.azure-pipelines.yml

# Python package
# Create and test a Python package on multiple Python versions.
# Add steps that analyze code, save the dist with the build record, publish to a PyPI-compatible index, and more:
# https://docs.microsoft.com/vsts/pipelines/languages/python

jobs:

- job: 'Testing'
  displayName: 'Unit Testing'
  pool:
    vmImage: 'Ubuntu 16.04'
  strategy:
    matrix:
      Python36:
        python.version: '3.6'
      Python37:
        python.version: '3.7'
    maxParallel: 3

  steps:
    # Specify to use the python version defined above
  - task: UsePythonVersion@0
    inputs:
      versionSpec: '$(python.version)'
      architecture: 'x64'

    # Additional info about the build
  - script: |
      uname -a
      free -m
      df -h
      ulimit -a

  # Setup python environment
  - script: |
      # Create test environment for package
      conda env create -n test python=$PYTHON_VER -f devtools/conda-envs/openff.yaml
      source activate test
      # Build and install package
      python setup.py develop --no-deps
    displayName: "Conda Env Creation"

  # Display MongoDB, Conda Environment, Python version
  - script: |
      sleep 5
      python -V
      mongod --version
      conda list
    displayName: 'Display'
    

  # Run tests
  - script: py.test -v --runslow --cov=qcfractal qcfractal/
    displayName: 'pytest'