Skip to content

Tool for the Automation of Cleanup and Analyses: tools for projects and data management at NGI Stockholm

License

Notifications You must be signed in to change notification settings

NationalGenomicsInfrastructure/TACA

 
 

Repository files navigation

Tool for the Automation of Cleanup and Analyses

PyPI version Documentation Status codecov

This package contains several tools for projects and data management in the National Genomics Infrastructure in Stockholm, Sweden.

Run tests in docker

git clone https://github.com/SciLifeLab/TACA.git
cd TACA
docker build -t taca_testing --target testing .
docker run -it taca_testing

Installation

Inside the repo, run pip install .

Development

Run pip install requirements-dev.txt to install packages used for development and pip install -e . to make the installation editable.

Automated linting

This repo is configured for automated linting. Linter parameters are defined in pyproject.toml.

As of now, we use:

  • ruff to perform automated formatting and a variety of lint checks.

    • Run with ruff check . and ruff format .
  • mypy for static type checking and to prevent contradictory type annotation.

    • Run with mypy **/*.py
  • pipreqs to check that the requirement files are up-to-date with the code.

    • This is run with a custom Bash script in GitHub Actions which will only compare the list of package names.

      # Extract and sort package names
      awk '{print $1}' $1 | sort -u > "$1".compare
      awk -F'==' '{print $1}' $2 | sort -u > "$2".compare
      
      # Compare package lists
      if cmp -s "$1".compare "$2".compare
      then
        echo "Requirements are the same"
        exit 0
      else
        echo "Requirements are different"
        exit 1
      fi
      
  • prettier to format common languages.

    • Run with prettier .
  • editorconfig-checker to enforce .editorconfig rules for all files not covered by the tools above.

    • Run with
      editorconfig-checker $(git ls-files | grep -v '.py\|.md\|.json\|.yml\|.yaml\|.html')
      

Configured in .github/workflows/lint-code.yml. Will test all commits in pushes or pull requests, but not change code or prevent merges.

Will prevent local commits that fail linting checks. Configured in .pre-commit-config.yml.

To set up pre-commit checking:

  1. Run pip install pre-commit
  2. Navigate to the repo root
  3. Run pre-commit install

This can be disabled with pre-commit uninstall

VS Code automation

To enable automated linting in VS Code, go the the user settings.json and include the following lines:

"[python]": {
    "editor.defaultFormatter": "charliermarsh.ruff",
}

This will run the ruff-mediated linting with the same parameters as the GitHub Actions and pre-commit every time VS Code is used to format the code in the repository.

To run formatting on save, include the lines:

"[python]": {
    "editor.formatOnSave": true,
}

Git blame suppression

When a non-invasive tool is used to tidy up a lot of code, it is useful to supress the Git blame for that particular commit, so the original author can still be traced.

To do this, add the hash of the commit containing the changes to .git-blame-ignore-revs, headed by an explanatory comment.

Deliver command

There is also a plugin for the deliver command. To install this in the same development environment:

# Install taca delivery plugin for development
git clone https://github.com/<username>/TACA.git
cd ../taca-ngi-pipeline
python setup.py develop
pip install -r ./requirements-dev.txt

# add required config files and env for taca delivery plugin
echo "foo:bar" >> ~/.ngipipeline/ngi_config.yaml
echo "foo:bar" >> ~/.ngipipeline/ngi_config.yaml
mkdir ~/.taca && cp tests/data/taca_test_cfg.yaml ~/.taca/taca.yaml
export CHARON_BASE_URL="http://tracking.database.org"
export CHARON_API_TOKEN="charonapitokengoeshere"

# Check that tests pass:
cd tests && nosetests -v -s

For a more detailed documentation please go to the documentation page.

About

Tool for the Automation of Cleanup and Analyses: tools for projects and data management at NGI Stockholm

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 98.8%
  • HTML 1.1%
  • Other 0.1%