Provide a default value for Interface.type #97
Conversation
Netbox and Nautobot have different choice options for the Interface.type field. This defaults to using the "other" type in Nautobot when the Netbox type is not supported by Nautobot.
|
I think it's a sane approach that supports any unexpected interface type not supported, and a pattern we can reuse in other cases where the enums are different. |
jmcgill298
force-pushed
the
bugfix-50/interface-type-choices
branch
from
ec75451 to
921f1ab
Compare
jmcgill298
requested review from
jathanism,
lampwins and
chadell
as code owners
|
@chadell @glennmatthews I changed this up to centralize this code some, and some of these models won't have an |
| The `_type_choices` variable should be a set of the valid choices for the type per | ||
| Nautobot. | ||
| """ | ||
| if "type" not in cls.__fields__ or "_type_choices" not in cls.__dict__: |
There was a problem hiding this comment.
Is this the cleanest way to do this, or would something like if not hasattr(cls, "type") or not hasattr(cls, "_type_choices") work?
There was a problem hiding this comment.
actually, the Pydantic definitions are not class attributes, and so hasattr(cls, "type") will fail. However, I think we can simplify this just to check for _type_choices and any implementation that opts in has to define a type field (which why define _type_choices without having a type?)
| # model: dcim.cable | ||
|
|
||
| #- fields: |
There was a problem hiding this comment.
Y'know, I don't recall why the cable instances are commented out in this validation data. Is it because of the use of GenericForeignKeys making it hard to know in advance what the termination_*_id value will be?
There was a problem hiding this comment.
I'm not sure, and I felt like I've spent too much time on this PR already and didn't want to go down another rabbit hole
Netbox and Nautobot have different choice options for the Interface.type field. This defaults to using the "other" type in Nautobot when the Netbox type is not supported by Nautobot.