pylightningfx

Python client for the bitFlyer Lightning API.

Two HTTP clients — Client and AsyncClient — cover every Public and Private endpoint and return validated Pydantic models. Two WebSocket clients — RealtimeClient and AsyncRealtimeClient — cover the streaming Realtime API.

Installation

pip install pylightningfx

Quick start

Public endpoints need no credentials; Private endpoints take an API key and secret. Use the context manager so the connection pool is closed for you.

from pylightningfx import ChildOrderType, Client, ProductCode, Side

with Client() as client:
    ticker = client.get_ticker(ProductCode.FX_BTC_JPY)
    print(ticker.ltp)

with Client(api_key, api_secret) as client:
    ack = client.send_child_order(
        ProductCode.BTC_JPY, ChildOrderType.LIMIT, Side.BUY, 0.001, price=5_000_000
    )
    print(ack.child_order_acceptance_id)

Streaming

Subscribe to any number of channels on one connection and iterate the messages. Payloads arrive parsed into the same models the HTTP clients return.

from pylightningfx import ProductCode, RealtimeClient, channels

with RealtimeClient() as rt:
    rt.subscribe(channels.ticker(ProductCode.FX_BTC_JPY))
    for message in rt.listen():
        print(message.channel, message.data.ltp)

The asyncio twin is the same shape with await and async for:

import asyncio

from pylightningfx import AsyncRealtimeClient, ProductCode, channels


async def main() -> None:
    async with AsyncRealtimeClient() as rt:
        await rt.subscribe(channels.executions(ProductCode.BTC_JPY))
        async for message in rt.listen():
            for trade in message.data:
                print(trade.price, trade.size)


asyncio.run(main())

Where to go next

• Clients — the HTTP clients and every endpoint method.
• Realtime — WebSocket clients and channel names.
• Models — the response models.
• Configuration — retries and client-side rate limiting.
• Errors — the exception hierarchy.
• Enums — product codes, order types, states.