app.py

# ============================================
# 📘 Study Supervisor RAG App (Colab Compatible)
# ============================================

# 🔧 STEP 1: Install Required Packages
!pip -q install gradio faiss-cpu sentence-transformers PyPDF2 requests

# 🔧 STEP 2: Imports
import gradio as gr
import faiss
import os, requests
from io import BytesIO
from PyPDF2 import PdfReader
from sentence_transformers import SentenceTransformer

# ✅ CONFIGURATION
GROQ_API_KEY = ""  # 🔐 Set your Groq API key here
GROQ_MODEL = "llama3-70b-8192"
GROQ_API_URL = "https://api.groq.com/openai/v1/chat/completions"
EMBED_MODEL = "all-MiniLM-L6-v2"
CHUNK_SIZE = 500
TOP_K = 5

# ✅ Load Embedding Model
embedder = SentenceTransformer(EMBED_MODEL)

# ✅ Global Variables
faiss_index = None
chunk_texts = []

# 🔄 STEP 3: Process PDF, Chunk, Embed, Index
def process_pdf(file_obj):
    global faiss_index, chunk_texts
    pdf_stream = BytesIO(file_obj.read())
    pdf = PdfReader(pdf_stream)
    
    full_text = ""
    for page in pdf.pages:
        text = page.extract_text()
        if text:
            full_text += text + "\n"

    if not full_text.strip():
        return "❌ No extractable text found."

    # Split into chunks
    chunk_texts = [full_text[i:i+CHUNK_SIZE] for i in range(0, len(full_text), CHUNK_SIZE)]
    embeddings = embedder.encode(chunk_texts)

    # FAISS index
    dim = embeddings.shape[1]
    faiss_index = faiss.IndexFlatL2(dim)
    faiss_index.add(embeddings)

    return f"✅ PDF processed: {len(chunk_texts)} chunks indexed."

# 🧠 STEP 4: Call Groq LLaMA 3 for Answer Generation
def call_groq_llm(context, question):
    if not GROQ_API_KEY:
        return "❌ API Key is missing. Set your GROQ_API_KEY."

    prompt = f"""You are an academic supervisor helping a student understand a research paper.

Context:
{context}

Student's Question:
{question}

Answer:"""

    headers = {
        "Authorization": f"Bearer {GROQ_API_KEY}",
        "Content-Type": "application/json"
    }
    data = {
        "model": GROQ_MODEL,
        "messages": [
            {"role": "system", "content": "You are a knowledgeable and supportive supervisor guiding a student through a research paper. Respond clearly and academically."},
            {"role": "user", "content": prompt}
        ]
    }

    try:
        response = requests.post(GROQ_API_URL, headers=headers, json=data, timeout=60)
        if response.status_code == 200:
            return response.json()['choices'][0]['message']['content']
        else:
            return f"❌ Groq API Error {response.status_code}: {response.text}"
    except Exception as e:
        return f"❌ Exception: {str(e)}"

# 💬 STEP 5: Query Handler
def ask_question(query):
    if faiss_index is None or not chunk_texts:
        return "❌ Please upload and process a PDF first."

    query_embed = embedder.encode([query])
    D, I = faiss_index.search(query_embed, TOP_K)
    retrieved_chunks = [chunk_texts[i] for i in I[0]]
    context = "\n---\n".join(retrieved_chunks)

    return call_groq_llm(context[:3000], query)

# 🎨 STEP 6: Gradio Interface
with gr.Blocks() as app:
    gr.Markdown("📘 **Student Study Assistant** - Upload a research paper and ask questions.")
    
    with gr.Row():
        file_input = gr.File(label="📎 Upload PDF")
        process_button = gr.Button("📥 Process Document")
    status_output = gr.Textbox(label="Processing Status")

    chatbot = gr.ChatInterface(
        fn=ask_question,
        title="🎓 Study Supervisor",
        description="Ask your supervisor questions about the uploaded paper.",
        theme="soft"
    )

    process_button.click(fn=process_pdf, inputs=file_input, outputs=status_output)

app.launch(share=True)