app.py

import gradio as gr
from PyPDF2 import PdfReader
from sentence_transformers import SentenceTransformer
import faiss
import os
from openai import OpenAI

# 🔐 Set your OpenAI API key securely
os.environ['OPENAI_API_KEY'] = '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'  # Replace with your actual key
client = OpenAI(api_key=os.environ['OPENAI_API_KEY'])

# ✅ Load the sentence-transformer embedding model
embedding_model = SentenceTransformer("all-MiniLM-L6-v2")

# ✅ Extract text from PDF using PyPDF2
def extract_text_from_pdf(pdf_file):
    try:
        reader = PdfReader(pdf_file)
        text = ""
        for page in reader.pages:
            page_text = page.extract_text() or ""
            text += page_text
        return text
    except Exception as e:
        return ""

# ✅ Create a FAISS index for semantic similarity search
def create_vector_index(sentences):
    embeddings = embedding_model.encode(sentences)
    index = faiss.IndexFlatL2(embeddings.shape[1])
    index.add(embeddings)
    return index, embeddings

# ✅ Main function to answer question using RAG + OpenAI
def answer_question(pdf, question):
    try:
        if pdf is None:
            return "⚠ Please upload a PDF first."

        # Extract and clean text
        text = extract_text_from_pdf(pdf)
        if not text.strip():
            return "⚠ No extractable text found in the PDF."

        # Simple sentence-based chunking
        chunks = [chunk.strip() for chunk in text.split(". ") if len(chunk.strip()) > 30]
        if len(chunks) < 3:
            return "⚠ Not enough meaningful content to generate an answer."

        # Create vector index
        index, _ = create_vector_index(chunks)

        # Embed question and search top 3 similar chunks
        q_embed = embedding_model.encode([question])
        _, I = index.search(q_embed, k=3)
        retrieved = [chunks[i] for i in I[0]]
        context = "\n".join(retrieved)

        # Create prompt and query OpenAI GPT
        prompt = f"Context:\n{context}\n\nQuestion: {question}\nAnswer:"

        response = client.chat.completions.create(
            model="gpt-3.5-turbo",
            messages=[{"role": "user", "content": prompt}]
        )

        return response.choices[0].message.content.strip()

    except Exception as e:
        return f"❌ Error: {str(e)}"

# ✅ Gradio Interface
with gr.Blocks() as demo:
    gr.Markdown("## 🧠 Resume Q&A Chatbot\nAsk questions about your uploaded resume.")

    with gr.Row():
        pdf_input = gr.File(label="Upload your Resume PDF", file_types=[".pdf"])
        question = gr.Textbox(label="Enter your question")

    with gr.Row():
        submit_btn = gr.Button("Submit")
        clear_btn = gr.Button("Clear")
    
    output = gr.Textbox(label="Answer")

    submit_btn.click(fn=answer_question, inputs=[pdf_input, question], outputs=output)
    clear_btn.click(lambda: "", inputs=[], outputs=output)

# ✅ Launch the app
demo.launch()