Fix: strip diacritics in step 3 CSV filter

scripts/filter_csv.py

@@ -13,6 +13,7 @@
 import csv
 import logging
 import sys
+import unicodedata
 from pathlib import Path
 
 logging.basicConfig(
@@ -34,8 +35,13 @@
 
 
 def normalize_artist(name: str) -> str:
-    """Normalize artist name for matching."""
-    return name.lower().strip()
+    """Normalize artist name for matching.
+
+    Strips diacritics so that Discogs "Björk" matches library "Bjork".
+    """
+    nfkd = unicodedata.normalize("NFKD", name)
+    stripped = "".join(c for c in nfkd if not unicodedata.combining(c))
+    return stripped.lower().strip()
 
 
 def load_library_artists(path: Path) -> set[str]:

tests/unit/test_filter_csv.py

@@ -40,8 +40,26 @@ class TestNormalizeArtist:
             ("RADIOHEAD", "radiohead"),
             ("  Mixed Case  ", "mixed case"),
             ("", ""),
+            ("Björk", "bjork"),
+            ("Sigur Rós", "sigur ros"),
+            ("Motörhead", "motorhead"),
+            ("Hüsker Dü", "husker du"),
+            ("Café Tacvba", "cafe tacvba"),
+            ("Zoé", "zoe"),
+        ],
+        ids=[
+            "lowercase",
+            "strip-spaces",
+            "all-caps",
+            "mixed-case-strip",
+            "empty",
+            "bjork",
+            "sigur-ros",
+            "motorhead",
+            "husker-du",
+            "cafe-tacvba",
+            "zoe",
         ],
-        ids=["lowercase", "strip-spaces", "all-caps", "mixed-case-strip", "empty"],
     )
     def test_normalize(self, raw: str, expected: str) -> None:
         assert normalize_artist(raw) == expected