[ENH] Add new query param for nodes that defaults to a prepopulated index

[alyssadai]

• Closes Add query parameter for node URLs #27
• Closes Add startup event to fetch public node API URLs #28
• Closes Create a template .env file #34

Changes proposed in this pull request:

Changes to existing code:

• NB_NODES environment variable renamed to LOCAL_NB_NODES
• .env file suggested to be named fed.env to differentiate from the node API .env file
• Updated constraints for LOCAL_NB_NODES format in README
  • see also Switching format of local nodes environment variable #32
• regex for matching nodes in LOCAL_NB_NODES updated to be whitespace-insensitive
• Refactored out code to add trailing / to URLs into a util function, reused below

New code:

• Added startup event to create an index of federation nodes that is a set of the public nodes from https://github.com/neurobagel/menu/blob/main/node_directory/neurobagel_public_nodes.json and the local nodes
  • If fetching public nodes fails, raise warning + just use the locally defined ones for federation
  • Existing /nodes/ endpoint updated to return values from federation index
• Added new query parameter that accepts a list of node URLs
• Validation added that queried node URL list is a subset of nodes known to the current (federation) API instance (i.e., a subset of the federation node index)
  • If not, throw an HTTPException
• Added processing for queried node URL list to remove empty strings and duplicates

Checklist

• PR has an interpretable title with a prefix ([ENH], [FIX], [REF], [TST], [CI], [MNT], [INF], [MODEL], [DOC]) (see https://neurobagel.org/contributing/pull_requests for more info)
• PR links to GitHub issue with mention Closes #XXXX
• Tests pass
• Checks pass

For new features:

• Tests have been added

For bug fixes:

• There is at least one test that would fail under the original bug conditions.

[coveralls]

Pull Request Test Coverage Report for Build 6938616496

• 148 of 148 (100.0%) changed or added relevant lines in 6 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+5.2%) to 98.127%

---

Totals
Change from base Build 6906927584:	5.2%
Covered Lines:	262
Relevant Lines:	267

---

💛 - Coveralls

[surchs]

Thanks @alyssadai for this big PR. I left a couple of comments, mostly for clarity. Take a look.

Good to go otherwise: 🧑‍🍳

[surchs]

does this mean that the get_terms request will always go out to all known nodes, regardless of any subselection the user might have made? in other words, wouldn't we also want to restrict this by

node_urls : list[str]
        List of Neurobagel nodes to send the query to.

as above in async def get?

[alyssadai]

I've opened a separate issue for this

• Allow selective /{data_element_URI} request based on specified nodes #36

as I think this could benefit from some further discussion.

Right now, I think only the query tool uses this endpoint (/attributes/{data_element_URI}) to get all the instances of a given controlled term like nb:Assessment (across the known graphs) to populate the dropdown options for the UI. I imagine this would be separate from/before the user actually selects specific nodes in the app to send a query to - not sure if we want the query tool to have dynamically changing dropdown options as a result.

[surchs]

isort? mind opening an issue to add this to the CI?

[alyssadai]

isort is already included in the precommit hooks for this repo! 🙂

federation-api/.pre-commit-config.yaml

Lines 25 to 29 in f85c44b

	- repo: https://github.com/pycqa/isort
	rev: 5.12.0
	hooks:
	- id: isort
	args: ["--profile", "black", "--filter-files"]

I think this is the expected import order following

1. standard library imports
2. related third party imports
3. local application/library specific imports

[surchs]

🚀 PR was released in v0.1.0 🚀