src/services/regex_pii_remover.py

import re
import logging

logger = logging.getLogger(__name__)


class RegexPIIRemover:
    """Remove PII using regex patterns"""
    
    def __init__(self):
        """Initialize PII removal patterns"""
        self.patterns = {
            # Social Security Numbers
            'ssn': re.compile(r'\b\d{3}-\d{2}-\d{4}\b'),
            
            # Phone numbers
            'phone': re.compile(r'\b\d{3}[-.]?\d{3}[-.]?\d{4}\b'),
            
            # Email addresses
            'email': re.compile(r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b'),
            
            # Dates (MM/DD/YYYY, MM-DD-YYYY, etc.)
            'date': re.compile(r'\b\d{1,2}[/-]\d{1,2}[/-]\d{2,4}\b'),
            
            # Medical Record Numbers (MRN)
            'mrn': re.compile(r'\b(MRN|Medical Record Number)[:\s]+\w+\b', re.IGNORECASE),
            
            # ZIP codes
            'zip': re.compile(r'\b\d{5}(-\d{4})?\b'),
            
            # Names (simple pattern - captures "Patient: John Doe" or "Name: Jane Smith")
            'patient_name': re.compile(r'(Patient|Name)[:\s]+([A-Z][a-z]+\s[A-Z][a-z]+)', re.IGNORECASE),
            
            # Date of Birth
            'dob': re.compile(r'(DOB|Date of Birth)[:\s]+\d{1,2}[/-]\d{1,2}[/-]\d{2,4}', re.IGNORECASE),
        }
        
        logger.info(f"RegexPIIRemover initialized with {len(self.patterns)} patterns")
    
    def remove_pii(self, text: str) -> tuple[str, int]:
        """
        Remove PII from text
        
        Args:
            text: Input text
            
        Returns:
            tuple: (cleaned_text, count_of_pii_removed)
        """
        cleaned_text = text
        total_removed = 0
        
        for pii_type, pattern in self.patterns.items():
            matches = pattern.findall(cleaned_text)
            count = len(matches)
            
            if count > 0:
                logger.debug(f"Found {count} instances of {pii_type}")
                total_removed += count
                
                # Replace with redacted placeholder
                if pii_type == 'patient_name':
                    cleaned_text = pattern.sub(r'\1: [REDACTED]', cleaned_text)
                elif pii_type in ['dob', 'mrn']:
                    cleaned_text = pattern.sub(r'\1: [REDACTED]', cleaned_text)
                else:
                    cleaned_text = pattern.sub('[REDACTED]', cleaned_text)
        
        logger.info(f"Removed {total_removed} PII entities")
        
        return cleaned_text, total_removed