build_test.yml

# You can have multiple workflows that are triggered by different
# events (eg a workflow that runs on branch pushes, versus a
# workflow that runs when a GitHub issue is created). Give your
# workflow a descriptive name so that you know what it is.
name: build_test

# What is the event that triggers this workflow? Options include
# push, pull_request, create, issues, etc. This configuration says
# that you should run on pushes to the main branch (ie after a PR
# is merged), or when a pull request is created to merge into main.
on:
  push:
    branches:
      - main
  pull_request:
    # Configuring this means that your PRs will get a GitHub comment
    # from Coveralls with your coverage.
    branches:
      - main

# A list of jobs to run as part of the workflow
jobs:

  # Each job gets a name. This one builds/tests all branches.
  build_test:

    # Provide a runner. Options are listed on GitHub, but include
    # windows, ubuntu, and macos (latest, as well as specific versions)
    runs-on: ubuntu-latest

    # In order to get Conda environments to work properly, we need to
    # tell the runner to start up the bash shell in a way that respects
    # the changes Conda makes to bash profile files.
    # More at:
    # https://github.com/marketplace/actions/setup-miniconda#use-a-default-shell
    defaults:
      run:
        shell: bash -el {0}

    # We specify all versions of Python that we want to validate.
    # GitHub will run the tests for each of these Python versions.
    strategy:
      matrix:
        python-version: ["3.9", "3.10", "3.11"]

    # The actual workflow steps!
    steps:
      # First, check out the repository for the current branch.
      - uses: actions/checkout@v3

      # Then set up Python with a python version from the matrix.
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}

      # This is an Action published by the Conda team to get your
      # Conda environment set up using your environment.yml file.
      # More at https://github.com/marketplace/actions/setup-miniconda
      - name: Set up conda
        uses: conda-incubator/setup-miniconda@v2
        with:
          # Use the environment name from your environment.yml file
          activate-environment: dinero_env
          # The name of the environment file you want conda to use
          environment-file: environment.yml
          # Whether you want conda to automatically update the conda
          # version (helpful when some runners have old versions of
          # conda)
          auto-update-conda: true
          # Which version of python you want conda to use
          python-version: ${{ matrix.python-version }}
          # We don't need the base environment, so don't activate it
          auto-activate-base: false

      # You don't normally need test coverage information when
      # developing locally - we just want it in CI - so we install
      # it here instead of in our environment file (which we also
      # use locally).
      - name: Install additional development dependencies
        run: |
          conda install --yes coverage

      # If there are any other steps in order for your tests to
      # run successfully, you can add those steps here!

      # Next step: run pylint. Anything less than 10/10 will fail.
      - name: Lint with pylint
        run: |
          pylint dinero/**/*.py 

      # Next step: run the unit tests with code coverage.
      - name: Unit tests
        run: |
          cd dinero
          coverage run -m unittest discover
          coverage report
          coverage lcov

      # Lastly: report code coverage to Coveralls.
      - name: Report code coverage to Coveralls
        uses: coverallsapp/github-action@master
        with:
          github-token: ${{ secrets.GITHUB_TOKEN }}
          base-path: dinero
          path-to-lcov: dinero/coverage.lcov
          # The following two options are needed for "parallel"
          # runs - ie if you have a matrix of python versions.
          # This holds off on reporting code coverage to
          # coveralls until all runs are done. If you don't have
          # a matrix, this isn't needed.
          flag-name: run ${{ join(matrix.*, ' - ') }}
          parallel: true

  # After all the parallel matrix runs for build_test are done
  # for the various python versions, we run this job to send the
  # final code coverage to Coveralls.
  finish:
    # This job requires that build_test be finished before it
    # can run.
    needs: build_test
    if: ${{ always() }}
    runs-on: ubuntu-latest
    steps:
    - name: Coveralls Finished
      uses: coverallsapp/github-action@master
      with:
        github-token: ${{ secrets.GITHUB_TOKEN }}
        parallel-finished: true
        # This says finish the parallel runs that we did above
        # and send to Coveralls.
        carryforward: "run-3.7,run-3.8,run-3.9,run-3.10"