Update app.py

app.py

@@ -12,6 +12,7 @@ import time
 import warnings
 import os
 import re
+import tempfile
 
 # Trafilatura imports
 from trafilatura import fetch_url, extract, bare_extraction
@@ -19,6 +20,27 @@ from trafilatura.downloads import fetch_url as trafilatura_fetch
 
 warnings.filterwarnings('ignore')
 
+# Global variable to store the latest vector data
+latest_vector_data = None
+
+def prepare_download(vector_df):
+    """Prepare the vector data for download"""
+    global latest_vector_data
+    if vector_df is not None and not vector_df.empty:
+        # Save to temporary file
+        temp_file = tempfile.NamedTemporaryFile(mode='w', suffix='.csv', delete=False, newline='', encoding='utf-8')
+        vector_df.to_csv(temp_file.name, index=False)
+        latest_vector_data = temp_file.name
+        return temp_file.name
+    return None
+
+def download_vector_data():
+    """Return the prepared vector data file"""
+    global latest_vector_data
+    if latest_vector_data:
+        return latest_vector_data
+    return None
+
 @dataclass
 class ContentChunk:
     content: str