yfinance_plus/yfinance_enhanced_cli.py

#!/usr/bin/env python3
"""
Command Line Interface for yfinance-enhanced
Provides cache management and utility functions
"""

import argparse
import sys
import json
import pickle
from pathlib import Path
from datetime import datetime

try:
    from yfinance_enhanced import get_cache_info, clear_cache, HistoricalDataCache, RequestConfig
except ImportError:
    print("Error: yfinance_enhanced module not found. Please install it first.")
    sys.exit(1)


def format_size(size_bytes):
    """Format size in bytes to human readable format"""
    if size_bytes == 0:
        return "0 B"
    
    for unit in ['B', 'KB', 'MB', 'GB']:
        if size_bytes < 1024.0:
            return f"{size_bytes:.1f} {unit}"
        size_bytes /= 1024.0
    return f"{size_bytes:.1f} TB"


def get_cache_file_details(cache_dir, filename):
    """Load and analyze cache file to get details about stored data"""
    try:
        cache_file = Path(cache_dir) / filename
        if not cache_file.exists():
            return None
            
        with open(cache_file, 'rb') as f:
            cache_data = pickle.load(f)
        
        # Extract details from cache data
        details = {
            'symbol': cache_data.get('symbol', 'Unknown'),
            'period': cache_data.get('period', 'Unknown'),
            'interval': cache_data.get('interval', '1d'),
            'rows': len(cache_data.get('data', [])),
            'date_range': None
        }
        
        # Try to get date range from the data
        data = cache_data.get('data')
        if data is not None and hasattr(data, 'index') and len(data) > 0:
            try:
                start_date = data.index[0].strftime('%Y-%m-%d')
                end_date = data.index[-1].strftime('%Y-%m-%d')
                details['date_range'] = f"{start_date} to {end_date}"
            except:
                pass
        
        return details
        
    except Exception as e:
        # If we can't read the cache file, return minimal info
        return {
            'symbol': 'Error',
            'period': f'({str(e)[:20]}...)',
            'interval': '?',
            'rows': 0,
            'date_range': None
        }


def cmd_cache_info(args):
    """Show cache information"""
    try:
        info = get_cache_info()
        
        print("🗂️  YFinance Plus Cache Information")
        print("=" * 50)
        print(f"Cache Directory: {info['cache_dir']}")
        print(f"Number of Files: {info['file_count']}")
        print(f"Total Size: {format_size(info['total_size_mb'] * 1024 * 1024)}")
        
        if info['file_count'] > 0:
            print(f"\n📁 Cache Files:")
            print("-" * 100)
            
            sorted_files = sorted(info['files'], key=lambda x: x['modified'], reverse=True)
            
            for file_info in sorted_files[:args.limit if args.limit else len(sorted_files)]:
                size_str = format_size(file_info['size_kb'] * 1024)
                modified_str = file_info['modified'].strftime('%Y-%m-%d %H:%M:%S')
                
                # Try to load and analyze cache file content
                cache_details = get_cache_file_details(info['cache_dir'], file_info['name'])
                
                print(f"  📄 {file_info['name'][:32]:<32} {size_str:>8} {modified_str}")
                if cache_details:
                    print(f"      Symbol: {cache_details['symbol']:<8} Period: {cache_details['period']:<6} "
                          f"Interval: {cache_details['interval']:<4} Rows: {cache_details['rows']}")
                    if cache_details['date_range']:
                        print(f"      Date Range: {cache_details['date_range']}")
                else:
                    print(f"      ⚠️  Could not read cache file details")
                print()
            
            if args.limit and len(sorted_files) > args.limit:
                print(f"  ... and {len(sorted_files) - args.limit} more files")
        else:
            print("\n💡 No cache files found")
            
    except Exception as e:
        print(f"❌ Error getting cache info: {e}")
        sys.exit(1)


def cmd_cache_clear(args):
    """Clear cache"""
    if not args.force:
        try:
            info = get_cache_info()
            if info['file_count'] > 0:
                response = input(f"⚠️  This will delete {info['file_count']} cache files "
                               f"({format_size(info['total_size_mb'] * 1024 * 1024)}). "
                               f"Continue? [y/N]: ")
                if response.lower() not in ['y', 'yes']:
                    print("❌ Operation cancelled")
                    return
        except Exception:
            pass
    
    try:
        clear_cache()
        print("✅ Cache cleared successfully")
    except Exception as e:
        print(f"❌ Error clearing cache: {e}")
        sys.exit(1)


def cmd_config_show(args):
    """Show current configuration"""
    config = RequestConfig()
    
    print("⚙️  YFinance Plus Configuration")
    print("=" * 40)
    print(f"Max Retries: {config.max_retries}")
    print(f"Base Delay: {config.base_delay}s")
    print(f"Max Delay: {config.max_delay}s")
    print(f"Jitter Enabled: {config.jitter}")
    print(f"Cache Enabled: {config.enable_cache}")
    print(f"Cache Directory: {config.cache_dir}")
    print(f"User Agents: {len(config.user_agents)} configured")


def cmd_test_connection(args):
    """Test connection to Yahoo Finance"""
    print("🔗 Testing connection to Yahoo Finance...")
    
    try:
        # Import here to avoid circular imports
        from yfinance_enhanced import Ticker
        
        print("  Creating test ticker...")
        ticker = Ticker("AAPL")
        
        print("  Fetching basic info...")
        info = ticker.info
        company_name = info.get('longName', 'Unknown')
        
        print("  Fetching historical data...")
        hist = ticker.history(period="5d")
        
        print(f"✅ Connection test successful!")
        print(f"   Company: {company_name}")
        print(f"   Historical data points: {len(hist)}")
        
        if ticker.enhanced_yf._cache:
            cache_info = ticker.get_cache_info()
            print(f"   Cache status: {cache_info['file_count']} files")
        
    except Exception as e:
        print(f"❌ Connection test failed: {e}")
        sys.exit(1)


def cmd_benchmark(args):
    """Run performance benchmark"""
    print("🏃 Running performance benchmark...")
    
    try:
        import time
        from yfinance_enhanced import Ticker, download
        
        symbols = ["AAPL", "GOOGL", "MSFT", "TSLA", "NVDA"]
        
        # Test individual tickers
        print(f"  Testing individual tickers ({len(symbols)} symbols)...")
        start_time = time.time()
        
        for symbol in symbols:
            ticker = Ticker(symbol)
            hist = ticker.history(period="1mo")
            print(f"    {symbol}: {len(hist)} data points")
        
        individual_time = time.time() - start_time
        
        # Test bulk download
        print(f"  Testing bulk download...")
        start_time = time.time()
        
        data = download(symbols, period="1mo")
        bulk_time = time.time() - start_time
        
        print(f"\n📊 Benchmark Results:")
        print(f"   Individual requests: {individual_time:.2f}s")
        print(f"   Bulk download: {bulk_time:.2f}s")
        print(f"   Bulk speedup: {individual_time/bulk_time:.2f}x")
        print(f"   Data shape: {data.shape}")
        
    except Exception as e:
        print(f"❌ Benchmark failed: {e}")
        sys.exit(1)


def main():
    """Main CLI entry point"""
    parser = argparse.ArgumentParser(
        description="YFinance Plus CLI - Cache management and utilities",
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog="""
Examples:
  yfp cache info                 # Show cache information
  yfp cache clear                # Clear cache (with confirmation)
  yfp cache clear --force        # Clear cache without confirmation
  yfp config show               # Show current configuration
  yfp test                       # Test connection to Yahoo Finance
  yfp benchmark                  # Run performance benchmark
        """
    )
    
    subparsers = parser.add_subparsers(dest='command', help='Available commands')
    
    # Cache commands
    cache_parser = subparsers.add_parser('cache', help='Cache management commands')
    cache_subparsers = cache_parser.add_subparsers(dest='cache_command')
    
    # Cache info
    info_parser = cache_subparsers.add_parser('info', help='Show cache information')
    info_parser.add_argument('--limit', type=int, help='Limit number of files to show')
    info_parser.set_defaults(func=cmd_cache_info)
    
    # Cache clear
    clear_parser = cache_subparsers.add_parser('clear', help='Clear cache')
    clear_parser.add_argument('--force', action='store_true', help='Clear without confirmation')
    clear_parser.set_defaults(func=cmd_cache_clear)
    
    # Config commands
    config_parser = subparsers.add_parser('config', help='Configuration commands')
    config_subparsers = config_parser.add_subparsers(dest='config_command')
    
    # Config show
    show_parser = config_subparsers.add_parser('show', help='Show current configuration')
    show_parser.set_defaults(func=cmd_config_show)
    
    # Test command
    test_parser = subparsers.add_parser('test', help='Test connection to Yahoo Finance')
    test_parser.set_defaults(func=cmd_test_connection)
    
    # Benchmark command
    benchmark_parser = subparsers.add_parser('benchmark', help='Run performance benchmark')
    benchmark_parser.set_defaults(func=cmd_benchmark)
    
    # Parse arguments
    args = parser.parse_args()
    
    # Handle no command
    if not args.command:
        parser.print_help()
        return
    
    # Handle cache subcommands
    if args.command == 'cache' and not args.cache_command:
        cache_parser.print_help()
        return
    
    # Handle config subcommands
    if args.command == 'config' and not args.config_command:
        config_parser.print_help()
        return
    
    # Execute command
    if hasattr(args, 'func'):
        try:
            args.func(args)
        except KeyboardInterrupt:
            print("\n❌ Operation cancelled by user")
            sys.exit(1)
    else:
        parser.print_help()


if __name__ == "__main__":
    main()