#!/usr/bin/env python3 import os import sys import json import re try: from jinja2 import Environment, FileSystemLoader except ImportError: print("FATAL: jinja2 is not installed. Please run 'pip install jinja2'.", file=sys.stderr) exit(1) import logging import ipaddress from typing import Optional # Configure logging logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s') def is_ip_address(address: str) -> bool: """Checks if a given string is a valid IP address (IPv4 or IPv6).""" if not address: return False try: ipaddress.ip_address(address) return True except ValueError: return False def load_dotenv(dotenv_path): """ Loads environment variables from a .env file. Does not override existing environment variables from the system. """ if not os.path.exists(dotenv_path): logging.warning(f".env file not found at {dotenv_path}. Using system environment variables or defaults.") return try: with open(dotenv_path) as f: for line in f: line = line.strip() if line and not line.startswith('#') and '=' in line: key, value = line.split('=', 1) key = key.strip() value = value.strip() # Remove surrounding quotes which are common in .env files if (value.startswith('"') and value.endswith('"')) or \ (value.startswith("'") and value.endswith("'")): value = value[1:-1] # os.environ only takes strings value = str(value) if key not in os.environ: os.environ[key] = value logging.info(f"Successfully loaded variables from {dotenv_path}") except Exception as e: logging.error(f"Failed to read or parse {dotenv_path}: {e}") # Continue, will use defaults or system env vars def _get_port_from_proxy_url(url: str) -> Optional[str]: """Extracts the port from a proxy URL string.""" if not url or not isinstance(url, str): return None match = re.search(r':(\d+)$', url.strip()) return match.group(1) if match else None def expand_env_vars(value: str) -> str: """ Expands environment variables in a string, including default values. Supports ${VAR} and ${VAR:-default}. """ if not isinstance(value, str): return value # Regex to find ${VAR:-default} or ${VAR} pattern = re.compile(r'\$\{(?P\w+)(?::-(?P.*?))?\}') def replacer(match): var_name = match.group('var') default_value = match.group('default') # Get value from os.environ, or use default, or empty string return os.getenv(var_name, default_value if default_value is not None else '') return pattern.sub(replacer, value) def generate_configs(): """ Generates envoy.yaml, docker-compose.camoufox.yaml, and camoufox_endpoints.json from Jinja2 templates and environment variables. """ try: # --- Load .env file --- script_dir = os.path.dirname(os.path.abspath(__file__)) dotenv_path = os.path.join(script_dir, '.env') load_dotenv(dotenv_path) # --- Common Configuration --- ytdlp_workers_str = os.getenv('YTDLP_WORKERS', '3').strip() try: # Handle empty string case by defaulting to 3, otherwise convert to int. worker_count = int(ytdlp_workers_str) if ytdlp_workers_str else 3 except (ValueError, TypeError): logging.warning(f"Invalid value for YTDLP_WORKERS: '{ytdlp_workers_str}'. Defaulting to 3.") worker_count = 3 if worker_count == 0: worker_count = os.cpu_count() or 1 logging.info(f"YTDLP_WORKERS is 0, auto-detected {worker_count} CPU cores for worker and camoufox config.") config_dir = os.path.join(script_dir, 'config') os.makedirs(config_dir, exist_ok=True) env = Environment(loader=FileSystemLoader(script_dir), trim_blocks=True, lstrip_blocks=True) # Get service role from environment to determine what to generate service_role = os.getenv('SERVICE_ROLE', 'all-in-one') logging.info(f"Service role for generation: '{service_role}'") # --- Camoufox Configuration (only for worker/all-in-one roles) --- camoufox_proxies = [] expanded_camoufox_proxies_str = "" if service_role != 'management': logging.info("--- Generating Camoufox (Remote Browser) Configuration ---") camoufox_proxies_str = os.getenv('CAMOUFOX_PROXIES') if not camoufox_proxies_str: logging.warning("CAMOUFOX_PROXIES environment variable not set. No camoufox instances will be generated.") else: # Expand environment variables within the string before splitting expanded_camoufox_proxies_str = expand_env_vars(camoufox_proxies_str) logging.info(f"Expanded CAMOUFOX_PROXIES from '{camoufox_proxies_str}' to '{expanded_camoufox_proxies_str}'") camoufox_proxies = [{'url': p.strip()} for p in expanded_camoufox_proxies_str.split(',') if p.strip()] logging.info(f"Found {len(camoufox_proxies)} proxy/proxies for Camoufox.") logging.info(f"Each Camoufox instance will support {worker_count} concurrent browser sessions.") logging.info(f"Total browser sessions supported on this worker: {len(camoufox_proxies) * worker_count}") vnc_password = os.getenv('VNC_PASSWORD', 'supersecret') base_vnc_port = int(os.getenv('CAMOUFOX_BASE_VNC_PORT', 5901)) camoufox_port = int(os.getenv('CAMOUFOX_PORT', 12345)) camoufox_backend_prefix = os.getenv('CAMOUFOX_BACKEND_PREFIX', 'camoufox-') # --- Generate docker-compose.camoufox.yaml --- compose_output_file = os.path.join(script_dir, 'docker-compose.camoufox.yaml') # Generate the compose file directly without template with open(compose_output_file, 'w') as f: f.write("# THIS FILE IS AUTO-GENERATED BY generate_envoy_config.py\n") f.write("# DO NOT EDIT MANUALLY.\n") f.write("#\n") f.write("# It contains the service definitions for the camoufox instances\n") f.write("# and adds the necessary dependencies to the main services.\n") f.write("services:\n\n") # Generate services for each proxy for i, proxy in enumerate(camoufox_proxies): service_name = f"camoufox-{i+1}" # Each container gets its own unique range of ports to avoid conflicts container_base_port = camoufox_port + i * worker_count host_base_port = container_base_port f.write(f" {service_name}:\n") f.write(f" build:\n") f.write(f" context: ./camoufox\n") f.write(f" dockerfile: Dockerfile\n") f.write(f" args:\n") f.write(f" VNC_PASSWORD: {vnc_password}\n") f.write(f" image: camoufox:latest\n") f.write(f" container_name: ytdlp-ops-{service_name}-1\n") f.write(f" restart: unless-stopped\n") f.write(f" shm_size: '2gb' # Mitigates browser crashes due to shared memory limitations\n") f.write(f" ports:\n") f.write(f" - \"{host_base_port}-{host_base_port + worker_count - 1}:{container_base_port}-{container_base_port + worker_count - 1}\"\n") f.write(f" environment:\n") f.write(f" - DISPLAY=:99\n") f.write(f" - MOZ_HEADLESS_STACKSIZE=2097152\n") f.write(f" - CAMOUFOX_MAX_MEMORY_MB=2048\n") f.write(f" - CAMOUFOX_MAX_CONCURRENT_CONTEXTS=8\n") f.write(f" - CAMOUFOX_RESTART_THRESHOLD_MB=1500\n") f.write(f" volumes:\n") f.write(f" - /tmp/.X11-unix:/tmp/.X11-unix:rw\n") f.write(f" - camoufox-data-{i+1}:/app/context-data\n") f.write(f" - camoufox-browser-cache:/root/.cache/ms-playwright # Persist browser binaries\n") f.write(f" command: [\n") f.write(f" \"--ws-host\", \"0.0.0.0\",\n") f.write(f" \"--port\", \"{container_base_port}\",\n") f.write(f" \"--num-instances\", \"{worker_count}\",\n") f.write(f" \"--ws-path\", \"mypath\",\n") f.write(f" \"--proxy-url\", \"{proxy['url']}\",\n") f.write(f" \"--headless\",\n") f.write(f" \"--monitor-resources\",\n") f.write(f" \"--memory-restart-threshold\", \"1800\",\n") f.write(f" \"--preferences\", \"layers.acceleration.disabled=true,dom.ipc.processCount=2,media.memory_cache_max_size=102400,browser.cache.memory.capacity=102400\"\n") f.write(f" ]\n") f.write(f" deploy:\n") f.write(f" resources:\n") f.write(f" limits:\n") f.write(f" memory: 2.5G\n") f.write(f" logging:\n") f.write(f" driver: \"json-file\"\n") f.write(f" options:\n") f.write(f" max-size: \"100m\"\n") f.write(f" max-file: \"3\"\n") f.write(f" networks:\n") f.write(f" - proxynet\n\n") # Add camoufox-group service that depends on all camoufox instances if camoufox_proxies: f.write(" camoufox-group:\n") f.write(" image: alpine:latest\n") f.write(" command: [\"echo\", \"Camoufox group ready.\"]\n") f.write(" restart: \"no\"\n") f.write(" depends_on:\n") for i in range(len(camoufox_proxies)): f.write(f" - camoufox-{i+1}\n") f.write(" networks:\n") f.write(" - proxynet\n\n") # Write volumes section f.write("volumes:\n") for i in range(len(camoufox_proxies)): f.write(f" camoufox-data-{i+1}:\n") if camoufox_proxies: f.write(" camoufox-browser-cache:\n") f.write("\n") # Write networks section f.write("networks:\n") f.write(" proxynet:\n") f.write(" name: airflow_proxynet\n") f.write(" external: true\n") logging.info(f"Successfully generated {compose_output_file} with {len(camoufox_proxies)} camoufox service(s).") logging.info("This docker-compose file defines the remote browser services, one for each proxy.") logging.info("----------------------------------------------------------") # --- Generate camoufox_endpoints.json --- endpoints_map = {} for i, proxy in enumerate(camoufox_proxies): proxy_port = _get_port_from_proxy_url(proxy['url']) if proxy_port: container_base_port = camoufox_port + i * worker_count endpoints = [] for j in range(worker_count): port = container_base_port + j endpoints.append(f"ws://{camoufox_backend_prefix}{i+1}:{port}/mypath") endpoints_map[proxy_port] = { "ws_endpoints": endpoints } else: logging.warning(f"Could not extract port from proxy URL: {proxy['url']}. Skipping for endpoint map.") endpoints_data = {"endpoints": endpoints_map} camoufox_dir = os.path.join(script_dir, 'camoufox') endpoints_output_file = os.path.join(camoufox_dir, 'camoufox_endpoints.json') with open(endpoints_output_file, 'w') as f: json.dump(endpoints_data, f, indent=2) logging.info(f"Successfully generated {endpoints_output_file} with {len(endpoints_map)} port-keyed endpoint(s).") logging.info("This file maps each proxy to a list of WebSocket endpoints for Camoufox.") logging.info("The token_generator uses this map to connect to the correct remote browser.") else: logging.info("Skipping Camoufox configuration generation for 'management' role.") # --- Generate docker-compose-ytdlp-ops.yaml --- ytdlp_ops_template = env.get_template('docker-compose-ytdlp-ops.yaml.j2') ytdlp_ops_output_file = os.path.join(script_dir, 'docker-compose-ytdlp-ops.yaml') # Combine all proxies (camoufox and general) into a single string for the server. all_proxies = [] if expanded_camoufox_proxies_str: all_proxies.extend([p.strip() for p in expanded_camoufox_proxies_str.split(',') if p.strip()]) general_proxies_str = os.getenv('GENERAL_PROXIES') if general_proxies_str: expanded_general_proxies_str = expand_env_vars(general_proxies_str) logging.info(f"Expanded GENERAL_PROXIES from '{general_proxies_str}' to '{expanded_general_proxies_str}'") general_proxies = [p.strip() for p in expanded_general_proxies_str.split(',') if p.strip()] all_proxies.extend(general_proxies) logging.info(f"Adding {len(general_proxies)} general purpose proxy/proxies.") # Also check for the SOCKS5_SOCK_SERVER_IP for backward compatibility with docs socks_server_ip = os.getenv('SOCKS5_SOCK_SERVER_IP', '172.17.0.1') if socks_server_ip: socks_server_port = os.getenv('SOCKS5_SOCK_SERVER_PORT', '1087') general_proxy_url = f"socks5://{socks_server_ip}:{socks_server_port}" if general_proxy_url not in all_proxies: all_proxies.append(general_proxy_url) logging.info(f"Adding general purpose proxy from SOCKS5_SOCK_SERVER_IP: {general_proxy_url}") combined_proxies_str = ",".join(all_proxies) logging.info(f"Combined proxy string for ytdlp-ops-service: '{combined_proxies_str}'") ytdlp_ops_config_data = { 'combined_proxies_str': combined_proxies_str, 'service_role': service_role, } rendered_ytdlp_ops_config = ytdlp_ops_template.render(ytdlp_ops_config_data) with open(ytdlp_ops_output_file, 'w') as f: f.write(rendered_ytdlp_ops_config) logging.info(f"Successfully generated {ytdlp_ops_output_file}") # --- Envoy Configuration --- envoy_port = int(os.getenv('ENVOY_PORT', 9080)) base_port = int(os.getenv('YTDLP_BASE_PORT', 9090)) envoy_admin_port = int(os.getenv('ENVOY_ADMIN_PORT', 9901)) # For local dev, ENVOY_BACKEND_ADDRESS is set to 127.0.0.1. For Docker, it's unset, so we default to the service name. backend_address = os.getenv('ENVOY_BACKEND_ADDRESS', 'ytdlp-ops-service') # Use STATIC for IP addresses, and STRICT_DNS for anything else (hostnames). envoy_cluster_type = 'STATIC' if is_ip_address(backend_address) else 'STRICT_DNS' # --- Generate envoy.yaml --- envoy_template = env.get_template('envoy.yaml.j2') envoy_output_file = os.path.join(script_dir, 'envoy.yaml') logging.info("--- Generating Envoy Configuration ---") logging.info(f"Envoy will listen on public port: {envoy_port}") logging.info(f"It will load balance requests across {worker_count} internal gRPC endpoints of the 'ytdlp-ops-service'.") logging.info(f"The backend service is located at: '{backend_address}' (type: {envoy_cluster_type})") envoy_config_data = { 'envoy_port': envoy_port, 'worker_count': worker_count, 'base_port': base_port, 'envoy_admin_port': envoy_admin_port, 'backend_address': backend_address, 'envoy_cluster_type': envoy_cluster_type, } rendered_envoy_config = envoy_template.render(envoy_config_data) with open(envoy_output_file, 'w') as f: f.write(rendered_envoy_config) logging.info(f"Successfully generated {envoy_output_file}") logging.info("--- Configuration Generation Complete ---") except Exception as e: logging.error(f"Failed to generate configurations: {e}", exc_info=True) exit(1) if __name__ == '__main__': generate_configs()