#!/usr/bin/env python3 """ Web-Based GUI for CSV Character Encoding Detection Tool Double-click to run - opens in your browser automatically! User Experience: 1. Double-click this file 2. Browser opens with modern interface 3. Drag and drop folders 4. See results in real-time """ import http.server import socketserver import webbrowser import subprocess import threading import os import sys import json import urllib.parse from pathlib import Path import tempfile import time import socket class CharsetAnalyzerHandler(http.server.SimpleHTTPRequestHandler): """Custom HTTP handler for the charset analyzer web GUI""" def __init__(self, *args, **kwargs): # Store reference to charset script self.charset_script = Path(__file__).parent / "check_csv_charset.py" super().__init__(*args, **kwargs) def do_GET(self): """Handle GET requests""" if self.path == '/' or self.path == '/index.html': self.serve_main_page() elif self.path == '/api/status': self.serve_status() else: super().do_GET() def do_POST(self): """Handle POST requests""" if self.path == '/api/analyze': self.handle_analyze() else: self.send_error(404) def serve_main_page(self): """Serve the main HTML interface""" html_content = self.get_html_interface() self.send_response(200) self.send_header('Content-type', 'text/html') self.send_header('Content-length', len(html_content.encode())) self.end_headers() self.wfile.write(html_content.encode()) def serve_status(self): """Serve status information""" status = { 'script_available': self.charset_script.exists(), 'script_path': str(self.charset_script) } self.send_response(200) self.send_header('Content-type', 'application/json') self.end_headers() self.wfile.write(json.dumps(status).encode()) def handle_analyze(self): """Handle analysis requests""" try: # Get content length content_length = int(self.headers['Content-Length']) post_data = self.rfile.read(content_length) # Parse JSON data data = json.loads(post_data.decode()) folder_path = data.get('folder_path', '') mode = data.get('mode', 'analyze') target_encoding = data.get('target_encoding', 'utf-8') dry_run = data.get('dry_run', True) fast_mode = data.get('fast_mode', False) # Validate folder path if not folder_path or not os.path.exists(folder_path): self.send_json_response({'error': 'Invalid folder path'}, 400) return # Check if user provided direct CSV folder vs parent folder folder_path_obj = Path(folder_path) if not folder_path_obj.is_dir(): self.send_json_response({'error': 'Path is not a directory'}, 400) return # Auto-detect folder structure (but now we support both!) csv_files_in_root = list(folder_path_obj.glob('*.csv')) + list(folder_path_obj.glob('*.CSV')) subfolders = [d for d in folder_path_obj.iterdir() if d.is_dir()] # Check if we have CSV files at all total_csv_in_subfolders = 0 for subfolder in subfolders: subfolder_csv = list(subfolder.rglob('*.csv')) + list(subfolder.rglob('*.CSV')) total_csv_in_subfolders += len(subfolder_csv) if not csv_files_in_root and total_csv_in_subfolders == 0: self.send_json_response({ 'error': f'No CSV files found in this directory or its subfolders.', 'help': 'Please select a folder that contains CSV files directly or has subfolders with CSV files.' }, 400) return if not self.charset_script.exists(): self.send_json_response({'error': 'check_csv_charset.py not found'}, 500) return # Build command cmd = [sys.executable, str(self.charset_script), folder_path] if mode == 'convert': cmd.extend(['--convert-to', target_encoding]) if dry_run: cmd.append('--dry-run') if fast_mode: cmd.append('--fast') # Always add summary for cleaner output cmd.append('--summary-only') # Run analysis try: result = subprocess.run(cmd, capture_output=True, text=True, timeout=300) # Clean ANSI codes from output clean_output = self.clean_ansi_codes(result.stdout) response = { 'success': result.returncode == 0, 'output': clean_output, 'error_output': result.stderr if result.stderr else None, 'command': ' '.join(cmd[1:]) # Don't include python path } self.send_json_response(response) except subprocess.TimeoutExpired: self.send_json_response({'error': 'Analysis timed out (5 minutes)'}, 500) except Exception as e: self.send_json_response({'error': f'Analysis failed: {str(e)}'}, 500) except Exception as e: self.send_json_response({'error': f'Request processing failed: {str(e)}'}, 500) def send_json_response(self, data, status_code=200): """Send JSON response""" response_data = json.dumps(data).encode() self.send_response(status_code) self.send_header('Content-type', 'application/json') self.send_header('Content-length', len(response_data)) self.end_headers() self.wfile.write(response_data) def clean_ansi_codes(self, text): """Remove ANSI color codes from text""" import re ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') return ansi_escape.sub('', text) def get_html_interface(self): """Generate the HTML interface""" return ''' CSV Charset Analyzer

šŸ” CSV Charset Analyzer

Easy drag-and-drop character encoding analysis

šŸ“ Select Your CSV Folder

šŸ“– Supported Folder Structures:

Option 1: Direct CSV Folder
šŸ“ csv_folder/ ā† Select this
    šŸ“„ file1.csv
    šŸ“„ file2.csv
    šŸ“„ data.csv
Option 2: Parent with Subfolders
šŸ“ parent_folder/ ā† Select this
    šŸ“ subfolder1/
        šŸ“„ file1.csv
    šŸ“ subfolder2/
        šŸ“„ data.csv

Flexible: The tool automatically detects which structure you're using!

šŸ“‚ Click to select your CSV folder

Choose the folder containing your CSV files

Modern browsers (Chrome/Edge) will show a folder picker.
Other browsers will show a path input dialog.
Selected folder:

āš™ļø Analysis Options

What would you like to do?

Advanced Options

''' def find_free_port(): """Find a free port to run the server on""" with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s: s.bind(('', 0)) s.listen(1) port = s.getsockname()[1] return port def main(): """Main function to start the web server""" # Check if charset script exists script_dir = Path(__file__).parent charset_script = script_dir / "check_csv_charset.py" if not charset_script.exists(): print("āŒ Error: check_csv_charset.py not found!") print(f"Please make sure check_csv_charset.py is in the same folder as this script.") print(f"Looking for: {charset_script}") input("Press Enter to exit...") return # Find a free port port = find_free_port() print(f"šŸš€ Starting CSV Charset Analyzer...") print(f"šŸ“ Script location: {script_dir}") print(f"šŸŒ Starting web server on port {port}...") try: # Start server with socketserver.TCPServer(("", port), CharsetAnalyzerHandler) as httpd: server_url = f"http://localhost:{port}" print(f"āœ… Server started successfully!") print(f"šŸ”— Opening browser: {server_url}") print(f"") print(f"šŸ’” Instructions:") print(f" 1. A browser window should open automatically") print(f" 2. If not, manually open: {server_url}") print(f" 3. Drag and drop or enter your CSV folder path") print(f" 4. Click 'Start Analysis' to run") print(f"") print(f"ā¹ļø To stop: Press Ctrl+C in this window") print(f"{'='*60}") # Open browser automatically def open_browser(): time.sleep(1) # Give server time to start try: webbrowser.open(server_url) except Exception as e: print(f"Could not auto-open browser: {e}") print(f"Please manually open: {server_url}") browser_thread = threading.Thread(target=open_browser) browser_thread.daemon = True browser_thread.start() # Serve forever httpd.serve_forever() except KeyboardInterrupt: print(f"\\nā¹ļø Server stopped by user") except Exception as e: print(f"āŒ Error starting server: {e}") input("Press Enter to exit...") if __name__ == "__main__": main()