Find Sensitive Information in Log Files with Python

import re
import os
from pathlib import Path

def find_sensitive_info(log_path):
    """Scans log files for patterns like emails, IPs, API keys, and passwords."""
    patterns = {
        'Email': r'[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}',
        'IP Address': r'\b(?:\d{1,3}\.){3}\d{1,3}\b',
        'API Key': r'(?i)(?:api[_-]?key|apikey)\s*[:=]\s*[\'"]?[a-zA-Z0-9]{16,}[\'"]?',
        'Password': r'(?i)(?:password|passwd|pwd)\s*[:=]\s*[\'"]?[^\s\'"]+[\'"]?'
    }
    findings = []
    log_file = Path(log_path)
    
    if not log_file.exists():
        print(f"File {log_path} not found.")
        return
    
    with open(log_file, 'r', encoding='utf-8', errors='ignore') as f:
        for line_num, line in enumerate(f, 1):
            for label, pattern in patterns.items():
                matches = re.findall(pattern, line)
                for match in matches:
                    findings.append((line_num, label, match))
    
    if findings:
        print(f"Sensitive information found in {log_path}:")
        for line_num, label, value in findings:
            print(f"  Line {line_num}: [{label}] {value}")
    else:
        print(f"No sensitive information found in {log_path}.")
    
    return findings

if __name__ == "__main__":
    # Example usage with a sample log file
    sample_log = "sample.log"
    with open(sample_log, 'w') as f:
        f.write("User email: john.doe@example.com\n")
        f.write("Server IP: 192.168.1.1\n")
        f.write("API key = sk-abc123def456ghi789\n")
        f.write("password: super_secret_123\n")
        f.write("Normal log entry at 2023-10-05\n")
    find_sensitive_info(sample_log)
    os.remove(sample_log)  # Clean up test file