import streamlit as st
from typing import Dict
from datetime import datetime
import anthropic
from src.utils.session import get_tutor_context
from src.utils.config import get_anthropic_api_key

class AITutorService:
    def __init__(self):
        self.initialize_client()
        self.load_avatar_assets()

    def initialize_client(self):
        """Initialize Anthropic client"""
        self.client = anthropic.Anthropic(api_key=get_anthropic_api_key())

    def load_avatar_assets(self):
        """Load avatar assets and animations"""
        self.avatar_states = {
            'neutral': """
                <svg width="100" height="100" viewBox="0 0 100 100">
                    <circle cx="50" cy="50" r="45" fill="#4A90E2"/>
                    <circle cx="35" cy="40" r="5" fill="white"/>
                    <circle cx="65" cy="40" r="5" fill="white"/>
                    <path d="M 30 60 Q 50 70 70 60" stroke="white" fill="none" stroke-width="3"/>
                </svg>
            """,
            'thinking': """
                <svg width="100" height="100" viewBox="0 0 100 100">
                    <circle cx="50" cy="50" r="45" fill="#4A90E2"/>
                    <circle cx="35" cy="40" r="5" fill="white"/>
                    <circle cx="65" cy="40" r="5" fill="white"/>
                    <path d="M 30 65 Q 50 65 70 65" stroke="white" fill="none" stroke-width="3"/>
                </svg>
            """,
            'happy': """
                <svg width="100" height="100" viewBox="0 0 100 100">
                    <circle cx="50" cy="50" r="45" fill="#4A90E2"/>
                    <circle cx="35" cy="40" r="5" fill="white"/>
                    <circle cx="65" cy="40" r="5" fill="white"/>
                    <path d="M 30 60 Q 50 80 70 60" stroke="white" fill="none" stroke-width="3"/>
                </svg>
            """
        }

    def display_avatar(self, state: str = 'neutral'):
        """Display the AI tutor avatar"""
        st.markdown(f"""
            <div style="display: flex; justify-content: center; margin: 20px 0;">
                {self.avatar_states.get(state, self.avatar_states['neutral'])}
            </div>
        """, unsafe_allow_html=True)

    def generate_response(self, user_input: str) -> str:
        """Generate contextualized response using Claude"""
        context = get_tutor_context()
        prompt = self.build_prompt(user_input, context)

        # Generate response using Claude
        message = self.client.messages.create(
            model="claude-3-opus-20240229",
            max_tokens=1024,
            temperature=0.7,
            system="You are an expert AI tutor, skilled at explaining complex concepts clearly and engaging students in their learning journey. Provide detailed, accurate explanations while maintaining an encouraging tone.",
            messages=[
                {
                    "role": "user",
                    "content": prompt
                }
            ]
        )
        
        response = message.content[0].text
        
        # Analyze sentiment (Claude can do this as part of its response)
        sentiment_score = 0.8 if any(word in response.lower() for word in ['great', 'excellent', 'good job', 'well done']) else 0.5
        
        # Update engagement metrics
        self.update_engagement_metrics(user_input, response, sentiment_score)
        
        return response

    def build_prompt(self, user_input: str, context: Dict) -> str:
        """Build a context-aware prompt for Claude"""
        topic_context = f"Current topic: {context['current_topic']}\n" if context['current_topic'] else ""
        
        chat_context = "\n".join([
            f"Student: {msg['content']}" if msg['role'] == 'user' else f"Tutor: {msg['content']}"
            for msg in context['chat_history'][-3:]
        ])
        
        return f"""
        You are an educational AI tutor helping a student learn.
        
        {topic_context}
        Difficulty level: {context['difficulty_level']}
        Learning style: {context['learning_style']}
        
        Previous conversation:
        {chat_context}
        
        Student: {user_input}
        
        Provide a clear, helpful response that:
        1. Addresses the student's question directly
        2. Explains concepts in an engaging way
        3. Uses examples when appropriate
        4. Encourages deeper understanding
        5. Maintains a supportive tone

        If the question is about code, include working code examples.
        If the topic is complex, break it down into simpler parts.
        """

    def update_engagement_metrics(self, user_input: str, response: str, sentiment_score: float):
        """Update student engagement metrics"""
        context = get_tutor_context()
        context['engagement_metrics'].append({
            'timestamp': datetime.now().isoformat(),
            'sentiment_score': sentiment_score,
            'interaction_length': len(user_input)
        })