use unidecode to convert unicode to ascii

requirements.txt

@@ -2,3 +2,4 @@ requests~=2.32.3
 python-decouple~=3.8
 websockets~=13.0
 xmltodict~=0.13
+unidecode~=1.3

x2gbfs/providers/cambio.py

@@ -3,7 +3,7 @@
 from typing import Any
 
 from x2gbfs.gbfs.base_provider import BaseProvider
-from x2gbfs.util import get
+from x2gbfs.util import get, unidecode_with_german_umlauts
 
 logger = logging.getLogger(__name__)
 
@@ -123,14 +123,6 @@ def _extract_vehicle_types_available(self, elem: dict[str, Any]) -> list[dict[st
         vehicle_classes_at_station = elem.get('vehicleClasses', [])
         return [{'vehicle_type_id': vehicle_class['id'], 'count': 1} for vehicle_class in vehicle_classes_at_station]
 
-    def _replace_umlauts(self, string: str) -> str:
-        substitions = (('ü', 'ue'), ('ä', 'ae'), ('ö', 'oe'), ('é', 'e'), ('ß', 'ss'))
-
-        for orig, subst in substitions:
-            string = string.replace(orig, subst)
-
-        return string
-
     def _extract_propulsion_type(self, elem: dict[str, str]) -> str:
         """
         Guesses the propulsion type from vehicle name.
@@ -148,7 +140,7 @@ def _extract_station_rental_uris(self, elem: dict[str, str]) -> dict[str, str]:
         """
         Guesses the propulsion type from vehicle name.
         """
-        station_name = self._replace_umlauts(elem['name'].lower())
+        station_name = unidecode_with_german_umlauts(elem['name'].lower())
         station_url = f"https://www.cambio-carsharing.de/stationen/station/{station_name}-{elem['id']}"
 
         return {

x2gbfs/util.py

@@ -2,6 +2,18 @@
 from typing import Any, Dict, Generator, Optional, Tuple
 
 import requests
+from unidecode import unidecode
+
+GERMAN_UMLAUTS_TRANSLATIONS = str.maketrans({'ä': 'ae', 'Ä': 'Ae', 'ö': 'oe', 'Ö': 'Oe', 'ü': 'ue', 'Ü': 'Ue'})
+
+
+def unidecode_with_german_umlauts(string: str) -> str:
+    """
+    Represents non-ascii Unicode string as their closest matching ascii variants.
+    As unidecode transliters German umlauts incorrectly (see the FAQ section of https://pypi.org/project/Unidecode/),
+    unidecode_with_german_umlauts first replaces them explicitly.
+    """
+    return unidecode(string.translate(GERMAN_UMLAUTS_TRANSLATIONS))
 
 
 def timestamp_to_isoformat(utctimestamp: datetime):