Spaces:
Running
Running
| from flask import Flask, render_template, request, jsonify | |
| import requests | |
| import base64 | |
| import markdown | |
| from bs4 import BeautifulSoup | |
| import os | |
| import mimetypes | |
| import json | |
| app = Flask(__name__) | |
| GITHUB_API = "https://api.github.com/repos/" | |
| def get_repo_contents(repo_url): | |
| """Extract contents from GitHub repo URL""" | |
| try: | |
| parts = repo_url.rstrip('/').split('/') | |
| owner, repo = parts[-2], parts[-1] | |
| api_url = f"{GITHUB_API}{owner}/{repo}/contents" | |
| response = requests.get(api_url) | |
| response.raise_for_status() | |
| return owner, repo, response.json() | |
| except Exception as e: | |
| return None, None, str(e) | |
| def process_file_content(file_info, owner, repo): | |
| """Process individual file content""" | |
| content = "" | |
| file_path = file_info['path'] | |
| if file_info['type'] == 'file': | |
| file_url = f"{GITHUB_API}{owner}/{repo}/contents/{file_path}" | |
| file_response = requests.get(file_url) | |
| file_data = file_response.json() | |
| if 'content' in file_data: | |
| file_extension = file_path.split('.')[-1] if '.' in file_path else '' | |
| mime_type, _ = mimetypes.guess_type(file_path) | |
| is_text = (mime_type and mime_type.startswith('text')) or file_extension in ['py', 'md', 'txt', 'js', 'html', 'css', 'json'] | |
| if is_text: | |
| try: | |
| decoded_content = base64.b64decode(file_data['content']).decode('utf-8') | |
| # Special handling for JSON files | |
| if file_extension == 'json': | |
| try: | |
| json_data = json.loads(decoded_content) | |
| formatted_json = json.dumps(json_data, indent=2) | |
| content = f"### File: {file_path}\n```json\n{formatted_json}\n```\n\n" | |
| except json.JSONDecodeError: | |
| content = f"### File: {file_path}\n```json\n{decoded_content}\n```\n[Note: Invalid JSON format]\n\n" | |
| else: | |
| content = f"### File: {file_path}\n```{(file_extension if file_extension else 'text')}\n{decoded_content}\n```\n\n" | |
| except UnicodeDecodeError: | |
| content = f"### File: {file_path}\n[Text content could not be decoded - possibly corrupted or non-UTF-8 text]\n\n" | |
| else: | |
| content = f"### File: {file_path}\n[Binary file - {file_data['size']} bytes]\n\n" | |
| return content | |
| def create_markdown_document(repo_url): | |
| """Create markdown document from repo contents""" | |
| owner, repo, contents = get_repo_contents(repo_url) | |
| if isinstance(contents, str): | |
| return f"Error: {contents}" | |
| markdown_content = f"# Repository: {owner}/{repo}\n\n" | |
| markdown_content += "Below are the contents of all files in the repository:\n\n" | |
| for item in contents: | |
| markdown_content += process_file_content(item, owner, repo) | |
| return markdown_content | |
| def index(): | |
| return render_template('index.html') | |
| def process_repo(): | |
| repo_url = request.json.get('repo_url') | |
| if not repo_url: | |
| return jsonify({'error': 'Please provide a repository URL'}), 400 | |
| markdown_content = create_markdown_document(repo_url) | |
| html_content = markdown.markdown(markdown_content) | |
| return jsonify({ | |
| 'markdown': markdown_content, | |
| 'html': html_content | |
| }) | |
| html_template = """ | |
| <!DOCTYPE html> | |
| <html> | |
| <head> | |
| <title>GitHub Repo to Markdown</title> | |
| <style> | |
| body { | |
| font-family: Arial, sans-serif; | |
| margin: 20px; | |
| max-width: 1200px; | |
| margin: 0 auto; | |
| } | |
| .container { | |
| padding: 20px; | |
| } | |
| textarea { | |
| width: 100%; | |
| height: 400px; | |
| margin-top: 20px; | |
| font-family: monospace; | |
| } | |
| button { | |
| padding: 10px 20px; | |
| background-color: #4CAF50; | |
| color: white; | |
| border: none; | |
| cursor: pointer; | |
| } | |
| button:hover { | |
| background-color: #45a049; | |
| } | |
| #output { | |
| margin-top: 20px; | |
| border: 1px solid #ddd; | |
| padding: 20px; | |
| background-color: #f9f9f9; | |
| } | |
| .spinner { | |
| display: none; | |
| border: 4px solid #f3f3f3; | |
| border-top: 4px solid #3498db; | |
| border-radius: 50%; | |
| width: 30px; | |
| height: 30px; | |
| animation: spin 1s linear infinite; | |
| margin: 20px auto; | |
| } | |
| @keyframes spin { | |
| 0% { transform: rotate(0deg); } | |
| 100% { transform: rotate(360deg); } | |
| } | |
| </style> | |
| </head> | |
| <body> | |
| <div class="container"> | |
| <h1>GitHub Repository to Markdown Converter</h1> | |
| <p>Enter a GitHub repository URL (e.g., https://github.com/username/repository)</p> | |
| <input type="text" id="repoUrl" style="width: 100%; padding: 8px;" placeholder="Enter GitHub repository URL"> | |
| <button onclick="processRepo()">Convert to Markdown</button> | |
| <div id="spinner" class="spinner"></div> | |
| <h2>Markdown Output:</h2> | |
| <textarea id="markdownOutput" readonly></textarea> | |
| <h2>Preview:</h2> | |
| <div id="output"></div> | |
| </div> | |
| <script> | |
| async function processRepo() { | |
| const repoUrl = document.getElementById('repoUrl').value; | |
| const spinner = document.getElementById('spinner'); | |
| const button = document.querySelector('button'); | |
| // Show spinner, disable button | |
| spinner.style.display = 'block'; | |
| button.disabled = true; | |
| try { | |
| const response = await fetch('/process', { | |
| method: 'POST', | |
| headers: { | |
| 'Content-Type': 'application/json', | |
| }, | |
| body: JSON.stringify({ repo_url: repoUrl }) | |
| }); | |
| const data = await response.json(); | |
| if (data.error) { | |
| alert(data.error); | |
| return; | |
| } | |
| document.getElementById('markdownOutput').value = data.markdown; | |
| document.getElementById('output').innerHTML = data.html; | |
| } catch (error) { | |
| alert('An error occurred: ' + error.message); | |
| } finally { | |
| // Hide spinner, enable button | |
| spinner.style.display = 'none'; | |
| button.disabled = false; | |
| } | |
| } | |
| </script> | |
| </body> | |
| </html> | |
| """ | |
| if not os.path.exists('templates'): | |
| os.makedirs('templates') | |
| with open('templates/index.html', 'w') as f: | |
| f.write(html_template) | |
| if __name__ == '__main__': | |
| app.run(host="0.0.0.0", port=7860, debug=True) |