app.py

import os
from dotenv import load_dotenv
import streamlit as st
import pandas as pd
import plotly.express as px
import cloudscraper
import warnings
import logging
# Charger les variables d'environnement (si vous utilisez un .env localement)
load_dotenv()
API_KEY = os.environ.get("API_KEY")
headers = {
    "Authorization": f"Bearer {API_KEY}",
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) " 
                  "AppleWebKit/537.36 (KHTML, like Gecko) " 
                  "Chrome/115.0.0.0 Safari/537.36"
}

url = "https://archeanvision.com/api/signals/available"

# Create a cloudscraper instance
scraper = cloudscraper.create_scraper()  # This will handle Cloudflare challenges
response = scraper.get(url, headers=headers)

print(response.status_code)
print(response.text)

# Load environment variables from .env
load_dotenv()

# Suppress deprecation warnings about experimental query params functions
warnings.filterwarnings(
    "ignore", 
    message="Please replace `st.experimental_get_query_params` with `st.query_params`"
)
warnings.filterwarnings(
    "ignore", 
    message="Please replace `st.experimental_set_query_params` with `st.query_params`"
)
warnings.filterwarnings("ignore", category=DeprecationWarning)

# Adjust Streamlit loggers to show only errors
logging.getLogger("streamlit.deprecation").setLevel(logging.ERROR)
logging.getLogger("streamlit.runtime.scriptrunner").setLevel(logging.ERROR)


# ---------------------------- #
#         AUTO-REFRESH         #
# ---------------------------- #
st.set_page_config(
    page_title="Dashboard Auto-Refresh",
    layout="wide"
)

REFRESH_INTERVAL = 260  # seconds
st.markdown(f"<meta http-equiv='refresh' content='{REFRESH_INTERVAL}'>", unsafe_allow_html=True)
# ---------------------------- #

LOGO_IMAGE_URL = "https://archeanvision.com/assets/archeanvision.png"
st.sidebar.image(LOGO_IMAGE_URL, use_container_width=True, caption="ArcheanVision")

# Get the API key from environment variables (stored in .env or Hugging Face Secrets)
if not API_KEY:
    st.error("API_KEY is not set. Please add it to your environment (e.g. .env file or Hugging Face Secrets).")
    st.stop()

# --- Helper Functions Using cloudscraper ---

def get_active_markets_cloudscraper(api_key):
    """Retrieves the list of active markets using cloudscraper to bypass Cloudflare."""
    headers = {
        "Authorization": f"Bearer {api_key}",
        "User-Agent": ("Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
                       "AppleWebKit/537.36 (KHTML, like Gecko) "
                       "Chrome/115.0.0.0 Safari/537.36")
    }
    url = "https://archeanvision.com/api/signals/available"
    scraper = cloudscraper.create_scraper()
    response = scraper.get(url, headers=headers)
    response.raise_for_status()  # Raises an exception for HTTP errors
    return response.json()  # Assuming the endpoint returns JSON

def get_market_data_cloudscraper(api_key, market):
    """Retrieves market data for the given market using cloudscraper."""
    headers = {
        "Authorization": f"Bearer {api_key}",
        "User-Agent": ("Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
                       "AppleWebKit/537.36 (KHTML, like Gecko) "
                       "Chrome/115.0.0.0 Safari/537.36")
    }
    # Endpoint for market data (1,440 points ~ 24h); adjust as per API docs
    url = f"https://archeanvision.com/api/signals/{market}/data"
    scraper = cloudscraper.create_scraper()
    response = scraper.get(url, headers=headers)
    response.raise_for_status()
    return response.json()

def get_market_signals_cloudscraper(api_key, market):
    """Retrieves market signals for the given market using cloudscraper."""
    headers = {
        "Authorization": f"Bearer {api_key}",
        "User-Agent": ("Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
                       "AppleWebKit/537.36 (KHTML, like Gecko) "
                       "Chrome/115.0.0.0 Safari/537.36")
    }
    url = f"https://archeanvision.com/api/signals/{market}/signals"
    scraper = cloudscraper.create_scraper()
    response = scraper.get(url, headers=headers)
    response.raise_for_status()
    return response.json()

# --- End Helper Functions ---

def get_selected_market(market_list):
    """
    Retourne le marché sélectionné à partir des paramètres d'URL ou, par défaut, le premier élément.
    Met à jour le paramètre de l'URL si l'utilisateur choisit un marché différent.
    """
    # Récupère les paramètres sous forme de dictionnaire-like
    params = st.query_params

    # Récupérer le paramètre "market" ou définir la valeur par défaut
    default_market = params.get("market", market_list[0])
    # Si "market" est une liste (clé répétée), on prend le dernier (ou le premier) élément
    if isinstance(default_market, list):
        default_market = default_market[0]

    # Trouver l'index correspondant
    default_index = market_list.index(default_market) if default_market in market_list else 0

    # Affiche un menu déroulant pour choisir le marché
    selected = st.selectbox("Select a market:", market_list, index=default_index)

    # Si l'utilisateur choisit un marché différent, on met à jour le paramètre dans l'URL
    if selected != default_market:
        st.query_params.market = selected  # Mise à jour via la notation par attribut
        # Vous pouvez également faire : st.query_params["market"] = selected

    return selected


def main():
    st.title("Active AI Crypto Markets - ArcheanVision")

    st.markdown("""
    ### What is ArcheanVision?
    **ArcheanVision** is an autonomous multi-market trading agent.  
    It operates simultaneously on multiple crypto assets, monitoring price movements
    in real time and delivering **data** as well as **signals** (BUY, SELL, etc.)
    to automate and optimize decision-making.
    - **AI Agent**: Continuously analyzes crypto markets.  
    - **Multi-Market**: Manages multiple assets at once.  
    - **Live Data**: Access to streaming data feeds (SSE).  
    - **Buy/Sell Signals**: Generated in real-time to seize market opportunities.  
    Below is a dashboard showcasing the active markets, their 24h data 
    (1,440 most recent data points), and their associated signals.
    ---
    **Join our Discord as a beta tester** to help improve the agent and the system.  
    - Official platform: [https://archeanvision.com](https://archeanvision.com)  
    - Discord link: [https://discord.gg/k9xHuM7Jr8](https://discord.gg/k9xHuM7Jr8)
    """)

    # Retrieve active markets using cloudscraper
    try:
        active_markets = get_active_markets_cloudscraper(API_KEY)
    except Exception as e:
        st.error(f"Error fetching active markets: {e}")
        return

    if not active_markets:
        st.error("No active markets found through the API.")
        return

    # Expecting active_markets to be a list of market names, e.g. ["BTC", "ETH", ...]
    market_list = []
    if isinstance(active_markets, list):
        for item in active_markets:
            # Depending on the response structure, adjust accordingly.
            if isinstance(item, dict) and "market" in item:
                market_list.append(item["market"])
            elif isinstance(item, str):
                market_list.append(item)
            else:
                st.warning(f"Item missing 'market' key: {item}")
    else:
        st.error("The structure of 'active_markets' is not a list as expected.")
        return

    if not market_list:
        st.error("The market list is empty or 'market' keys not found.")
        return

    selected_market = get_selected_market(market_list)
    if not selected_market:
        st.error("No market selected.")
        return

    st.subheader(f"Selected Market: {selected_market}")
    st.write(f"Fetching data for **{selected_market}** ...")

    # Retrieve market data using cloudscraper
    try:
        market_data = get_market_data_cloudscraper(API_KEY, selected_market)
    except Exception as e:
        st.error(f"Error fetching market data for {selected_market}: {e}")
        return

    if not market_data:
        st.error(f"No data found for market {selected_market}.")
        return

    df = pd.DataFrame(market_data)
    if "close_time" in df.columns:
        df['close_time'] = pd.to_datetime(df['close_time'], unit='ms', errors='coerce')
    else:
        st.error("The 'close_time' column is missing from the retrieved data.")
        return

    st.write("### Market Data Overview")
    st.dataframe(df.head())

    required_cols = {"close", "last_predict_15m", "last_predict_1h"}
    if not required_cols.issubset(df.columns):
        st.error(
            f"The required columns {required_cols} are not all present. "
            f"Available columns: {list(df.columns)}"
        )
        return

    fig = px.line(
        df,
        x='close_time',
        y=['close', 'last_predict_15m', 'last_predict_1h'],
        title=f"{selected_market} : Close Price & Predictions",
        labels={
            'close_time': 'Time',
            'value': 'Price',
            'variable': 'Metric'
        }
    )
    st.plotly_chart(fig, use_container_width=True)

    st.write(f"### Signals for {selected_market}")
    try:
        signals = get_market_signals_cloudscraper(API_KEY, selected_market)
    except Exception as e:
        st.error(f"Error fetching signals for {selected_market}: {e}")
        return

    if not signals:
        st.warning(f"No signals found for market {selected_market}.")
    else:
        df_signals = pd.DataFrame(signals)
        if 'date' in df_signals.columns:
            df_signals['date'] = pd.to_datetime(df_signals['date'], unit='s', errors='coerce')
        for col in df_signals.columns:
            if df_signals[col].apply(lambda x: isinstance(x, dict)).any():
                df_signals[col] = df_signals[col].apply(lambda x: str(x) if isinstance(x, dict) else x)
        if 'date' in df_signals.columns:
            df_signals = df_signals.sort_values('date', ascending=False)
        st.write("Total number of signals:", len(df_signals))
        st.write("Preview of the last 4 signals:")
        st.dataframe(df_signals.head(4))

if __name__ == "__main__":
    main()