import streamlit as st
from transformers import pipeline
import yfinance as yf
import requests
from bs4 import BeautifulSoup
import xml.etree.ElementTree as ET
import numpy as np

# Set up the models for Named Entity Recognition, Sentiment Analysis, and Text Generation
ner_model = "Cassie-0121/fin-bert-finetuned-ner"
sentiment_model = "yiyanghkust/finbert-tone"
text_gen_model = "gpt2"

ner = pipeline("ner", model=ner_model)
sentiment_analyzer = pipeline("sentiment-analysis", model=sentiment_model)
text_generator = pipeline("text-generation", model=text_gen_model)

# App title
st.title("AI-Powered Investment Advisor")

# Sidebar for selecting stock
st.sidebar.header("Select Stock for Analysis")
top_100_stocks = [
    "AAPL", "MSFT", "GOOGL", "AMZN", "TSLA", "META", "NVDA", "BRK-B", "JNJ", "JPM",
    "V", "PG", "UNH", "DIS", "MA", "HD", "BAC", "XOM", "VZ", "PFE",
    "KO", "PEP", "MRK", "WMT", "CSCO", "ABT", "CMCSA", "ADBE", "NKE", "CRM",
    "T", "INTC", "ORCL", "IBM", "AVGO", "TXN", "LLY", "NFLX", "QCOM", "COST",
    "MCD", "HON", "PM", "BMY", "WBA", "DHR", "UNP", "RTX", "ACN", "UPS",
    "LOW", "CAT", "SCHW", "SPGI", "LIN", "NEE", "CVS", "MDT", "INTU", "BLK",
    "GS", "AMT", "MS", "DE", "CI", "BA", "C", "ZTS", "PLD", "GE",
    "NOW", "MO", "SBUX", "AMGN", "SYK", "MMM", "ADP", "TGT", "DUK", "SO",
    "BKNG", "CB", "ANTM", "AXP", "BDX", "ISRG", "TJX", "MMC", "APD", "NSC",
    "USB", "ECL", "CL", "CCI", "WM", "PGR", "MDLZ", "EQIX", "LMT", "ATVI"
]
selected_stock = st.sidebar.selectbox("Choose a stock for analysis:", top_100_stocks)

# Function to fetch latest stock data using yfinance
def get_stock_data(ticker):
    stock = yf.Ticker(ticker)
    stock_info = stock.history(period="1d")
    return stock_info

# Function to fetch latest news from Yahoo Finance or fallback to Google News RSS feed
def get_latest_news(ticker):
    try:
        # Attempt to fetch from Yahoo Finance
        url = f"https://finance.yahoo.com/quote/{ticker}?p={ticker}"
        page = requests.get(url)
        soup = BeautifulSoup(page.content, "html.parser")
        headlines = soup.find_all('h3', class_="Mb(5px)")
        news = [headline.text for headline in headlines[:5]]
        if news:
            return news
    except Exception as e:
        st.warning("Yahoo Finance news not available. Switching to Google News.")
    
    # Fallback to Google News RSS if Yahoo Finance fails
    rss_url = f"https://news.google.com/rss/search?q={ticker}"
    response = requests.get(rss_url)
    root = ET.fromstring(response.content)
    news = [item.find('title').text for item in root.findall(".//item")][:5]
    return news

# Display stock data
st.header(f"Latest Stock Data for {selected_stock}")
stock_data = get_stock_data(selected_stock)
st.write(stock_data)

# Display latest news
st.header(f"Latest News for {selected_stock}")
news = get_latest_news(selected_stock)
if news:
    for article in news:
        st.write(f"- {article}")
else:
    st.write("No news available for this stock at the moment.")

# Extract key financial entities with filtering
st.header("Extracted Key Financial Entities")
if news:
    entities = ner(" ".join(news))
    
    # Set a confidence threshold for filtering
    confidence_threshold = 0.8
    extracted_entities = []
    
    for entity in entities:
        # Filter entities with score above threshold and check for sub-word tokens
        if entity['score'] >= confidence_threshold and not entity['word'].startswith("##"):
            # Append only unique entities
            extracted_entities.append((entity['word'], entity['entity'], entity['score']))
    
    if extracted_entities:
        for word, label, score in extracted_entities:
            st.write(f"Entity: {word}, Label: {label}, Score: {score:.2f}")
    else:
        st.write("No high-confidence financial entities detected.")
else:
    st.write("Entity extraction requires news data.")

# Sentiment analysis on news
st.header("Sentiment Analysis")
if news:
    sentiment = sentiment_analyzer(" ".join(news))
    if sentiment:
        for result in sentiment:
            st.write(f"Sentiment: {result['label']}, Score: {result['score']:.2f}")
    else:
        st.write("No sentiment detected.")
else:
    st.write("Sentiment analysis requires news data.")

# Investment advice for a specific stock
st.header("Investment Advice or Strategy")

# Fetch stock data for the last month to calculate volatility
stock = yf.Ticker(selected_stock)
data = stock.history(period="1mo")

# Ensure data is available
if len(data) < 2:
    st.warning(f"Not enough data available for {selected_stock}. Please check the stock symbol.")
else:
    # Get the latest closing price
    latest_close = data['Close'][-1]

    # Get the 50-day moving average (if enough data is available)
    moving_avg_50 = data['Close'].rolling(window=50).mean().iloc[-1] if len(data) > 50 else None

    # Calculate daily returns
    data['Daily Return'] = data['Close'].pct_change()

    # Calculate daily volatility (standard deviation of daily returns)
    daily_volatility = data['Daily Return'].std()

    # Annualize the daily volatility to match Yahoo Finance's convention
    annualized_volatility = daily_volatility * np.sqrt(252)

    # Prepare the investment advice based on stock data and trends
    investment_advice = f"""
    Investment Strategy for {selected_stock}:
    - Latest Closing Price: ${latest_close:.2f}
    """
    
    # Add the 50-day moving average only if it's available
    if moving_avg_50 is not None:
        investment_advice += f"- 50-Day Moving Average: ${moving_avg_50:.2f}\n"
    else:
        investment_advice += "- 50-Day Moving Average: Not available (insufficient data)\n"

    investment_advice += f"""
    - Stock Volatility (Annualized 1-Month): {annualized_volatility:.2%}
    Based on recent stock data, {selected_stock} is currently trading at ${latest_close:.2f}. The stock has shown a recent uptrend and is supported by strong fundamentals, making it a solid investment for long-term growth. However, the volatility indicates some price fluctuations in the short term, so investors should be cautious of potential short-term risks.
    ### Suggested Actions:
    1. **Consider Buying** if the stock is trading below its 50-day moving average, as this may indicate an undervalued opportunity for long-term investors.
    2. **Watch for Volatility**: If the stock experiences significant price swings, it may indicate market uncertainty or external factors affecting the stock.
    3. **Monitor Earnings Reports**: {selected_stock}'s quarterly earnings report will provide important insights into its revenue streams and future prospects.
    4. **Long-term Investment**: Due to its strong market position and potential growth, {selected_stock} remains a good candidate for long-term portfolios.
    ### Risks:
    - **Market Volatility**: As indicated by the volatility of returns, there could be some short-term price swings.
    - **Competition**: The competitive landscape in the industry is constantly evolving, which could impact future performance.
    ### Expected Benefits:
    - **Growth Potential**: {selected_stock}'s growth in its core business or new ventures should continue to drive long-term stock price appreciation.
    - **Strong Cash Flow**: The company’s significant cash reserves may allow for innovation, stock buybacks, and dividend payments to shareholders.
    In conclusion, {selected_stock} is a strong stock with long-term growth potential, but investors should monitor market trends and the stock's volatility to time their investments effectively.
    """

    # Display the generated investment strategy
    st.write(investment_advice)