app.py

import streamlit as st 
import sqlite3
import uuid
from langchain_google_genai import GoogleGenerativeAI
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_core.output_parsers import StrOutputParser
from langchain_community.chat_message_histories import SQLChatMessageHistory
from langchain_core.runnables.history import RunnableWithMessageHistory
import time

# Set page config
st.set_page_config(page_title="Gemini Code Reviewer", layout="wide")

# Animated text function
def animated_text(text, speed=0.05):
    placeholder = st.empty()
    displayed_text = ""
    for letter in text:
        displayed_text += letter
        placeholder.markdown(f"""
            <h1 style="text-align:center; color: #FF4500;">{displayed_text} 🚀</h1>
        """, unsafe_allow_html=True)  # Corrected f-string formatting
        time.sleep(speed)

# Display animated text
animated_text("Conversational AI Data Science Tutor", speed=0.1)

# Streamlit Ui
st.write("Ask me anything about Data Science!")
# Load API key
GOOGLE_API_KEY = st.secrets.get("GOOGLE_API_KEY")

# Set up the Gemini 1.5 Pro model
llm = GoogleGenerativeAI(api_key=GOOGLE_API_KEY, model="gemini-1.5-pro")

# Initialize SQLite database
conn = sqlite3.connect("chat_history.db", check_same_thread=False)
cursor = conn.cursor()
cursor.execute("""
CREATE TABLE IF NOT EXISTS chat (
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    session_id TEXT,
    role TEXT,
    content TEXT
)
""")
conn.commit()

# Function to save messages
def save_message(session_id, role, content):
    cursor.execute("INSERT INTO chat (session_id, role, content) VALUES (?, ?, ?)", (session_id, role, content))
    conn.commit()

# Function to load chat history
def load_chat_history(session_id):
    cursor.execute("SELECT role, content FROM chat WHERE session_id = ?", (session_id,))
    return cursor.fetchall()

# Chat history instance
def chat_history(session_id):
    return SQLChatMessageHistory(
        session_id=session_id,
        connection="sqlite:///chat_history.db"
    )

# Generate unique session ID
if "session_id" not in st.session_state:
    st.session_state.session_id = str(uuid.uuid4())

# "New Chat" button
if st.button("🆕 New Chat"):
    st.session_state.session_id = str(uuid.uuid4())  # Generate new session ID
    st.session_state.messages = []  # Clear chat history
    st.rerun()  # Refresh the app

session_id = st.session_state.session_id
chat_history_instance = chat_history(session_id)

# Define Chat Prompt Template
chat_prompt = ChatPromptTemplate(
    messages=[
        ('system', """You are an AI assistant specialized in Data Science tutoring. 
                      You will only answer questions related to Data Science. 
                      If asked anything outside this topic, politely decline and request a Data Science-related question.
                   """),
        MessagesPlaceholder(variable_name="history", optional=True),
        ('human', '{prompt}')
    ]
)

# Define output parser
out_parser = StrOutputParser()

# Create a chain
chain = chat_prompt | llm | out_parser

# Define Runnable with message history
chat = RunnableWithMessageHistory(
    chain,
    lambda session: SQLChatMessageHistory(session, "sqlite:///chat_history.db"),
    input_messages_key="prompt",
    history_messages_key="history"
)


# 🔹 Chat history container
chat_container = st.container()

# 🔹 Always keep input box at the bottom
input_container = st.empty()

# Load chat history and keep it sticky
if "messages" not in st.session_state:
    st.session_state.messages = load_chat_history(session_id)


# Display chat messages in order
with chat_container:
    for role, content in st.session_state.messages:
        with st.chat_message(role):
            st.markdown(content)
# Display chat history

for role, content in st.session_state.messages:
    with st.chat_message(role):
        st.markdown(content)

with input_container:
    # User input at the bottom
    user_input = st.text_input("Type your message here:", key="user_message")

if user_input:
    # Save user message
    save_message(session_id, "user", user_input)
    st.session_state.messages.append(("user", user_input))
    
    # Invoke AI model
    config = {'configurable': {'session_id': session_id}}
    response = chat.invoke({'prompt': user_input}, config)
    
    # Save AI response
    save_message(session_id, "assistant", response)
    st.session_state.messages.append(("assistant", response))
    
    # Display AI response
    with st.chat_message("assistant"):
        st.markdown(response)

    # ✅ Clear input field
    st.session_state.pop("user_message")
    st.session_state["user_message"] = ""  
    st.rerun()  # Refresh the app