Update app.py

app.py

@@ -2,6 +2,7 @@ import PyPDF2
 import xml.etree.ElementTree as ET
 import gradio as gr
 import tempfile
+import os
 
 def pdf_to_xml(pdf_path, xml_path):
     """
@@ -12,7 +13,11 @@ def pdf_to_xml(pdf_path, xml_path):
         root = ET.Element("document")
         
         for page_num, page in enumerate(pdf_reader.pages, start=1):
-            page_text = page.extract_text()
+            try:
+                page_text = page.extract_text()
+            except Exception as e:
+                # Log or handle extraction errors per page if needed.
+                page_text = ""
             page_element = ET.SubElement(root, "page", number=str(page_num))
             
             if page_text:
@@ -27,22 +32,31 @@ def pdf_to_xml_interface(pdf_file):
     """
     Gradio interface function that accepts an uploaded PDF file and returns the converted XML file.
     """
-    # Create a temporary file to store the output XML
+    # Handle case when no file is uploaded.
+    if pdf_file is None:
+        return "No file uploaded."
+    
+    # Create a temporary file to store the output XML.
     with tempfile.NamedTemporaryFile(delete=False, suffix=".xml") as tmp:
         output_path = tmp.name
-        
-    # Convert the uploaded PDF (pdf_file.name gives the file path)
-    pdf_to_xml(pdf_file.name, output_path)
+    
+    try:
+        pdf_to_xml(pdf_file.name, output_path)
+    except Exception as e:
+        # You can return a more detailed error message if desired.
+        return f"An error occurred during conversion: {e}"
+    
     return output_path
 
-# Define the Gradio interface
+# Define the Gradio interface.
 iface = gr.Interface(
     fn=pdf_to_xml_interface,
-    inputs=gr.components.File(label="Upload PDF File"),
-    outputs=gr.components.File(label="Download XML File"),
+    inputs=gr.File(label="Upload PDF File"),
+    outputs=gr.File(label="Download XML File"),
     title="PDF to XML Extractor",
     description="Upload a PDF file to extract its text into a structured XML format."
 )
 
 if __name__ == "__main__":
+    # Launch the Gradio app.
     iface.launch()