Polymarket/ws_rtds.py

import asyncio
import json
import logging
import socket
import traceback
from datetime import datetime
from typing import AsyncContextManager

import numpy as np
import pandas as pd
import requests.packages.urllib3.util.connection as urllib3_cn  # type: ignore
from sqlalchemy import text
import websockets
from sqlalchemy.ext.asyncio import create_async_engine
import valkey

### Allow only ipv4 ###
def allowed_gai_family():
    return socket.AF_INET
urllib3_cn.allowed_gai_family = allowed_gai_family

### Database ###
USE_DB: bool = True
USE_VK: bool = True
VK_CHANNEL = 'poly_rtds_cl_btcusd'
CON: AsyncContextManager | None = None
VAL_KEY = None


### Logging ###
LOG_FILEPATH: str = '/root/logs/Polymarket_RTDS.log'

### Globals ###
WSS_URL = "wss://ws-live-data.polymarket.com"
# HIST_TRADES = np.empty((0, 2))

### Database Funcs ###
async def create_rtds_btcusd_table(
        CON: AsyncContextManager,
        engine: str = 'mysql', # mysql | duckdb
    ) -> None:
    if CON is None:
        logging.info("NO DB CONNECTION, SKIPPING Create Statements")
    else:
        if engine == 'mysql':
            logging.info('Creating Table if Does Not Exist: poly_rtds_cl_btcusd')
            await CON.execute(text("""
            CREATE TABLE IF NOT EXISTS poly_rtds_cl_btcusd (
                timestamp_arrival BIGINT,
                timestamp_msg BIGINT,
                timestamp_value BIGINT,
                value DOUBLE
            );
            """))
            await CON.commit()
        else:
            raise ValueError('Only MySQL engine is implemented')

async def insert_rtds_btcusd_table(
        timestamp_arrival: int,
        timestamp_msg: int,
        timestamp_value: int,
        value: int,
        CON: AsyncContextManager,
        engine: str = 'mysql', # mysql | duckdb
    ) -> None:
    params={
        'timestamp_arrival': timestamp_arrival,
        'timestamp_msg': timestamp_msg,
        'timestamp_value': timestamp_value,
        'value': value,
    }
    if CON is None:
        logging.info("NO DB CONNECTION, SKIPPING Insert Statements")
    else:
        if engine == 'mysql':
            await CON.execute(text("""
            INSERT INTO poly_rtds_cl_btcusd 
            (
                timestamp_arrival,
                timestamp_msg,
                timestamp_value,
                value
            )
            VALUES 
                (
                :timestamp_arrival,
                :timestamp_msg,
                :timestamp_value,
                :value
                )
            """),
            parameters=params
            )
            await CON.commit()
        else:
            raise ValueError('Only MySQL engine is implemented')


### Websocket ###
async def rtds_stream():
    global HIST_TRADES

    async for websocket in websockets.connect(WSS_URL):
        logging.info(f"Connected to {WSS_URL}")

        subscribe_msg = {
            "action": "subscribe",
            "subscriptions": [
                {
                "topic": "crypto_prices_chainlink",
                "type": "*",
                "filters": "{\"symbol\":\"btc/usd\"}"
                }
            ]
        }

        await websocket.send(json.dumps(subscribe_msg))

        try:
            async for message in websocket:
                if isinstance(message, str):
                    try: 
                        data = json.loads(message)
                        if data['payload'].get('value', None) is not None:
                            ts_arrival = round(datetime.now().timestamp()*1000)
                            print(f'🤑 BTC Chainlink Last Px: {data['payload']['value']:_.4f}; TS: {pd.to_datetime(data['payload']['timestamp'], unit='ms')}')
                            VAL_KEY_OBJ = json.dumps({
                                'timestamp_arrival': ts_arrival,
                                'timestamp_msg': data['timestamp'],
                                'timestamp_value': data['payload']['timestamp'],
                                'value': data['payload']['value'],
                            })
                            VAL_KEY.publish(VK_CHANNEL, VAL_KEY_OBJ)
                            VAL_KEY.set(VK_CHANNEL, VAL_KEY_OBJ)
                            await insert_rtds_btcusd_table(
                                CON=CON,
                                timestamp_arrival=ts_arrival,
                                timestamp_msg=data['timestamp'],
                                timestamp_value=data['payload']['timestamp'],
                                value=data['payload']['value'],
                            )
                        else:
                            # logging.info(f'Initial or unexpected data struct, skipping: {data}')
                            logging.info('Initial or unexpected data struct, skipping')
                            continue
                    except (json.JSONDecodeError, ValueError):
                        logging.warning(f'Message not in JSON format, skipping: {message}')
                        continue
                else:
                    raise ValueError(f'Type: {type(data)}  not expected: {message}')
        except websockets.ConnectionClosed as e:
            logging.error(f'Connection closed: {e}')
            logging.error(traceback.format_exc())
            continue
        except Exception as e:
            logging.error(f'Connection closed: {e}')
            logging.error(traceback.format_exc())


async def main():
    global VAL_KEY
    global CON
    
    if USE_VK:
        VAL_KEY = valkey.Valkey(host='localhost', port=6379, db=0)
        published_count = VAL_KEY.publish(VK_CHANNEL,f"Hola, starting to publish to valkey: {VK_CHANNEL} @ {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
        logging.info(f"Valkey message published to {published_count} subscribers of {VK_CHANNEL}")
    else:
        VAL_KEY = None
        logging.warning("VALKEY NOT BEING USED, NO DATA WILL BE PUBLISHED")

    if USE_DB:
        engine = create_async_engine('mysql+asyncmy://root:pwd@localhost/polymarket')
        async with engine.connect() as CON:
            await create_rtds_btcusd_table(CON=CON)
            await rtds_stream()
    else:
        CON = None
        logging.warning("DATABASE NOT BEING USED, NO DATA WILL BE RECORDED")
        await rtds_stream()


if __name__ == '__main__':
    START_TIME = round(datetime.now().timestamp()*1000)

    logging.info(f'Log FilePath: {LOG_FILEPATH}')

    logging.basicConfig(
        force=True,
        filename=LOG_FILEPATH,
        level=logging.INFO,
        format='%(asctime)s - %(levelname)s - %(message)s',
        filemode='w'
    )
    logging.info(f"STARTED: {START_TIME}")
    
    try:
        asyncio.run(main())
    except KeyboardInterrupt:
        logging.info("Stream stopped")