from typing import Dict, List, Optional
from datetime import datetime, timedelta
import numpy as np
from .gemini_analyzer import GeminiAnalyzer
from .database import DatabaseManager

class PredictionEngine:
    def __init__(self, gemini_analyzer: GeminiAnalyzer, db_manager: DatabaseManager):
        self.gemini_analyzer = gemini_analyzer
        self.db_manager = db_manager
        
    def generate_predictions(self) -> List[Dict]:
        """Generate predictions for current narratives"""
        try:
            # Get active high-risk narratives
            narratives = self.db_manager.get_high_risk_narratives(threshold=60)
            
            predictions = []
            for narrative in narratives:
                scenarios = self.gemini_analyzer.generate_prediction_scenarios(narrative['content'])
                
                for scenario in scenarios:
                    prediction = {
                        'narrative_id': narrative['id'],
                        'narrative_title': narrative['title'][:50] + "..." if len(narrative['title']) > 50 else narrative['title'],
                        'risk_score': narrative['risk_score'],
                        'probability': scenario.get('probability', 50),
                        'predicted_reach': scenario.get('predicted_reach', 10000),
                        'timeline': f"{scenario.get('timeline_hours', 24)} hours",
                        'best_case_scenario': scenario.get('description', 'No description available')
                    }
                    predictions.append(prediction)
            
            return predictions[:10]  # Limit to top 10
            
        except Exception as e:
            print(f"❌ Prediction generation failed: {e}")
            return []
    
    def calculate_spread_probability(self, narrative_content: str, historical_data: List[Dict] = None) -> float:
        """Calculate probability of viral spread"""
        # This would use ML models in production
        # For now, return a calculated estimate
        base_probability = 0.3
        
        # Adjust based on content characteristics
        if any(term in narrative_content.lower() for term in ['breaking', 'urgent', 'exclusive']):
            base_probability += 0.2
            
        if any(term in narrative_content.lower() for term in ['government', 'conspiracy', 'cover-up']):
            base_probability += 0.15
            
        return min(base_probability, 0.95)
    
    def predict_timeline(self, narrative_content: str) -> Dict:
        """Predict spread timeline"""
        return {
            'initial_detection': 0,  # hours
            'community_spread': 6,   # hours  
            'cross_platform': 12,   # hours
            'mainstream_pickup': 24, # hours
            'peak_virality': 48      # hours
        }