fix: hk-* actions migrated to using requirements file

[RobPasMue]

Closes #1115

@ansys/pyansys-core - should we consider removing the Python version for actions such as this one? This was proposed by @greschd in #1115 as well. I haven't got a strong opinion on this: I agree with @greschd that it's not needed and users shouldn't have to worry about it, but it also provided a way for a quick fix in case somebody needed to "solve" this issue right away.

[RobPasMue]

Successful run in : https://github.com/ansys/pyansys-geometry/actions/runs/21132251692

[RobPasMue]

Works in https://github.com/ansys/pyansys-geometry/actions/runs/21132567335/job/60766921963

[moe-ad]

Closes #1115

@ansys/pyansys-core - should we consider removing the Python version for actions such as this one? This was proposed by @greschd in #1115 as well. I haven't got a strong opinion on this: I agree with @greschd that it's not needed and users shouldn't have to worry about it, but it also provided a way for a quick fix in case somebody needed to "solve" this issue right away.

I think it should be left exposed exactly for the reason you mentioned i.e. fixing convenience. Exposing the python version (and having a default value) in actions where python scripting is used has also been a convenient way to quickly identify such actions (especially during general minimum version bumps). There can be other convenience I might not be thinking about atm too.

But I don't have strong opinions either (but mine is definitely stronger than yours 😉).

[greschd]

I think it should be left exposed exactly for the reason you mentioned i.e. fixing convenience.

There are numerous ways the action could break. Only a small number of them are fixed by changing the Python version.

To me, the Python version used is an internal of the action, and shouldn't be exposed.

The downside of having a user-overridden Python version is that:

• it's less well-tested compared to the default version; this could hide subtle bugs
• it increases maintenance overhead on the actions repository if it needs to implicitly support multiple Python versions / dropping Python versions means a breaking change
• inevitably, the user will eventually need to update or remove the override again

[SMoraisAnsys]

In the case of hk-* actions, I don't mind removing the python version used and would advocate to remove it on other actions as long as there is no direct correlation between the action and the repo it is used on. For example, testing, doc build, ... require interaction with the project and it's dependencies. So, no modifications on those :)

Also, the dependency changes align with the effort we discussed on pinning dependencies and the on going POC (stopped ftm) #991

[RobPasMue]

Let me remove it in that case. I'll pin it to Python 3.13

[RobPasMue]

Successful run in https://github.com/ansys/pyansys-geometry/actions/runs/21172889164

[RobPasMue]

Can I get a review and final approval @ansys/pyansys-core if you agree with this?

[AlejandroFernandezLuces]

LGTM, minor comments

[ansys-cla-bot]

The following people have not signed the Contributors License Agreement (CLA):

• @<21alex295

Read the CLA in the link above and sign it by clicking the link below:

Login with GitHub

You will receive a confirmation as soon as your signature is captured.

[RobPasMue]

I'll be merging then thankyou all