create docker image via CI

[danthe1st]

This PR creates a docker image using a Java 21 JRE automatically on commits to main.

Specifically, it contains the following changes:

• Gradle and Lombok were updated to work with Java 21 (for a future update)
• The default CI script was changed to run on every branch
• A new CI script was added.
  This script builds a docker image that allows running the bot.

The docker image can be executed like that (where USERNAME/REPONAME is the repository of the docker image, e.g. dan1st/javabot-test)

docker run --rm -v "$PWD/config:/work/config" -v "$PWD/logs:/work/logs" -v "$PWD/db:/work/db" -v "$PWD/purgeArchives:/work/purgeArchives" -p 9000:9000 -p 9001:9001 USERNAME/REPONAME

This command forwards both the HTTP and AJP port.

Publishing the image to Docker Hub requires the following secrets to be present:

• DOCKERHUB_USERNAME: The username (or organization name) for Docker Hub, e.g. dan1st
• DOCKERHUB_REPONAME: The repository name for Docker Hub, e.g. javabot
• DOCKERHUB_TOKEN The Docker Hub token for uploading

If DOCKERHUB_USERNAME is missing, the image will not be uploaded but the build should still work.

For testing, a repository is available here.
An example run can be found here

Alternatives

• It would be possible to use one script for both verifying the build and uploading the Docker image and only build the image on the main branch. This way, there wouldn't be 2 similar scripts running on every push to main. This was implemented.
• It would be possible to require Java 21 for building. This can be done in a future PR.
• A different base image/JRE vendor could be used instead Eclipse Adoptium/Temurin.
• It would be possible to use GraalVM's native-image for a Docker image not requiring a JRE. This was not done due to it being pretty fragile in my experience/quite some work to maintain (I don't want to write more bugreports to Spring/etc. due to issues with that).
• It would be possible to use jlink to create the JRE.
• It may be a good idea to change server.address=127.0.0.1 to server.address=0.0.0.0 in the application.properties. See the merge notes.

Merge notes - check before merging/deploying/using

• The branch main should be protected! Only PRs should be allowed. Make sure this is the case. This seems to be the case already.
• Using this requires changes to the deployment process.
  For example, it would be possible to execute docker pull USERNAME/REPONAME and the aforementioned docker run command in order to (re)start the bot
  This could be done every time the bot is stopped with /redeploy.
• This requires the DB to be located in a directory called db in the directory where the docker run command is executed.
  • This requires moving the DB and changing the systems.json config.
• a full backup of the bot should be created before deploying!
  • It might also be a good idea to automatically create a backup of the config and DB after every restart (or periodically) if this isn't done already.
• To make the API accessible, it is necessary to add an application.properties with the content server.address=0.0.0.0 to the config directory. Something similar may be necessary for AJP though it seems to be accessible from outside anyways. If AJP is used without requiring the HTTP endpoint, this is not necessary.

[danthe1st]

This is an alternative to the effort in #330

[danthe1st]

We might also want to update the README to add instructions for this. I can add that to the PR if necessary/wanted.