Coding environment¶

This page explains how to set up your computer to develop on the RS-Python Python repositories (rs-server, rs-client-libraries, rs-dpr-service, rs-testmeans...). It assumes a Linux or macOS computer. No specific IDE is required, Visual Studio Code and PyCharm are both used by the team.

Prerequisites¶

Tool	Usage
git	Get the source code.
Docker	Run the local mode, build the Docker images.
pyenv (recommended)	Install the Python versions used by the project.
pipx (recommended)	Install the Python tools in isolated environments.
Poetry	Manage the dependencies and the virtual environments of the projects.
pre-commit	Run the code checks before each commit.
Trivy	Security and license checks, run by pre-commit and the CI.

# Install the Python version used by the project (see the "python" constraint of each pyproject.toml)
pyenv install 3.13.16
pyenv global 3.13.16

# Install Poetry (use the version of the CI, see the RS-PYTHON/actions repository)
pipx install poetry

Get the source code¶

Clone the repositories in a common parent folder, e.g.:

mkdir ~/rspy && cd ~/rspy
for repo in rs-server rs-client-libraries rs-dpr-service rs-demo rs-testmeans rs-documentation; do
  git clone https://github.com/RS-PYTHON/${repo}.git
done

The default branch of the repositories is develop.

Set up a project¶

From each repository:

cd ~/rspy/rs-server

# Create the virtual environment and install the dependencies, including the development ones
poetry install --with dev

# Install the OpenTelemetry instrumentation of the installed libraries
poetry run opentelemetry-bootstrap -a install

# Install the git hooks
poetry run pre-commit install

Configure your IDE to use the Poetry virtual environment (poetry env info --path), and to run the formatters (black, isort) when you save a file.

Run the tests¶

The tests are written with pytest and categorised with markers:

poetry run pytest -m unit         # fast unit tests, with mockup servers
poetry run pytest -m integration  # integration tests

Tips:

In rs-client-libraries, the tests start a Prefect test server, which is slow. Instead, start a Prefect server from the local mode with docker compose up prefect-server -d and set SKIP_PREFECT_TEST_HARNESS=1.
The local mode runs the whole system on your computer, and allows to test your local source code without rebuilding the Docker images.

Code quality¶

The checks run by pre-commit and by the CI are described in Contributing. Run them on all the files with:

poetry run pre-commit run --all-files

If pre-commit fails after a dependency update or a git conflict resolution, reinstall it with poetry run pre-commit uninstall && poetry run pre-commit install.

Develop on the cluster¶

You can also develop from your JupyterHub session on a cluster: clone the repositories in your home directory, and add them to the Python path of your notebooks:

import sys
sys.path.insert(0, "/home/rspy/rs-demo/notebooks")
sys.path.insert(0, "/home/rspy/rs-client-libraries")
import rs_client, resources
print(rs_client.__path__, resources.__path__)

See Create a custom flow for the development of Prefect flows.