Italian Driver License Recognizer Fix (#1651)

Author: K3y5tr0ke

CHANGELOG.md

@@ -3,6 +3,10 @@
 All notable changes to this project will be documented in this file.
 
 ## [unreleased]
+### Analyzer
+#### Changed
+- Implement Fix for Italian Duplicate Driver license
+    - Update Test cases to account for changed recognizer
 
 ## [2.2.359] - 2025-07-06
 ### Analyzer

presidio-analyzer/presidio_analyzer/predefined_recognizers/it_driver_license_recognizer.py

@@ -18,7 +18,7 @@ class ItDriverLicenseRecognizer(PatternRecognizer):
             "Driver License",
             (
                 r"\b(?i)(([A-Z]{2}\d{7}[A-Z])"
-                r"|(^[U]1[BCDEFGHLMNPRSTUWYXZ]\w{6}[A-Z]))\b"
+                r"|(^[U]1[BCDEFGHLJKMNPRSTUWYXZ0-9]{7}[A-Z]))\b"
             ),
             0.2,
         ),
@@ -31,7 +31,7 @@ def __init__(
         context: Optional[List[str]] = None,
         supported_language: str = "it",
         supported_entity: str = "IT_DRIVER_LICENSE",
-    ):
+    ) -> None:
         patterns = patterns if patterns else self.PATTERNS
         context = context if context else self.CONTEXT
         super().__init__(

presidio-analyzer/tests/test_it_driver_license_recognizer.py

@@ -5,12 +5,12 @@
 
 
 @pytest.fixture(scope="module")
-def recognizer():
+def recognizer() -> ItDriverLicenseRecognizer:
     return ItDriverLicenseRecognizer()
 
 
 @pytest.fixture(scope="module")
-def entities():
+def entities() -> list[str]:
     return ["IT_DRIVER_LICENSE"]
 
 
@@ -25,8 +25,10 @@ def entities():
         2,
         ((0, 10), (15, 25),),
         ((0.1, 0.4), (0.1, 0.4),),),
-        # Test with old Driver License
-        ("U1H00A000B", 1, ((0, 10),), ((0.1, 0.4),),),
+        # Test with valid duplicate Driver License
+        ("U1H00B000C", 1, ((0, 10),), ((0.1, 0.4),),),
+        # Test with invalid duplicate Driver License
+        ("U1H00A000B", 0, (), (),),
         # Test with invalid Driver License
         ("990123456B", 0, (), (),),
         # fmt: on
@@ -40,7 +42,7 @@ def test_when_driver_licenses_in_text_then_all_it_driver_licenses_found(
     recognizer,
     entities,
     max_score,
-):
+) -> None:
     results = recognizer.analyze(text, entities)
     assert len(results) == expected_len
     for res, (st_pos, fn_pos), (st_score, fn_score) in zip(