fix(ml): handle workspace stuck in soft-deleted state for permanently-delete and create commands

[Copilot]

Related command

az ml workspace delete --permanently-delete / az ml workspace create

Problem

When a workspace is soft-deleted, the ARM GET endpoint returns 404 for it. The SDK's begin_delete() calls self.get(name) first — so az ml workspace delete --permanently-delete on an already-soft-deleted workspace always fails with "workspace not found" before the actual purge REST call is made. This also blocks az ml workspace create with the same name, with no guidance on how to unblock it.

Changes

workspace.py — core fix:

• ml_workspace_delete: catch ResourceNotFoundError from begin_delete separately. When --permanently-delete is set and the workspace is not reachable via get() (already soft-deleted), fall back to calling the underlying REST operation directly with force_to_purge=True, bypassing the SDK's pre-flight get() check:

# Before: SDK's begin_delete() called self.get(name) first → 404 → failed
# After: falls through to direct REST call when workspace is already soft-deleted
del_result = ml_client.workspaces._operation.begin_delete(
    resource_group_name=resource_group_name,
    workspace_name=name,
    force_to_purge=True,
)

• ml_workspace_create: on failure, emit a targeted WARNING with the exact purge command when the error resembles a soft-delete conflict (keywords: soft, deleted, conflict, already exists):

WARNING: If a workspace named 'my-ws' already exists in a soft-deleted state and is blocking
this creation, you can permanently delete (purge) it with:
  az ml workspace delete -g my-rg -n my-ws --permanently-delete

_help/_workspace_help.py / _params/_workspace_params.py — updated help text for ml workspace delete and --permanently-delete / -p to document the soft-deleted workspace purge scenario.

---

This checklist is used to make sure that common guidelines for a pull request are followed.

General Guidelines

• Have you run azdev style <YOUR_EXT> locally? (pip install azdev required)
• Have you run python scripts/ci/test_index.py -q locally? (pip install azdev required)
• My extension version conforms to the Extension version schema

For new extensions:

• My extension description/summary conforms to the Extension Summary Guidelines.

About Extension Publish

There is a pipeline to automatically build, upload and publish extension wheels.
Once your pull request is merged into main branch, a new pull request will be created to update src/index.json automatically.
You only need to update the version information in file setup.py and historical information in file HISTORY.rst in your PR but do not modify src/index.json.

[azure-client-tools-bot-prd]

Hi @copilot,
Please write the description of changes which can be perceived by customers into HISTORY.rst.
If you want to release a new extension version, please update the version in setup.py as well.

[azure-pipelines]

Azure Pipelines:
There may be pipelines that require an authorized user to comment /azp run to run.

[yonzhan]

ML