Normalization of ä, ö,ü and ß in filters and search of duplicates

app/main.py

@@ -4868,16 +4868,48 @@ def _software_job_filter_values(request: Request) -> dict[str, str]:
     }
 
 
+def _normalize_filter_text(value: Any) -> str:
+    """Normalize user-facing text for accent- and German umlaut-aware search."""
+    import unicodedata
+
+    text = str(value or "").translate(
+        str.maketrans(
+            {
+                "Ä": "Ae",
+                "Ö": "Oe",
+                "Ü": "Ue",
+                "ä": "ae",
+                "ö": "oe",
+                "ü": "ue",
+                "ẞ": "SS",
+                "ß": "ss",
+            }
+        )
+    )
+    return "".join(
+        character
+        for character in unicodedata.normalize("NFKD", text.casefold())
+        if not unicodedata.combining(character)
+    )
+
+
+def _sql_normalized_filter_text(column: Any):
+    """Return the SQL equivalent of :func:`_normalize_filter_text`."""
+    expression = func.lower(func.coalesce(cast(column, String), ""))
+    for source, target in (("ä", "ae"), ("ö", "oe"), ("ü", "ue"), ("ß", "ss")):
+        expression = func.replace(expression, source, target)
+    return expression
+
+
 def _sql_text_contains(column: Any, value: str):
-    """Case-insensitive literal substring search for SQL text expressions."""
+    """Literal substring search with umlaut and ASCII spelling equivalence."""
     escaped = (
-        str(value or "")
-        .casefold()
+        _normalize_filter_text(value)
         .replace("\\", "\\\\")
         .replace("%", "\\%")
         .replace("_", "\\_")
     )
-    return func.lower(func.coalesce(cast(column, String), "")).like(
+    return _sql_normalized_filter_text(column).like(
         f"%{escaped}%",
         escape="\\",
     )
@@ -4894,7 +4926,7 @@ def _translated_filter_codes(
     The translation dictionaries are cached on the request so one filter
     request does not open a database session for every possible status value.
     """
-    needle = str(search_value or "").strip().casefold()
+    needle = _normalize_filter_text(str(search_value or "").strip())
     if not needle:
         return []
 
@@ -4920,7 +4952,7 @@ def _translated_filter_codes(
     for code in codes:
         key = f"{prefix}.{code}"
         label = cache["own"].get(key) or cache["en"].get(key) or code
-        if needle in code.casefold() or needle in str(label).casefold():
+        if needle in _normalize_filter_text(code) or needle in _normalize_filter_text(label):
             matches.append(code)
     return matches