[Python] Add generated resource code

[kanterov]

Changes

Add Python code generated from databricks bundle schema

Tests

Add a few tests for special cases in the generated code

[pietern]

Two things:

• Are you planning to move the codegen into this repository as well?
• The distinction between the "jobs" and "compute" package will go away in the next version of the SDKs. We should plan for the migration path.

[kanterov]

@pietern

1. Yes
2. We already don't export any types from "compute" package. All necessary types are re-exported from "jobs" package, so I can write:

from databricks.bundles.jobs import ClusterSpec, PyPiLibrary, ...

We can manually add overrides to codegen to generate code in the "jobs" package, but then it will introduce divergence from the bundle schema. I think we should update the bundle schema, and codegen will naturally follow. Because of exports, it isn't going to be a breaking change. The only thing that will change is the __module__ value on classes.