Update auto-generation of query result examples

[alyssadai]

• Related to Merge aggregate and non-aggregate dataset examples in query result TSVs neurobagel_examples#45

We need to update the current workflow for auto-generating query result examples, to ensure that the examples are intuitive for a user and represent a typical federated query result structure.

Ideally, we would be able to auto-generate an example that:

• Represents aggregated and non-aggregated dataset results in the same file = one human-readable result TSV example, one machine-optimized result TSV example (this is what we currently have, but the examples were generated manually)

Steps to implement

• ensure that neurobagel_examples submodule exists in query tool repo
• repurpose our f-API success response example (which contains 2 aggregated datasets and 1 non-aggregated dataset) https://github.com/neurobagel/neurobagel_examples/blob/main/api-responses/fapi_query_success_200.json to mock the f-API response to auto-generate the query result TSVs and to avoid having to stand up two separate nodes in the query tool repo
• create workflow to regenerate query result TSVs using above
  • wf should run:
    • whenever the f-API example outputs in neurobagel_examples is updated
    • whenever the query result TSV format itself changes (e.g., columns renamed)

[surchs]

This might be a bit complex, so we should

• think through how to do this automatically and discuss it with the team
• then implement the automatic generation

[rmanaem]

One idea could be to use write an e2e test and mock the response using the f-api success response example save the files as artifact (I think cypress has a setting for this) and then make a PR to the Neurobagel example with the artifacts using the Neurobagel bot.

[alyssadai]

Updated the issue desc with steps to implement. @rmanaem, let me know if the tasks make sense to you!