axioms can reappear after filtering them out

[cmungall]

Not sure this is really something that robot can/should fix but wanted to park somewhere

@caufieldjh noticed this odd situation

1. we run robot filter over an application ontology O to remove logical domain constraints
2. when the ontology is reloaded, due to AP/DP punning annotation domain axioms are induced back into logical domain constraints (source: SEPIO defining several dc properties as DPs, causing illegal punning when integrating monarch-initiative/SEPIO-ontology#38)

For now we get around this by changing our filter from just domain to domain AnnotationPropertyDomain in the initial filter

We are working on a smaller test case...

[cmungall]

OK, I don't think this is something robot has to worry about

I got myself confused as I had removed disjointness axioms, which I thought would lead to excluding incoherency with the profile I was using. However, I wasn't actually removing domains. The punning inducement from SEPIO caused some kind of disjointness violation between literals and objects... I think