#!/usr/bin/env python3

import os
import re
import sqlite3
from datetime import datetime, timezone

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+)'
)
DATE_FMT = "%d/%b/%Y:%H:%M:%S %z"

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
)
"""
)
cursor.execute("CREATE INDEX IF NOT EXISTS idx_logs_time ON logs(time)")

conn.commit()

cursor.execute("SELECT time FROM logs ORDER BY id DESC LIMIT 1")
row = cursor.fetchone()
last_dt = None
if row and row[0]:
    # Support both legacy log date format and ISO timestamps
    for fmt in ("%Y-%m-%d %H:%M:%S", DATE_FMT):
        try:
            parsed = datetime.strptime(row[0], fmt)
            if fmt == DATE_FMT:
                parsed = parsed.astimezone(timezone.utc).replace(tzinfo=None)
            last_dt = parsed
            break
        except ValueError:
            continue

try:
    log_files = []
    for f in os.listdir(LOG_DIR):
        match = LOG_FILE_PATTERN.match(f)
        if match:
            suffix = match.group(1)
            number = int(suffix.lstrip(".")) if suffix else 0
            log_files.append((number, os.path.join(LOG_DIR, f)))
    log_files = [path for _, path in sorted(log_files, key=lambda x: x[0], reverse=True)]
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()
                try:
                    entry_dt = datetime.strptime(data["time"], DATE_FMT)
                except ValueError:
                    continue
                entry_dt = entry_dt.astimezone(timezone.utc).replace(tzinfo=None)
                if last_dt and entry_dt <= last_dt:
                    continue
                cursor.execute(
                    """
                    INSERT INTO logs (
                        ip, host, time, request, status, bytes_sent,
                        referer, user_agent, cache_status
                    ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
                    """,
                    (
                        data["ip"],
                        data["host"],
                        entry_dt.strftime("%Y-%m-%d %H:%M:%S"),
                        data["request"],
                        int(data["status"]),
                        int(data["bytes_sent"]),
                        data["referer"],
                        data["user_agent"],
                        data["cache_status"],
                    ),
                )
                last_dt = entry_dt
                inserted += 1

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