Rewriting a SPARQL Conjunctive Query according to an OWL2 QL ontology

Question

I have an OWL2 QL ontology like the following one:

Prefix(:=<http://example.org/my-ontology#>)
Ontology(<http://example.org/my-ontology>
  Declaration(Class(:Student))
  Declaration(Class(:BachelorStudent))
  Declaration(Class(:Teacher))
  Declaration(ObjectProperty(:EnrolledIn))
  SubClassOf(:BachelorStudent :Student)
  SubClassOf(ObjectSomeValuesFrom(:EnrolledIn owl:Thing) :Student)
)

My goal is to rewrite a Conjunctive Query (CQ) with respect to the above intensional rules. I can express a CQ in SPARQL as follows:

PREFIX : <http://example.org/my-ontology#>
SELECT ?x WHERE {
  ?x a :Student ;
     a :Teacher .
}

The expected output is a Union of Conjunctive Queries (UCQ), like the following one:

PREFIX : <http://example.org/my-ontology#>
SELECT ?x WHERE {
  { ?x a :Student         ; a :Teacher . } UNION
  { ?x a :BachelorStudent ; a :Teacher . } UNION
  { ?x :EnrolledIn ?y     ; a :Teacher . }
}

(or, equivalently, a set of CQs)

I know there is some theoretical algorithm that accomplishes this task, like PerfectRef (see [1]).

Is there any already-implemented open-source Java library for achieving this or I have to implement this from scratch?