app.py

import os
import gradio as gr
import tempfile
from dotenv import load_dotenv

from langchain_community.document_loaders import PyPDFLoader
from langchain_text_splitters import RecursiveCharacterTextSplitter
from langchain_huggingface import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS

from groq import Groq

# ================= ENVIRONMENT =================
load_dotenv()
GROQ_API_KEY = os.getenv("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")

client = None
if GROQ_API_KEY:
    client = Groq(api_key=GROQ_API_KEY)

vector_db = None

# ================= LLM FUNCTION =================
def groq_llm(prompt):
    if client is None:
        return "❌ GROQ API key not set. Set it in Hugging Face Secrets."

    response = client.chat.completions.create(
        model="llama-3.3-70b-versatile",
        messages=[{"role": "user", "content": prompt}],
    )

    return response.choices[0].message.content


# ================= PROCESS PDF =================
def process_pdf(file):
    global vector_db

    if file is None:
        return "❌ Please upload a PDF."

    with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
        tmp.write(file)
        pdf_path = tmp.name

    loader = PyPDFLoader(pdf_path)
    documents = loader.load()

    splitter = RecursiveCharacterTextSplitter(
        chunk_size=500,
        chunk_overlap=100
    )
    docs = splitter.split_documents(documents)

    embeddings = HuggingFaceEmbeddings(
        model_name="sentence-transformers/all-MiniLM-L6-v2"
    )

    vector_db = FAISS.from_documents(docs, embeddings)

    return f"✅ PDF processed successfully! {len(docs)} chunks created."


# ================= ASK QUESTION =================
def ask_question(question):
    global vector_db

    if vector_db is None:
        return "❌ Please upload and process a document first."

    retriever = vector_db.as_retriever(search_kwargs={"k": 3})
    docs = retriever.get_relevant_documents(question)

    context = "\n\n".join([doc.page_content for doc in docs])

    prompt = f"""
You are an intelligent assistant.
Answer ONLY using the provided context.

Context:
{context}

Question:
{question}

Answer:
"""

    return groq_llm(prompt)


# ================= GRADIO UI =================
with gr.Blocks(title="📄 RAG PDF QA App") as demo:

    gr.Markdown("# 📄 RAG (Retrieval-Augmented Generation) PDF QA")
    gr.Markdown("Upload a PDF and ask questions about its content.")

    if client is None:
        gr.Markdown("⚠️ GROQ_API_KEY not set. Set it in Hugging Face Secrets to enable answering.")

    with gr.Row():
        pdf_upload = gr.File(label="Upload PDF", file_types=[".pdf"])
        process_btn = gr.Button("Process PDF")

    status = gr.Textbox(label="Status", interactive=False)

    question = gr.Textbox(label="Ask Question")
    answer = gr.Textbox(label="Answer", lines=10)

    process_btn.click(process_pdf, inputs=pdf_upload, outputs=status)
    question.submit(ask_question, inputs=question, outputs=answer)

demo.launch()