import re

# This script extracts ip, mac, and hostname from a dhcp server's log file to a csv file.
# This works only if the log files match the format in the example log file in this directory

def extract_info_from_logs(log_file_list, output_file_path):
    # Regular expression to match IP address, MAC address, and hostname
    regex = re.compile(
        r'\b(?:DHCPACK|DHCPREQUEST|DHCPOFFER)\b.*?on\s+([\d.]+)\s+to\s+([\da-f:]+)(?:\s+\((.*?)\))?'
    )
    
    # A list of (ip, mac, host) data extracted from matched lines
    extracted_data = []

    for logfile in log_file_list:
        with open(logfile, 'r') as logfile:
            for line in logfile:
                match = regex.search(line)
                if match:
                    ip = match.group(1)
                    mac = match.group(2)
                    host = match.group(3) if match.group(3) else "Unknown"
                    data = ip + "," + mac + "," + host  # Customizable as per requirements
                    if data not in extracted_data:      # Ensure uniqueness
                        # Add other conditions here as per requirements
                        extracted_data.append(data)     # Add extratcted data to the list

    # Write the extracted data to the output file
    with open(output_file_path, 'w') as output_file:
        output_file.write("IP,MAC,HOST\n")  # Header row
        output_file.write('\n'.join(extracted_data))    # Data, as row+'\n'

    print(f"Extraction complete. Data written to {output_file_path}.")

logfilelist = ['/var/log/syslog', '/var/log/syslog.1', '/var/log/dhcp.log']      # Specify the list of /path/log_file
outputfile = 'output.csv'   # Specify the output file /path/name.csv

extract_info_from_logs(logfilelist, outputfile)