Firstly, I would suggest hat don't use ./<filename> - it's cleaner to just do !include Person.puml and to specify plantuml.include.path with the folder containing that file.
Assuming you are ok with that, the following should work for you
The following method should only be used IF you are running plantuml server on your local laptop via docker and like to use it instead of the jar with say visual studio code simply due to the speed boost it provides over the jar. I wouldn't suggest using this on a remote server.
It works on the principle that docker allows you to mount folders on the host as volumes , and thus we are basically mounting the folder with the files to be included as /include inside the container
You probably want to start one container per plantuml project and change the path and port on host accordingly
first start a container - edit the path to the include folder that contains the files you want to include - i usually point it to an include folder in my project
docker run -d -p 8080:8080 -v "<path-to-your-local-include-folder>":/include --name plantuml plantuml/plantuml-server:jetty
This creates /include inside the container pointing to the folder in your host.
I have deliberately used --name plantuml so that the container is called plantuml for the steps below. Use any name as required, e.g. your project /workspace name for easy reference.
we then need to edit the jetty.start file to pickup this folder
create a new file called jetty.start with the following content (single line)- the bits I have highlighted in red are the options that I have added to the existing jetty.start that comes with the container- you only need the -Dplantuml.include.path=/include option in there - the rest are various useful tunings that i tend to use - edit as required (e.g. i notice in docker stats that the container tends to go higher than 256M on first invocation while on startup is around 133M, so hence set the minimal to 256M accordingly.
use docker cp to update jetty.start
docker cp <path to jetty start on your local>/jetty.start plantuml:/var/lib/jetty/.
docker stop plantuml
docker start plantuml
to restart the container to pickup the changes
Once done - test it - on VSCode, plantuml preview picked it up in a second and is lightning fast now
if you are happy with this then
docker commit plantuml plantuml/plantuml-server:jetty
so that this is preserved inside your local image- so that when you spin additional containers for other projects it picks up the include folder specified in the volume
Obviously if you messup your image, simply delete the local dockermage using docker image rm (remember to docker stop <containername>, docker rm <containername> first!) and use docker run again so that the original image is fetched from dockerhub. Repeat and rinse and hopefully it works this time
If you need to add folders in your include path than just one you should be able to specify it like this seperated by a colon in the java args (inside jetty.start) - I haven't tried this but this is referred from https://forum.plantuml.net/2410/multiple-include-directorie
and for each path you can map it as a volume in your docker startup using one -v argument per volume
Hope this helps!
P.S : I am probably breaking a couple of rules here - this link
does suggest that i can pass environment variables to plantuml running in the docker image using -e , however i couldn't get it to work unfortunately hence had to resort to editing jetty.start - i assume that only specific environent variables are allowed by default so it wouldn't pickup the plantuml.include.path by default without me tweaking the dockerfile script.