feat: Require SCA for registration

[m-horky]

• Card ID: CCT-612

[github-actions]

Coverage (computed on Fedora latest) •

File	Stmts	Miss	Cover	Missing
rhsmlib/services
register.py	113	23	79%	31, 67, 75–76, 78–79, 81–82, 84–85, 117, 170, 192, 203–204, 206, 221, 223, 235, 248, 268, 275, 277
TOTAL	18339	4668	74%

Tests	Skipped	Failures	Errors	Time
2630	14 💤	0 ❌	0 🔥	36.829s ⏱️

[jirihnidek]

Overall it is good idea. I have small request that is related to the case, when user is not member of any organization.

[m-horky]

@jirihnidek The no-org state is already handled on higher levels. For both CLI and D-Bus, the determine_owner_key has a no_owner_cb callback that stops the registration.

CLI:

subscription-manager/src/subscription_manager/cli_command/register.py

Lines 347 to 349 in a01dabd

	owner_key = service.determine_owner_key(
	username=self.username, get_owner_cb=self._get_owner_cb, no_owner_cb=self._no_owner_cb
	)

subscription-manager/src/subscription_manager/cli_command/register.py

Lines 496 to 501 in a01dabd

	def _no_owner_cb(username):
	"""
	Method called, when there it no owner in the list of owners for given user
	:return: None
	"""
	system_exit(1, _("{name} cannot register with any organizations.").format(name=username))

D-Bus:

subscription-manager/src/rhsmlib/dbus/objects/register.py

Lines 203 to 207 in a01dabd

	organization: str = service.determine_owner_key(
	username=connection_options["username"],
	get_owner_cb=self._on_multiple_organizations,
	no_owner_cb=self._on_no_organization,
	)

subscription-manager/src/rhsmlib/dbus/objects/register.py

Lines 253 to 258 in a01dabd

	def _on_no_organization(self, username: str) -> None:
	"""A function to call when a member is not part of any organization.
	:raises NoOrganizationException:
	"""
	raise NoOrganizationException(username=username)

[jirihnidek]

It looks OK to me, but it would be to modify cockpit integration tests ... I can see that you already try to do it here: candlepin/subscription-manager-cockpit#70 but some tests are still failing...

[m-horky]

I've tried to fix the Cockpit tests, but it got to the limits of the timebox I set up for it, see downstream CCT-624 for details.

[ptoscano]

generally good, few things to change

[ptoscano]

Thank you for all the updates, now it LGTM. After all the cockpit tests were fixed/adapted, now this works for cockpit too :D

@jirihnidek would you like to take a second look at it, after the latest changes?

[jirihnidek]

LGTM