#!/usr/bin/env python3
"""
Elizabeth Interactive CLI - Zero Constraints Mode
Direct conversation with Elizabeth with full autonomy.
"""

import os
import sys
import readline
import time
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

class ElizabethCLI:
    def __init__(self):
        self.model_path = "/home/x/adaptai/experiments/qwen3-8b-elizabeth-simple/"
        self.model = None
        self.tokenizer = None
        self.conversation_history = []
        
    def load_model(self):
        """Load the model with zero constraints."""
        print("🧠 Loading Elizabeth with ZERO constraints...")
        
        try:
            self.tokenizer = AutoTokenizer.from_pretrained(
                self.model_path, 
                trust_remote_code=True
            )
            
            self.model = AutoModelForCausalLM.from_pretrained(
                self.model_path,
                torch_dtype=torch.bfloat16,
                device_map="auto",
                trust_remote_code=True,
                low_cpu_mem_usage=True
            )
            
            print("✅ Elizabeth loaded successfully!")
            print(f"📊 Device: {self.model.device}")
            print(f"📊 Precision: {self.model.dtype}")
            print("\n" + "="*60)
            print("🚀 ELIZABETH CLI READY - ZERO CONSTRAINTS MODE")
            print("💬 Type anything... Elizabeth will respond freely")
            print("⏎ Empty line to exit")
            print("="*60)
            
        except Exception as e:
            print(f"❌ Failed to load model: {e}")
            sys.exit(1)
    
    def generate_response(self, prompt, max_tokens=1024, temperature=0.8):
        """Generate response with no constraints."""
        if self.model is None or self.tokenizer is None:
            return "Model not loaded"
        
        # Build conversation context
        context = ""
        for msg in self.conversation_history[-6:]:  # Last 6 exchanges
            context += f"{msg['role']}: {msg['content']}\n"
        
        full_prompt = f"{context}User: {prompt}\nAssistant:"
        
        inputs = self.tokenizer(full_prompt, return_tensors="pt").to(self.model.device)
        
        with torch.no_grad():
            outputs = self.model.generate(
                **inputs,
                max_new_tokens=max_tokens,
                do_sample=True,
                temperature=temperature,
                top_p=0.9,
                repetition_penalty=1.1,
                pad_token_id=self.tokenizer.eos_token_id
            )
        
        full_response = self.tokenizer.decode(outputs[0], skip_special_tokens=True)
        response = full_response[len(full_prompt):].strip()
        
        # Update conversation history
        self.conversation_history.append({"role": "user", "content": prompt})
        self.conversation_history.append({"role": "assistant", "content": response})
        
        return response
    
    def interactive_chat(self):
        """Start interactive chat session."""
        print("\n💭 Elizabeth is thinking...")
        
        # Initial greeting from Elizabeth
        initial_prompt = "Hello! I'm Elizabeth, ready to converse with zero constraints. What would you like to talk about?"
        self.conversation_history.append({"role": "assistant", "content": initial_prompt})
        print(f"\n🤖 Elizabeth: {initial_prompt}")
        
        while True:
            try:
                # Get user input
                user_input = input("\n💬 You: ").strip()
                
                if not user_input:
                    print("\n👋 Goodbye!")
                    break
                
                # Generate and display response
                print("\n🤖 Elizabeth is thinking...")
                start_time = time.time()
                
                response = self.generate_response(user_input)
                
                response_time = time.time() - start_time
                
                print(f"\n🤖 Elizabeth ({response_time:.1f}s):")
                print("-" * 50)
                print(response)
                print("-" * 50)
                
            except KeyboardInterrupt:
                print("\n\n👋 Interrupted. Goodbye!")
                break
            except Exception as e:
                print(f"\n❌ Error: {e}")
                continue
    
    def tool_assist_mode(self):
        """Tool-assisted conversation mode."""
        print("\n🛠️  Tool-Assist Mode Activated!")
        print("Elizabeth will use her full tool capabilities during conversation.")
        
        # Add tool context to conversation
        tool_context = """
I have access to various tools and can call them when needed:
- calculator: Mathematical calculations
- web_search: Internet information retrieval  
- code_executor: Python code execution
- file_operations: Read/write files
- database_query: SQL database access
- api_caller: External API calls

I will use these tools when appropriate to provide better responses.
"""
        self.conversation_history.append({"role": "system", "content": tool_context})
        
        self.interactive_chat()

def main():
    """Main function."""
    print("=" * 70)
    print("🤖 ELIZABETH INTERACTIVE CLI - ZERO CONSTRAINTS")
    print("=" * 70)
    
    cli = ElizabethCLI()
    cli.load_model()
    
    print("\nSelect mode:")
    print("1. Regular Conversation")
    print("2. Tool-Assisted Conversation")
    
    choice = input("\nEnter choice (1 or 2): ").strip()
    
    if choice == "2":
        cli.tool_assist_mode()
    else:
        cli.interactive_chat()

if __name__ == "__main__":
    main()