ENH+BF: Format species names consistently in species_map, make matching more stringent for common names, improved testing

dandi/metadata/util.py

@@ -332,7 +332,7 @@ def extract_cellLine(metadata: dict) -> str | None:
 
 NCBITAXON_URI_TEMPLATE = "http://purl.obolibrary.org/obo/NCBITaxon_{}"
 
-# common_names, prefix, uri, name
+# common_names, prefix, uri, name ({current name} - {GenBank common name})
 species_map = [
     (
         ["mouse"],
@@ -386,25 +386,31 @@ def extract_cellLine(metadata: dict) -> str | None:
         ["c. elegans", "caenorhabditis elegans"],
         "caenorhabditis",
         NCBITAXON_URI_TEMPLATE.format("6239"),
-        "Caenorhabditis elegans",
+        "Caenorhabditis elegans - Roundworm",
     ),
     (
         ["pig-tailed macaque", "pigtail monkey", "pigtail macaque"],
         None,
         NCBITAXON_URI_TEMPLATE.format("9545"),
-        "Macaca nemestrina",
+        "Macaca nemestrina - Pig-tailed macaque",
+    ),
+    (
+        ["bonnet macaque", "bonnet monkey", "radiata"],
+        None,
+        NCBITAXON_URI_TEMPLATE.format("9548"),
+        "Macaca radiata - Bonnet macaque",
     ),
     (
         ["mongolian gerbil", "mongolian jird"],
         None,
         NCBITAXON_URI_TEMPLATE.format("10047"),
-        "Meriones unguiculatus",
+        "Meriones unguiculatus - Mongolian gerbil",
     ),
     (
         ["common paper wasp"],
         None,
         NCBITAXON_URI_TEMPLATE.format("30207"),
-        "Polistes fuscatus",
+        "Polistes fuscatus - Common paper wasp",
     ),
 ]
 
@@ -484,14 +490,24 @@ def extract_species(metadata: dict) -> models.SpeciesType | None:
         else:
             lower_value = value_orig.lower()
             for common_names, prefix, uri, name in species_map:
+                scientific_name, _, common_name = name.partition(" - ")
                 if (
                     lower_value == name.lower()
-                    or any(key in lower_value for key in common_names)
-                    or (prefix is not None and lower_value.startswith(prefix))
+                    or lower_value == scientific_name.lower()
+                    or (common_name and lower_value == common_name.lower())
                 ):
                     value_id = uri
                     value = name
                     break
+            else:
+                for common_names, prefix, uri, name in species_map:
+                    if (
+                        any(key in lower_value for key in common_names)
+                        or (prefix is not None and lower_value.startswith(prefix))
+                    ):
+                        value_id = uri
+                        value = name
+                        break
         if value_id is None:
             raise ValueError(
                 f"Cannot interpret species field: {value_orig}. Please "

dandi/tests/test_metadata.py

@@ -770,7 +770,7 @@ def test_species_all_possible(species: str) -> None:
     assert species_rec.model_dump(mode="json", exclude_none=True) == {
         "identifier": "http://purl.obolibrary.org/obo/NCBITaxon_10047",
         "schemaKey": "SpeciesType",
-        "name": "Meriones unguiculatus",
+        "name": "Meriones unguiculatus - Mongolian gerbil",
     }
 
 
@@ -782,9 +782,15 @@ def test_extract_unknown_species():
 
 def test_species_map():
     # all alternative names should be lower case
-    for common_names, *_ in species_map:
+    for common_names, _, uri, name in species_map:
         for key in common_names:
             assert key.lower() == key
+        assert " - " in name
+        for species in name.split(" - "):
+            species_rec = extract_species({"species": species})
+            assert species_rec
+            assert str(species_rec.identifier) == uri
+            assert species_rec.name == name
 
 
 @pytest.mark.parametrize(