This implements a JupyterLab kernel to interpret Dockerfiles.

• Line-based execution of Dockerfiles in JupyterLab
• Conveniently launch shell sessions running in a container created from the current image by clicking on a button.
• Command auto-completion
• Syntax Highlighting
• Help text integration with the ? operation
• Export and Import from Dockerfiles
• Custom magics for e.g.
  • Tag the current Docker image with %tag myimage:latest
  • Easily follow best practices of installation with %install apt <packagename> (will expand to a RUN command doing all the required setup and cleanup). Also implemented for pip and npm.
  • Manipulate build arguments with %arg
  • Get an overview of existing build stages with %stages
  • Manipulate the build context with %context

• Currently, only Linux is supported
• Docker needs to be installed
• The user needs to be in the docker group (see e.g. here)
• JupyterLab >= 4.0.0 needs to be installed

With pip:

python -m pip install dockerfile-kernel
python -m dockerfile_kernel.install

With conda (soon!):

conda install -c conda-forge dockerfile-kernel

jupyter lab