hyper/base_strategy.py

from abc import ABC, abstractmethod
import pandas as pd
import json
import os
import logging
from datetime import datetime, timezone
import sqlite3
import multiprocessing
import time

from logging_utils import setup_logging
from hyperliquid.info import Info
from hyperliquid.utils import constants

class BaseStrategy(ABC):
    """
    An abstract base class that defines the blueprint for all trading strategies.
    It provides common functionality like loading data, saving status, and state management.
    """
    
    def __init__(self, strategy_name: str, params: dict, trade_signal_queue: multiprocessing.Queue = None, shared_status: dict = None):
        self.strategy_name = strategy_name
        self.params = params
        self.trade_signal_queue = trade_signal_queue
        # Optional multiprocessing.Manager().dict() to hold live status (avoids file IO)
        self.shared_status = shared_status
        
        self.coin = params.get("coin", "N/A")
        self.timeframe = params.get("timeframe", "N/A")
        self.db_path = os.path.join("_data", "market_data.db")
        self.status_file_path = os.path.join("_data", f"strategy_status_{self.strategy_name}.json")

        self.current_signal = "INIT"
        self.last_signal_change_utc = None
        self.signal_price = None

        # Note: Logging is set up by the run_strategy function

    def load_data(self) -> pd.DataFrame:
        """Loads historical data for the configured coin and timeframe."""
        table_name = f"{self.coin}_{self.timeframe}"
        
        periods = [v for k, v in self.params.items() if 'period' in k or '_ma' in k or 'slow' in k or 'fast' in k]
        limit = max(periods) + 50 if periods else 500

        try:
            with sqlite3.connect(f"file:{self.db_path}?mode=ro", uri=True) as conn:
                query = f'SELECT * FROM "{table_name}" ORDER BY datetime_utc DESC LIMIT {limit}'
                df = pd.read_sql(query, conn, parse_dates=['datetime_utc'])
                if df.empty: return pd.DataFrame()
                df.set_index('datetime_utc', inplace=True)
                df.sort_index(inplace=True)
                return df
        except Exception as e:
            logging.error(f"Failed to load data from table '{table_name}': {e}")
            return pd.DataFrame()

    @abstractmethod
    def calculate_signals(self, df: pd.DataFrame) -> pd.DataFrame:
        """The core logic of the strategy. Must be implemented by child classes."""
        pass
        
    def calculate_signals_and_state(self, df: pd.DataFrame) -> bool:
        """
        A wrapper that calls the strategy's signal calculation, determines
        the last signal change, and returns True if the signal has changed.
        """
        df_with_signals = self.calculate_signals(df)
        df_with_signals.dropna(inplace=True)
        if df_with_signals.empty:
            return False

        df_with_signals['position_change'] = df_with_signals['signal'].diff()
        
        last_signal_int = df_with_signals['signal'].iloc[-1]
        new_signal_str = "HOLD"
        if last_signal_int == 1: new_signal_str = "BUY"
        elif last_signal_int == -1: new_signal_str = "SELL"
        
        signal_changed = False
        if self.current_signal == "INIT":
            if new_signal_str == "BUY": self.current_signal = "INIT_BUY"
            elif new_signal_str == "SELL": self.current_signal = "INIT_SELL"
            else: self.current_signal = "HOLD"
            signal_changed = True
        elif new_signal_str != self.current_signal:
            self.current_signal = new_signal_str
            signal_changed = True
        
        if signal_changed:
            last_change_series = df_with_signals[df_with_signals['position_change'] != 0]
            if not last_change_series.empty:
                last_change_row = last_change_series.iloc[-1]
                self.last_signal_change_utc = last_change_row.name.tz_localize('UTC').isoformat()
                self.signal_price = last_change_row['close']
        
        return signal_changed

    def _save_status(self):
        """Saves the current strategy state to its JSON file."""
        status = {
            "strategy_name": self.strategy_name,
            "current_signal": self.current_signal,
            "last_signal_change_utc": self.last_signal_change_utc,
            "signal_price": self.signal_price,
            "last_checked_utc": datetime.now(timezone.utc).isoformat()
        }
        # If a shared status dict is provided (Manager.dict()), update it instead of writing files
        try:
            if self.shared_status is not None:
                try:
                    # store the status under the strategy name for easy lookup
                    self.shared_status[self.strategy_name] = status
                except Exception:
                    # Manager proxies may not accept nested mutable objects consistently; assign a copy
                    self.shared_status[self.strategy_name] = dict(status)
            else:
                with open(self.status_file_path, 'w', encoding='utf-8') as f:
                    json.dump(status, f, indent=4)
        except IOError as e:
            logging.error(f"Failed to write status file for {self.strategy_name}: {e}")

    def run_polling_loop(self):
        """
        The default execution loop for polling-based strategies (e.g., SMAs).
        """
        while True:
            df = self.load_data()
            if df.empty:
                logging.warning("No data loaded. Waiting 1 minute...")
                time.sleep(60)
                continue
            
            signal_changed = self.calculate_signals_and_state(df.copy())
            self._save_status() 
            
            if signal_changed or self.current_signal == "INIT_BUY" or self.current_signal == "INIT_SELL":
                logging.warning(f"New signal detected: {self.current_signal}")
                self.trade_signal_queue.put({
                    "strategy_name": self.strategy_name,
                    "signal": self.current_signal,
                    "coin": self.coin,
                    "signal_price": self.signal_price,
                    "config": {"agent": self.params.get("agent"), "parameters": self.params}
                })
                if self.current_signal == "INIT_BUY": self.current_signal = "BUY"
                if self.current_signal == "INIT_SELL": self.current_signal = "SELL"

            logging.info(f"Current Signal: {self.current_signal}")
            time.sleep(60)

    def run_event_loop(self):
        """
        A placeholder for event-driven (WebSocket) strategies.
        Child classes must override this.
        """
        logging.error("run_event_loop() is not implemented for this strategy.")
        time.sleep(3600) # Sleep for an hour to prevent rapid error loops
        
    def on_fill_message(self, message):
        """
        Placeholder for the WebSocket callback.
        Child classes must override this.
        """
        pass