app.py

import PyPDF2
import chromadb
import streamlit as st
from langchain_openai import OpenAI

# Function to extract text from PDF
def extract_text_from_pdf(pdf_path):
    text = ""
    with open(pdf_path, "rb") as file:
        pdf_reader = PyPDF2.PdfFileReader(file)
        for page_num in range(pdf_reader.numPages):
            text += pdf_reader.getPage(page_num).extractText()
    return text

# Function to create chunks of text
def create_text_chunks(text, chunk_size=1000, overlap_size=100):
    chunks = []
    for i in range(0, len(text), chunk_size - overlap_size):
        chunks.append(text[i:i + chunk_size])
    return chunks

# Function to save chunks to chromadb vector database
def save_to_chromadb(chunks, quiz_name, quiz_topic):
    # Assume you have a ChromaDB instance named 'db'
    db = chromadb.ChromaDB("your_chromadb_url")
    for i, chunk in enumerate(chunks):
        vector = langchain_openai.get_vector(chunk)
        db.add_vector(quiz_name, quiz_topic, i, vector)

# Function to generate questions using ChatGPT-3.5-turbo-16k
def generate_questions(topic):
    prompt = f"Generate questions on the topic: {topic}"
    response = langchain_openai.complete(prompt)
    return response.choices[0].text.strip()

# Streamlit interface
def main():
    st.title("Quiz Generator")

    # User inputs
    quiz_name = st.text_input("Enter Quiz Name:")
    quiz_topic = st.text_input("Enter Quiz Topic:")
    num_questions = st.number_input("Number of Questions:", value=5, min_value=1)
    pdf_path = st.file_uploader("Upload PDF File:", type=["pdf"])

    if pdf_path:
        # Extract text from PDF
        pdf_text = extract_text_from_pdf(pdf_path)

        # Create and save text chunks to ChromaDB
        text_chunks = create_text_chunks(pdf_text)
        save_to_chromadb(text_chunks, quiz_name, quiz_topic)

        # User input for query
        user_query = st.text_input("Enter Query for Question Generation:")

        # Search for the topic in the vector database
        if quiz_topic in db.get_topics(quiz_name):
            # Generate questions using ChatGPT-3.5-turbo-16k
            generated_questions = generate_questions(user_query)

            st.subheader("Generated Questions:")
            st.write(generated_questions)
        else:
            st.warning("Specified topic not found in the document.")

if __name__ == "__main__":
    main()