[Python] Generate each resource in separate namespace

[kanterov]

Changes

Change code generation to make each namespace (e.g., databricks.bundles.jobs) isolated from another, following the design of Databricks SDK v1.

Since we already exported all necessary classes from namespace packages, imports don't change. There are 2 copies of the cluster-related classes, and they can't be used interchangeably anymore when constructing jobs and pipelines. One can be converted into another through dictionaries:

import databricks.bundles.jobs as jobs
import databricks.bundles.pipelines as pipelines

job_init_scripts = jobs.InitScriptInfo(...)
pipeline_init_scripts = pipelines.InitScriptInfo.from_dict(job_init_scripts.to_dict())

Why

It uncouples resources from each other and allows us to evolve their schema independently

Tests

Updating existing tests and reviewing codegen diff

[eng-dev-ecosystem-bot]

Run: 17046048154

	Env	✅‌pass	🙈‌skip
✅‌	aws linux	310	482
✅‌	aws windows	311	481
✅‌	aws-ucws linux	423	381
✅‌	aws-ucws windows	424	380
✅‌	azure linux	310	481
✅‌	azure windows	311	480
✅‌	azure-ucws linux	425	378
✅‌	azure-ucws windows	426	377
✅‌	gcp linux	309	483
✅‌	gcp windows	310	482