Kylin API

This API is set up for use with Python >= 3.7 and Docker. You can set-up your local environment manually or compose up with docker to launch a containerised version of the API.

git clone https://github.com/Kylin-Network/kylin-api.git

Running with Docker

To run the server with Docker, you'll need to install Docker if you havent already. Then, you can run:

docker-compose up -d

This will start two Docker containers:

• kylin-api: Gunicorn server that wraps the Flask app defined in api/app.py
• postgres: PostgreSQL database

You should now be able to send:

curl localhost:8080/health

And receive the response OK and status code 200. You can see other example calls, here.

Your server and database are running in a detached state as indicated by -d. When you are ready to bring down your server, run:

docker-compose down

Running Locally

To run the server locally, you'll need to install a few requirements. To do this, run:

pip install -r requirements/common.txt

If you are running a local PostgreSQL instance, create an 'SQLALCHEMY_DATABASE_URI' environment variable:

export SQLALCHEMY_DATABASE_URI="YOUR CONNECTION STRING"

If you plan on writing to the database, you'll need to create a parachain_data table as defined in schemas.sql.

Finally, to boot up the server, run:

bash bin/run.sh

You should now be able to interact with your server as described above.

Example Calls

Get current price data:

curl "http://localhost:8080/prices?currency_pairs=btc_usd"

Get historical price data:

curl "http://api.kylin-node.co.uk/prices/hist?currency_pair=eth_usd&before=1642438800&after=1642352400&period=300"

• currency_pair: currency pair to query price data for
• after: Timestamp to return candles opening after this time
• before: Timestamp to return candles opening before this time
• periods: Comma separated integers representing seconds. Only return these time periods. (60,180,300)

Write to database:

curl -d '{"data": "This is json serializable data", "feed": "demo_feed", "block": "1", "hash": "demo_hash"}' -H "Content-Type: application/json" http://localhost:8080/submit

Query database:

# select all data
curl http://localhost:8080/query/all

# select by hash
curl "http://localhost:8080/query?hash=demo_hash"

# select by feed
curl "http://localhost:8080/query?feed=demo_feed"

Testing the API

Testing the API is set up using pytest. To execute tests you can install the project's development dependencies with:

pip install -r requirements/develop.txt

Then from the root directory, run:

pytest

This runs tests/test_api.py which contains test functions.

Accessing the Swagger

With the application running, use the browser to search the following:

http://localhost:8080/

You can see the API's specification and try it directly from the swagger UI.

Inside the default namespace you will see the list of the endpoints available. You can test them using the try_out button.