app.py

import gradio as gr
import os
import tempfile
import cv2
import numpy as np
import urllib.parse
from screencoder.main import generate_html_for_demo
from PIL import Image
import shutil
import html
import base64
from bs4 import BeautifulSoup
from pathlib import Path

# Predefined examples
examples_data = [
    [
        "screencoder/data/input/test1.png",
        "",
        "",
        "",
        "",
        "screencoder/data/input/test1.png"
    ],
    [
        "screencoder/data/input/test2.png",
        "",
        "",
        "",
        "",
        "screencoder/data/input/test2.png"
    ],
    [
        "screencoder/data/input/test3.png",
        "",
        "",
        "",
        "",
        "screencoder/data/input/test3.png"
    ],
]

TAILWIND_SCRIPT = "<script src='https://cdn.jsdelivr.net/npm/@tailwindcss/browser@4'></script>"

def image_to_data_url(image_path):
    """Convert an image file to a data URL for embedding in HTML."""
    try:
        with open(image_path, 'rb') as img_file:
            img_data = img_file.read()
            # Detect image type from file extension
            ext = os.path.splitext(image_path)[1].lower()
            mime_type = {
                '.png': 'image/png',
                '.jpg': 'image/jpeg',
                '.jpeg': 'image/jpeg',
                '.gif': 'image/gif',
                '.webp': 'image/webp'
            }.get(ext, 'image/png')
            
            encoded = base64.b64encode(img_data).decode('utf-8')
            return f'data:{mime_type};base64,{encoded}'
    except Exception as e:
        print(f"Error converting image to data URL: {e}")
        return None

def render_preview(code: str, width: int, height: int, scale: float) -> str:
    """
    Preview renderer with both width and height control for the inner canvas.
    """
    # Clean up the HTML code to remove problematic script and CSS references
    soup = BeautifulSoup(code, 'html.parser')
    
    # Remove any script tags that reference local assets or iframeResizer
    for script in soup.find_all('script'):
        src = script.get('src', '')
        if src and ('assets/' in src or 'index-' in src or 'iframeResizer' in src):
            script.decompose()
    
    # Remove any link tags that reference local CSS assets
    for link in soup.find_all('link'):
        href = link.get('href', '')
        if href and ('assets/' in href or 'index-' in href):
            link.decompose()
    
    # Also remove any inline scripts that might contain problematic content
    for script in soup.find_all('script'):
        if script.string and ('assets/' in script.string or 'index-' in script.string):
            script.decompose()
    
    # Get the cleaned HTML
    cleaned_code = str(soup)
    safe_code = html.escape(cleaned_code).replace("'", "&apos;")
    
    iframe_html = f"""
    <div style="width: 100%; max-width: 1920px; margin: 0 auto; overflow-x: auto; overflow-y: hidden;">
        <div style="
            width: 1920px;
            height: 1000px;
            margin: 0 auto;
            display: flex;
            justify-content: center;
            align-items: center;
            border: 1px solid #ddd;
            overflow: hidden;
            background: #f9fafb;
            position: relative;
            box-shadow: 0 4px 12px rgba(0,0,0,0.1);">
            <div style="
                width: {width}px;
                height: {height}px;
                transform: scale({scale});
                transform-origin: top left;
                border: none;
                position: relative;">
                <iframe
                    style="width: 100%; height: 100%; border: none; display: block;"
                    srcdoc='{safe_code}'>
                </iframe>
            </div>
        </div>
    </div>
    """
    return iframe_html

def process_and_generate(image_np, image_path_from_state, sidebar_prompt, header_prompt, navigation_prompt, main_content_prompt):
    """
    Main processing pipeline: takes an image, generates code, creates a downloadable
    package, and returns the initial preview and code outputs.
    """
    final_image_path = ""
    is_temp_file = False
    if image_path_from_state:
        final_image_path = image_path_from_state
    elif image_np is not None:
        is_temp_file = True
        with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as tmp:
            cv2.imwrite(tmp.name, cv2.cvtColor(image_np, cv2.COLOR_RGB2BGR))
            final_image_path = tmp.name
    else:
        return "No image provided.", "Please upload or select an image.", gr.update(visible=False), None

    instructions = {
        "sidebar": sidebar_prompt, "header": header_prompt,
        "navigation": navigation_prompt, "main content": main_content_prompt
    }
    html_content, run_id = generate_html_for_demo(final_image_path, instructions)

    if not run_id: # Handle potential errors from the generator
        return "Generation failed.", f"Error: {html_content}", gr.update(visible=False), None

    # Rewrite image paths to be absolute for Gradio serving
    # HF Spaces: Use Path objects for robust path handling
    base_dir = Path(__file__).parent.resolve()
    soup = BeautifulSoup(html_content, 'html.parser')
    
    print(f"Processing HTML for run_id: {run_id}")
    
    # Clean up problematic script and CSS references
    for script in soup.find_all('script'):
        src = script.get('src', '')
        if src and ('assets/' in src or 'index-' in src or 'iframeResizer' in src):
            print(f"Removing problematic script: {src}")
            script.decompose()
    
    for link in soup.find_all('link'):
        href = link.get('href', '')
        if href and ('assets/' in href or 'index-' in href):
            print(f"Removing problematic CSS link: {href}")
            link.decompose()
    
    # Also remove any inline scripts that might contain problematic content
    for script in soup.find_all('script'):
        if script.string and ('assets/' in script.string or 'index-' in script.string):
            print(f"Removing problematic inline script")
            script.decompose()
    
    for img in soup.find_all('img'):
        if img.get('src') and not img['src'].startswith(('http', 'data:')):
            original_src = img['src']
            print(f"Processing image: {original_src}")
            
            # In HF Spaces, paths can be tricky. We'll rely on the fact
            # that the image replacer creates a predictable structure.
            img_path = base_dir / 'screencoder' / 'data' / 'output' / run_id / original_src
            
            if img_path.exists():
                print(f"Found image at: {img_path}")
                # Convert to base64 data URL for better iframe compatibility
                data_url = image_to_data_url(str(img_path))
                if data_url:
                    print(f"Converted to data URL: {original_src}")
                    img['src'] = data_url
                else:
                    # Fallback to Gradio file path (might not work in all Spaces configs)
                    img['src'] = f'/file={str(img_path)}'
            else:
                print(f"Image not found at expected path: {img_path}")
                # Keep original path as fallback
                img['src'] = original_src
    
    html_content = str(soup)

    output_dir = base_dir / 'screencoder' / 'data' / 'output' / run_id
    packages_dir = base_dir / 'screencoder' / 'data' / 'packages'
    packages_dir.mkdir(exist_ok=True)
    
    shutil.make_archive(str(packages_dir / run_id), 'zip', str(output_dir))
    package_path = packages_dir / f'{run_id}.zip'
    package_url = f'/file={str(package_path)}'
    
    if is_temp_file:
        os.unlink(final_image_path)
    
    initial_preview = render_preview(html_content, 1280, 600, 0.7)

    return initial_preview, html_content, gr.update(value=package_url, visible=True)

with gr.Blocks(head=TAILWIND_SCRIPT, theme=gr.themes.Soft(primary_hue="blue", secondary_hue="sky")) as demo:
    gr.Markdown("# ScreenCoder: Screenshot to Code")
    with gr.Row():
        with gr.Column(scale=1):
            gr.Markdown("### Step 1: Provide an Image")
            active_image = gr.Image(type="numpy", height=400, value=examples_data[0][0])
            upload_button = gr.UploadButton("Click to Upload", file_types=["image"], variant="primary")
            
            gr.Markdown("### Step 2: Write Prompts (Optional)")
            with gr.Accordion("Component-specific Prompts", open=False):
                sidebar_prompt = gr.Textbox(label="Sidebar", placeholder="Instructions for the sidebar...")
                header_prompt = gr.Textbox(label="Header", placeholder="Instructions for the header...")
                navigation_prompt = gr.Textbox(label="Navigation", placeholder="Instructions for the navigation...")
                main_content_prompt = gr.Textbox(label="Main Content", placeholder="Instructions for the main content...")

            generate_btn = gr.Button("Generate HTML", variant="primary")

        with gr.Column(scale=2):
            with gr.Tabs():
                with gr.TabItem("Preview"):
                    with gr.Row():
                        scale_slider = gr.Slider(0.2, 1.5, value=0.7, step=0.05, label="Zoom")
                        width_slider = gr.Slider(400, 1920, value=1280, step=100, label="Canvas Width (px)")
                        height_slider = gr.Slider(300, 1080, value=600, step=50, label="Canvas Height (px)")
                    
                    html_preview = gr.HTML(label="Rendered HTML", show_label=False)

                with gr.TabItem("Code"):
                    html_code_output = gr.Code(label="Generated HTML", language="html")

            download_button = gr.Button("⬇️ Download Package", visible=False, variant="secondary")

    gr.Examples(
        examples=examples_data,
        fn=lambda *args: args[0],
        inputs=[gr.State(examples_data[0][0])],
        outputs=[active_image],
        cache_examples=False,
    )
    
    active_image_path_state = gr.State(examples_data[0][5])

    def handle_example_click(img_path):
        return img_path, img_path
    
    demo.load(
        lambda: (examples_data[0][0], examples_data[0][5]), None, [active_image, active_image_path_state]
    )
    
    def handle_upload(uploaded_image_np):
        return uploaded_image_np, None, gr.update(visible=False)
    
    upload_button.upload(handle_upload, upload_button, [active_image, active_image_path_state, download_button])

    generate_btn.click(
        process_and_generate,
        [active_image, active_image_path_state, sidebar_prompt, header_prompt, navigation_prompt, main_content_prompt],
        [html_preview, html_code_output, download_button],
        show_progress="full"
    )

    preview_controls = [scale_slider, width_slider, height_slider]
    for control in preview_controls:
        control.change(
            render_preview,
            [html_code_output, width_slider, height_slider, scale_slider],
            html_preview,
            show_progress=True
        )

    download_button.click(None, download_button, None, js= \
        "(url) => { const link = document.createElement('a'); link.href = url; link.download = ''; document.body.appendChild(link); link.click(); document.body.removeChild(link); }")

base_dir = Path(__file__).parent.resolve()
allowed_paths = [
    str(base_dir),
    str(base_dir / 'screencoder' / 'data' / 'output'),
    str(base_dir / 'screencoder' / 'data' / 'packages')
]

# Add all example file paths to allowed_paths to ensure they are accessible
for example in examples_data:
    # HF Spaces: Ensure the path is absolute by joining with base_dir
    # The example path is relative, so we join it with the base_dir to make it absolute.
    example_abs_path = (base_dir / example[0]).resolve()
    example_dir = example_abs_path.parent
    if str(example_dir) not in allowed_paths:
        allowed_paths.append(str(example_dir))

print("Allowed paths for file serving:")
for path in allowed_paths:
    print(f"  - {path}")

if __name__ == "__main__":
    demo.launch(
        allowed_paths=allowed_paths,
        show_error=True,
        quiet=False,
        favicon_path=None,
        prevent_thread_lock=True
    )