[uss_qualifier] oir_simple: check oir can be created without sub and attached to explicit sub #804
Conversation
Shastick
force-pushed
the
oir-simple-1-check-explicit-can-be-added-to-none
branch
from
9c1d145 to
358cde9
Compare
…attached to explicit sub
Shastick
force-pushed
the
oir-simple-1-check-explicit-can-be-added-to-none
branch
from
358cde9 to
6c9828a
Compare
| @@ -34,6 +34,14 @@ | |||
| ) | |||
| from monitoring.uss_qualifier.suites.suite import ExecutionContext | |||
|
|
|||
| # The official DSS implementation will set an OIR's subscription ID to 00000000-0000-4000-8000-000000000000 | |||
There was a problem hiding this comment.
Not relevant to this PR, but IMO that means the OpenAPI is wrong and the subscription ID should not be a mandatory field. We should consider fixing that.
There was a problem hiding this comment.
Yes, that would be the correct approach.
I think this was used as an immediate stop-gap because the DSS needed to be made compliant (by allowing ACCEPTED OIRs without subscriptions, which the standard explicitly allows) without breaking the OpenAPI spec, but we lost track of this.
I created #806 to keep track of it
| @@ -8,3 +8,10 @@ Check query succeeds | |||
|
|
|||
| If the OIR returned by the DSS under test is not attached to the expected subscription, | |||
| it is in violation of **[astm.f3548.v21.DSS0005,1](../../../../../../requirements/astm/f3548/v21.md)** | |||
|
|
|||
| ## [Get referenced Subscription](../sub/crud/read_query.md) | |||
There was a problem hiding this comment.
Do I understand correctly that this should succeed if the the subscription does not exist? That warrants a separate check, or usage of a dedicated fragment 'check subscription does not exist'.
|
|
||
| This step verifies that an OIR can be created in the ACCEPTED state without providing any subscription information (implicit or explicit) in the request. | ||
|
|
||
| ### [OIR is not attached to any subscription test step](./fragments/oir/oir_has_expected_subscription.md) |
There was a problem hiding this comment.
Sounds like this shouldn't be a step but rather included as a fragment under the above step.
|
|
||
| ## Validate explicit subscription on OIR creation test case | ||
|
|
||
| Ensures that the explicit subscription provided upon creation of an OIR is properly validated and attached to the OIR. | ||
|
|
||
| ### [Ensure clean workspace test step](./clean_workspace.md) |
There was a problem hiding this comment.
Same comment as other PR: a 'clean workspace' step in middle of scenario is not a good idea IMO.
| If the DSS under test allows the qualifier to attach an insufficient explicit subscription to a subscription-free OIR, | ||
| it is in violation of **[astm.f3548.v21.DSS0005,1](../../../../requirements/astm/f3548/v21.md)** | ||
|
|
||
| ### [OIR is not attached to any subscription test step](./fragments/oir/oir_has_expected_subscription.md) |
There was a problem hiding this comment.
Sounds like this shouldn't be a step but rather included as a fragment under the above step.
| ### [Attach explicit subscription to OIR test step](./fragments/oir/crud/update_query.md) | ||
|
|
||
| This step verifies that an explicit subscription covering the OIR can be attached to an OIR that currently has no subscription. | ||
|
|
||
| ### [OIR is attached to expected subscription test step](./fragments/oir/oir_has_expected_subscription.md) | ||
|
|
||
| This step verifies that the OIR is attached to the subscription provided upon creation. |
There was a problem hiding this comment.
I don't feel like those belong in this test case?
There was a problem hiding this comment.
The test case's intention is to:
- check that a subscription that does not cover the OIR cannot be attached to it
- check that a subscription that covers the OIR can be attached to it
The first step demonstrates that some validation indeed occurs. The second step demonstrates that the validation logic is correct (at least superficially). In other words, the test case demonstrates that validation occurs and is not buggy.
Possibly I could rename the test case to Properly validate explicit subscription being attached to OIR without subscription?
Thanks. I think we should first figure out the open points on #803 before diving deeper into this PR anyway. |
|
Moved to draft while we figure out things on the parent PR |
Extend
OIRSimplewith a test case that:This covers point (4) on interuss/dss#1088
Note: this builds on top of #803, only the last commit of this PR is relevant.