xwidget enabled Emacs for WSLg.
This docker image can launch emacs with GUI and xwidget.
- WSL2
- Ubuntu 22.04
launch emacs container with mounts and environments for WSLg.
This sample additionally mounts home directory into /root
.
- reuse container
run with sleep infinity
docker run \
--rm \
--name emacs-wslg \
-d \
-v ~:/root \
-v /tmp/.X11-unix:/tmp/.X11-unix \
-v /mnt/wslg:/mnt/wslg \
-e DISPLAY \
-e WAYLAND_DISPLAY \
-e XDG_RUNTIME_DIR \
-e PULSE_SERVER \
peccu/emacs-wslg:latest
and exec emacs in it.
docker exec -it emacs emacs &
- launch container each time
docker run \
--rm \
--name emacs-wslg \
-it \
--entrypoint emacs \
-v ~:/root \
-v /tmp/.X11-unix:/tmp/.X11-unix \
-v /mnt/wslg:/mnt/wslg \
-e DISPLAY \
-e WAYLAND_DISPLAY \
-e XDG_RUNTIME_DIR \
-e PULSE_SERVER \
peccu/emacs-wslg:latest
- bash scripts
- use exists container or reuse already running container (copy
emacs-wslg.sh
into/usr/local/bin/emacs-wslg
and add execute permission)- this automatically mounts current working directory (
pwd
) into/app
- this automatically mounts current working directory (
- use exists container or reuse already running container (copy
#!/bin/bash
# -*- shell-script -*-
function inShortTime(){
local launch=$1
local restarttime=$2
[ $(($(date +%s) - launch)) -lt $restarttime ]
}
# try exec or up and exec if failed in 10 seconds
# ignore when execed after 10 seconds
restarttime=10
launch=$(date "+%s")
# automatically mounts / into /mnt/host
# workdir is corresponds directory in /mnt/host
function docker_run_d(){
docker run \
--rm \
--name emacs-wslg \
-d \
--entrypoint /usr/bin/sleep \
-v /:/mnt/host \
-w "$(pwd | sed 's:^/:/mnt/host/:')" \
-v ~:/root \
-v /tmp/.X11-unix:/tmp/.X11-unix \
-v /mnt/wslg:/mnt/wslg \
-e DISPLAY \
-e WAYLAND_DISPLAY \
-e XDG_RUNTIME_DIR \
-e PULSE_SERVER \
peccu/emacs-wslg:latest \
infinity
}
docker exec emacs-wslg emacs "$@" \
|| (\
inShortTime $launch $restarttime \
&& docker_run_d \
&& docker exec emacs-wslg emacs "$@" \
)
- reuse the container with docker-compose version
- copy
emacs-wslg_docker-compose.sh
into/usr/local/bin/emacs-wslg
- add execute permission
sudo chmod +x /usr/local/bin/emacs-wslg
- change docker-compose.yml path in the file
- host's root dir (
/
) is mounted in/mnt/host
- copy
#!/bin/bash
# -*- shell-script -*-
# automatically mounts / into /mnt/host
# workdir is corresponds directory in /mnt/host
# docker-compose.yml's path
COMPOSE_FILE_PATH=~/Codes/emacs-wslg/docker-compose.yml
function docker_compose_up-d(){
docker compose \
-f ${COMPOSE_FILE_PATH} \
up -d
}
function docker_exec(){
docker exec \
-it \
-w "$(pwd | sed 's:^/:/mnt/host/:')" \
emacs-wslg \
/usr/local/bin/emacs "$@"
}
# try exec or up and exec if failed
docker_exec "$@" && : || (docker_compose_up-d && docker_exec "$@")
Based version is Emacs 29 (emacs-29 branch from emacs mirror git repo).
WSLg related document is here
- emacs server and emacsclient
- change CMD or ENTRYPOINT into emacs server
- more emacs versions
- script sample
- test host dependencies
- emacs can run without runtime libs like libgtk
- change user in container from root to user