Update US_SSN CONTEXT and unit test (#1455)

microsoft · Sep 29, 2024 · 89ccadb · 89ccadb
1 parent c54ce2b
commit 89ccadb
Show file tree

Hide file tree

Showing 3 changed files with 14 additions and 8 deletions.
diff --git a/presidio-analyzer/presidio_analyzer/predefined_recognizers/us_ssn_recognizer.py b/presidio-analyzer/presidio_analyzer/predefined_recognizers/us_ssn_recognizer.py
@@ -27,8 +27,8 @@ class UsSsnRecognizer(PatternRecognizer):
         # "sec", # Task #603: Support keyphrases ("social sec")
         "ssn",
         "ssns",
-        "ssn#",
-        "ss#",
+        # "ssn#",  # iss:1452 - a # does not work with LemmaContextAwareEnhancer
+        # "ss#",  # iss:1452 - a # does not work with LemmaContextAwareEnhancer
         "ssid",
     ]
 

diff --git a/presidio-analyzer/tests/data/context_sentences_tests.txt b/presidio-analyzer/tests/data/context_sentences_tests.txt
@@ -8,16 +8,22 @@ IP_ADDRESS
 my ip: 192.168.0.1
 
 US_SSN
-my ssn is 078-051120 07805-1120
+my ssn is 078-051121
 
 US_SSN
 my social security number is 078051120
 
 US_SSN
-my social security number is 078-05-1120
+my social security number is 078-05-1121
 
 US_SSN
-my social security number is 078051120
+my social security number is 078051121
+
+US_SSN
+my ssns is 078-05-1121
+
+US_SSN
+my ssid is 078-05-1121
 
 PHONE_NUMBER
 my phone number is (425) 882-9090

diff --git a/presidio-analyzer/tests/test_context_support.py b/presidio-analyzer/tests/test_context_support.py
@@ -70,9 +70,9 @@ def dataset(recognizers_map):
             raise ValueError(f"bad entity type {entity_type}")
 
         test_items.append((item, recognizer, [entity_type]))
-    # Currently we have 31 sentences, this is a sanity check
-    if not len(test_items) == 32:
-        raise ValueError(f"expected 31 context sentences but found {len(test_items)}")
+    # Currently we have 34 sentences, this is a sanity check
+    if not len(test_items) == 34:
+        raise ValueError(f"expected 34 context sentences but found {len(test_items)}")
 
     yield test_items
-Original file line number
+Diff line change
@@ Expand Up / @@ -27,8 +27,8 @@ class UsSsnRecognizer(PatternRecognizer): @@
             # "sec", # Task #603: Support keyphrases ("social sec")
             "ssn",
             "ssns",
-            "ssn#",
-            "ss#",
+            # "ssn#",  # iss:1452 - a # does not work with LemmaContextAwareEnhancer
+            # "ss#",  # iss:1452 - a # does not work with LemmaContextAwareEnhancer
             "ssid",
         ]
@@ Expand Down @@