Andela Resource Tracker (ART) Backend

Description

This repository contains the API endpoints and models for the ART project implemented using Django.

• Local Development
  • Env setup
  • Project setup
    • Installation script
    • Manual setup
    • Development using Docker
  • Running the app
• CI / CD

Local Development

Env setup

• Clone the repository: git clone https://github.com/AndelaOSP/art-backend.git
• Open the project directory: cd art-backend
• Create a .env file. We currrently use the following env variables.

Variable	Description	Default
DATABASE_URL	Required - Used by dj_database_url to connect to the database. Format: DATABASE_URL=postgresql://:@localhost:5432/.
SECRET_KEY	Required - String of random characters used to provide cryptographic signing for Django projects.
PROJECT_ID	Required - The Firebase project ID (We use Firebase for authentication).
PRIVATE_KEY	Required - The Firebase private key.
CLIENT_EMAIL	Required - The firebase client email value.
DJANGO_SETTINGS_MODULE	Required (if running the app using gunicorn gunicorn art.wsgi) - settings.prod for prod, settings.dev optional for dev/staging.	settings.dev
SLACK_TOKEN	Optional - The token to authenticate/authorize the slack app used to send slack notifications.
SLACK_LIMIT	Optional - The number of results per page for slack calls.	1000
SLACK_CALLS	Optional - The number of calls to make to clask API before exiting.	10
ASSET_LIMIT	Optional - A number representing the minimum number of allowed available assets to trigger notification on shortage to slack.	0
AIS_URL	Optional - Needed to sync users from AIS.
AIS_TOKEN	Optional - Needed to sync users from AIS.
AIS_LIMIT	Optional - Number of records to fetch from AIS per request (call it pagination).	5000
ART_BUILDS_CHANNEL	Optional - Slack channel when logs will be posted.	#art-builds
OPS_CHANNEL	Optional - Slack channel when art bot posts ops related messages if recipient isn't defined.	#art-test
RETRIES	Optional - Number of times to retry an external request (currently to AIS) if an error other than 401 is received.	3
RETRY_TIMEOUT	Optional - Number of seconds to wait before retrying an external request (currently to AIS) if an error other than 401 is received.	10
LOGLEVEL	Optional - Default log level - error, warning, info, debug.	info
ADMINS	Optional - Email addresses to send error logs to.	art:[email protected],art_group:[email protected]

Project setup

Installation script

The easiest way to set up is to run . ./install_art.sh which does the following:

• Installs the following if they don't exist:
  • HomeBrew
  • Python 3
  • Pipenv
  • dependencies
  • ShellCheck
  • PostgreSQL
• Creates a .env file if it doesn't exist
• Gives you an option to set up the database

Manual setup

• Create a database and add its url to your project environment. eg DATABASE_URL=postgresql://<user>:<password>@localhost:5432/<db>
• Create and activate a virtual environment - we recommend using pipenv for this by running pipenv shell
• Install the project dependencies stored in Pipfile. Run pipenv install --dev.
• Run migrations - python manage.py migrate
• Create cache table python manage.py createcachetable

Development using Docker

To use the Docker setup, ensure you have Docker installed then run the following commands:

• make compose to create the docker-compose file from the template already in the repository. Open the file with your preferred editor and update the environment section under the art-backend service. Replace the values of the varaibles whose values are enclosed in angle brackets with values provided by a fellow team member or team lead.

• make start to build and start the app services. Migrations will be ran here.

• make exec to run any other commands as necessary. eg python manage.py createsuperuser

• make open to run the application

Running the app

• Create a Superuser account: python manage.py createsuperuser
• To create cache tables python manage.py createcachetable
• To run tests: pytest
• Run the app: python manage.py runserver
• You can now log into the admin dashboard on http://127.0.0.1:8000/admin/

CI / CD

We use CircleCI for this. Merging to develop deploys to staging, and merging to master deploys to production.

To ensure consistency we have automated checks for a couple of things. (To run these commands activate the env first pipenv shell, or use pipenv's run command. eg pipenv run pytest):

• Project tests - pytest
• Python (pep8) styling checks - flake8 .
• Black formatter checks - black --diff -S --exclude="migrations|.venv" . (include the -S or --skip-string-normalization option to allow single quotes on strings.)
• Shell scripts styling checks - for file in $(find . -type f -name "*.sh"); do shellcheck --format=gcc $file; done;
• Imports sorting - isort -rc --diff --atomic .. Using --check-only will perform a dry-run
  • Imports should be in the following order:
    • Future imports
    • Python standard library imports
    • Third party packages
    • Local app imports - absolute
    • Local app imports - relative

You can set up the pre-commit Git hooks with the standard styling conventions as per the guide inn the Wiki.