import requests
from fastapi import FastAPI, Query
from fastapi.responses import FileResponse
import matplotlib.pyplot as plt
from datetime import datetime, timedelta
import pandas as pd
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import StandardScaler
from fastapi.middleware.cors import CORSMiddleware
import uvicorn
from fastapi.responses import HTMLResponse

app = FastAPI()

# Configure CORS
origins = [
    "*",  # Allows all origins
]

app.add_middleware(
    CORSMiddleware,
    allow_origins=origins,  # Allows all origins
    allow_credentials=True,
    allow_methods=["*"],  # Allows all methods
    allow_headers=["*"],  # Allows all headers
)

# Constants for API access
API_KEY = 'U9ER11OA4VGEWV9K'
STOCK_SYMBOL = 'AAPL'
API_URL = f"https://www.alphavantage.co/query?function=TIME_SERIES_DAILY&symbol={STOCK_SYMBOL}&apikey={API_KEY}"

# Function to fetch stock data
def fetch_stock_data(symbol):
    url = f"https://www.alphavantage.co/query?function=TIME_SERIES_DAILY&symbol={symbol}&apikey={API_KEY}"
    response = requests.get(url)
    data = response.json()
    return data['Time Series (Daily)']

# Function to update stock data and plot
def fetch_and_update(symbol=STOCK_SYMBOL, graph_type="line"):
    daily_data = fetch_stock_data(symbol)
    
    dates = []
    close_prices = []
    for date, daily_info in sorted(daily_data.items()):
        dates.append(datetime.strptime(date, '%Y-%m-%d'))
        close_prices.append(float(daily_info['4. close']))
    
    plt.figure(figsize=(14, 7))
    
    if graph_type == "line":
        plt.plot(dates, close_prices, marker='o', linestyle='-')
    elif graph_type == "bar":
        plt.bar(dates, close_prices)
    elif graph_type == "scatter":
        plt.scatter(dates, close_prices)
    elif graph_type == "buy_sell":
        stock_data = pd.DataFrame({'Date': dates, 'Close': close_prices})
        stock_data['Short_MA'] = stock_data['Close'].rolling(window=40).mean()
        stock_data['Long_MA'] = stock_data['Close'].rolling(window=100).mean()

        buy_signals = stock_data[(stock_data['Short_MA'] > stock_data['Long_MA']) & (stock_data['Short_MA'].shift(1) <= stock_data['Long_MA'].shift(1))].index
        sell_signals = stock_data[(stock_data['Short_MA'] < stock_data['Long_MA']) & (stock_data['Short_MA'].shift(1) >= stock_data['Long_MA'].shift(1))].index

        plt.plot(stock_data['Date'], stock_data['Close'], label='Closing Price', alpha=0.5)
        plt.plot(stock_data['Date'], stock_data['Short_MA'], label='40-Day Moving Average', alpha=0.75)
        plt.plot(stock_data['Date'], stock_data['Long_MA'], label='100-Day Moving Average', alpha=0.75)
        plt.scatter(stock_data.loc[buy_signals]['Date'], stock_data.loc[buy_signals]['Close'], marker='^', color='g', label='Buy Signal', alpha=1)
        plt.scatter(stock_data.loc[sell_signals]['Date'], stock_data.loc[sell_signals]['Close'], marker='v', color='r', label='Sell Signal', alpha=1)

    plt.title(f'{symbol} Stock Prices Over Time')
    plt.xlabel('Date')
    plt.ylabel('Close Price ($)')
    plt.gcf().autofmt_xdate()
    plt.legend()
    plt.savefig("stock.png")  # Save the plot as a PNG file
    plt.close()
    return dates, close_prices

# Preprocess data for prediction
def preprocess_data(dates, close_prices):
    df = pd.DataFrame({'Date': dates, 'Close': close_prices})
    df['Date_ordinal'] = pd.to_datetime(df['Date']).apply(lambda date: date.toordinal())
    
    # Handle missing values (if any)
    df.fillna(method='ffill', inplace=True)
    df.fillna(method='bfill', inplace=True)
    
    # Feature scaling
    scaler = StandardScaler()
    df['Close_scaled'] = scaler.fit_transform(df[['Close']])
    
    return df, scaler

# Machine Learning Model for Prediction
def predict_stock_prices(dates, close_prices, interval_days=7):
    df, scaler = preprocess_data(dates, close_prices)
    
    model = LinearRegression()
    model.fit(df[['Date_ordinal']], df['Close_scaled'])
    
    last_date = df['Date'].max()
    future_dates = [last_date + timedelta(days=i) for i in range(1, interval_days+1)]
    future_ordinal = [date.toordinal() for date in future_dates]
    scaled_predictions = model.predict(pd.DataFrame(future_ordinal, columns=['Date_ordinal']))
    
    predictions = scaler.inverse_transform(scaled_predictions.reshape(-1, 1)).flatten()
    
    return future_dates, predictions


@app.get("/", response_class=HTMLResponse)
async def read_root():
    html= open("index.html","r")
    return HTMLResponse(content=html.read())

# FastAPI endpoint to serve the graph image
@app.get("/graph")
async def get_graph(symbol: str = STOCK_SYMBOL, graph_type: str = Query("line", enum=["line", "bar", "scatter", "buy_sell"])):
    dates, close_prices = fetch_and_update(symbol, graph_type)
    return FileResponse("stock.png")

# FastAPI endpoint to predict stock prices
@app.get("/predict")
async def predict(symbol: str = STOCK_SYMBOL, interval: int = 7):
    dates, close_prices = fetch_and_update(symbol)
    future_dates, predictions = predict_stock_prices(dates, close_prices, interval)
    
    prediction_data = {str(date): float(pred) for date, pred in zip(future_dates, predictions)}
    return prediction_data

# if __name__ == "__main__":
#     # Run the FastAPI app using uvicorn with automatic reloading
#     uvicorn.run(app, host="127.0.0.1", port=8000)