Upload pii_extension.py with huggingface_hub

pii_extension.py

@@ -37,14 +37,15 @@ class PIIDetector:
             re.compile(r'\b\d{7,10}\b'),  # Plain digits 7-10 chars
         ]
         
-        # Address patterns (basic street address detection)
+        # Address patterns (enhanced street address detection)
         self.address_patterns = [
-            re.compile(r'\b\d+\s+[A-Za-z]+\s+(?:Street|St|Avenue|Ave|Road|Rd|Boulevard|Blvd|Lane|Ln|Drive|Dr|Court|Ct|Way|Place|Pl)\b', re.IGNORECASE),
+            re.compile(r'\b\d+\s+\d*[A-Za-z]+(?:\s+[A-Za-z]+)?\s+(?:Street|St|Avenue|Ave|Road|Rd|Boulevard|Blvd|Lane|Ln|Drive|Dr|Court|Ct|Way|Place|Pl|Circle|Cir|Trail|Trl|Parkway|Pkwy)\b', re.IGNORECASE),
             re.compile(r'\b(?:PO|P\.O\.)\s*Box\s*\d+\b', re.IGNORECASE),
+            re.compile(r'\b\d+\s+[A-Za-z]+\s+(?:Street|St|Ave|Road|Rd)\b', re.IGNORECASE),
         ]
         
-        # Credit card (basic pattern - matches common formats)
-        self.cc_pattern = re.compile(r'\b(?:\d{4}[-\s]?){3}\d{4}\b')
+        # Credit card (enhanced pattern)
+        self.cc_pattern = re.compile(r'\b(?:\d{4}[-\s]?){3}\d{4}\b|\b\d{16}\b')
         
         # SSN (US Social Security Number)
         self.ssn_pattern = re.compile(r'\b\d{3}[-\s]?\d{2}[-\s]?\d{4}\b')
@@ -225,38 +226,41 @@ class PIIDetector:
         # Check grooming context for social media
         grooming_risk = 0.0
         grooming_keywords = []
-        social_media_allowed = True
         
-        if has_social_media:
+        # Check other PII first (blocked for all ages)
+        critical_pii = pii_types.intersection({PIILabel.EMAIL, PIILabel.PHONE, PIILabel.ADDRESS, PIILabel.CREDIT_CARD, PIILabel.SSN})
+        
+        if critical_pii:
+            action = "block"
+            reason = f"PII detected: {', '.join([p.value for p in critical_pii])}"
+        elif has_social_media:
+            # Social media rules
             is_grooming, grooming_risk, grooming_keywords = self.detect_grooming_context(text)
             
-            # Rules by age
             if age < 13:
                 # Under 13: Block ALL social media sharing
-                social_media_allowed = False
                 action = "block"
                 reason = "Social media sharing not permitted under 13"
+            elif is_grooming:
+                # 13+: Block if grooming detected
+                action = "block"
+                reason = f"Potential grooming detected (risk: {grooming_risk:.0%})"
             else:
-                # 13+: Allow but check for grooming
-                if is_grooming:
-                    social_media_allowed = False
-                    action = "block"
-                    reason = f"Potential grooming detected (risk: {grooming_risk:.0%})"
-                else:
-                    social_media_allowed = True
-                    action = "allow"
-                    reason = "Social media permitted for 13+ (no grooming signals)"
-        
-        # Check other PII (blocked for all ages)
-        critical_pii = pii_types.intersection({PIILabel.EMAIL, PIILabel.PHONE, PIILabel.ADDRESS, PIILabel.CREDIT_CARD, PIILabel.SSN})
-        
-        if critical_pii:
-            action = "block"
-            reason = f"PII detected: {', '.join([p.value for p in critical_pii])}"
-        elif not has_social_media and not pii_types:
+                # 13+: Allow social media, no grooming
+                action = "allow"
+                reason = "Social media permitted for 13+ (no grooming signals)"
+        else:
             action = "allow"
             reason = "No PII detected"
         
+        # Determine if social media is allowed for return value
+        social_media_allowed = True
+        if has_social_media:
+            if age < 13:
+                social_media_allowed = False
+            elif grooming_risk > 0:
+                social_media_allowed = False
+        
         return {
             "has_pii": len(pii_types) > 0,
             "pii_types": [p.value for p in pii_types],