Another Update on Publishing my Emacs Configuration

2022-02-20

424 words, ~3min reading time

A few weeks ago I wrote a post about how I experimented with publishing my Emacs configuration (which is written in Org) using org-publish. Kashual Modi, the creator of ox-hugo, replied to me and asked me if I thought about publishing the configuration using ox-hugo. I didn't! And it turned out that it was done by just adding three lines at the top of my Emacs configuration file as I wrote in a follow-up post a few days later. I was really astonished and didn't know what to do. Should I choose the org-publish or the ox-hugo path?

Well, after writing the blog post I didn't invest much time in thinking about what solution I should use and just got on with other stuff. Until I made some changes to my Emacs configuration last week and wanted to display these changes online. At this point, I wanted some CI/CD solution so that I don't need to take care of the building and publishing manually.

For some reason, it seemed a little bit easier for me to use the solution I wrote using org-publish instead of importing my dot-emacs repository into the GitLab pipeline (for the sake of completeness: I know that this is not only possible but also quite easy but decisions don't need to be rational all the time ;) ). So I decided to quickly set up my own pipeline for the dot-emac repository using a slightly adjusted version of the pipeline the builds and publishes my website.

The resulting GitLab CI pipeline configuration (.gitlab-ci.yml) is quite easy (well at least the script for the build stage, admittedly the before_script is not that obvious).

before_script:
  - apk add --no-cache openssh
  - eval $(ssh-agent -s)
  - echo "$SSH_PRIVATE_KEY" | tr -d '\r' | ssh-add -
  - mkdir ~/.ssh
  - chmod 700 ~/.ssh
  - echo "$SSH_KNOWN_HOSTS" | tr -d '\r' >> ~/.ssh/known_hosts
  - chmod 644 ~/.ssh/known_hosts

I first define a before_script for setting up the SSH configuration for uploading the published files to my server.

build:
  image: silex/emacs:27.2-alpine-ci
  stage: build
  script:
    - emacs -Q --script publish/publish.el
    - apk add --no-cache rsync
    - rsync --archive --verbose --chown=gitlab-ci:www-data --delete --progress -e"ssh -p "$SSH_PORT"" public/ "$SSH_USER"@mmk2410.org:/var/www/config.mmk2410.org/

Using the Emacs Docker image from silex I run the publish Emacs Lisp script I wrote earlier, install rsync and upload the resulting website files in the public folder to my webserver.

As you can see I again defined four SSH related variables:

• $SSH_PRIVATE_KEY: The private key for uploading to the server.
• $SSH_KNOWN_HOSTS: The server public keys for host authentication. These can be found by executing ssh-keyscan [-p $MY_PORT] $MY_DOMAIN (from a trusted environment, if possible from the server itself).
• $SSH_PORT: The port at which the SSH server on my server listens
• $SSH_USER: The user as which the GitLab CI runner should upload the files.

After a few stupid mistakes regarding the place of the publish.el script, the paths in the script and the public/ folder I got it running quite fast and now always have my config.mmk2410.org page up-to-date.

Regarding ox-hugo: As long as the scripts I wrote for using org-publish work I will probably continue using this solution. But if it fails someday in the future and/or I would need to make some larger adjustments I will more likely switch to ox-hugo.

Day 15 of the #100DaysToOffload challenge.