from flask import Flask, render_template, request, send_from_directory
from datetime import datetime
from langchain_community.llms import HuggingFaceHub
from langchain.prompts import PromptTemplate
import requests
import json
import nltk
from textblob import TextBlob
from nltk.tokenize import word_tokenize
from nltk.stem import PorterStemmer
from nltk.stem import WordNetLemmatizer
import tensorflow as tf
from tensorflow import keras
import spacy
from bs4 import BeautifulSoup

nltk.download('punkt')
nltk.download('wordnet')

def download_spacy_model():
    import spacy  # Import spacy within the function scope
    try:
        spacy.load("en_core_web_sm")
    except OSError:
        import spacy.cli
        spacy.cli.download("en_core_web_sm")

download_spacy_model()

nlp = spacy.load("en_core_web_sm")

app = Flask(__name__)


# Load the JSON data from the file
with open('ai_chatbot_data.json', 'r') as file:
    json_data = json.load(file)

    


template = "Message: {message}\n\nSentiment Analysis: {sentiment}\n\nConversation Now Between you and user: {history}\n\nDate and Time: {date_time}\n\nBitcoin Price: ${bitcoin_price}\n\nBitcoin history from 1-jan-2024 to today the tidy is date-open-high-low-close-adj close-volum: {database_tag}\n\nYour system: {json_data}.\n\nResponse:"
prompt = PromptTemplate(template=template, input_variables=["message", "sentiment", "history", "date_time", "bitcoin_price", "database_tag", "json_data"])
conversation_history = []

MAX_HISTORY_LENGTH = 55

url = "https://dooratre-info.hf.space/"

response = requests.get(url)
soup = BeautifulSoup(response.content, 'html.parser')

div_content = soup.find('div', {'id': '45'})
if div_content:
    print(div_content)
else:
    print("No div with id=45 found on the page.")
database_tag=div_content

def update_conversation_history(message):
    if len(conversation_history) >= MAX_HISTORY_LENGTH:
        conversation_history.pop(0)
    conversation_history.append(message)


def get_bitcoin_price():
    current_time = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
    url = 'https://api.coindesk.com/v1/bpi/currentprice.json'
    response = requests.get(url)

    if response.status_code == 200:
        data = response.json()
        bitcoin_price = data['bpi']['USD']['rate']
        return bitcoin_price, current_time
    else:
        return 'Error fetching data', current_time

@app.route('/assets/<path:path>')
def send_static(path):
    return send_from_directory('assets', path)

@app.route('/')
def index():
    global conversation_history
    return render_template('index.html', conversation=conversation_history)

@app.route('/submit', methods=['POST'])
def submit():
    user_input = request.json.get('user_input')

    doc = nlp(user_input)
    tokens = [token.text for token in doc]

    sentiment = TextBlob(user_input).sentiment
    
    # Add Spacy NLP processing here
    
    ps = PorterStemmer()
    stemmed_tokens = [ps.stem(token) for token in tokens]

    lemmatizer = WordNetLemmatizer()
    lemmatized_tokens = [lemmatizer.lemmatize(token) for token in tokens]

    sentiment = TextBlob(user_input).sentiment

    bitcoin_price, current_time = get_bitcoin_price()

    conversation_history.append("User: " + user_input)
    
    # NLTK processing for conversation history
    history_tokens = word_tokenize("<br>".join(conversation_history))
    history_stemmed_tokens = [ps.stem(token) for token in history_tokens]
    history_lemmatized_tokens = [lemmatizer.lemmatize(token) for token in history_tokens]

    model_input = prompt.format(message=user_input, sentiment=sentiment, history="<br>".join(conversation_history), database_tag=div_content, date_time=current_time, bitcoin_price=bitcoin_price, json_data=json_data,history_tokens=history_tokens,history_stemmed_tokens=history_stemmed_tokens,history_lemmatized_tokens=history_lemmatized_tokens)

    response = llm(model_input, context="<br>".join(conversation_history))

    bot_response = response.split('Response:')[1].strip()
    bot_response = bot_response.strip().replace('\n', '<br>')

    # Update the conversation history with bot's response
    update_conversation_history("You " + bot_response)

    conversation_html = '<br>'.join(conversation_history)

    return bot_response

@app.route('/clear_history')
def clear_history():
    global conversation_history
    conversation_history = []
    return 'Conversation history cleared'

with open('i.txt', 'r') as file:
    data = file.read()

if __name__ == "__main__":
    repo_id = "mistralai/Mixtral-8x7B-Instruct-v0.1"
    huggingfacehub_api_token = "hf" + data

    llm = HuggingFaceHub(huggingfacehub_api_token=huggingfacehub_api_token,
                         repo_id=repo_id,
                         model_kwargs={
                             "temperature": 0.5,
                             "max_new_tokens": 256,
                             "top_p": 0.5,
                             "repetition_penalty": 1.2,
                             "num_beams": 3,
                             "length_penalty": 1.2,
                             "no_repeat_ngram_size": 2,
                             "early_stopping": True,
                             "num_return_sequences": 1,
                             "use_cache": True,
                             "task": "predictions",
                             "data_source": "financial_markets",
                             "historical_data_fetch": True,
                             "real-time_data_integration": True,
                             "feature_engineering": ["technical_indicators", "sentiment_analysis", "volume_analysis"],
                             "machine_learning_models": ["LSTM", "Random Forest", "ARIMA", "Gradient Boosting"],
                             "prediction_horizon": "short-term",
                             "evaluation_metrics": ["accuracy", "MSE", "MAE", "RMSE"],
                             "model_fine-tuning": True,
                             "interpretability_explanation": True,
                             "ensemble_methods": ["voting", "stacking"],
                             "hyperparameter_optimization": True,
                             "cross-validation": True,
                             "online_learning": True,
                         }
                         )

    app.run(host="0.0.0.0", port=7860)