check that we deal with obsoletion workflows and dangling classes

[nlharris]

(from @cmungall)

[gouttegd]

Not sure what Chris meant with that comment, but this might be related to #52, about what to do with axioms/edges that refer to a to-be-obsoleted class.

[nlharris]

Chris added this comment in response to my question about Table 1 in the KGCL paper (Types of changes supported by KGCL, grouped into Node Changes and Edge Changes). I asked whether Node Deletion also deletes the child nodes, or what it does, and Chris noted that that needed to be addressed.

[nlharris]

See also #71