app.py

import streamlit as st
import PyPDF2
import os
import requests
import json
from dotenv import load_dotenv
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.vectorstores import FAISS
from langchain_community.embeddings import HuggingFaceEmbeddings
import tempfile

# Load environment variables
load_dotenv()
GROQ_API_TOKEN = os.getenv("GROQ_API_TOKEN")

# Function to extract text from PDF
def extract_text_from_pdf(file):
    with tempfile.NamedTemporaryFile(delete=False) as temp_file:
        temp_file.write(file.getvalue())
        temp_file_path = temp_file.name

    try:
        with open(temp_file_path, 'rb') as file:
            pdf_reader = PyPDF2.PdfReader(file)
            text = ""
            for page in pdf_reader.pages:
                text += page.extract_text()
    except Exception as e:
        st.error(f"Error processing PDF: {str(e)}")
        text = ""
    finally:
        os.unlink(temp_file_path)
    return text

# Function to extract text from TXT
def extract_text_from_txt(file):
    return file.getvalue().decode("utf-8")

# Function to query GROQ API
def query_groq(prompt, context, temperature, max_tokens):
    headers = {
        "Authorization": f"Bearer {GROQ_API_TOKEN}",
        "Content-Type": "application/json"
    }
    
    data = {
        "model": "mixtral-8x7b-32768",
        "messages": [
            {"role": "system", "content": "You are a helpful assistant. Answer questions based only on the provided context."},
            {"role": "user", "content": f"Context: {context}\n\nQuestion: {prompt}"}
        ],
        "temperature": temperature,
        "max_tokens": max_tokens
    }
    
    try:
        response = requests.post("https://api.groq.com/openai/v1/chat/completions", headers=headers, json=data)
        response.raise_for_status()
        return response.json()["choices"][0]["message"]["content"]
    except requests.exceptions.RequestException as e:
        st.error(f"Error querying GROQ API: {str(e)}")
        return None

# Function to create vector store
def create_vector_store(text):
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
    chunks = text_splitter.split_text(text)
    
    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    vector_store = FAISS.from_texts(chunks, embeddings)
    
    return vector_store

# Streamlit UI
st.set_page_config(layout="wide")

# Custom CSS for scrollable chat container
st.markdown("""
<style>
    .chat-container {
        height: 600px;
        display: flex;
        flex-direction: column;
        border: 1px solid #ccc;
        border-radius: 5px;
    }
    .chat-messages {
        flex: 1;
        overflow-y: auto;
        padding: 10px;
    }
    .chat-input {
        border-top: 1px solid #ccc;
        padding: 10px;
    }
</style>
""", unsafe_allow_html=True)

st.title("Enhanced Document Query System")

# Create two columns for the split-screen layout
left_column, right_column = st.columns(2)

# Left column: Document upload and processing
with left_column:
    st.header("Document Upload")
    uploaded_file = st.file_uploader("Choose a file", type=["pdf", "txt"])
    doc_type = st.selectbox("Select document type", ["PDF", "TXT"])

    # Model parameters
    st.subheader("Model Parameters")
    temperature = st.slider("Temperature", 0.0, 1.0, 0.5, 0.1)
    max_tokens = st.slider("Max Tokens", 100, 2000, 1000, 100)

    if uploaded_file is not None:
        # Extract text based on document type
        if doc_type == "PDF":
            doc_text = extract_text_from_pdf(uploaded_file)
        else:
            doc_text = extract_text_from_txt(uploaded_file)
        
        if doc_text:
            st.success("File uploaded and processed successfully!")
            
            # Create vector store
            vector_store = create_vector_store(doc_text)
            st.session_state.vector_store = vector_store
        else:
            st.error("Failed to extract text from the document. Please try again.")

    # Clear chat history button
    if st.button("Clear Chat History"):
        st.session_state.messages = []
        st.rerun()

# Right column: Chat interface
with right_column:
    st.header("Chat Interface")

    # Chat history
    if "messages" not in st.session_state:
        st.session_state.messages = []

    # Scrollable chat container
    chat_container = st.container()
    with chat_container:
        st.markdown('<div class="scrollable-chat">', unsafe_allow_html=True)
        for message in st.session_state.messages:
            with st.chat_message(message["role"]):
                st.markdown(message["content"])
        st.markdown('</div>', unsafe_allow_html=True)

    # # Display chat history
    # for message in st.session_state.messages:
    #     with st.chat_message(message["role"]):
    #         st.markdown(message["content"])

    # User query input
    user_query = st.chat_input("Enter your question about the document:")
    
    if user_query and 'vector_store' in st.session_state:
        # Add user message to chat history
        st.session_state.messages.append({"role": "user", "content": user_query})
        with chat_container:
            with st.chat_message("user"):
                st.markdown(user_query)

        # Retrieve relevant context
        relevant_docs = st.session_state.vector_store.similarity_search(user_query, k=3)
        context = "\n".join([doc.page_content for doc in relevant_docs])

        # Query GROQ API
        response = query_groq(user_query, context, temperature, max_tokens)
        
        if response:
            # Add assistant message to chat history
            st.session_state.messages.append({"role": "assistant", "content": response})
            with st.chat_message("assistant"):
                st.markdown(response)
    elif user_query:
        st.warning("Please upload and process a document first.")