UniProt importer: use abbrev, not id field

src/bioregistry/external/uniprot/__init__.py

@@ -26,43 +26,55 @@
 RAW_PATH = RAW_DIRECTORY / "uniprot.json"
 PROCESSED_PATH = DIRECTORY / "processed.json"
 
+# the field to use as the prefix in UniProt dblist records
+PREFIX_FIELD = "abbrev"
+
 #: resources with these UniProt prefixes don't exist anymore
-skip_prefixes = {
+UNIPROT_SKIP_PREFIXES = {
     "UniPathway",  # doesn't exist anymore
     "BRENDA",  # has bad format string contains EC, UniProt, and taxon
     "eggNOG",  # not sure what this does
     "PlantReactome",  # incomprehensible URLs
     "Reactome",  # incomprehensible URLs
-    "DB-0180",  # genewiki
+    "GeneWiki",  # genewiki abbrev
 }
 
 HAS_BAD_URI = {
-    "DB-0148",  # ensembl.fungi
-    "DB-0147",  # ensembl.bacteria
+    "EnsemblFungi",  # ensembl.fungi
+    "EnsemblBacteria",  # ensembl.bacteria
 }
 
 
 def process_uniprot_raw(path: Path) -> dict[str, Record]:
     """Process UniProt raw JSON."""
     rv = {}
     for record in json.loads(path.read_text())["results"]:
-        prefix = record.pop("id")
-        if prefix in skip_prefixes:
-            continue
-        processed_record = _process_record(prefix, record)
-        if processed_record is None:
+        prefix = record.pop(PREFIX_FIELD)
+        if prefix in UNIPROT_SKIP_PREFIXES:
             continue
-        rv[prefix] = processed_record
+        if processed_record := _process_record(prefix, record):
+            rv[prefix] = processed_record
     return rv
 
 
 def _process_record(prefix: str, record: dict[str, Any]) -> Record | None:
     rv = {
         "name": record.pop("name"),
-        "abbreviation": record.pop("abbrev"),
         "homepage": record.pop("servers")[0],
         "keywords": [record.pop("category")],
     }
+
+    value = record.pop("dbUrl")
+    if "%s" in value and "%u" in value:
+        logger.debug("has both formats: %s", value)
+        return None
+
+    value = value.replace("%s", "$1").replace("%u", "$1")
+    if "$1" in value and prefix not in HAS_BAD_URI:
+        rv[URI_FORMAT_KEY] = value
+    else:
+        logger.debug("no annotation in %s", prefix)
+
     publication = {}
     if doi := record.pop("doiId", None):
         doi = doi.lower().rstrip(".")
@@ -75,19 +87,9 @@ def _process_record(prefix: str, record: dict[str, Any]) -> Record | None:
     if publication:
         rv["publications"] = [publication]
 
-    del record["linkType"]
-    del record["statistics"]
+    for key in ["id", "linkType", "statistics"]:
+        del record[key]
 
-    value = record.pop("dbUrl")
-    if "%s" in value and "%u" in value:
-        logger.debug(f"has both formats: {value}")
-        return None
-    else:
-        value = value.replace("%s", "$1").replace("%u", "$1")
-        if "$1" in value and prefix not in HAS_BAD_URI:
-            rv[URI_FORMAT_KEY] = value
-        else:
-            logger.debug("no annotation in %s", prefix)
     if record:
         logger.debug("forgot something: %s", record)
     return Record.model_validate(rv)
@@ -105,9 +107,8 @@ class UniProtAligner(Aligner):
     """Aligner for UniProt."""
 
     key = "uniprot"
-    alt_key_match = "abbreviation"
     getter = get_uniprot
-    curation_header: ClassVar[Sequence[str]] = ("abbreviation", "name", URI_FORMAT_KEY, "keywords")
+    curation_header: ClassVar[Sequence[str]] = ("name", URI_FORMAT_KEY, "keywords")
 
 
 if __name__ == "__main__":