feat(bigquery-storage): split GAPIC from the bigquery-storage handwritten package

[quirogas]

Decouple the handwritten BigQuery Storage wrapper from the auto-generated GAPIC client layer by importing a standalone version of the GAPIC library. Standalone -api packages are now generated and maintained independently.

• Removed embedded GAPIC generated code (src/v1/, src/v1beta2/) and OwlBot files (owlbot.py, .OwlBot.yaml).
• Added dependency on standalone @google-cloud/bigquery-storage-api (^0.2.0).
• Re-exported BigQueryReadClient, BigQueryWriteClient, adapt, reader, and protos from @google-cloud/bigquery-storage-api.
• Maintained runtime protobuf descriptor extensions required by the adapt and reader sub-modules.

Internal: b/531788771

[gemini-code-assist]

Code Review

This pull request removes locally generated client files, configurations, and tests for various API versions, migrating the handwritten wrapper to use the external @google-cloud/bigquery-storage-api package instead. Feedback on the changes points out that importing the entire package namespace as v1 is incorrect and suggests importing v1 as a named export alongside the other clients to avoid namespace nesting issues.

[gemini-code-assist]

Importing the entire @google-cloud/bigquery-storage-api package as v1 via import * as v1 is incorrect because it binds the entire package namespace to v1. This results in the exported v1 containing a nested v1 property (i.e., v1.v1 would be the actual v1 namespace), and exports the entire API package under the name v1 instead of just the v1 namespace. Instead, import v1 as a named export from @google-cloud/bigquery-storage-api along with the clients in a single import statement.

Suggested change

	import * as v1 from '@google-cloud/bigquery-storage-api';
	import {BigQueryReadClient, BigQueryWriteClient} from '@google-cloud/bigquery-storage-api';
	import {v1, BigQueryReadClient, BigQueryWriteClient} from '@google-cloud/bigquery-storage-api';