#!/usr/bin/env python3

import os
import sqlite3
import re
from pathlib import Path

LOG_DIR = "/var/log/nginx"
DB_FILE = "database/ngxstat.db"
LOG_FILE_PATTERN = re.compile(r'access\.log(\.\d+)?$')
LOG_FORMAT_REGEX = re.compile(
    r'(?P<ip>\S+) - (?P<host>\S+) \[(?P<time>.*?)\] "(?P<request>.*?)" '
    r'(?P<status>\d{3}) (?P<bytes_sent>\d+) "(?P<referer>.*?)" "(?P<user_agent>.*?)" (?P<cache_status>\S+)'
)

os.makedirs("database", exist_ok=True)
conn = sqlite3.connect(DB_FILE)
cursor = conn.cursor()

cursor.execute('''
CREATE TABLE IF NOT EXISTS logs (
    id INTEGER PRIMARY KEY,
    ip TEXT,
    host TEXT,
    time TEXT,
    request TEXT,
    status INTEGER,
    bytes_sent INTEGER,
    referer TEXT,
    user_agent TEXT,
    cache_status TEXT
)
''')

conn.commit()

try:
    log_files = sorted([
        os.path.join(LOG_DIR, f)
        for f in os.listdir(LOG_DIR)
        if LOG_FILE_PATTERN.match(f)
    ])
except FileNotFoundError:
    print(f"[ERROR] Log directory not found: {LOG_DIR}")
    exit(1)

print(f"[INFO] Found {len(log_files)} log files.")

inserted = 0
for log_file in log_files:
    print(f"[INFO] Parsing {log_file}...")
    with open(log_file, 'r', encoding='utf-8', errors='ignore') as f:
        for line in f:
            match = LOG_FORMAT_REGEX.match(line.strip())
            if match:
                data = match.groupdict()
                cursor.execute('''
                    INSERT INTO logs (
                        ip, host, time, request, status, bytes_sent,
                        referer, user_agent, cache_status
                    ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
                ''', (
                    data["ip"], data["host"], data["time"], data["request"],
                    int(data["status"]), int(data["bytes_sent"]),
                    data["referer"], data["user_agent"], data["cache_status"]
                ))
                inserted += 1

conn.commit()
conn.close()
print(f"[DONE] Inserted {inserted} entries into {DB_FILE}.")