Normalize instances of UTF to Unicode

Signed-off-by: Jono Yang <[email protected]>
aboutcode-org · May 8, 2024 · ef2403b · ef2403b
1 parent 3b2e2f7
commit ef2403b
Showing 1 changed file with 2 additions and 0 deletions.
diff --git a/src/scancode/cli_test_utils.py b/src/scancode/cli_test_utils.py
@@ -335,6 +335,8 @@ def streamline_scanned_file(scanned_file, remove_file_date=False, normalize_file
     if normalize_file_type and (file_type := scanned_file.get('file_type')):
         file_type_segments = file_type.split()
         normalized_file_type = file_type_segments[0]
+        if 'UTF' in normalized_file_type:
+            normalized_file_type = 'Unicode'
         scanned_file['file_type'] = normalized_file_type