Work with Other Formats¶

Ontoplexis reads and writes exactly one document format: OWL/XML, because it is the OWL 2 structural specification serialized as XML. Everything else — format conversion, reasoning, profile checking — is a pre- or post-processing step with the OWLAPI-based tool you already use. This page shows the common recipes.

Convert to OWL/XML¶

With ROBOT, the OBO community's OWLAPI CLI:

robot convert --input pizza.ttl --format owx --output pizza.owx

Or export from Protégé (File → Save As… → OWL/XML Syntax), or call OWLAPI directly from Java/JPype. Then:

from pathlib import Path
from ontoplexis import Ontology

ont = Ontology.from_owlxml(Path("pizza.owx").read_text())

Convert from OWL/XML¶

The same in reverse — serialize, then convert:

robot convert --input out.owx --format ttl --output out.ttl

Path("out.owx").write_text(ont.to_owlxml())

Reason before loading¶

If you want inferred axioms in the graph, materialize them first:

robot reason --reasoner hermit --input pizza.owx \
      --axiom-generators "SubClass ClassAssertion" \
      convert --format owx --output pizza-inferred.owx

robot reason also fails on inconsistent ontologies, covering the consistency-check use case; robot validate-profile reports OWL 2 profile membership.

Why no built-in converter?¶

Earlier versions bundled OWLAPI as a subprocess bridge. It made every install carry a JVM dependency and a fat jar for what is a one-time preprocessing step in most pipelines. The reference implementation still guards the walker — the round-trip fidelity tests run OWLAPI as a development-only oracle — it just no longer ships.