Open Property Data Association
Smart Property Data Trust Framework
Become a member Join a working group

    Running & validating

    Everything on this page runs from source/03-standards/rml/. It needs a JDK (RMLMapper self-provisions into a repo-root .rmlmapper/ cache on first use) and the tools/opda-gen Python virtualenv, which the harness scripts already invoke directly (../opda-gen/.venv/bin/python) — no separate setup.

    Make targets

    cd source/03-standards/rml
    TargetWhat it does
    make provenance-testPrimary gate. RML↔OWL consistency + resource↔schema-location tracing (Jena arq). No instance data.
    make dct-auditAudits every ontology dct:source citation against the canonical schema dictionary.
    make rml-testFull instance-materialisation pipeline: 3 conformant fixtures (sound + layer-1-complete) + 1 negative (must SHACL-violate) + the acyclicity regression guard.
    make rml-mapMaterialise the default fixture → build/out.nt.
    make rml-shaclValidate build/out.nt against opda-shapes-merged.ttl.
    make rml-completeLayer-1 completeness gate over the materialised output.
    make rml-pytestRuns tests/test_rml_mapping.py (sound / complete / negative assertions).
    python3 build/final_scope2.pyRegenerates build/final-gap.json — the domain-module resource coverage figure cited on the section overview and Coverage & gaps.

    Run against any conformant PDTF instance directly: 

    ../opda-gen/.venv/bin/python harness/run_mapping.py \
  --mapping mapping/opda-pdtf.rml.ttl \
  --data ../../source/03-standards/schemas/src/examples/v3/exampleTransaction.json \
  --out build/out.nt
bash harness/validate_shacl.sh build/out.nt

    File layout

    PathRole
    mapping/opda-pdtf.rml.ttlThe mapping itself — the canonical deliverable.
    mapping/functions/FnO functions.ttl + the Java UDF (OpdaFunctions.java) RMLMapper loads for the enum → SKOS-concept-IRI function and date truncation.
    CONTRACT.mdThe build contract — ground-truth inputs, namespace law, layer-1/layer-2 definitions, "definitions of done."
    README.mdUsage, the scope/coverage ledger, SHACL traps handled.
    provenance-index.json / .mdThe leaf→predicate inversion — machine and human form.
    harness/run_mapping.py, validate_shacl.sh, check_completeness.py, audit_dct_source.py, validate_provenance.py.
    testdata/4 fixtures (3 conformant + 1 negative) + MANIFEST.md describing what each exercises.
    tests/test_rml_mapping.pyThe pytest suite — sound / complete / negative assertions.
    build/Generated output — N-Triples + JSON reports, including final-gap.json. Gitignored; regenerate locally, don't expect it fresh on a clone.

    Two documents in this directory, gap-register.md and ONTOLOGY-COVERAGE.md, are point-in-time audits from earlier in the gap-closing effort and now carry their own superseded-notices pointing at build/final-gap.json as current truth — the same reason Coverage & gaps on this site cites that file and not them.

    ← Back to Ontology · Back to Schema →

    Comments

    Loading comments…