Update app.py

app.py

@@ -1,6 +1,15 @@
 import streamlit as st
 import re
 import os
+import bs4
+from sentence_transformers import SentenceTransformer
+import torch
+from langchain.chains import ChatGroq
+from langchain.document_loaders import UnstructuredURLLoader  # Updated import
+from langchain.prompts import RunnablePassthrough
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.vectorstores import Chroma
+from langchain.output_parsers import StrOutputParser
 
 # Sidebar Style with Multicolored Background
 sidebar_bg_style = """
@@ -81,14 +90,13 @@ if 'chat_history' not in st.session_state:
 
 # Submit button for chat
 if st.button("Submit Query"):
-    if query and url_input:
+    if not query:
+        st.markdown('<p style="color:red; font-weight:bold;">Please enter a query before submitting</p>', unsafe_allow_html=True)
+    elif not url_input:
+        st.markdown('<p style="color:red; font-weight:bold;">Please enter a valid URL before submitting</p>', unsafe_allow_html=True)
+    else:
         # Blog loading logic based on user input URL
-        loader = WebBaseLoader(
-            web_paths=(url_input,),  # Use the user-input URL
-            bs_kwargs=dict(
-                parse_only=bs4.SoupStrainer()  # Adjust based on the user's URL structure
-            ),
-        )
+        loader = UnstructuredURLLoader(urls=[url_input])  # Using UnstructuredURLLoader instead of WebBaseLoader
         docs = loader.load()
 
         text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)