Update app.py

app.py

@@ -83,14 +83,29 @@ def predict(input_text, pdf_file):
     return extract_result or "Too many users. Please wait a moment!"
 
 
-def view_pdf(pdf_file):
+def view_pdf(pdf_file, max_pages=3):
     if pdf_file is None:
         return "Please upload a PDF file to view."
 
-    with open(pdf_file.name, 'rb') as f:
-        pdf_data = f.read()
-    b64_data = base64.b64encode(pdf_data).decode('utf-8')
-    return f"<embed src='data:application/pdf;base64,{b64_data}' type='application/pdf' width='100%' height='700px' />"
+    try:
+        # Open the PDF file
+        doc = fitz.open(pdf_file.name)
+
+        # Only read up to `max_pages` pages to reduce size for large PDFs
+        preview_pdf = fitz.open()  # Create an empty PDF for the preview
+        for page_num in range(min(max_pages, doc.page_count)):
+            preview_pdf.insert_pdf(doc, from_page=page_num, to_page=page_num)
+
+        # Save the preview as a temporary in-memory file
+        pdf_data = preview_pdf.tobytes()
+
+        # Encode as base64 for embedding in HTML
+        b64_data = base64.b64encode(pdf_data).decode('utf-8')
+        return f"<embed src='data:application/pdf;base64,{b64_data}' type='application/pdf' width='100%' height='700px' />"
+
+    except Exception as e:
+        print(f"Error displaying PDF: {e}")
+        return "Error displaying PDF. Please try re-uploading."
 
 
 en_1 = """Could you please help me extract the information of 'title'/'journal'/'year'/'author'/'institution'/'email' from the previous content in a markdown table format?