Refactor: Uncouple from HTTP and use data models

lornajane commented 5 months ago

What/Why/How?

When we created the Museum API, we stopped iterating at the HTTP data transfer level. This PR doesn't change much of the API description (removed phone numbers from ticket records) but takes a more RESTful approach of modelling entities.

Still needs work: since the ID fields are readonly, I ran into https://github.com/Redocly/redocly-cli/issues/1416 which marks the examples as invalid.

Check yourself

[x] Code is linted
[ ] Tested with redoc/reference-docs/workflows (internal)
[ ] All new/updated code is covered with tests

Security

[x] Security impact of change has been considered
[x] Code follows company security practices and guidelines

lornajane commented 1 month ago

@rudi23 Finally got back to this! I don't think the 404 responses make sense for the collection endpoints because they exist, I've defined them .... does that make sense? I picked up most of your other comments and removed the readOnly fields since our linting doesn't handle these yet. I've left optional fields in place an we can iterate.

rudi23 commented 1 month ago

@rudi23 Finally got back to this! I don't think the 404 responses make sense for the collection endpoints because they exist, I've defined them .... does that make sense? I picked up most of your other comments and removed the readOnly fields since our linting doesn't handle these yet. I've left optional fields in place an we can iterate.

👍 Yes, I totally agree, not sure why I recommended 404 for the collection endpoint 😅

Redocly / museum-openapi-example