Feature request: set UID and GID of Shaarli (nginx) process inside container

[cerebrate]

This is in many ways a dupe of #1349, #1127 and #1031, as it's the same underlying issue, but I'd like to make a feature request to hopefully address the issue permanently.

Essentially, I ran into the same problem they had with Shaarli being unable to write to the data and cache directories, which I tracked down to it needing to be able to access them as its internal nginx:nginx (or, as my admin host knows them, _rpc:crontab). This was fixable by setting the appropriate permissions, but since I'm running Shaarli on a k3s cluster and mounting its volumes from a NAS that have their own requirements where security policy, backups, etc., etc., and concerned, having to set arbitrary file ownership makes for a decent workaround, but an administratively painful long-term solution.

Therefore, I'd like to request the ability to set the UID/GID of the relevant Shaarli processes (I think just nginx) in the deployment, perhaps in a similar way to the various linuxserver.io container images (i.e., set PUID and PGID environmental variables on the container), so that it's easy to configure the container to run in a way that fits existing user/group infrastructure.

Sorry I can't just PR this, but it's not a stack I'm familiar with, I'm afraid.

[nodiscc]

Somehow related to #1882 (#1874) since all installation difficulties/workarounds required when using Shaarli through podman stem from hardcoded UID/GID mapping.