fix(analyzer): normalize case before fi hetu and sg uen checksums

presidio-analyzer/presidio_analyzer/predefined_recognizers/country_specific/finland/fi_personal_identity_code_recognizer.py

@@ -81,7 +81,10 @@ def validate_result(self, pattern_text: str) -> Optional[bool]:
         except ValueError:
             return False
         individual_number = pattern_text[7:10]
-        control_character = pattern_text[-1]
+        # The control character is defined in upper case; the pattern is matched
+        # case-insensitively, so upper-case it before comparing or a valid code
+        # written with a lower-case control character would be rejected.
+        control_character = pattern_text[-1].upper()
         valid_control_characters = "0123456789ABCDEFHJKLMNPRSTUVWXY"
         number_to_check = int(date_part + individual_number)
         return valid_control_characters[number_to_check % 31] == control_character

presidio-analyzer/presidio_analyzer/predefined_recognizers/country_specific/singapore/sg_uen_recognizer.py

@@ -105,6 +105,10 @@ def validate_result(self, pattern_text: str) -> Optional[bool]:
         Only the part in text that was detected by the regex engine
         :return: A bool indicating whether the validation was successful.
         """
+        # The prefix, entity-type and check letters are defined in upper case;
+        # the pattern is matched case-insensitively, so normalize before the
+        # checksum or a valid UEN written in lower case would be rejected.
+        pattern_text = pattern_text.upper()
 
         if len(pattern_text) == 9:
             # Checksum validation for UEN format A

presidio-analyzer/tests/test_fi_personal_identity_code_recognizer.py

@@ -147,6 +147,11 @@ def entities():
         ),
         # Valid HETU with the canonical "-" century separator.
         ("131052-308T", 1, ((0, 11),)),
+        # Regression: a valid code whose control character is written in lower
+        # case must still be detected. The pattern matches case-insensitively,
+        # so the control character is upper-cased before the checksum.
+        ("131052-308t", 1, ((0, 11),)),
+        ("020504a902e", 1, ((0, 11),)),
         # invalid Personal Identity Codes scores
         ("111111-111A", 0, ()),
         ("111111+110G", 0, ()),

presidio-analyzer/tests/test_sg_uen_recognizer.py

@@ -33,6 +33,10 @@ def entities():
         ("53125226D 201434292D S57TU0392K", 3, [(0, 9), (10, 20), (21, 31)],),
         # Test with valid UEN in a sentence
         ("UEN 53125226D was processed", 1, [(4, 13)],),
+        # Lower-case UENs must still validate; the prefix, entity-type and
+        # check letters are upper-cased before the checksum. Format A and C:
+        ("53125226d", 1, [(0, 9)],),
+        ("t16rf0037c", 1, [(0, 10)],),
 
         ## No match
         # Test with invalid length