fix: use the proper logs for filebrowser and add generic testBaseLine function

[phorcys420]

Follow-up to #400 that fixes the tests.

• Added the bash and curl deps
• Made the script stricter via set -euo pipefail (previously, the script was completely failing due to unknown commands but exit code was 0, hence why tests passed even though our testing setup was broken)
• Removed the toEqual calls and replaced them with toContains to avoid the tests breaking if the log lines aren't exactly what is set.

The new script version doesn't log command (e.g "Running 'filebrowser --noauth --root /home/coder/project --port 13339 --baseurl ' ") so I also removed that, but I could reimplement this because we only test for the exit code and generic logs now, but we don't actually check if the proper command is ran.

In the long-term, we should overhaul our testing to do proper e2e testing (e.g actually runnign healthchecks on the modules' backend), but in the mean-time we should also make all of our scripts stricter to be able to actually catch these issues.

[kylecarbs]

I'm a bit confused isn't this just formatting?

[kylecarbs]

Oh, oops. My diff just showed the formatting, but looks good.