main.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

"""
Okey Google - Main Launcher
Capture voice input and send to the agent
"""

import time
import sys
import os
from agent import agent
from speech_to_text import listen_and_recognize
from text_to_speech import speak
from langchain_core.messages import HumanMessage
from dotenv import load_dotenv

load_dotenv()


def main():
    """Main program - Capture voice commands and send to the agent"""
    
    # API Key check
    if not os.getenv("GOOGLE_API_KEY"):
        print("\n❌ ERROR: GOOGLE_API_KEY environment variable is not set!")
        print("   Please create a .env file and add your API key")
        sys.exit(1)
    
    print("\n" + "="*60)
    print("✨ Okey Google - Voice-Controlled Assistant")
    print("="*60)
    print("\n🛑 Press Ctrl+C to exit\n")
    
    try:
        while True:
            try:
                print("🎤 I'm ready... Please speak")
                
                # Get voice command
                listened_text = listen_and_recognize()
                print(f"📝 Recognized: {listened_text}\n")
                
                if not listened_text or "hatası" in listened_text.lower():
                    print("⚠️ Speech not recognized, please try again...\n")
                    speak("Ses tanınamadı, lütfen tekrar deneyin")
                    time.sleep(1)
                    continue
                
                # Run agent - send listened_text as HumanMessage
                print("🤖 Agent running...\n")
                result = agent.invoke({
                    "messages": [HumanMessage(content=listened_text)],
                    "request": listened_text,
                    "installed_programs": [],
                    "desktop_files": [],
                    "screen_texts": [],
                    "is_action_successful": False,
                    "actions": [],
                    "current_action_idx": 0,
                    "action_result": ""
                }, {"configurable": {"thread_id": "1"}})
                
                # Get agent response and speak it
                agent_response = result.get("messages", [])
                if agent_response:
                    last_message = agent_response[-1]
                    # AIMessage içinden content'i çıkar
                    if hasattr(last_message, 'content'):
                        response_text = last_message.content
                    else:
                        response_text = str(last_message)
                    
                    print(f"\n🤖 Response: {response_text}\n")
                    
                    # Play spoken response
                    print("🔊 Playing spoken response...")
                    speak(response_text)
                
                print("✅ Completed!\n")
                
                # Listen for additional commands - refresh after each
                print("⏳ Waiting for additional commands (refreshes after each)...\n")
                
                # Listen for additional commands within 2 minutes
                additional_timeout_start = time.time()
                additional_timeout_limit = 120  # 2 dakika
                
                while time.time() - additional_timeout_start < additional_timeout_limit:
                    try:
                        print("🎤 Please say an additional command (listening for 10 seconds)...\n")
                        
                        # Listen with 10-second timeout
                        extra_command = listen_and_recognize(
                            timeout=10, 
                            phrase_time_limit=None, 
                            require_activation=False
                        )
                        
                        if extra_command and "hatası" not in extra_command.lower() and extra_command.strip():
                            print(f"\n📝 Additional command: {extra_command}\n")
                            
                            # Run additional command
                            print("🤖 Agent running...\n")
                            result = agent.invoke({
                                "messages": [HumanMessage(content=extra_command)],
                                "request": extra_command,
                                "installed_programs": [],
                                "desktop_files": [],
                                "screen_texts": [],
                                "is_action_successful": False,
                                "actions": [],
                                "current_action_idx": 0,
                                "action_result": ""
                            }, {"configurable": {"thread_id": "1"}})
                            
                            # Play the response
                            agent_response = result.get("messages", [])
                            if agent_response:
                                last_message = agent_response[-1]
                                if hasattr(last_message, 'content'):
                                    response_text = last_message.content
                                else:
                                    response_text = str(last_message)
                                
                                print(f"\n🤖 Response: {response_text}\n")
                                print("🔊 Playing spoken response...")
                                speak(response_text)
                            
                            # After processing, start listening again (timeout resets)
                            print("\n" + "="*60)
                            print("Ready for the next command...\n")
                        
                        else:
                            # Timeout - continue if silence
                            remaining = additional_timeout_limit - (time.time() - additional_timeout_start)
                            if remaining > 0:
                                print(f"⏱️  Still listening ({remaining:.0f}s left)...\n")
                    
                    except KeyboardInterrupt:
                        print("\n🛑 Additional command wait cancelled...\n")
                        break
                    except Exception as e:
                        remaining = additional_timeout_limit - (time.time() - additional_timeout_start)
                        if remaining > 0:
                            print(f"⚠️  Still listening ({remaining:.0f}s left)...\n")
                        else:
                            break
                
                print("\n" + "="*60)
                print("Additional command wait time expired\n")
                
                print("\n" + "="*60)
                print("Waiting for Ok Google...\n")
                
            except KeyboardInterrupt:
                print("\n🛑 Loop interrupted...\n")
                time.sleep(1)
            except Exception as e:
                print(f"\n❌ Error: {e}\n")
                import traceback
                traceback.print_exc()
                time.sleep(2)
                
    except KeyboardInterrupt:
        print("\n👋 Shutting down...")
        sys.exit(0)


if __name__ == "__main__":
    main()