Refactor reference class usage

src/pyobo/api/alts.py

@@ -11,7 +11,7 @@
 from .utils import _get_pi, get_version_from_kwargs
 from ..constants import GetOntologyKwargs, check_should_cache, check_should_force
 from ..getters import get_ontology
-from ..identifier_utils import wrap_norm_prefix
+from ..identifier_utils import Reference, wrap_norm_prefix
 from ..utils.cache import cached_multidict
 from ..utils.path import CacheArtifact, get_cache_path
 
@@ -67,7 +67,7 @@ def get_primary_reference(
     identifier: str | None = None,
     /,
     **kwargs: Unpack[GetOntologyKwargs],
-) -> curies.ReferenceTuple | None:
+) -> Reference | None:
     """Get the primary reference for an entity."""
     reference = _get_pi(prefix, identifier)
     try:
@@ -77,7 +77,7 @@ def get_primary_reference(
             raise
         # this happens on invalid prefix. maybe revise?
         return None
-    return curies.ReferenceTuple(reference.prefix, primary_identifier)
+    return Reference(prefix=reference.prefix, identifier=primary_identifier)
 
 
 def get_primary_curie(

src/pyobo/api/hierarchy.py

@@ -17,8 +17,8 @@
 from .properties import get_literal_properties
 from .utils import _get_pi
 from ..constants import GetOntologyKwargs
+from ..identifier_utils import Reference
 from ..struct import has_member, has_part, is_a, member_of, part_of
-from ..struct.reference import Reference
 from ..struct.struct_utils import ReferenceHint, _ensure_ref
 
 __all__ = [

src/pyobo/api/relations.py

@@ -22,8 +22,7 @@
     check_should_use_tqdm,
 )
 from ..getters import get_ontology
-from ..identifier_utils import wrap_norm_prefix
-from ..struct.reference import Reference
+from ..identifier_utils import Reference, wrap_norm_prefix
 from ..struct.struct_utils import ReferenceHint, _ensure_ref
 from ..utils.cache import cached_df
 from ..utils.path import CacheArtifact, get_cache_path, get_relation_cache_path

src/pyobo/api/utils.py

@@ -1,11 +1,9 @@
 """Utilities for high-level API."""
 
-import warnings
-
 import curies
-from bioregistry import NormalizedNamableReference as Reference
 from curies import ReferenceTuple
 
+from ..identifier_utils import Reference
 from ..utils.ver import (
     VersionError,
     get_version,
@@ -24,21 +22,14 @@
 
 
 def _get_pi(
-    prefix: str | curies.Reference | ReferenceTuple, identifier: str | None = None, /
+    reference: str | curies.Reference | ReferenceTuple, _unused: str | None = None, /
 ) -> Reference:
-    if isinstance(prefix, ReferenceTuple | curies.Reference):
-        if identifier is not None:
-            raise ValueError("unexpected non-none value passed as second positional argument")
-        return Reference(prefix=prefix.prefix, identifier=prefix.identifier)
-    if isinstance(prefix, str) and identifier is None:
-        return Reference.from_curie(prefix)
-    if identifier is None:
-        raise ValueError(
-            "prefix was given as a string, so an identifier was expected to be passed as a string as well"
-        )
-    warnings.warn(
-        "Passing a prefix and identifier as seperate arguments is deprecated. Please pass a curies.Reference or curies.ReferenceTuple in the first positional-only argument instead.",
-        DeprecationWarning,
-        stacklevel=4,  # this is 4 since this is (always?) called from inside a decorator
-    )
-    return Reference(prefix=prefix, identifier=identifier)
+    if _unused is not None:
+        raise ValueError("unexpected non-none value passed as second positional argument")
+    if isinstance(reference, Reference):
+        return reference
+    if isinstance(reference, ReferenceTuple | curies.Reference):
+        return Reference.from_reference(reference)
+    if isinstance(reference, str):
+        return Reference.from_curie(reference)
+    raise TypeError(f"unexpected type {type(reference)}")

src/pyobo/cli/lookup.py

@@ -21,7 +21,7 @@
 from ..constants import GetOntologyKwargs
 
 if TYPE_CHECKING:
-    from curies import Reference
+    import curies
 
 __all__ = [
     "lookup",
@@ -306,7 +306,7 @@ def descendants(curie: str, **kwargs: Unpack[GetOntologyKwargs]) -> None:
 
 
 def _list_curies(
-    references: Iterable[Reference] | None, **kwargs: Unpack[GetOntologyKwargs]
+    references: Iterable[curies.Reference] | None, **kwargs: Unpack[GetOntologyKwargs]
 ) -> None:
     if not references:
         return

src/pyobo/identifier_utils/__init__.py

@@ -8,6 +8,7 @@
     NotCURIEError,
     ParseError,
     ParseValidationError,
+    Reference,
     UnparsableIRIError,
     UnregisteredPrefixError,
     _is_valid_identifier,
@@ -24,6 +25,7 @@
     "NotCURIEError",
     "ParseError",
     "ParseValidationError",
+    "Reference",
     "UnparsableIRIError",
     "UnregisteredPrefixError",
     "_is_valid_identifier",

src/pyobo/identifier_utils/api.py

@@ -9,13 +9,13 @@
 
 import bioregistry
 import click
-from bioregistry import NormalizedNamableReference as Reference
 from bioregistry.constants import FailureReturnType
 from curies.preprocessing import BlocklistError, PreprocessingConverter
 from curies_processing import get_rules
 from pydantic import ValidationError
 from typing_extensions import Doc
 
+from .reference import Reference
 from .relations import ground_relation
 
 __all__ = [
@@ -24,6 +24,7 @@
     "NotCURIEError",
     "ParseError",
     "ParseValidationError",
+    "Reference",
     "UnparsableIRIError",
     "UnregisteredPrefixError",
     "_parse_str_or_curie_or_uri_helper",

src/pyobo/identifier_utils/reference.py

@@ -0,0 +1,5 @@
+"""Defines a PyOBO reference."""
+
+from bioregistry import NormalizedNamableReference as Reference
+
+__all__ = ["Reference"]

src/pyobo/identifier_utils/relations/api.py

@@ -6,9 +6,10 @@
 from pathlib import Path
 
 import requests
-from bioregistry import NormalizedNamedReference
 from tqdm import tqdm
 
+from ..reference import Reference
+
 __all__ = [
     "get_normalized_label",
     "ground_relation",
@@ -65,7 +66,7 @@ def _norm(s: str) -> str:
     return s.replace(" ", "").replace("_", "").replace(":", "").lower()
 
 
-def ground_relation(s: str) -> NormalizedNamedReference | None:
+def ground_relation(s: str) -> Reference | None:
     """Ground a string to a RO property."""
     return get_lookups().get(_norm(s))
 
@@ -82,14 +83,14 @@ def get_normalized_label(curie_or_uri: str) -> str | None:
 
 
 @lru_cache(1)
-def get_lookups() -> Mapping[str, NormalizedNamedReference]:
+def get_lookups() -> Mapping[str, Reference]:
     """Get lookups for relation ontology properties."""
     d = {}
     for record in json.loads(PATH.read_text()):
         prefix, identifier, label = record["prefix"], record["identifier"], record["label"]
-        d[_norm(label)] = NormalizedNamedReference(prefix=prefix, identifier=identifier, name=label)
+        d[_norm(label)] = Reference(prefix=prefix, identifier=identifier, name=label)
         for s in record.get("synonyms", []):
-            d[_norm(s)] = NormalizedNamedReference(prefix=prefix, identifier=identifier, name=label)
+            d[_norm(s)] = Reference(prefix=prefix, identifier=identifier, name=label)
     return d
 
 

src/pyobo/ner/normalizer.py

@@ -58,5 +58,5 @@ def ground(
         #  can be type annotated with the right reference
         return Reference.from_reference(match.reference)
     if strict_match:
-        raise ValueError
+        raise ValueError(f"no match found for query: {query} against prefixes: {prefix}")
     return None

src/pyobo/struct/__init__.py

@@ -1,12 +1,6 @@
 """Data structures for OBO."""
 
-from .reference import (
-    OBOLiteral,
-    Reference,
-    Referenced,
-    _parse_str_or_curie_or_uri,
-    default_reference,
-)
+from .reference import OBOLiteral, Referenced, _parse_str_or_curie_or_uri, default_reference
 from .struct import (
     CHARLIE_TERM,
     HUMAN_TERM,
@@ -41,6 +35,7 @@
     transcribes_to,
     translates_to,
 )
+from ..identifier_utils import Reference
 
 __all__ = [
     "CHARLIE_TERM",
@@ -50,6 +45,7 @@
     "OBOLiteral",
     "Obo",
     "Reference",
+    "Reference",
     "Referenced",
     "Stanza",
     "StanzaType",

src/pyobo/struct/functional.py

@@ -14,8 +14,9 @@
 from rdflib import XSD
 
 from . import vocabulary as pv
-from .reference import OBOLiteral, Reference, _parse_datetime
+from .reference import OBOLiteral, _parse_datetime
 from .struct import get_iris
+from ..identifier_utils import Reference
 
 if TYPE_CHECKING:
     from .reference import Referenced

src/pyobo/struct/obo/reader.py

@@ -14,7 +14,6 @@
 
 import bioregistry
 import networkx as nx
-from bioregistry import NormalizedNamableReference as Reference
 from curies import ReferenceTuple
 from curies.preprocessing import BlocklistError
 from curies.vocabulary import SynonymScope
@@ -45,6 +44,7 @@
 from ...identifier_utils import (
     NotCURIEError,
     ParseError,
+    Reference,
     UnparsableIRIError,
     _is_valid_identifier,
     _parse_str_or_curie_or_uri_helper,

src/pyobo/struct/obo/reader_utils.py

@@ -8,10 +8,10 @@
 from collections.abc import Mapping, Sequence
 
 import click
-from bioregistry import NormalizedNamableReference as Reference
 from curies import ReferenceTuple
 from curies import vocabulary as v
 
+from pyobo.identifier_utils import Reference
 from pyobo.struct.reference import (
     OBOLiteral,
     _obo_parse_identifier,

src/pyobo/struct/obograph/utils.py

@@ -4,10 +4,10 @@
 from typing import cast
 
 import obographs as og
-from curies import Reference
 from obographs import StandardizedGraph, StandardizedMeta
 
 from pyobo.constants import TypeDefType
+from pyobo.identifier_utils import Reference
 
 __all__ = [
     "assert_graph_equal",

src/pyobo/struct/reference.py

@@ -14,22 +14,21 @@
 import curies
 import dateutil.parser
 import pytz
-from bioregistry import NormalizedNamableReference as Reference
 from curies import ReferenceTuple
 from curies import vocabulary as v
 from curies.preprocessing import BlocklistError
 
 from ..identifier_utils import (
     NotCURIEError,
     ParseError,
+    Reference,
     UnparsableIRIError,
     _is_valid_identifier,
     _parse_str_or_curie_or_uri_helper,
 )
 
 __all__ = [
     "OBOLiteral",
-    "Reference",
     "Referenced",
     "default_reference",
     "get_preferred_curie",
@@ -158,9 +157,13 @@ def default_reference(prefix: str, identifier: str, name: str | None = None) ->
 
     >>> default_reference("chebi", "conjugate_base_of")
     Reference(prefix="obo", identifier="chebi#conjugate_base_of", name=None)
+
+    >>> default_reference("CHEBI", "conjugate_base_of")
+    Reference(prefix="obo", identifier="chebi#conjugate_base_of", name=None)
     """
     if not identifier.strip():
         raise ValueError("default identifier is empty")
+    prefix = bioregistry.normalize_prefix(prefix) or prefix.lower()
     return Reference(prefix="obo", identifier=f"{prefix}#{identifier}", name=name)
 
 

src/pyobo/struct/skos/export.py

@@ -6,7 +6,8 @@
 from typing import TYPE_CHECKING
 
 if TYPE_CHECKING:
-    from curies import Converter, Reference
+    import curies
+    from curies import Converter
     from rdflib import Graph, Node, URIRef
 
     from pyobo.struct import Obo
@@ -29,7 +30,7 @@ def write_skos(
     graph.serialize(path, format=format or "ttl")
 
 
-def _expand_rdflib(converter: Converter, reference: Reference) -> URIRef:
+def _expand_rdflib(converter: Converter, reference: curies.Reference) -> URIRef:
     import rdflib
 
     return rdflib.URIRef(converter.expand_reference(reference, strict=True))

src/pyobo/struct/skos/reader.py

@@ -6,14 +6,12 @@
 
 import curies
 import rdflib
-from bioregistry import NormalizedNamableReference, NormalizedNamedReference
-from curies import Reference
 from curies import vocabulary as v
 from rdflib import DCTERMS, RDF, RDFS, SKOS, VANN, Graph, Node, URIRef
 from tqdm import tqdm
 
 from pyobo import Annotation
-from pyobo.identifier_utils import get_converter
+from pyobo.identifier_utils import Reference, get_converter
 from pyobo.struct import Obo, Term, build_ontology
 from pyobo.struct.vocabulary import has_source
 
@@ -73,12 +71,10 @@ def _get_scheme_object_literal(p: Node) -> str | None:
         raise ValueError(f"no prefix given nor found using {VANN.preferredNamespacePrefix}")
 
     root_terms = [
-        NormalizedNamableReference.from_reference(
-            converter.parse_uri(str(subject), strict=True).to_pydantic()
-        )
+        Reference.from_reference(converter.parse_uri(str(subject), strict=True).to_pydantic())
         for subject in graph.objects(scheme, SKOS.hasTopConcept)
     ]
-    terms: dict[Reference, Term] = {}
+    terms: dict[curies.Reference, Term] = {}
     for concept in tqdm(
         graph.subjects(RDF.type, SKOS.Concept),
         desc=f"[{prefix}] SKOS concepts to OWL",
@@ -129,10 +125,10 @@ def _get_scheme_object_literal(p: Node) -> str | None:
     )
 
 
-def _cleanup_narrow_matches(terms: dict[Reference, Term]) -> None:
+def _cleanup_narrow_matches(terms: dict[curies.Reference, Term]) -> None:
     for parent in terms.values():
         for child in parent.get_property_objects(v.narrow_match):
-            if not isinstance(child, Reference):
+            if not isinstance(child, curies.Reference):
                 continue
             if child not in terms:
                 continue
@@ -174,7 +170,7 @@ def get_term(
     labels = _literal_objects(graph, node, SKOS.prefLabel, language_priority)
     definitions = _literal_objects(graph, node, SKOS.definition, language_priority)
     term = Term(
-        reference=NormalizedNamedReference(
+        reference=Reference(
             prefix=reference_tuple.prefix,
             identifier=reference_tuple.identifier,
             name=labels[0] if labels else None,