app.py

import os
from dotenv import load_dotenv
load_dotenv()

import streamlit as st
from src.scraper import scrape
from src.chunker import chunk_text
from src.embedder import embed
from src.retriever import Retriever
from src.generator import generate

# Build or load index on first run (simple demo)
@st.cache_resource
def load_retriever():
    url = "https://example.com"  # replace with your corpus source
    html = scrape(url)
    chunks = chunk_text(html)
    embeddings = embed(chunks)
    retriever = Retriever()
    metas = [{"path": url, "chunk_index": i} for i in range(len(chunks))]
    retriever.build(embeddings, metas)
    return retriever, chunks

retriever, all_chunks = load_retriever()

st.title("RAG Chatbot")
question = st.text_input("Ask a question:")
if st.button("Generate"):
    if not question:
        st.warning("Please enter a question.")
    else:
        query_emb = embed([question])[0]
        top_meta = retriever.retrieve(query_emb, k=5)
        top_indices = [m["chunk_index"] for m in top_meta]
        context_chunks = [all_chunks[i] for i in top_indices]
        answer = generate(question, context_chunks)
        st.write(answer)