Support learner data export in non-LIF formats

[cbeach47]

Currently, the LIF system consumes learner data in disparate schemas and outputs that data in the "Org LIF" format via the LIF API .

There is a desire for the LIF ecosystem to export learner data in formats other than the Org LIF data model.

Requires two new abilities:

• An endpoint needs to consume a learnerId and dataModel, and return data, for that learner, in that data model format
• We need a way for an adopter to enumerate the available "Org LIF to other data models" that have a transformation group setup in MDR.

Notes:

• MDR already supports managing the definitions of various data models
• MDR already supports managing the transformations to convert learner data from a source data model to a target data model.

Design:

• Leave the LIF API endpoint as-is:
  • A learner ID is passed in, the data is returned in the Org LIF data model format.
• Create a new REST-like API micro-service LIF Learner Data Export that is open to the external web:
  • Depends at least on LIF Query Planner, MDR API, and LIF Translator.
  • AuthN / AuthZ for both endpoints:
    • Uses the authorization header whose Bearer token is a JWT token.
    • This token is created from a call to the MDR or Advisor.
    • Data access is dependent on the logged in user's tenant schema.
  • Endpoints:
    • GET [lif-learner-data-export]/available-data-formats?urlEscaped=true|false
      • Enumerates the available data models that have the Org LIF as a transformation source.
      • The response are sets of tuples. All values are URL escaped based on the query parameter urlEscaped:

[
  {
    "name": str,
    "version": str,
    "contributorOrganization":str,
    "transformationVersions": list[str],
  },
  ...
]

- `GET [lif-learner-data-export]?learnerId={LEARNER_ID}&dataModelName={DATA_MODEL_NAME}&dataModelVersion={DATA_MODEL_VERSION}&dataModelContributorOrganization={DATA_MODEL_CONTRIBUTOR_ORGANIZATION}&transformationId={TRANSFORMATION_ID}`
    - Return available data, for that learner, for that data model, using the given transformation version.
    - Does not filter out parts of the schema (as opposed to the LIF API (graphql) endpoint)
    - Internal call flow:
        - Use `MDR API` to confirm the data model is known for the tenant.
        - Use `MDR API` to confirm there is a transformation from Org LIF to the desired data model for that tenant.
        - Call `LIF Query Planner`to retrieve the LIF formatted data 
        - Capture the response and use `LIF Translator` to translate the data into the target schema
        - Return the result

[cbeach47]

Implementation plan:

• Create APIs, build out docker services
• Connect GET [lif-learner-data-export]/available-data-formats endpoint to the MDR API
  • Needs to be target_schema aware
• Connect `GET [lif-learner-data-export]/ to LIF Query Planner, MDR API, and Translator
  • Needs to be target_schema aware
• Setup GHA and AWS config files
• Deploy to AWS Dev, ensure external access
• Deploy to AWS Demo, ensure external access