[tfe_workspace_run] Allow configuration without apply and destroy blocks

[tmatilai]

Description

Allow configuring the tfe_workspace_run resource without apply and destroy blocks. This makes it possible to remove the destroy block from a resource without an apply block, to avoid a destroy run to be scheduled like when removing the whole resource.

This would be especially useful when managing workspaces and workspace runs with automation, allowing to leave the empty resources in state without worrying about accidental destroy runs.

Remember to:

• Update the Change Log
• Update the Documentation [No changes, the documentation doesn't say that at least one block is required]

Testing plan

Example configuration, which errors out without this change.

resource "tfe_workspace" "test" {
  name = "foo"
}

resource "tfe_workspace_run" "test" {
  workspace_id = tfe_workspace.test.id
}

Output from acceptance tests

[The test setup seems to have changed, disallowing running the test suite using own HCP Terraform organization]

$ TESTARGS="-run TestAccTFEWorkspaceRun" make testacc

...

[tmatilai]

This would be especially useful when changing the execution mode of a workspace from remote/agent to local. For locally executed workspaces the API obviously fails if runs are tried to be created. So the only option now would be to remove the tfe_workspace_run resource, but that would trigger a DESTROY run for the workspace, assuming that the destroy block was specified.