[IT-4485] Refactor to prep for adding new endpoint

[ConsoleCatzirl]

Refactor the lambda into multiple submodules in preparation of adding a new API endpoint for getting the current balances. The new API endpoint will use the chart of accounts for a different segment (not "Program") and collate with data from an additional upstream API endpoint.

Also rename 'mips' to 'mip' to accurately reflect the name of the upstream service.

[coveralls]

coverage: 95.076% (-0.9%) from 95.968%
when pulling 51c5912 on ConsoleCatzirl:it-4485-refactor
into bb18342 on Sage-Bionetworks-IT:master.

[brucehoff]

Typically long lists are handled by pagination, i.e. the client can limit the (page) size but then it can also request later page(s) to get all the content. What's the implication of having truncation without a way to get the rest of the content?

[ConsoleCatzirl]

Interesting, I hadn't realized that this is basically half of a pagination implementation. The use case this is supporting is truncating the chart to the first 25 entries for configuring Service Catalog tag options: https://github.com/Sage-Bionetworks-IT/organizations-infra/blob/master/sceptre/scipool/config/prod/sc-tag-options.yaml#L16

[brucehoff]

Thank you for providing more context. So it's critical here that the items in the chart are prioritized, so that if a client can only handle n items, it can say, "give me the n most important ones." (Maybe it's more analogous to string truncation than to pagination.) With this understanding I think the code is fine, but perhaps you can just add a comment like, "limit the response to the chosen number of high priority items."

[Copilot]

Pull Request Overview

This PR refactors the lambda codebase by breaking apart a monolithic mips_api module into multiple focused submodules in preparation for adding a new API endpoint. The refactoring improves code organization and maintainability while also correcting the service name from 'mips' to 'mip' throughout the codebase.

Key changes include:

• Splitting the single mips_api/__init__.py file into separate modules (chart.py, upstream.py, ssm.py, s3.py, util.py)
• Renaming the package from mips_api to mip_api to accurately reflect the upstream service name
• Adding S3 lifecycle configuration for cleaning up old object versions

Reviewed Changes

Copilot reviewed 11 out of 11 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
tests/unit/test_handler.py	Updates all test imports and function calls to use the new mip_api module structure
template.yaml	Changes Lambda handler reference and adds S3 lifecycle policy for version cleanup
mips_api/init.py	Removes the original monolithic module (543 lines deleted)
mip_api/util.py	New utility module containing helper functions for parameter parsing and validation
mip_api/upstream.py	New module handling all upstream API interactions with MIP Cloud
mip_api/ssm.py	New module for AWS Systems Manager parameter retrieval
mip_api/s3.py	New module for S3 cache operations
mip_api/chart.py	New module for chart of accounts processing and business logic
mip_api/init.py	New main module containing only the Lambda handler function
.github/workflows/test.yaml	Removes pre-commit job from the workflow
.coveragerc	Updates source path to use new mip_api module name

Comments suppressed due to low confidence (1)

mip_api/util.py:50

• [nitpick] The variable 'number' is not descriptive. Consider renaming to 'limit_value' or 'parsed_limit' to better indicate its purpose.

            number = int(params["limit"])

[brucehoff]

I left one follow up suggestion (documentation) but I don't think that should hold up approving.