Update app.py

app.py

@@ -1,33 +1,63 @@
 import gradio as gr
-from docling.parsers.pdf_parser import PDFParser
+from docling.document_converter import DocumentConverter
 import tempfile
 import os
+import shutil
+import requests
 
-def pdf_to_markdown(pdf_file):
-    if pdf_file is None:
-        return "No file uploaded."
+converter = DocumentConverter()
 
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
-        tmp.write(pdf_file.read())
-        tmp_path = tmp.name
+def convert_input(pdf_file, pdf_url):
+    temp_pdf_path = None
 
     try:
-        parser = PDFParser(tmp_path)
-        doc = parser.parse()
-        markdown = doc.to_markdown()
-        return markdown
+        if pdf_file:
+            with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
+                tmp.write(pdf_file.read())
+                temp_pdf_path = tmp.name
+                source = temp_pdf_path
+        elif pdf_url:
+            source = pdf_url
+        else:
+            raise gr.Error("Please upload a PDF or provide a URL.")
+
+        doc = converter.convert(source).document
+        markdown = doc.export_to_markdown()
+
+        # Save markdown to a file
+        with tempfile.NamedTemporaryFile(delete=False, suffix=".md", mode="w", encoding="utf-8") as tmp_md:
+            tmp_md.write(markdown)
+            markdown_path = tmp_md.name
+
+        return pdf_file.name if pdf_file else None, markdown, markdown_path
+
     except Exception as e:
-        return f"Error parsing PDF: {str(e)}"
+        return None, f"Error: {str(e)}", None
     finally:
-        os.remove(tmp_path)
+        if temp_pdf_path and os.path.exists(temp_pdf_path):
+            os.remove(temp_pdf_path)
+
+with gr.Blocks() as demo:
+    gr.Markdown("# 📄 PDF to Markdown with `docling`")
+    gr.Markdown("Upload a PDF or enter a URL (e.g., from arXiv). View Markdown and download it.")
+
+    with gr.Row():
+        pdf_input = gr.File(label="📤 Upload PDF", file_types=[".pdf"])
+        url_input = gr.Textbox(label="🌐 Or enter PDF URL (e.g., https://arxiv.org/pdf/...)")
+
+    convert_btn = gr.Button("Convert to Markdown")
+
+    with gr.Row():
+        pdf_preview = gr.File(label="📘 PDF Preview", interactive=False)
+        md_file = gr.File(label="📝 Download Markdown", interactive=False)
+
+    md_output = gr.Markdown(label="📄 Markdown Output")
 
-iface = gr.Interface(
-    fn=pdf_to_markdown,
-    inputs=gr.File(label="Upload PDF", file_types=[".pdf"]),
-    outputs=gr.Markdown(label="Markdown Output"),
-    title="PDF to Markdown with docling",
-    description="Upload a PDF file. This app parses it using `docling` and displays the Markdown version."
-)
+    convert_btn.click(
+        fn=convert_input,
+        inputs=[pdf_input, url_input],
+        outputs=[pdf_preview, md_output, md_file]
+    )
 
 if __name__ == "__main__":
-    iface.launch()
+    demo.launch()