fix: NA county_code is valid, not Nan

refs: symerio#73

pgeocode.py

@@ -141,6 +141,26 @@
     "ZA",
 ]
 
+NA_VALUES = [
+    "", 
+    "#N/A", 
+    "#N/A N/A", 
+    "#NA", 
+    "-1.#IND", 
+    "-1.#QNAN", 
+    "-NaN", 
+    "-nan", 
+    "1.#IND", 
+    "1.#QNAN", 
+    "<NA>", 
+    "N/A", 
+    # NA", 
+    "NULL", 
+    "NaN", 
+    "n/a", 
+    "nan", 
+    "null",
+]
 
 @contextlib.contextmanager
 def _open_extract_url(url: str, country: str) -> Any:
@@ -231,7 +251,7 @@ def _get_data(country: str) -> Tuple[str, pd.DataFrame]:
 
         data_path = os.path.join(STORAGE_DIR, country.upper() + ".txt")
         if os.path.exists(data_path):
-            data = pd.read_csv(data_path, dtype={"postal_code": str})
+            data = pd.read_csv(data_path, dtype={"postal_code": str}, na_values=NA_VALUES)
         else:
             download_urls = [
                 val.format(country=country) for val in DOWNLOAD_URL
@@ -246,7 +266,6 @@ def _get_data(country: str) -> Tuple[str, pd.DataFrame]:
                 )
             os.makedirs(STORAGE_DIR, exist_ok=True)
             data.to_csv(data_path, index=None)
-
         return data_path, data
 
     def _index_postal_codes(self) -> pd.DataFrame:

test_pgeocode.py

@@ -277,6 +277,12 @@ def test_query_location_exact():
     assert isinstance(res, pd.DataFrame)
     assert res["state_name"].unique().tolist() == ["Île-de-France"]
 
+def test_location_naples():
+    # https://github.com/symerio/pgeocode/issues/73
+    nomi = Nominatim("it")
+    res = nomi.query_location("Napoli")
+    assert res["county_name"].unique().tolist() == ["Napoli"]
+    assert res["county_code"].unique().tolist() == ["NA"]
 
 def test_query_location_fuzzy():
     pytest.importorskip("thefuzz")