src/database.py

"""
Database Models and Connection Management
SQLite database for users, predictions, and feedback
"""

import os
import sqlite3
from datetime import datetime, timedelta
from typing import Dict, List, Optional, Tuple
from contextlib import contextmanager
import json

class Database:
    """Database management class for SQLite"""
    
    def __init__(self, db_path: str = '../data/app.db'):
        self.db_path = os.path.abspath(db_path)
        db_dir = os.path.dirname(self.db_path)
        os.makedirs(db_dir, exist_ok=True)
        self._init_database()
    
    @contextmanager
    def get_connection(self):
        """Get database connection with automatic commit/rollback"""
        conn = sqlite3.connect(self.db_path)
        conn.row_factory = sqlite3.Row  # Return rows as dictionaries
        try:
            yield conn
            conn.commit()
        except Exception as e:
            conn.rollback()
            raise e
        finally:
            conn.close()
    
    def _init_database(self):
        """Initialize database tables"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            
            # Users table
            cursor.execute('''
                CREATE TABLE IF NOT EXISTS users (
                    id INTEGER PRIMARY KEY AUTOINCREMENT,
                    user_id TEXT UNIQUE NOT NULL,
                    email TEXT,
                    phone TEXT,
                    baby_name TEXT NOT NULL,
                    baby_gender TEXT,
                    baby_birthday TEXT,
                    platform TEXT DEFAULT 'web',
                    language TEXT DEFAULT 'ar',
                    registration_date TEXT NOT NULL,
                    created_at TEXT DEFAULT CURRENT_TIMESTAMP,
                    updated_at TEXT DEFAULT CURRENT_TIMESTAMP
                )
            ''')
            
            # Predictions table (recordings)
            cursor.execute('''
                CREATE TABLE IF NOT EXISTS predictions (
                    id INTEGER PRIMARY KEY AUTOINCREMENT,
                    user_id TEXT,
                    prediction TEXT NOT NULL,
                    confidence REAL NOT NULL,
                    audio_id TEXT UNIQUE NOT NULL,
                    model_type TEXT,
                    timestamp TEXT NOT NULL,
                    created_at TEXT DEFAULT CURRENT_TIMESTAMP,
                    FOREIGN KEY (user_id) REFERENCES users(user_id)
                )
            ''')
            
            # Feedback table
            cursor.execute('''
                CREATE TABLE IF NOT EXISTS feedback (
                    id INTEGER PRIMARY KEY AUTOINCREMENT,
                    audio_id TEXT NOT NULL,
                    user_id TEXT,
                    predicted_label TEXT NOT NULL,
                    correct_label TEXT NOT NULL,
                    is_correct INTEGER NOT NULL,
                    confidence REAL,
                    category TEXT,
                    timestamp TEXT NOT NULL,
                    created_at TEXT DEFAULT CURRENT_TIMESTAMP,
                    FOREIGN KEY (user_id) REFERENCES users(user_id)
                )
            ''')
            
            # Analytics events table
            cursor.execute('''
                CREATE TABLE IF NOT EXISTS analytics_events (
                    id INTEGER PRIMARY KEY AUTOINCREMENT,
                    event_name TEXT NOT NULL,
                    user_id TEXT,
                    event_data TEXT,
                    timestamp TEXT NOT NULL,
                    created_at TEXT DEFAULT CURRENT_TIMESTAMP
                )
            ''')
            
            # Create indexes for better query performance
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_users_user_id ON users(user_id)')
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_predictions_user_id ON predictions(user_id)')
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_predictions_timestamp ON predictions(timestamp)')
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_feedback_user_id ON feedback(user_id)')
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_feedback_audio_id ON feedback(audio_id)')
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_feedback_timestamp ON feedback(timestamp)')
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_analytics_user_id ON analytics_events(user_id)')
            cursor.execute('CREATE INDEX IF NOT EXISTS idx_analytics_timestamp ON analytics_events(timestamp)')
            
            conn.commit()
    
    # ==================== USERS CRUD ====================
    
    def create_user(self, user_data: Dict) -> str:
        """Create a new user"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            
            registration_date = user_data.get('registration_date', datetime.now().isoformat())
            
            cursor.execute('''
                INSERT INTO users (
                    user_id, email, phone, baby_name, baby_gender, baby_birthday,
                    platform, language, registration_date
                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
            ''', (
                user_data.get('user_id'),
                user_data.get('email'),
                user_data.get('phone'),
                user_data.get('baby_name'),
                user_data.get('baby_gender'),
                user_data.get('baby_birthday'),
                user_data.get('platform', 'web'),
                user_data.get('language', 'ar'),
                registration_date
            ))
            
            return user_data.get('user_id')
    
    def get_user(self, user_id: str) -> Optional[Dict]:
        """Get user by user_id"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            cursor.execute('SELECT * FROM users WHERE user_id = ?', (user_id,))
            row = cursor.fetchone()
            return dict(row) if row else None
    
    def get_all_users(self, limit: Optional[int] = None, offset: int = 0) -> List[Dict]:
        """Get all users with pagination"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            if limit:
                cursor.execute('SELECT * FROM users ORDER BY registration_date DESC LIMIT ? OFFSET ?', (limit, offset))
            else:
                cursor.execute('SELECT * FROM users ORDER BY registration_date DESC')
            return [dict(row) for row in cursor.fetchall()]
    
    def update_user(self, user_id: str, user_data: Dict) -> bool:
        """Update user information"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            
            updates = []
            values = []
            
            for key in ['email', 'phone', 'baby_name', 'baby_gender', 'baby_birthday', 'platform', 'language']:
                if key in user_data:
                    updates.append(f'{key} = ?')
                    values.append(user_data[key])
            
            if not updates:
                return False
            
            updates.append('updated_at = ?')
            values.append(datetime.now().isoformat())
            values.append(user_id)
            
            cursor.execute(f'''
                UPDATE users SET {", ".join(updates)}
                WHERE user_id = ?
            ''', values)
            
            return cursor.rowcount > 0
    
    def delete_user(self, user_id: str) -> bool:
        """Delete user"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            cursor.execute('DELETE FROM users WHERE user_id = ?', (user_id,))
            return cursor.rowcount > 0
    
    def count_users(self) -> int:
        """Count total users"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            cursor.execute('SELECT COUNT(*) as count FROM users')
            return cursor.fetchone()['count']
    
    # ==================== PREDICTIONS ====================
    
    def create_prediction(self, prediction_data: Dict):
        """Save a prediction (recording)"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            cursor.execute('''
                INSERT INTO predictions (
                    user_id, prediction, confidence, audio_id, model_type, timestamp
                ) VALUES (?, ?, ?, ?, ?, ?)
            ''', (
                prediction_data.get('user_id'),
                prediction_data.get('prediction'),
                prediction_data.get('confidence', 0),
                prediction_data.get('audio_id'),
                prediction_data.get('model_type'),
                prediction_data.get('timestamp', datetime.now().isoformat())
            ))
    
    def count_predictions(self, user_id: Optional[str] = None) -> int:
        """Count total predictions/recordings"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            if user_id:
                cursor.execute('SELECT COUNT(*) as count FROM predictions WHERE user_id = ?', (user_id,))
            else:
                cursor.execute('SELECT COUNT(*) as count FROM predictions')
            return cursor.fetchone()['count']
    
    # ==================== FEEDBACK ====================
    
    def create_feedback(self, feedback_data: Dict):
        """Save feedback"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            cursor.execute('''
                INSERT INTO feedback (
                    audio_id, user_id, predicted_label, correct_label,
                    is_correct, confidence, category, timestamp
                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?)
            ''', (
                feedback_data.get('audio_id'),
                feedback_data.get('user_id'),
                feedback_data.get('predicted_label'),
                feedback_data.get('correct_label'),
                1 if feedback_data.get('is_correct', False) else 0,
                feedback_data.get('confidence'),
                feedback_data.get('correct_label'),  # category
                feedback_data.get('timestamp', datetime.now().isoformat())
            ))
    
    def count_feedback(self, is_correct: Optional[bool] = None) -> int:
        """Count feedback (optionally filtered by is_correct)"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            if is_correct is not None:
                cursor.execute(
                    'SELECT COUNT(*) as count FROM feedback WHERE is_correct = ?',
                    (1 if is_correct else 0,)
                )
            else:
                cursor.execute('SELECT COUNT(*) as count FROM feedback')
            return cursor.fetchone()['count']
    
    # ==================== STATISTICS ====================
    
    def get_user_stats(self) -> Dict:
        """Get comprehensive user statistics"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            
            # Total users
            cursor.execute('SELECT COUNT(*) as count FROM users')
            total_users = cursor.fetchone()['count']
            
            # Active users (last 7/30 days) - users who made predictions
            seven_days_ago = (datetime.now() - timedelta(days=7)).isoformat()
            thirty_days_ago = (datetime.now() - timedelta(days=30)).isoformat()
            
            cursor.execute('''
                SELECT COUNT(DISTINCT user_id) as count FROM predictions
                WHERE timestamp >= ? AND user_id IS NOT NULL
            ''', (seven_days_ago,))
            active_7d = cursor.fetchone()['count']
            
            cursor.execute('''
                SELECT COUNT(DISTINCT user_id) as count FROM predictions
                WHERE timestamp >= ? AND user_id IS NOT NULL
            ''', (thirty_days_ago,))
            active_30d = cursor.fetchone()['count']
            
            # New users
            cursor.execute('''
                SELECT COUNT(*) as count FROM users
                WHERE registration_date >= ?
            ''', (seven_days_ago,))
            new_users_7d = cursor.fetchone()['count']
            
            cursor.execute('''
                SELECT COUNT(*) as count FROM users
                WHERE registration_date >= ?
            ''', (thirty_days_ago,))
            new_users_30d = cursor.fetchone()['count']
            
            return {
                'total_users': total_users,
                'active_users_7d': active_7d,
                'active_users_30d': active_30d,
                'new_users_7d': new_users_7d,
                'new_users_30d': new_users_30d
            }
    
    def get_feedback_stats(self) -> Dict:
        """Get feedback statistics"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            
            # Total feedback
            total_feedback = self.count_feedback()
            
            # Positive feedback (is_correct = 1)
            positive_feedback = self.count_feedback(is_correct=True)
            
            # Negative feedback (is_correct = 0)
            negative_feedback = self.count_feedback(is_correct=False)
            
            # Feedback by category
            cursor.execute('''
                SELECT category, COUNT(*) as count
                FROM feedback
                GROUP BY category
            ''')
            category_counts = {row['category']: row['count'] for row in cursor.fetchall()}
            
            # Total recordings/predictions
            total_recordings = self.count_predictions()
            
            return {
                'total_feedback': total_feedback,
                'positive_feedback': positive_feedback,
                'negative_feedback': negative_feedback,
                'category_counts': category_counts,
                'total_recordings': total_recordings
            }
    
    def get_platform_distribution(self) -> Dict:
        """Get platform distribution"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            cursor.execute('''
                SELECT platform, COUNT(*) as count
                FROM users
                GROUP BY platform
            ''')
            return {row['platform']: row['count'] for row in cursor.fetchall()}
    
    def get_language_distribution(self) -> Dict:
        """Get language distribution"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            cursor.execute('''
                SELECT language, COUNT(*) as count
                FROM users
                GROUP BY language
            ''')
            return {row['language']: row['count'] for row in cursor.fetchall()}
    
    def get_user_timeline(self, days: int = 30) -> List[Dict]:
        """Get user registration timeline"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            from datetime import timedelta
            days_ago = (datetime.now() - timedelta(days=days)).isoformat()
            cursor.execute('''
                SELECT DATE(registration_date) as date, COUNT(*) as count
                FROM users
                WHERE registration_date >= ?
                GROUP BY DATE(registration_date)
                ORDER BY date ASC
            ''', (days_ago,))
            return [{'date': row['date'], 'count': row['count']} for row in cursor.fetchall()]
    
    def get_feedback_timeline(self, days: int = 30) -> List[Dict]:
        """Get feedback submission timeline"""
        with self.get_connection() as conn:
            cursor = conn.cursor()
            from datetime import timedelta
            days_ago = (datetime.now() - timedelta(days=days)).isoformat()
            cursor.execute('''
                SELECT DATE(timestamp) as date, COUNT(*) as count
                FROM feedback
                WHERE timestamp >= ?
                GROUP BY DATE(timestamp)
                ORDER BY date ASC
            ''', (days_ago,))
            return [{'date': row['date'], 'count': row['count']} for row in cursor.fetchall()]