide feedback of whether I was doing it right. The RDF* suite provides me no feedback as to whether I'm doing it right." https://lists.w3.org/Archives/Public/semantic-web/2018Nov/0229.html

[joshuef]

Toootally agree.

I've played around with a javascript based proof of concept over here: https://github.com/joshuef/rdf-check-mate

Using 'schemas' from npm packages to provide on the fly validation of objects.

(All of which is part of a toolchain I've been playing with, described over here: https://forum.safedev.org/t/javascript-rdf-tooling-ideas/2239 if anyone's interested).

[chiarcos]

For mere data validation, I sometimes recommend http://ttl.summerofcode.be/ to people not used to work on the command line. It's old, basic, and restricted to Turtle, of course, but robust and proven to be helpful to beginners. There are some RDF converter sites which should provide validation, too. (I just tried http://www.easyrdf.org/converter but it broke for my test sample :( )

As a side-note, I also introduce RDF via Turtle because it most naturally leads over to querying RDF.

For vocabulary-specific validation, the situation is much more complex. For example, I am not aware of any satisfying* SKOS validator/visualization available online, for example. (* There are several, but visualizers can be restricted to a fragment of the language and/or the error reports are not necessarily helpful.)

[KonradHoeffner]

On a local machine, I have been very happy with the rapper (redland raptor) tool for many years. Validating a file is as simple as rapper -i turtle -c myfile.ttl and it even gives helpful error messages. As for validating data against a schema, pySHACL works very well and if you want to integrate it into a CI pipeline, you can reuse this GitHub action workflow.