Turn generate_neurobagel_example_jsonlds.sh into makefile

[alyssadai]

The script operations may be better handled with a makefile. Mainly because you have to make sure that the script is being run only from the repo root dir (and a bunch of other stuff is also needed). But we're all human, so we might run it from the wrong dir, and then the docker commands are going to create a bunch of root-ownder folders in the repo - which is annoying.

makefile would make this all a lot easier. And it could also do the submodule getting / branching and so on for you. So I suggest we do that. If this is too big a move, let's split this off into its own PR

Also, we should simplify the internal mountpoint from /data/neurobagel/bagel-cli to /data.

Originally posted by @surchs in #410 (comment)

[alyssadai]

Hey @surchs, following up on this issue - I think the original context for this suggestion was to automate the following steps which are currently needed before you can run the script:

bagel-cli/helper_scripts/generate_neurobagel_example_jsonlds.sh

Lines 4 to 7 in 5a5709c

	# 1. cd into the tests/neurobagel_examples submodule and create a new branch that will contain the updated example files
	# 2. Navigate back to the bagel-cli repository root directory and run this script from there to regenerate the example synthetic JSONLD files inside of the tests/neurobagel_examples submodule
	# in neurobagel_examples.
	# 3. Navigate again to tests/neurobagel_examples and from there, commit the changes, push the changes to the submodule origin, and open a PR there to merge the updated examples.

However, as mentioned today, I think these operations could be integrated into the same existing shell script rather than moving to a Makefile, since the whole script will always need to be run together.

Wdyt?

[surchs]

Hey @alyssadai, that sounds good to me!