cdisc-org/usdm ? reverse-engineered prompt

Build me a Python package for the CDISC and TransCelerate USDM model that can take an Excel workbook and turn it into the JSON study definition used by the USDM API.

I want the app to read the study sheets in the workbook, understand the CDISC codes, external terminology, timing values, person names, addresses, and the cross references between sheets, then build a complete structured study model. It should also include the core model classes for the USDM objects, plus a simple way to run an import from a file and inspect the output.

Please make it work with the sample workbook format described in the docs, and add or improve tests using the example files in the test folder. If anything about the current USDM or CDISC terminology needs checking, look up the current docs online if you need to. Keep it practical and easy to use from a script.