app.py

import gradio as gr
import subprocess
import os
import base64
import uuid

def process_conversion(base64_input):
    # Setup paths in /tmp for Hugging Face environment
    unique_id = uuid.uuid4().hex
    temp_docx = f"/tmp/input_{unique_id}.docx"
    output_dir = "/tmp/outputs"
    
    if not os.path.exists(output_dir):
        os.makedirs(output_dir)

    try:
        # 1. Handle Base64 Input
        if not base64_input or not base64_input.strip():
            return None, "Error: No Base64 data provided."

        # Clean base64 string (remove data header if present)
        if "," in base64_input:
            base64_input = base64_input.split(",")[1]
        
        # Write the binary data to a temporary .docx file
        with open(temp_docx, "wb") as f:
            f.write(base64.b64decode(base64_input))
        
        # 2. Run LibreOffice Conversion
        result = subprocess.run([
            'libreoffice', '--headless', '--convert-to', 'pdf', 
            '--outdir', output_dir, temp_docx
        ], capture_output=True, text=True, timeout=30)

        if result.returncode != 0:
            return None, f"LibreOffice Error: {result.stderr}"

        # 3. Locate and Read the Resulting PDF
        pdf_filename = f"input_{unique_id}.pdf"
        pdf_path = os.path.join(output_dir, pdf_filename)

        if os.path.exists(pdf_path):
            with open(pdf_path, "rb") as f:
                pdf_bytes = f.read()
                pdf_base64 = base64.b64encode(pdf_bytes).decode('utf-8')
            
            # Cleanup the temporary docx file
            if os.path.exists(temp_docx):
                os.remove(temp_docx)
                
            # Return File object for download and the Base64 string for API use
            return gr.FileData(path=pdf_path, mime_type="application/pdf"), pdf_base64
        
        return None, "Error: Conversion failed - PDF not generated."

    except Exception as e:
        return None, f"System Error: {str(e)}"

# UI Layout
with gr.Blocks() as demo:
    gr.Markdown("# 📄 DOCX to PDF (Base64 Only)")
    gr.Markdown("Paste the Base64 string of a binary `.docx` file to convert it to `.pdf`.")
    
    with gr.Row():
        with gr.Column():
            text_in = gr.Textbox(
                label="Paste Base64 DOCX", 
                placeholder="Input base64 string here...", 
                lines=8
            )
            submit_btn = gr.Button("Convert to PDF", variant="primary")
            
        with gr.Column():
            file_out = gr.File(label="Download PDF File")
            # Removed 'show_copy_button' to fix the TypeError
            text_out = gr.Textbox(label="Result Base64 PDF", lines=8)

    submit_btn.click(
        fn=process_conversion,
        inputs=[text_in],
        outputs=[file_out, text_out]
    )

if __name__ == "__main__":
    demo.launch()