app.py

import os, shutil, subprocess, tempfile, json
import gradio as gr

def process_upload(pdf_file, output_format):
    # Create temp output directory
    out_dir = tempfile.mkdtemp()
    # Run Marker CLI: outputs files marker-0000.md or marker-0000.json
    fmt = "markdown" if output_format=="markdown" else "json"
    cmd = [
        "marker_single",
        pdf_file.name,
        "--output_format", fmt,
        "--output_dir", out_dir,
        "--paginate_output"  # page separators
    ]
    subprocess.run(cmd, check=True)
    # Read and combine results
    results = []
    for fname in sorted(os.listdir(out_dir)):
        path = os.path.join(out_dir, fname)
        with open(path, 'r', encoding='utf-8') as f:
            results.append(f.read())
    # Cleanup
    shutil.rmtree(out_dir)
    if output_format == "markdown":
        return "\n\n---\n\n".join(results)
    else:
        # If JSON, combine into list of pages
        return json.dumps({"pages": results}, indent=2, ensure_ascii=False)

demo = gr.Interface(
    fn=process_upload,
    inputs=[
        gr.File(label="Upload PDF", file_types=[".pdf"]),
        gr.Radio(["markdown","json"], value="markdown", label="Output format")
    ],
    outputs=gr.Code(label="Output"),
    title="PDF → Markdown/JSON with LaTeX Support",
    description=(
        "Uploads a PDF and uses Marker to extract text, structure, and LaTeX math. "
        "Choose Markdown to get a single .md with `$...$`/`$$...$$` math, "
        "or JSON for a page-by-page array."
    )
)

if __name__=="__main__":
    demo.launch(server_name="0.0.0.0", server_port=7860)