Getting started#

Scitacean is developed using uv to manage Python, dependencies, etc. (see also Dependency management) and Just as a task runner.

Setting up#

Dependencies#

  • Install uv using any of the methods described here.

  • You can install Just as described in its documentation. Or as a uv tool using

    uv tool install rust-just

  • Optional: Install pre-commit to run the same lints and formatters as on CI. See the documentation or use

    uv tool install pre-commit --with pre-commit-uv

  • Install Pandoc with is required building the documentation. pandoc is not on PyPI and needs to be installed through other means. (E.g. with your OS package manager.)

  • Optional: If you want to run tests against a real backend or SFTP server, you also need docker-compose. See Testing for what this is good for and why.

Set up git hooks#

The CI pipeline runs a number of code formatting and static analysis tools. If they fail, a build is rejected. To avoid that, you can run the same tools locally. This can be done conveniently using pre-commit:

pre-commit install

Alternatively, most checks can also be run manually through just. Take a look at .pre-commit-config.yaml for a list of all checks.

Running tests#

The tests can be run using

just test

If you also want to run backend and SFTP tests (requires Docker), use

just test-all

The test and test-all rules accept arbitrary parameters which are passed on to pytest. E.g., filter tests by name and run 4 worker processes:

just test -k <test_name> -n 4

The examples above automatically use the correct dependencies. If you don’t want to use just, take a look at justfile and pyproject.toml to find the relevant commands.

Building the docs#

Build the documentation using

just docs-build

This will build the documentation pages and store them in a folder called html.

You can also automatically check the documentation using

just docs

This takes longer but is recommended before committing changes.

Static type checking#

Run Mypy using

just mypy

Linting the codebase#

Lint the codebase using

just lint

This will run a number of static analysis tools.

Formatting the codebase#

Run a number of formatters on the codebase using

just format