app.py

import os
import gdown
import streamlit as st
from PyPDF2 import PdfReader
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS
from langchain.text_splitter import RecursiveCharacterTextSplitter
from groq import Groq

# Initialize Groq client
client = Groq(api_key=os.environ['GROQ_API_KEY'])

# Download and save PDF using gdown
def download_pdf(url):
    output_path = "/tmp/drive_doc.pdf"
    try:
        gdown.download(url=url, output=output_path, quiet=True, fuzzy=True)
        return output_path
    except Exception as e:
        st.error(f"❌ Download failed: {e}")
        return None

# Extract text from PDF
def extract_text(pdf_path):
    reader = PdfReader(pdf_path)
    text = ""
    for page in reader.pages:
        content = page.extract_text()
        if content:
            text += content
    return text

# Split text into chunks
def chunk_text(text):
    splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=50)
    return splitter.split_text(text)

# Create embeddings and store in FAISS
def build_vector_db(chunks):
    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    return FAISS.from_texts(chunks, embedding=embeddings)

# Query the vector DB and get response from Groq
def query_groq(query, vector_db):
    docs = vector_db.similarity_search(query, k=3)
    context = "\n".join([doc.page_content for doc in docs])
    response = client.chat.completions.create(
        model="llama3-8b-8192",
        messages=[
            {"role": "system", "content": f"Use the following context:\n{context}"},
            {"role": "user", "content": query},
        ]
    )
    return response.choices[0].message.content

# --- Streamlit App ---

st.title("📄 RAG QA from Google Drive PDF")

link = "https://drive.google.com/file/d/1SGXNLO841VyHnGiX81oo6x2RHIrTmP5S/view?usp=sharing"

st.write("📥 Downloading and processing document...")
pdf_path = download_pdf(link)

if pdf_path:
    try:
        text = extract_text(pdf_path)
        chunks = chunk_text(text)
        vector_db = build_vector_db(chunks)
        st.success("✅ Document processed successfully.")
    except Exception as e:
        st.error(f"❌ Error processing PDF: {e}")
        vector_db = None
else:
    vector_db = None

query = st.text_input("🔍 Enter your query:")
if query and vector_db:
    answer = query_groq(query, vector_db)
    st.subheader("💬 Answer:")
    st.write(answer)
elif query:
    st.warning("⚠️ Document not ready yet.")