owlery server for KB

[dosumis]

Range enforcement (which classes are allowed in which columns or with which relations) currently works by rolling name lookups using regex against short_forms.
This is rather crude. We could achieve much more fine-grained control with OWL queries, but I'm reluctant to use the staging owlery server for this in case the lag between current server state and KB is too great.

Options:

1. Automate pipeline to staging so that it always reflects current state of KB (note - DataSets don't need to be up-to-date as we only care about classes)
2. Run a separate owlery server for KB. This would just have the latest ontologies (FBbt, VFB_ext...) - maybe plus EP & split class nodes from KB + their classifiers, although this would require a custom OWL build job...

[Robbie1977]

@dosumis Is this still, needed? Owlery can be deployed after dumps so a day before the PDB.p2 is created.