You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
GET /get_normalized_nodes declares drug_chemical_conflate: bool = fastapi.Query(True, ...) (node_normalizer/server.py:277), but the POST body model CurieList declares drug_chemical_conflate: bool = Field(False, ...) (node_normalizer/model/input.py:29-32).
So the same set of CURIEs normalized via GET and via POST can return different cliques, and nothing in the documentation or the OpenAPI schema warns about it.
v2.4.1 set DrugChemical conflation active by default (#347); it looks like only the GET path was updated.
Changing the POST default is a behaviour change for existing callers, so it probably wants a release note.
Also worth noting: CurieList.Config.schema_extra sets drug_chemical_conflate: True in its example, contradicting the declared default, and omits individual_types and include_taxa entirely.
Documented as a known divergence in documentation/API.md in the meantime.
GET /get_normalized_nodesdeclaresdrug_chemical_conflate: bool = fastapi.Query(True, ...)(node_normalizer/server.py:277), but the POST body modelCurieListdeclaresdrug_chemical_conflate: bool = Field(False, ...)(node_normalizer/model/input.py:29-32).So the same set of CURIEs normalized via GET and via POST can return different cliques, and nothing in the documentation or the OpenAPI schema warns about it.
v2.4.1 set DrugChemical conflation active by default (#347); it looks like only the GET path was updated.
Two things to decide:
True, to match GET and Minor changes to the API #347.Also worth noting:
CurieList.Config.schema_extrasetsdrug_chemical_conflate: Truein its example, contradicting the declared default, and omitsindividual_typesandinclude_taxaentirely.Documented as a known divergence in
documentation/API.mdin the meantime.