Skip to content
This repository has been archived by the owner on Sep 25, 2023. It is now read-only.

Latest commit

 

History

History
175 lines (108 loc) · 5.86 KB

README.md

File metadata and controls

175 lines (108 loc) · 5.86 KB
Raw

Azure Pipeline Templates

This repository is the counterpart to the Jenkins Shared Library repository and contains various templates that can be used by Azure Pipelines.

Templates

Jobs

python.setup.py.yml

This job is based on setuptools and does the following:

  • builds the Python module
  • validates the module (eg. linter, tests)
  • deploys the module to a PyPI index
  • deploys the module as a Docker image to a Docker registry

Before using the pipeline please create a new variable group with the name GitHub. This group contains variables required during the release/deploy process.

Add the following variables to the group:

  • GH_TOKEN

    The personal access token is used when pushing the version bump commits to the master branch during a release (see details below in Variables -> release). Change the variable type to secret.

  • GH_USER_MAIL

    The email of the user assigned to the personal access token (GH_TOKEN).

    Example: [email protected]

  • GH_USER_NAME

    The name of the user assigned to the personal access token.

    Example: my-ci-user

Variables
release

The Python pipeline allows only on-demand releases. By default the project is only built and validated. If a new release of the Python module and Docker image should be released a new build has to be queued.

In order to initiate a new release the variable release: true needs to be defined during the queuing. This will result in bumping the version (snapshot->release) and result in a deployed PyPI package, a Docker image and two version commits.

dockerSnapshot

By default Docker images are built only during the release process. In case a snapshot image is needed one can queue a new build and set the variable dockerSnapshot: true. This will result in building a Docker image with a snapshot version.

Parameters
dockerDeploy

Flag to enable/disable building and publishing a Docker image with the Python module.

Default: true

dockerDeployFile

Filename of the Dockerfile used for defining the content of the image to be pushed to the Docker registry.

Default: Dockerfile

dockerRegistryConnector

Azure DevOps service connector that defines the Docker registry where the Docker image should be pushed to.

Default: tomtom-docker-registry-bintray

dockerRepo

Name of the repository an image should be pushed to. Also called a namespace in Docker terminology.

A Docker image with a Python module will be called <dockerRepo>/<name-of-module>.

Default: python

pypiConnector

Azure DevOps service connector that defines the PyPI index (eg. pypi.org) where PyPi modules should be deployed to.

Default: testpypi-tomtom-dev

pythonDeployVersion

The Python version used for creating the source distribution package deployed to the specified PyPI index.

Default: 3.6

pythonTestVersions

Specify the Python versions to be used in all jobs other than deploy (which uses just pythonDeployVersion).

Default:

  - name: Python35
    version: "3.5"
  - name: Python36
    version: "3.6"
  - name: Python37
    version: "3.7"
jobs

Specify the job names and VM images. By default this runs on all three of the supported platforms, Linux, macOS and Windows.

Default:

  - name: Linux
    vmImage: ubuntu-16.04
  - name: macOS
    vmImage: macos-10.13
  - name: Windows
    vmImage: vs2017-win2016

Creating Azure Pipelines

This step-by-step guide shows how to make use of Azure Pipelines and these templates in your own project. It shows you how to create a Python project and use the Python template job from this repository.

The following steps will create a Python project generated with the tomtom-international/cookiecutter-python template.

  1. Create a project using cookiecutter

    The cookiecutter will ask you at some point for a project name (project_name). For this tutorial we will use the name example-project.

    cookiecutter gh:tomtom-international/cookiecutter-python

  2. Create a new repository on GitHub Create repo

  3. Push the generated project to GitHub

    # Enter the previously generated project
cd example-project

# Initialize project for Git
git init

# Configure the remote to the newly create Github project
git remote add origin https://github.com/tomtom-international/example-project.git

git add -A
git commit -am "Initial commit"
git push -u origin master

  4. Go to your Azure DevOps account (eg. https://dev.azure.com/your_organization/your_project/_build)

  5. Create a new build pipeline for the newly created repository Create pipeline

  6. Select Github: Create pipeline

  7. Scroll down to the bottom of the list and click on the connection link (Search for 'You may also select a connection'): Create pipeline

  8. Select the service connection for your GitHub account/organization (A detailed description on how to create service connections can be found in the official Azure Devops documentation) Create pipeline

  9. Select the newly created repository: Create pipeline

  10. Click on Run or Save and run.. to trigger the first build and wait until the build finishes to see the results: Create pipeline