import streamlit as st
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_core.runnables import RunnablePassthrough
import os
from dotenv import load_dotenv
load_dotenv()
from langchain_community.document_loaders import WebBaseLoader
from langchain.vectorstores import FAISS
from langchain_core.output_parsers import StrOutputParser
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.prompts import ChatPromptTemplate
from langchain_huggingface import HuggingFaceEmbeddings

st.title("Ask From Website")

user_url = st.text_input("Enter your URL here")
user_question = st.text_input("Enter your question from the website:")
query_button = st.button("View Results")

def data_ingestion(path):
    loader = WebBaseLoader(path)  
    data = loader.load()
    return data

def VectorStore(data, embedding):
    splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=500)
    chunks = splitter.split_documents(data)
    vector = FAISS.from_documents(chunks, embedding)
    retriever = vector.as_retriever()
    return retriever

def prompt_helper():
    template = """ Answer Based on the following context:
    {context}
    Question: {question}
    Provide only helpful information.
    """
    prompt = ChatPromptTemplate.from_template(template)
    return prompt

def main():
    if user_url and user_question:
        data = data_ingestion(user_url)
        retriever = VectorStore(data, embedding)
        prompt = prompt_helper()
        chain = (
            {'context': retriever, 'question': RunnablePassthrough()}
            | prompt
            | llm
            | StrOutputParser()
        )
        response = chain.invoke(user_question)
        return response

llm = ChatGoogleGenerativeAI(model='gemini-1.5-flash')
embedding = HuggingFaceEmbeddings(model_name='sentence-transformers/all-MiniLM-L6-v2')

if query_button:
    response = main()
    if response:
        st.markdown(response)