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

def process_upload(pdf_file, output_format):
    out_dir = tempfile.mkdtemp()
    fmt = "markdown" if output_format == "markdown" else "json"
    cmd = [
        "marker_single",
        pdf_file.name,
        "--output_format", fmt,
        "--output_dir", out_dir,
        "--paginate_output"
    ]
    subprocess.run(cmd, check=True)

    # Recursively find only .md/.json files
    collected = []
    for root, _, files in os.walk(out_dir):
        for fname in sorted(files):
            if fmt == "markdown" and fname.lower().endswith(".md"):
                collected.append(os.path.join(root, fname))
            elif fmt == "json" and fname.lower().endswith(".json"):
                collected.append(os.path.join(root, fname))

    pages = []
    for path in collected:
        with open(path, 'r', encoding='utf-8') as f:
            pages.append(f.read())

    shutil.rmtree(out_dir)

    if output_format == "markdown":
        return "\n\n---\n\n".join(pages)
    else:
        return json.dumps({"pages": pages}, 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="Converted Output"),
    title="PDF → Markdown/JSON with LaTeX Support",
    description=(
        "Upload a PDF and get back Markdown or structured JSON, "
        "with math preserved as LaTeX."
    )
)

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