Pydantic model should be generated from LinkML model

[matentzn]

The diff when running

make src/linkml_map/datamodel/transformer_model.py

Is huge, and there are many errors as a result of some mismatch between pydantic and linkml.

I tried updating some things in the LinkML model, but I cannot seem to be able to get the tests to pass. I can already see that, for example, SlotDerivation class does not have a "value" slot:

https://github.com/linkml/linkml-map/pull/81/files#r2276961844

Can we make the make command above work again please so I can do my model edit?

[matentzn]

@amc-corey-cox It would be great if you could guesstimate an ETA for this because I have some dependencies on this.. Sorry about the shove, I am fine to try and do it myself if need be!

[amc-corey-cox]

@matentzn Sorry for the slow response. I am starting to work on this now. I'd hope to finish it up by the end of the day. I'll push my changes in a draft PR in case you want to take a look or take over from me if I'm not able to get it when you need.

Update: I just got started on this at the end of the day. I'll be working on this on Wednesday, tomorrow. I'll update the eta tomorrow after I've made some progress.

[matentzn]

No worries at all, I am glad you are looking into it and it takes however long it takes :) Much appreciated!

[amc-corey-cox]

Okay, now that I've had a chance to get into it. First, I fixed the issues I had created with adding the object_derivations. As I mentioned, this is something I will be changing how we implement later but I wanted to do it properly anyway.

Now, I'm digging into the make process on the pydantic model. Unfortunately, this is complicated by our recent switch to uv to manage the project. I'm having trouble going back to a working version of the pydantic model that was created using the make process. I'm not certain but it appears the pydantic model was edited manually at least a couple times.

I'm continuing to work on this but I wanted to give an update on what is slowing this down.

[matentzn]

iiiiii that sounds nasty! I am sorry I opened that hornets nest!

[amc-corey-cox]

I've now switched from trying to find a previously working pydantic model build to just fixing the current model. This is going to take some time but it is definitely worth it. When I finish, I'll also make the rebuild part of the standard CI and we should avoid having this problem in the future. Look for more work on this early next week.

[matentzn]

Good luck @amc-corey-cox that there are no major gotchas in there and THANKS for doing it!