A simple web server streaming up-to-date exchange rates to the subscribed clients via websocket streaming

Usage

Subscribe to the server by sending a websocket connection message

Launching

General setting

1. Rename compose/envs/sample.common.env into compose/envs/common.env
2. Set the missing .env variables with values required for the project to run, e.g., EMCONT_EXCHANGE_RATES_URL

Using Docker

Pull the images
docker compose pull

Build the images
docker compose build

Run the docker compose configuration
docker compose up

Asynchronous tasks

Run asynchornous tasks manually:
docker compose exec backend bash
poetry run python async_tasks/async_periodic_tasks.py

Mongo

Connect to the Mongo DB instance:
docker compose exec db sh
mongosh $MONGO_CONNECTION_URI
show dbs to list all the databases
use db to switch to the backend database

Websocket actions

The only application endpoint is Endpoint: "/"

1. Retreive the available assets

message: "{"action": "assets", "message": {}}" Response sample:

{
    "action": "assets",
    "message": {
        "assets": [
            {
                "id": 1,
                "name": "EURUSD"
            },
            {
                "id": 2,
                "name": "USDJPY"
            }
        ]
    }
}

2. Subscribe to the specified asset

Description:
Subscribe to the incoming Exchange Rates records live data by:
1. Retreiving Exchange Rates for the last 30 minutes;
2. Receiving Exchange Rate data live.
Endpoint: "/"
Message: "{"action": "subscribe", "message": {"assetId": 1}}"
Response:

1. The initial response is exchange rate records per second for last 30 minutes or 1800 seconds.
   Response sample:

{
  "message": {
      "points": [
          {
              "assetName": "EURUSD",
              "time": 1455883484,
              "assetId": 1,
              "value": 1.110481
          },
          {
              "assetName": "EURUSD",
              "time": 1455883485,
              "assetId": 1,
              "value": 1.110948
          }
      ]
  },
  "action": "asset_history"
}

2. The incoming live Exchange Rate records:
   Response per each row:

{

    "action": "point",
    "message": {
        "assetName": "EURUSD",
        "time": 1455883484,
        "assetId": 1,
        "value": 1.110481
    }
}

Technical details

Stack

• Python
  • Poetry
  • FastAPI
  • Beanie
    • Motor
  • pytest / pytest-asyncio
• MongoDB
• Docker / Docker-compose

General components

The app components are technically described by the docker-compose.yaml file.

Back end

The main FastAPI web server to handle incomming connections.
The only websocket URI it serves is ws://0.0.0.0:8080/ - the websocket connection function described in Websocket actions.

Async periodic tasks

The asynchronous tasks running in a single thread to:

1. fetch the exchange rates data from an externcal resource;
2. save the results into the databasse.
   The primary goal of the asynchonorous tasks is to provide fresh exchange rate records per each second.

Data Base

A MongoDB DBMS instance to store and serve the application data in form of documents.

Contribute

Install pre-commit
poetry add pre-commit

Init pre-commit by installing the hooks
poetry run pre-commit install

Test with Mypy against the code (src/) depending on the automatically imported mypy.ini file:
poetry run mypy src