This repo contains tested reference examples of using Python with Azure Pipelines.

This sample contains several Azure Pipelines for Python developers that showcase useful end-to-end patterns of varying complexity. All pipelines are in the .azure-pipelines folder and have been fully annotated and validated.

Pipeline results and sample Artifacts for the definitions contained in this repo can be viewed at the following Azure DevOps organization: az-samples

simple_package is a pure Python package with no external dependencies. It exists to give just-enough structure to show how to use Azure Pipelines with Python.

Build a simple package against a single Python version

Concepts:

• Building for a project contained in a repo subfolder (separate src/tests folders)
• Choosing a Python version
• Installing build dependencies
• Linting (pylint and flake8)
• Running tests (pytest)
• Building a source archive and built distribution
• Capturing build artifacts

Build a simple package against multiple Python versions

Concepts:

• Building for multiple Python versions
• Capturing multiple artifacts per build

Build a simple package against multiple Python versions, then upload it to Azure Artifacts (private PyPI)

Concepts:

• Setting job dependency order
• Conditionally running pipeline jobs
• Disabling code checkout to control the pipeline environment
• Downloading previously uploaded build artifacts
• Authenticating to Azure Artifacts
• Publishing to a private Artifacts feed

simple_server is a minimal Flask application that takes a dependency on simple_package from the Azure Artifacts feed. It can be run as a standalone Python application, or can be built as a Docker container image.

Build a Python application that has a dependency on a package sourced from Azure Artifacts

Concepts:

• Authenticating to Azure Artifacts for pip
• Installing private dependencies
• Distinguish between Python versions in pytest output

Build a Docker image using an application that pulls dependencies from Azure Artifacts

Concepts:

• Customizing the build artifact download path
• Building a containerized Python app with Azure Artifacts
• Using multi-stage builds with pip

This project has been configured for Python development with VS Code using the following extensions:

• Python
• Azure Pipelines
• Docker

To get started, clone the repo, and run the following commands to get started with a new environment

# Clone the repo
git clone https://github.com/Azure-Samples/azure-pipelines-python.git
cd azure-pipelines-python

# Create a virtual environment at .venv
python3 -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements.dev.txt -U --upgrade-strategy eager

# Install packages in editable mode
python -m pip install -e src/simple_package
python -m pip install -e src/simple_server

# Start coding! :)
code .

• Build Python apps
• Publish Python packages in Azure Pipelines
• Azure Pipelines YAML schema