Spaces:
Sleeping
Sleeping
| from flask import Flask, request, render_template_string, send_file, jsonify | |
| import markdown | |
| import os | |
| import traceback | |
| from io import BytesIO | |
| import re | |
| import base64 | |
| from pygments.lexers import get_lexer_by_name | |
| from pygments.formatters import HtmlFormatter | |
| from pygments.styles import get_all_styles | |
| # New import for the headless browser | |
| os.system("playwright install") | |
| from playwright.sync_api import sync_playwright | |
| app = Flask(__name__) | |
| # Corrected line to use the universally writable /tmp directory | |
| TEMP_DIR = "/tmp/markdown_temp" | |
| os.makedirs(TEMP_DIR, exist_ok=True) | |
| def parse_repo2markdown(text): | |
| components = [] | |
| pattern = re.compile(r'### File: (.*?)\n([\s\S]*?)(?=\n### File:|\Z)', re.MULTILINE) | |
| first_match = pattern.search(text) | |
| if first_match: | |
| intro_text = text[:first_match.start()].strip() | |
| if intro_text: | |
| components.append({'type': 'intro', 'filename': 'Introduction', 'content': intro_text, 'is_code_block': False, 'language': ''}) | |
| for match in pattern.finditer(text): | |
| filename = match.group(1).strip() | |
| raw_content = match.group(2).strip() | |
| code_match = re.search(r'^```(\w*)\s*\n([\s\S]*?)\s*```$', raw_content, re.DOTALL) | |
| if code_match: | |
| components.append({'type': 'file', 'filename': filename, 'content': code_match.group(2).strip(), 'is_code_block': True, 'language': code_match.group(1)}) | |
| else: | |
| components.append({'type': 'file', 'filename': filename, 'content': raw_content, 'is_code_block': False, 'language': ''}) | |
| return components | |
| def parse_standard_readme(text): | |
| components = [] | |
| parts = re.split(r'^(## .*?)$', text, flags=re.MULTILINE) | |
| intro_content = parts[0].strip() | |
| if intro_content: | |
| components.append({'type': 'intro', 'filename': 'Introduction', 'content': intro_content}) | |
| for i in range(1, len(parts), 2): | |
| components.append({'type': 'section', 'filename': parts[i].replace('##', '').strip(), 'content': parts[i+1].strip()}) | |
| return components | |
| def parse_changelog(text): | |
| components = [] | |
| parts = re.split(r'^(## \[\d+\.\d+\.\d+.*?\].*?)$', text, flags=re.MULTILINE) | |
| intro_content = parts[0].strip() | |
| if intro_content: | |
| components.append({'type': 'intro', 'filename': 'Changelog Header', 'content': intro_content}) | |
| for i in range(1, len(parts), 2): | |
| components.append({'type': 'version', 'filename': parts[i].replace('##', '').strip(), 'content': parts[i+1].strip()}) | |
| return components | |
| def parse_endpoint(): | |
| text = request.form.get('markdown_text', '') | |
| if 'markdown_file' in request.files and request.files['markdown_file'].filename != '': | |
| text = request.files['markdown_file'].read().decode('utf-8') | |
| if not text: return jsonify({'error': 'No text or file provided.'}), 400 | |
| try: | |
| if "## File Structure" in text and "### File:" in text: | |
| format_name, components = "Repo2Markdown", parse_standard_readme(text) | |
| elif re.search(r'^## \[\d+\.\d+\.\d+.*?\].*?$', text, flags=re.MULTILINE): | |
| format_name, components = "Changelog", parse_changelog(text) | |
| elif text.strip().startswith("#") and re.search(r'^## ', text, flags=re.MULTILINE): | |
| format_name, components = "Standard README", parse_standard_readme(text) | |
| else: | |
| format_name, components = "Unknown", [{'type': 'text', 'filename': 'Full Text', 'content': text}] | |
| return jsonify({'format': format_name, 'components': components}) | |
| except Exception as e: | |
| return jsonify({'error': f'Failed to parse: {e}'}), 500 | |
| def build_full_html(markdown_text, styles, include_fontawesome): | |
| wrapper_id = "#output-wrapper" | |
| font_family = styles.get('font_family', "'Arial', sans-serif") | |
| google_font_name = font_family.split(',')[0].strip("'\"") | |
| google_font_link = "" | |
| if " " in google_font_name and google_font_name not in ["Times New Roman", "Courier New"]: | |
| google_font_link = f'<link href="https://fonts.googleapis.com/css2?family={google_font_name.replace(" ", "+")}:wght@400;700&display=swap" rel="stylesheet">' | |
| highlight_theme = styles.get('highlight_theme', 'default') | |
| pygments_css = "" | |
| if highlight_theme != 'none': | |
| formatter = HtmlFormatter(style=highlight_theme, cssclass="codehilite") | |
| pygments_css = formatter.get_style_defs(f' {wrapper_id}') | |
| scoped_css = f""" | |
| {wrapper_id} {{ | |
| font-family: {font_family}; font-size: {styles.get('font_size', '16')}px; | |
| color: {styles.get('text_color', '#333')}; background-color: {styles.get('background_color', '#fff')}; | |
| }} | |
| {wrapper_id} table {{ border-collapse: collapse; width: 100%; }} | |
| {wrapper_id} th, {wrapper_id} td {{ border: 1px solid #ddd; padding: 8px; text-align: left; }} | |
| {wrapper_id} th {{ background-color: #f2f2f2; }} | |
| {wrapper_id} img {{ max-width: 100%; height: auto; }} | |
| {wrapper_id} pre {{ padding: {styles.get('code_padding', '15')}px; border-radius: 5px; white-space: pre-wrap; word-wrap: break-word; }} | |
| {wrapper_id} h1, {wrapper_id} h2, {wrapper_id} h3 {{ border-bottom: 1px solid #eee; padding-bottom: 5px; margin-top: 1.5em; }} | |
| {wrapper_id} :not(pre) > code {{ font-family: 'Courier New', monospace; background-color: #eef; padding: .2em .4em; border-radius: 3px; }} | |
| {pygments_css} {styles.get('custom_css', '')} | |
| """ | |
| md_extensions = ['fenced_code', 'tables', 'codehilite'] | |
| html_content = markdown.markdown(markdown_text, extensions=md_extensions, extension_configs={'codehilite': {'css_class': 'codehilite'}}) | |
| final_html_body = f'<div id="output-wrapper">{html_content}</div>' | |
| fontawesome_link = '<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.5.1/css/all.min.css">' if include_fontawesome else "" | |
| full_html = f"""<!DOCTYPE html> | |
| <html><head><meta charset="UTF-8">{google_font_link}{fontawesome_link}<style> | |
| body {{ background-color: transparent; margin: 0; padding: 0; }} | |
| #output-wrapper {{ background-color: {styles.get('background_color', '#fff')}; padding: 25px; display: inline-block;}} | |
| {scoped_css} | |
| </style></head><body>{final_html_body}</body></html>""" | |
| return full_html | |
| def convert_endpoint(): | |
| data = request.json | |
| try: | |
| full_html = build_full_html( | |
| markdown_text=data.get('markdown_text', ''), | |
| styles=data.get('styles', {}), | |
| include_fontawesome=data.get('include_fontawesome', False) | |
| ) | |
| # --- REPLACEMENT FOR IMGKIT --- | |
| with sync_playwright() as p: | |
| browser = p.chromium.launch() | |
| page = browser.new_page() | |
| page.set_content(full_html) | |
| # Locate the specific element we want to capture | |
| element = page.locator("#output-wrapper") | |
| # Take a screenshot of just that element | |
| png_bytes = element.screenshot(type="png") | |
| browser.close() | |
| # --- END REPLACEMENT --- | |
| if data.get('download', False): | |
| download_type = data.get('download_type', 'png') | |
| if download_type == 'html': | |
| return send_file(BytesIO(full_html.encode("utf-8")), as_attachment=True, download_name="output.html", mimetype="text/html") | |
| else: | |
| return send_file(BytesIO(png_bytes), as_attachment=True, download_name="output.png", mimetype="image/png") | |
| else: | |
| png_base64 = base64.b64encode(png_bytes).decode('utf-8') | |
| return jsonify({'preview_html': full_html, 'preview_png_base64': png_base64}) | |
| except Exception as e: | |
| traceback.print_exc() | |
| return jsonify({'error': f'Failed to convert content: {str(e)}'}), 500 | |
| def index(): | |
| highlight_styles = sorted(list(get_all_styles())) | |
| return render_template_string(""" | |
| <!DOCTYPE html> | |
| <html lang="en"> | |
| <head> | |
| <meta charset="UTF-8"> | |
| <title>Intelligent Markdown Converter</title> | |
| <style> | |
| body { font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; max-width: 1200px; margin: 0 auto; padding: 20px; background-color: #f9f9f9; } | |
| h1, h2 { text-align: center; color: #333; } | |
| form { background: #fff; padding: 20px; border-radius: 8px; box-shadow: 0 2px 4px rgba(0,0,0,0.1); margin-bottom: 20px; } | |
| textarea { width: 100%; box-sizing: border-box; border: 1px solid #ccc; border-radius: 4px; padding: 10px; font-family: monospace; } | |
| fieldset { border: 1px solid #ddd; padding: 15px; border-radius: 5px; margin-top: 20px; } | |
| legend { font-weight: bold; color: #555; padding: 0 10px; } | |
| select, input[type="number"], input[type="color"] { width: 100%; padding: 8px; border: 1px solid #ccc; border-radius: 4px; box-sizing: border-box;} | |
| button { padding: 10px 15px; border: none; border-radius: 4px; cursor: pointer; font-size: 14px; transition: background-color 0.2s; } | |
| .action-btn { background-color: #007BFF; color: white; font-size: 16px; padding: 12px 20px;} | |
| .generate-btn { background-color: #5a32a3; color: white; font-size: 16px; padding: 12px 20px; } | |
| .download-btn { background-color: #28a745; color: white; display: none; } | |
| .main-actions { display: flex; flex-wrap: wrap; gap: 15px; align-items: center; margin-top: 20px; } | |
| .preview-header { display: flex; justify-content: space-between; align-items: center; border-bottom: 2px solid #eee; padding-bottom: 10px; margin-bottom: 15px; } | |
| .preview-container { border: 1px solid #ddd; padding: 10px; margin-top: 20px; background: #fff; box-shadow: 0 2px 4px rgba(0,0,0,0.05); min-height: 100px; } | |
| .preview-container img { max-width: 100%; } | |
| .error { color: #D8000C; background-color: #FFD2D2; padding: 10px; border-radius: 5px; margin-top: 15px; display: none; } | |
| .info { color: #00529B; background-color: #BDE5F8; padding: 10px; border-radius: 5px; margin: 10px 0; display: none;} | |
| .style-grid { display: grid; grid-template-columns: repeat(auto-fill, minmax(220px, 1fr)); gap: 15px; align-items: end; } | |
| .component-grid { display: grid; grid-template-columns: repeat(auto-fill, minmax(300px, 1fr)); gap: 15px; } | |
| .component-container { border: 1px solid #e0e0e0; border-radius: 5px; background: #fafafa; } | |
| .component-header { background: #f1f1f1; padding: 8px 12px; border-bottom: 1px solid #e0e0e0; display: flex; align-items: center; gap: 10px; } | |
| .component-content textarea { height: 150px; } | |
| .selection-controls { margin: 15px 0; display: flex; gap: 10px; } | |
| </style> | |
| </head> | |
| <body> | |
| <h1>Intelligent Markdown Converter</h1> | |
| <form id="main-form" onsubmit="return false;"> | |
| <fieldset><legend>1. Load Content</legend><div id="info-box" class="info"></div><textarea id="markdown-text-input" name="markdown_text" rows="8"></textarea><div style="margin-top: 10px; display: flex; align-items: center; gap: 10px;"><label for="markdown-file-input">Or upload a file:</label><input type="file" id="markdown-file-input" name="markdown_file" accept=".md,.txt,text/markdown"></div><div style="margin-top: 15px;"><button type="button" id="load-btn" class="action-btn">Load & Analyze</button></div></fieldset> | |
| <fieldset id="components-fieldset" style="display:none;"><legend>2. Select Components</legend><div class="selection-controls"><button type="button" onclick="toggleAllComponents(true)">Select All</button><button type="button" onclick="toggleAllComponents(false)">Deselect All</button></div><div id="components-container" class="component-grid"></div></fieldset> | |
| <fieldset><legend>3. Configure Styles</legend><div class="style-grid"><div><label>Font Family:</label><select id="font_family"><optgroup label="Sans-Serif"><option value="'Arial', sans-serif">Arial</option><option value="'Roboto', sans-serif">Roboto</option></optgroup><optgroup label="Serif"><option value="'Times New Roman', serif">Times New Roman</option><option value="'Georgia', serif">Georgia</option></optgroup></select></div><div><label>Font Size (px):</label><input type="number" id="font_size" value="16"></div><div><label>Highlight Theme:</label><select id="highlight_theme"><option value="none">None</option>{% for style in highlight_styles %}<option value="{{ style }}" {% if style == 'default' %}selected{% endif %}>{{ style }}</option>{% endfor %}</select></div><div><label>Text Color:</label><input type="color" id="text_color" value="#333333"></div><div><label>Background Color:</label><input type="color" id="background_color" value="#ffffff"></div><div><label>Code Padding (px):</label><input type="number" id="code_padding" value="15"></div></div><div><input type="checkbox" id="include_fontawesome"><label for="include_fontawesome">Include Font Awesome</label></div><div><label for="custom_css">Custom CSS:</label><textarea id="custom_css" rows="3"></textarea></div></fieldset> | |
| <div class="main-actions"><button type="button" id="generate-btn" class="generate-btn">Generate Preview</button></div> | |
| </form> | |
| <div id="error-box" class="error"></div> | |
| <div id="preview-section" style="display:none;"> | |
| <h2>Preview</h2> | |
| <div class="preview-header"> | |
| <h3>HTML Output</h3> | |
| <button type="button" id="download-html-btn" class="download-btn">Download HTML</button> | |
| </div> | |
| <div id="html-preview-container" class="preview-container"></div> | |
| <div class="preview-header" style="margin-top: 30px;"> | |
| <h3>PNG Output</h3> | |
| <button type="button" id="download-png-btn" class="download-btn">Download PNG</button> | |
| </div> | |
| <div id="png-preview-container" class="preview-container"></div> | |
| </div> | |
| <script> | |
| const loadBtn = document.getElementById('load-btn'), generateBtn = document.getElementById('generate-btn'), | |
| downloadHtmlBtn = document.getElementById('download-html-btn'), downloadPngBtn = document.getElementById('download-png-btn'), | |
| markdownTextInput = document.getElementById('markdown-text-input'), markdownFileInput = document.getElementById('markdown-file-input'), | |
| componentsFieldset = document.getElementById('components-fieldset'), componentsContainer = document.getElementById('components-container'), | |
| previewSection = document.getElementById('preview-section'), htmlPreviewContainer = document.getElementById('html-preview-container'), | |
| pngPreviewContainer = document.getElementById('png-preview-container'), errorBox = document.getElementById('error-box'), | |
| infoBox = document.getElementById('info-box'); | |
| function toggleAllComponents(checked) { componentsContainer.querySelectorAll('.component-checkbox').forEach(cb => cb.checked = checked); } | |
| function displayError(message) { errorBox.textContent = message; errorBox.style.display = 'block'; previewSection.style.display = 'none'; } | |
| function buildPayload() { | |
| let finalMarkdown = ""; | |
| if (componentsFieldset.style.display === 'block') { | |
| const parts = []; | |
| componentsContainer.querySelectorAll('.component-container').forEach(div => { | |
| if (div.querySelector('.component-checkbox').checked) { parts.push(div.dataset.reconstructed || div.dataset.content); } | |
| }); | |
| finalMarkdown = parts.join('\\n\\n---\\n\\n'); | |
| } else { finalMarkdown = markdownTextInput.value; } | |
| return { | |
| markdown_text: finalMarkdown, | |
| styles: { | |
| font_family: document.getElementById('font_family').value, font_size: document.getElementById('font_size').value, | |
| text_color: document.getElementById('text_color').value, background_color: document.getElementById('background_color').value, | |
| code_padding: document.getElementById('code_padding').value, highlight_theme: document.getElementById('highlight_theme').value, | |
| custom_css: document.getElementById('custom_css').value | |
| }, | |
| include_fontawesome: document.getElementById('include_fontawesome').checked, | |
| }; | |
| } | |
| generateBtn.addEventListener('click', async () => { | |
| generateBtn.textContent = 'Generating...'; generateBtn.disabled = true; errorBox.style.display = 'none'; | |
| const payload = buildPayload(); | |
| payload.download = false; | |
| try { | |
| const response = await fetch('/convert', { method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify(payload) }); | |
| const result = await response.json(); | |
| if (!response.ok) throw new Error(result.error || `Server error ${response.status}`); | |
| previewSection.style.display = 'block'; | |
| htmlPreviewContainer.innerHTML = result.preview_html; | |
| pngPreviewContainer.innerHTML = `<img src="data:image/png;base64,${result.preview_png_base64}" alt="PNG Preview">`; | |
| downloadHtmlBtn.style.display = 'inline-block'; downloadPngBtn.style.display = 'inline-block'; | |
| } catch (err) { displayError('Error generating preview: ' + err.message); } | |
| finally { generateBtn.textContent = 'Generate Preview'; generateBtn.disabled = false; } | |
| }); | |
| async function handleDownload(fileType) { | |
| const button = fileType === 'html' ? downloadHtmlBtn : downloadPngBtn; | |
| button.textContent = 'Preparing...'; const payload = buildPayload(); | |
| payload.download = true; payload.download_type = fileType; | |
| try { | |
| const response = await fetch('/convert', { method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify(payload) }); | |
| if (!response.ok) throw new Error(`Download failed: ${response.statusText}`); | |
| const blob = await response.blob(); const url = window.URL.createObjectURL(blob); | |
| const a = document.createElement('a'); a.style.display = 'none'; a.href = url; a.download = 'output.' + fileType; | |
| document.body.appendChild(a); a.click(); window.URL.revokeObjectURL(url); a.remove(); | |
| } catch (err) { displayError('Error preparing download: ' + err.message); } | |
| finally { button.textContent = `Download ${fileType.toUpperCase()}`; } | |
| } | |
| downloadHtmlBtn.addEventListener('click', () => handleDownload('html')); | |
| downloadPngBtn.addEventListener('click', () => handleDownload('png')); | |
| loadBtn.addEventListener('click', async () => { | |
| loadBtn.textContent = 'Loading...'; loadBtn.disabled = true; errorBox.style.display = 'none'; infoBox.style.display = 'none'; componentsFieldset.style.display = 'none'; componentsContainer.innerHTML = ''; | |
| const formData = new FormData(); | |
| if (markdownFileInput.files.length > 0) { formData.append('markdown_file', markdownFileInput.files[0]); } else { formData.append('markdown_text', markdownTextInput.value); } | |
| try { | |
| const response = await fetch('/parse', { method: 'POST', body: formData }); | |
| const result = await response.json(); | |
| if (!response.ok) throw new Error(result.error || `Server error`); | |
| infoBox.innerHTML = `Detected Format: <strong>${result.format}</strong>`; infoBox.style.display = 'block'; | |
| if (result.format !== 'Unknown') { | |
| componentsFieldset.style.display = 'block'; | |
| result.components.forEach((comp, index) => { | |
| const div = document.createElement('div'); div.className = 'component-container'; div.dataset.type = comp.type; div.dataset.filename = comp.filename; div.dataset.content = comp.content; | |
| let reconstructedContent = comp.content; | |
| if (comp.is_code_block) { div.dataset.isCodeBlock = 'true'; div.dataset.language = comp.language || ''; reconstructedContent = "```" + (comp.language || '') + "\\n" + comp.content + "\\n```"; } | |
| if (comp.type === 'section') div.dataset.reconstructed = `## ${comp.filename}\\n${comp.content}`; if (comp.type === 'version') div.dataset.reconstructed = `## ${comp.filename}\\n${comp.content}`; | |
| div.innerHTML = `<div class="component-header"><input type="checkbox" id="comp-check-${index}" class="component-checkbox" checked><label for="comp-check-${index}">${comp.filename}</label></div><div class="component-content"><textarea readonly>${comp.content}</textarea></div>`; | |
| componentsContainer.appendChild(div); | |
| }); | |
| } | |
| if(markdownFileInput.files.length > 0) { markdownTextInput.value = await markdownFileInput.files[0].text(); } | |
| } catch (err) { displayError('Error parsing content: ' + err.message); } finally { loadBtn.textContent = 'Load & Analyze'; loadBtn.disabled = false; } | |
| }); | |
| </script> | |
| </body> | |
| </html> | |
| """, highlight_styles=highlight_styles) | |
| if __name__ == "__main__": | |
| app.run(host="0.0.0.0", port=7860) |