Docker image for Python development

Main Objective

To provide execution environment during Python development

Why do I want this?

When I prepare test environment for multiple Python projects, it was a bit troublesome to prepare / update each environment.

Solution

By using Docker, provide consistent way to prepare test environment.

Usage

1. Copy build_pydev.bat and Dockerfile to your module's directory
   • see "Requirements" for details
2. Build image (whennever you update code, do this step again)
   • run build_pydev.bat ${module_name}
3. Example usage
   1. Run test
      • docker run pydev_${module_name} pytest
   2. Jupyter (note notebooks are just created in the container and will not be saved by default)
      • docker run -p ${port}:8888 pydev_${module_name}
   3. General interactive shell
      • docker run -i -t pydev_${module_name} /bin/bash
      • e.g. upload built wheel to pypi python3 -m twine upload --repository pypi dist/*
   4. Try Flask app
      • docker run -p ${port}:5000 pydev_${module_name} /bin/bash
      • Then use /home/neo/venv/bin/flask to launch your Flask app (requirements.txt should have flask)

How it works?

• Copy everything in current directory in container
• Install dependent packages by your requirements.txt
  • We split dependent module installation and your module installation to save building time
• Install your package by your setup.py
  • Tips: You can use pkg_resources.parse_requirements to specify dependent packages from requirements.txt

Requirements

When you build image, current directory must have

• Dockerfile and build_pydev.bat in this repository
• Following files for your Python package
  • ./requirements.txt which describes all the dependent Python libraries
  • ./setup.py and ./README.md to install your module