DataOps-platform/tests/data_research/test_ontology_exchange.py

from __future__ import annotations

import pytest

from tests.data_research.test_ontology_validation import base_graph


def test_json_and_owl_round_trip_preserve_stable_uids_and_hashes():
    from app.core.data_research.ontology.exchange import OntologyExchange
    from app.core.data_research.ontology.repository import canonical_graph_hash

    exchange = OntologyExchange(max_import_bytes=100_000)
    graph = base_graph()
    json_bytes = exchange.export_json("ontology-1", 3, graph)
    owl_bytes = exchange.export_owl("ontology-1", 3, graph)

    imported_json = exchange.import_json(json_bytes)
    imported_owl = exchange.import_owl(owl_bytes)
    assert imported_json["ontology_uid"] == "ontology-1"
    assert imported_owl["ontology_uid"] == "ontology-1"
    assert imported_owl["graph_document"] == graph
    assert canonical_graph_hash(imported_json["graph_document"]) == canonical_graph_hash(graph)
    assert canonical_graph_hash(imported_owl["graph_document"]) == canonical_graph_hash(graph)


def test_owl_import_rejects_unsafe_xml_and_size_limits():
    from app.core.data_research.ontology.exchange import ExchangeInvalid, OntologyExchange

    exchange = OntologyExchange(max_import_bytes=80)
    with pytest.raises(ExchangeInvalid, match="unsafe XML"):
        exchange.import_owl(b'<!DOCTYPE x [<!ENTITY e SYSTEM "file:///etc/passwd">]><x>&e;</x>')
    with pytest.raises(ExchangeInvalid, match="size limit"):
        exchange.import_json(b"{" + b" " * 100 + b"}")


def test_exports_redact_secret_like_fields_deterministically():
    from app.core.data_research.ontology.exchange import OntologyExchange

    graph = base_graph()
    graph["classes"][0]["password"] = "must-not-export"
    first = OntologyExchange().export_json("ontology-1", 1, graph)
    second = OntologyExchange().export_json("ontology-1", 1, graph)
    assert first == second
    assert b"must-not-export" not in first
    assert b"[redacted]" in first