#!/usr/bin/env python3
import re
import ipaddress
from datetime import datetime, timezone
import os
from local import *  # Make sure this defines ACCESS_LOG, LOG_DIR, etc.

# ==== CONFIGURATION ====
MAX_LOG_LINES = 50000  # adjust as needed
LOG_DIR = os.path.join(os.path.dirname(__file__), "logs")
os.makedirs(LOG_DIR, exist_ok=True)

INTERNAL_NETWORKS = [
    ipaddress.ip_network("10.0.0.0/8"),
    ipaddress.ip_network("192.168.0.0/16"),
    ipaddress.ip_network("172.16.0.0/12"),
]

log_line_re = re.compile(
    r'(?P<ip>\S+) - - \[(?P<time>[^\]]+)\] "(?P<method>\S+) (?P<path>\S+) \S+"'
)

def is_external(ip):
    ip_addr = ipaddress.ip_address(ip)
    return not any(ip_addr in net for net in INTERNAL_NETWORKS)

def parse_nginx_line(line):
    match = log_line_re.match(line)
    if not match:
        return None
    data = match.groupdict()
    if not is_external(data["ip"]):
        return None
    dt = datetime.strptime(data["time"], "%d/%b/%Y:%H:%M:%S %z")
    dt_utc = dt.astimezone(timezone.utc)
    iso_time = dt_utc.strftime("%Y-%m-%dT%H:%M:%SZ")
    return f'{iso_time} {data["ip"]} {data["method"]} {data["path"]}'

def get_current_logfile():
    """Find or create the latest log file with line limit."""
    today = datetime.now().strftime("%Y-%m-%d")
    base_name = os.path.join(LOG_DIR, f"filtered_{today}")
    
    index = 1
    while True:
        log_file = f"{base_name}_{index}.log"
        if not os.path.exists(log_file):
            return log_file
        # Check line count
        with open(log_file, "r") as f:
            line_count = sum(1 for _ in f)
        if line_count < MAX_LOG_LINES:
            return log_file
        index += 1

def process_log():
    output_file = get_current_logfile()
    buffer = []

    with open(ACCESS_LOG, "r") as f:
        for line in f:
            parsed = parse_nginx_line(line)
            if parsed:
                buffer.append(parsed)
    if buffer:
        with open(output_file, "a") as out:
            out.write("\n".join(buffer) + "\n")
def flush_access_log():
    """Safely truncate the access log after processing."""
    with open(ACCESS_LOG, "w"):
        pass  # Opening with 'w' truncates file

def main():
    process_log()
    #flush_access_log()

if __name__ == "__main__":
    main()