Handle special letters in by_name() ignore_accents

remove_accents() only stripped accents that Unicode can split off from their base letter (NFD normalization), so letters with no such decomposition (ß, œ, æ, ø, ł, đ, etc.) passed through unchanged.
Add an explicit translation map, applied after the NFD strip so composed forms (e.g. "ǿ") are also handled correctly. Updated test.

Author: gefleury

pipeline/src/additional_methods/by_name.py.txt

@@ -26,7 +26,9 @@
             all (bool, optional): Whether to return all objects that match the name, or only the first. Defaults to False.
             case_sensitive (bool, optional): Whether the search should be case-sensitive. Defaults to True.
             ignore_accents (bool, optional): Whether to ignore accents (acute, grave, circumflex) and
-                other diacritical marks (cedilla, tilde, ring, etc.) when matching. Defaults to False.
+                other diacritical marks (cedilla, tilde, ring, etc.) when matching. Also treat
+                special letters (ß, œ, æ, ø, ł, etc.) as their closest plain-letter equivalents
+                (e.g. "ß" as "ss"). Defaults to False.
         """
         namelike_properties = ("name", "lookup_label", "family_name", "full_name", "short_name", "abbreviation")
         if cls._instance_lookup is None:
@@ -49,8 +51,21 @@
         def remove_accents(s):
             import unicodedata
 
+            special = str.maketrans({
+                "Ł": "L", "ł": "l",
+                "Ø": "O", "ø": "o",
+                "Đ": "D", "đ": "d",
+                "Ð": "D", "ð": "d",
+                "Þ": "Th", "þ": "th",
+                "Æ": "AE", "æ": "ae",
+                "Œ": "OE", "œ": "oe",
+                "ß": "ss", "ẞ": "SS",
+                "Ə": "E", "ə": "e",
+                "ı": "i",
+            })
             nfd_form = unicodedata.normalize("NFD", s)
-            return "".join(c for c in nfd_form if not unicodedata.combining(c))
+            stripped = "".join(c for c in nfd_form if not unicodedata.combining(c))
+            return stripped.translate(special)
 
         def normalize(s):
             if not case_sensitive:

pipeline/tests/test_regressions.py

@@ -749,3 +749,17 @@ def test_pr0103_by_name_ignore_accents(om):
     for query, case_sensitive, ignore_accents, should_match in cases:
         match = SovereignState.by_name(query, case_sensitive=case_sensitive, ignore_accents=ignore_accents)
         assert (match is not None and match.name == "France") == should_match
+
+    # ignore_accents also has to map special letters
+    special_letter_cases = [
+        # (query, ignore_accents, expected_country_or_None)
+        ("Azərbaycan Respublikası", False, "Azerbaijan"),     # exact
+        ("Azerbaycan Respublikasi", True, "Azerbaijan"),
+        ("Azerbaycan Respublikasi", False, None),
+        ("Wááshindoon Bikéyah Ałhidadiidzooígíí", False, "United States"),  # exact
+        ("Waashindoon Bikeyah Alhidadiidzooigii", True, "United States"),
+        ("Waashindoon Bikeyah Alhidadiidzooigii", False, None),
+    ]
+    for query, ignore_accents, expected_name in special_letter_cases:
+        match = SovereignState.by_name(query, ignore_accents=ignore_accents)
+        assert (match.name if match else None) == expected_name