Update app.py

app.py

@@ -5,6 +5,7 @@ import gradio as gr
 import pandas as pd
 import requests
 import random
+import feedparser
 import urllib.parse
 from tempfile import NamedTemporaryFile
 from typing import List
@@ -277,7 +278,11 @@ def process_google_news_rss(query, temperature, top_p, repetition_penalty):
         processed_articles.append(processed_article)
     
     # Add processed articles to the database
-    docs = [Document(page_content=article["summary"], metadata={"url": article["url"], "title": article["title"], "published_date": article["published_date"]}) for article in processed_articles]
+    docs = [Document(page_content=article["summary"], metadata={
+        "source": article["url"],
+        "title": article["title"],
+        "published_date": article["published_date"]
+    }) for article in processed_articles]
     
     if os.path.exists("faiss_database"):
         database = FAISS.load_local("faiss_database", embed, allow_dangerous_deserialization=True)
@@ -341,13 +346,14 @@ def ask_question(question, temperature, top_p, repetition_penalty, web_search, g
         """
         prompt_val = ChatPromptTemplate.from_template(prompt_template)
         formatted_prompt = prompt_val.format(context=context_str, question=question)
+    
     elif google_news_rss:
         if database is None:
             return "No news articles available. Please fetch news articles first."
 
         retriever = database.as_retriever()
         relevant_docs = retriever.get_relevant_documents(question)
-        context_str = "\n".join([f"Title: {doc.metadata['title']}\nURL: {doc.metadata['url']}\nSummary: {doc.page_content}" for doc in relevant_docs])
+        context_str = "\n".join([f"Title: {doc.metadata.get('title', 'N/A')}\nURL: {doc.metadata.get('source', 'N/A')}\nSummary: {doc.page_content}" for doc in relevant_docs])
 
         prompt_template = """
         Answer the question based on the following news summaries: