Conversation
shanmukhkalasamudram
commented
Mar 19, 2026
| Query | Checks |
|---|---|
| CCO_classes_have_BFO_superclass | Fixed — old namespace was dead, always passed silently |
| no_cco_elucidation | No terms using deprecated v1 elucidation property |
| missing_definition_source | All terms have a definition source annotation |
| exactly_1_label | Every term has exactly one rdfs:label |
| exactly_1_curated_in | Every term has exactly one "curated in ontology" annotation |
| annotation_language_tag | All natural-language annotations carry @en tag |
| iri_format_check | All IRIs match ont + 8-digit format |
| curated_in_matches_ontology | "Curated in" values point to a valid CCO module IRI |
| no_duplicate_declarations | No term declared as two different OWL types |
| duplicate_iri_number | No two terms share the same 8-digit numeric fragment across namespaces |
- Fix CCO_classes_have_BFO_superclass.sparql namespace (v1 -> v2) - Add validate-profile DL targets to Makefile (Step 6b) - Add build/artifacts/ and build/lib/ to .gitignore - New: no_cco_elucidation.sparql (Step 6j) - New: missing_definition_source.sparql (Step 6h) - New: exactly_1_label.sparql (Step 6c) - New: exactly_1_curated_in.sparql (Step 6c) - New: annotation_language_tag.sparql (Step 6i) - New: iri_format_check.sparql (Step 6e) - New: curated_in_matches_ontology.sparql (Step 6f) - New: no_duplicate_declarations.sparql (Step 6d) All queries tested against CommonCoreOntologiesMerged.ttl and verified with synthetic TTL files to confirm non-silent failure. Pre-release gaps surfaced: 815 missing definition sources, 1 missing curated-in (ont00001660), 9 missing @en language tags.
- Add .github/release.yml to categorize PRs by label in release notes - Add 'Generate Release Notes' step to manage_release.yml that calls the GitHub API to auto-generate notes and injects them into the existing draft release via gh release edit --notes-file
|
@shanmukhkalasamudram Did you check that these return the expected results? If so, can you provide the results of your testing in the comments here? |
|
Hi @johnbeve — yes, tested locally by running all 19 SPARQL queries via Impact summary: After this PR, every push runs 19 checks. Seven pass completely clean against the current ontology, giving us a reliable release gate for those properties. The remaining queries surfaced pre-existing issues that were previously invisible — none were introduced by this PR. Queries passing clean (0 violations)
Pre-existing issues surfaced by the new queries
The remaining results contain only test-fixture IRIs ( |
|
Full CSV results are attached as |