Add None to choices of optional enum in metadata.py#118
Add None to choices of optional enum in metadata.py#118rjoussen merged 1 commit into4C-multiphysics:mainfrom
None to choices of optional enum in metadata.py#118Conversation
|
@gilrrei maybe you could have a quick look :) |
|
@davidrudlstorfer What is the usual workflow for issues like this? Are the fourcipp metadata.py and 4C metadata.py expected to be the same? (or at least almost the same, in regard to functionality) Because it seems like they are, except for very small differences due to some recent changes. Even the |
|
@rjoussen Thanks for the fix :)
tl;dr No current workflow, fix as we go Optimally, they should be identical. There were some ideas in the past to keep them synced; however, it's not so easy without creating cyclic dependencies. In combination with a lack of consensus on what is/isn't part of 4C and, therefore, a lack of a combined testing infrastructure, fourcipp is in this limbo where mismatches are only caught during nightly checks or manually in 4C PRs by fourcipp enthusiasts like @davidrudlstorfer :) Afterwards, they have to be fixed manually in fourcipp. |
|
@gilrrei Thanks for the explanation. It could be worth amending the copilot instructions for reviews by something like: "Changes in metadata.py need to be ported to fourcipp" |
|
@rjoussen could you actually open a PR in 4C for your proposal regarding your addition to the copilot review instructions?:) |
|
@davidrudlstorfer done. |
This should fix #117.
The latest metadata change in 4C (#1725) fixed that optional enums need None (null in yaml) as a valid choice additionally. #1688 used this feature for the first time, that's why it failed now.
This PR basically copies the changes made in 4C #1725
@davidrudlstorfer