tornado_tree

A microservice application written in Python with Tornado and SQLAlchemy, which implements an asynchronous REST API for adding and retrieving values from the tree-like data structure.

• Nodes have non-unique names, optional text data
• Each node has exactly one parent
• Each node can have unlimited number of children nodes
• A node GET request returns its name, data, number of ancestors and list of their IDs
• A node POST request returns the created node's data, along with number of ancestors and their IDs.

API endpoints

Route	Method	Type	Description
/	GET	Text	Get the whole tree (for testing purposes)
/api/v1/get/{id}	GET	JSON	Get a single node data
/api/v1/add	POST	JSON	Add a node on the tree. 'name' parameter is required. A request without a 'parent' ID provided will fail if the tree is not empty.

Installation

Attention: this package is not indexed on PyPI, use the included setup.py in order to build and install it. Using a separate virtualenv is advised. For manual installation, get the dependencies and install the package using setup.py:

$ cd tornado_tree
$ pip install -r requirements.txt
$ pip install ..

tornado_tree makes use of PostgreSQL (9.5+) and psycopg2 as a DB adapter. If you install it manually, you'd have to create the application's DB with PGSQL extension 'ltree' - instead of tampering with 'template1', entrypoint.sh only creates the extension on the database specified by APP_DB_NAME env variable. Please ensure that you have appropriate versions installed, along with postgresql-client if you use psql. Assuming local PGSQL installation and that sudo works on your system:

# User postgres must have password set

$ sudo -u postgres psql -c "CREATE DATABASE $APP_DB_NAME;"
$ sudo -u postgres psql $APP_DB_NAME -c "CREATE EXTENSION IF NOT EXISTS LTREE;"

Alternatively, you can deploy the service with docker-compose, which will handle this automatically. After you've set up the database, you may proceed with schema migrations and running the app.

Migrations and running the app

Environment variables

These variables are provided by the .env script, adjust these settings in the file if necessary before running migrations and launching the app:

Name	Default	Description
APP_DB_USERNAME	'treeuser'	DB account's username.
APP_DB_PASSWORD	'tree123'	DB account's password.
APP_DB_NAME	'treedb'	Name of app's database.
APP_DB_HOSTNAME	'localhost'	DB server hostname, used for app instances outside a container.
APP_DB_PORT	5432	DB server port, exposed in docker-compose configuration.
APP_PORT	3000	DB server port, exposed in docker-compose configuration.
MIGRATE	1	Run migrations before starting the app.
DEBUG	1	Enable debug mode (pretty logging).

Before you run migrations or start the application, export these variables in your shell environment or source the .env file:

$ . ../.env

Migrations

tornado_tree uses Alembic to detect major DB schema changes and generate migration scripts. If you're using a separate virtualenv, it should be available on your path at this point. In order to apply the latest migrations, execute:

$ alembic upgrade head

Running

After you have installed the required packages and successfully ran DB migrations, you should be able to run the service:

$ python entry.py

Make sure that the service is working:

$ curl localhost:$APP_PORT
{"error": "The tree is empty yet"}

Docker deployment

The default docker-compose configuration is intended for development use. Please refer to Environment variables section for additional information. You can also provide custom settings in docker-compose.yml or pass them with -e key when calling docker-compose run. Remember that docker-compose run overrides the entrypoint command defined in the app configuration.

Building images and deploying is easy:

$ docker-compose up