codekoala/sikuli repository overview

⁠Sikuli In a Container

This container should allow you to quickly and easily get Sikuli working on any Linux system that is a Docker host.

docker run -it --rm -u $UID -v $(pwd):/root -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=unix$DISPLAY codekoala/sikuli

In this command, we're mapping the current directory to /root in the container. This is how we can share our existing Sikuli tests with a new container, or save new tests that can be reused later.

Next, we have to mount our X11 socket inside of our container so Sikuli can access our existing X11 session. Along with that, we set an environment variable (DISPLAY) to point to this socket in the container.

If you simply wish to run an existing Sikuli script, use the command above and add a -r relative/path/to/test.sikuli.

docker run -it --rm -u $UID -v $(pwd):/root -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=unix$DISPLAY -w /root codekoala/sikuli -r test.sikuli

⁠Simplify

You can simplify using containerized Sikuli quite a bit if you create some aliases. For example:

_dsikuli='docker run -it --rm -u $UID -v `pwd`:/root -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=unix$DISPLAY '
alias sikuli="$_dsikuli codekoala/sikuli "
alias sikuli-run="$_dsikuli -w /root codekoala/sikuli -r "

You can put this alias in your ~/.bashrc or ~/.zshrc, depending on which shell you use. Note the use of single quotes versus double quotes, as well as the trailing space inside of the alias definitions. The single quotes will make it so things like pwd are evaluated when the alias is invoked rather than when it's defined. The space at the end allows us to pass additional arguments to our container, such as sikuli-run test.sikuli.

⁠Troubleshooting

If you get errors about not being able to connect to an X11 session, you may need to run something like xhost + on your Docker host.