app.py

import gradio as gr
import os
import hashlib
import pickle
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.document_loaders import PyPDFLoader
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS
from langchain.chains import RetrievalQA
from langchain_groq import ChatGroq

# Load Groq API key
GROQ_API_KEY = os.getenv("GROQ_API_KEY")

# Directory to cache vectorstores
CACHE_DIR = "vector_cache"
os.makedirs(CACHE_DIR, exist_ok=True)


def get_pdf_hash(pdf_path: str) -> str:
    """Generate a hash for the PDF file to use as cache key"""
    with open(pdf_path, "rb") as f:
        data = f.read()
    return hashlib.md5(data).hexdigest()


def build_vectorstore(pdf_path: str):
    """Load PDF, chunk it, embed, and create FAISS index"""
    loader = PyPDFLoader(pdf_path)
    documents = loader.load()

    # Chunking strategy
    text_splitter = RecursiveCharacterTextSplitter(
        chunk_size=1000,
        chunk_overlap=200,
        separators=["\n\n", "\n", " ", ""]
    )
    chunks = text_splitter.split_documents(documents)

    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    vectorstore = FAISS.from_documents(chunks, embeddings)
    return vectorstore


def get_vectorstore(pdf_path: str):
    """Return cached FAISS index if available, else build new one"""
    pdf_hash = get_pdf_hash(pdf_path)
    cache_file = os.path.join(CACHE_DIR, f"{pdf_hash}.pkl")

    if os.path.exists(cache_file):
        with open(cache_file, "rb") as f:
            return pickle.load(f)

    # Build and cache
    vectorstore = build_vectorstore(pdf_path)
    with open(cache_file, "wb") as f:
        pickle.dump(vectorstore, f)
    return vectorstore


def rag_bot(question: str, pdf_path: str):
    """Answer user queries using the uploaded PDF"""
    if not pdf_path:
        return "⚠️ Please upload a PDF first."

    try:
        # Load or build vectorstore
        vectorstore = get_vectorstore(pdf_path)
        retriever = vectorstore.as_retriever(search_kwargs={"k": 3})

        # Use Groq LLM
        llm = ChatGroq(
            groq_api_key=GROQ_API_KEY,
            model_name="llama-3.3-70b-versatile",  # The updated model name
        )

        qa = RetrievalQA.from_chain_type(
            llm=llm,
            chain_type="stuff",
            retriever=retriever,
        )
        result = qa.run(question)
        return result
    except Exception as e:
        return f"❌ Error: {e}"


# ------------------ Gradio UI ------------------
with gr.Blocks() as demo:
    gr.Markdown("## 📖 RAG Q&A Bot – Powered by Groq + HuggingFace Embeddings")

    with gr.Row():
        pdf_file = gr.File(label="Upload PDF", type="filepath", file_types=[".pdf"])
    with gr.Row():
        question = gr.Textbox(label="Ask a Question")
    with gr.Row():
        answer = gr.Textbox(label="Answer", interactive=False)

    submit = gr.Button("Submit")
    submit.click(fn=rag_bot, inputs=[question, pdf_file], outputs=answer)

if __name__ == "__main__":
    demo.launch(server_name="0.0.0.0", server_port=7860)