import gradio as gr
import pandas as pd
import numpy as np
from NSEDownload import stocks
from time import sleep

## Get 10 year data from NSE
def get_data(symb):
    l = []
    year = 2015
    for i in range(0,10,5):
        year = i + year
        df = stocks.get_data(stock_symbol=symb, start_date=f'1-1-{year}', end_date=f'1-1-{year+5}')
        df.reset_index(drop=False,inplace=True)
        # print(df.head())
        l.append(df)
        sleep(15)
    dff = pd.concat(l,ignore_index=True)
    dff.to_csv(f'{symb}.csv',index=False,encoding='utf-8')
    return

# Calculate profit/loss based on stock price movement after condition is met
def calculate_profit_loss(stock_data):
    buy_sell_actions = []
    
    for i in range(len(stock_data)):
        if stock_data['condition'].iloc[i] == 1:  # Trigger condition met
            buy_price = stock_data['Open Price'].iloc[i+1]  # Buy on the next day's open price
            monitored_prices = stock_data.iloc[i+1:i+9]  # Monitor the next 8 days
            
            sell_price = None
            for j in range(len(monitored_prices)):
                no_trigger = 0
                open_price = monitored_prices['Open Price'].iloc[j]
                close_price = monitored_prices['Close Price'].iloc[j]
                change_percent = (close_price - buy_price) / buy_price * 100

                # Check for the +2%, +3%, +5%, +8% thresholds and set stop loss
                if change_percent >= 8:
                    sell_price = close_price
                    break
                elif change_percent >= 5:
                    sell_price = max(sell_price or 0, close_price)
                    if change_percent <= 5:
                        break
                elif change_percent >= 3:
                    sell_price = max(sell_price or 0, close_price)
                    if change_percent <= 3:
                        break
                elif change_percent >= 2:
                    sell_price = max(sell_price or 0, close_price)
                    if change_percent <= 2:
                        break

                # Stop-loss at -3%
                elif change_percent <= -3:
                    sell_price = close_price
                    break

            # If no triggers happen, sell at the 8th day's closing price
            if sell_price is None:
                sell_price = monitored_prices['Close Price'].iloc[-1]
                no_trigger = 1
            
            # Calculate profit/loss percentage
            profit_loss_percent = (sell_price - buy_price) / buy_price * 100
            buy_sell_actions.append({
                'Buy Date': stock_data['Date'].iloc[i+1],
                'Sell Date': monitored_prices['Date'].iloc[j] if sell_price != monitored_prices['Close Price'].iloc[-1] else monitored_prices['Date'].iloc[-1],
                'Buy Price': buy_price,
                'Sell Price': sell_price,
                'Profit/Loss (%)': profit_loss_percent,
                'No trigger': no_trigger
            })
    
    return pd.DataFrame(buy_sell_actions)

# Example function to simulate loading and processing stock data
def get_stock_data(stock_name):

    get_data(stock_name)
    
    stock_data = pd.read_csv(f'{stock_name}.csv')

    # Ensure the 'Date' column is in datetime format
    stock_data['Date'] = pd.to_datetime(stock_data['Date'])

    # Sort data by date
    stock_data = stock_data.sort_values(by='Date')
    
    # Calculate daily RSI
    def calculate_rsi(data, window):
        delta = data['Close Price'].diff(1)
        # print(delta)
        gain = np.where(delta > 0, delta, 0)
        loss = np.where(delta < 0, -delta, 0)

        avg_gain = pd.Series(gain).rolling(window=window).mean()
        avg_loss = pd.Series(loss).rolling(window=window).mean()

        rs = avg_gain / avg_loss
        rsi = 100 - (100 / (1 + rs))
        
        return rsi

    # Add a new column 'Daily RSI' for 14-day RSI
    stock_data['Daily RSI'] = calculate_rsi(stock_data, window=14)

    # Function to calculate sliding weekly RSI
    def calculate_sliding_weekly_rsi(data):
        global weekly_rsi
        weekly_rsi = []
        for i in range(7):
            stock_data1 = stock_data.iloc[i::7].reset_index(drop=True)
            stock_data1['weekly RSI'] = calculate_rsi(stock_data1, window=14)
            weekly_rsi.append(stock_data1)
        
        stock_data2 = pd.concat(weekly_rsi,ignore_index=True)
        stock_data.drop_duplicates(subset=['Date'],keep='first',inplace=True)
        stock_data2 = stock_data2.sort_values(by='Date')
        
        return stock_data2

    # Calculate sliding weekly RSI
    stock_data = calculate_sliding_weekly_rsi(stock_data)
    stock_data.reset_index(drop=True,inplace=True)

    ## Applying the condition
    for i in range(30, len(stock_data)):
        prev_30_days_rsi = stock_data['Daily RSI'][i-30:i]
        
        if all(prev_30_days_rsi < 65) and 65 <= stock_data['Daily RSI'].iloc[i] <= 70:
            stock_data.at[i, 'condition'] = 1
    
    fstock = stock_data[stock_data['condition']==1].reset_index(drop=True)

    profit_data = calculate_profit_loss(stock_data)

    # Returning two dataframes: One for the full stock data, another for RSI values
    return fstock, profit_data

# Function to save CSV file and return its path
def save_to_csv(stock_name):
    fstock, profit_data = get_stock_data(stock_name)
    csv_file_path = f'{stock_name}.csv'

    return fstock, profit_data, csv_file_path

# Create the Gradio interface
with gr.Blocks() as demo:
    gr.Markdown(
        """
        <h1 style="text-align: center; color: #4CAF50;">Stock Analysis Interface</h1>
        <p style="text-align: center;">Enter a stock Symbol and Calculate the algo returns.</p>
        """
    )
    
    with gr.Row():
        with gr.Column():
            stock_input = gr.Textbox(label="Enter Stock Symbol", placeholder="e.g., CANBK", lines=1)
            submit_button = gr.Button("Submit", variant="primary")
        
        with gr.Column():
            gr.Markdown("<h3 style='text-align: center;'>Output</h3>")
            output_stock_data = gr.DataFrame(label="Dates where Conditions met", interactive=False)
            output_rsi_data = gr.DataFrame(label="Profit and Loss Statement", interactive=False)
            csv_download = gr.File(label="Download CSV")

    # When the button is clicked, show the two dataframes and provide a downloadable CSV
    submit_button.click(save_to_csv, inputs=stock_input, outputs=[output_stock_data, output_rsi_data, csv_download])

# Launch the Gradio interface
demo.launch()