armada-client 0.2.0

Armada gRPC API python client

Armada Python Client

---

How To Use This Application

---

These instructions are intended for end-users. If you wish to develop against armada_client, please see our documentation on contributing

1. Click on armada_client.
2. Read the file CONTRIBUTING.md
3. Click here get an overview of Armada client libraries

Python client wrapping the gRPC services defined in submit.proto and events.proto; allows for

• submitting, cancelling, and re-prioritising jobs, and
• watching for job events.

Build

Prerequisites:

For building the python client: In the repo level, run make python then cd client/python and lastly run pip install . to install the client from git with all necessary dependencies.

1. pyenv
   • Sets up local python environment for supporting multiple python environments
   • Set up a local python 3.9 environment
2. pip
   • Package is defined by pyproject.toml
   • pip install . will pull dependencies and install based on pyproject.toml
3. tox
   • tox -e format will check formatting/linter for your code according to default black settings and flake8
   • tox -e py39 will run unit tests with your default python 3.9 environment
4. Auto formatting
   • tox -e format-code will run black formatter on client and testing code.

CI

We use tox for running our formatting and testing jobs in github actions.

Testing

gRPC requires a server to start so our unit tests are not true unit tests. We start a grpc server and then our unit tests run against that server.

poetry run pytest tests/unit/test_client.py

This is just a simple test that starts a grpc server in the background and verifies that we can call the client.

Releasing

This is to be automated.

Manual Release:

Log into the Armada-GROSS account on PyPI. Generate the API tokens and copy those tokens to ~/.pypirc. poetry build poetry run twine upload dist/*