AlexsLemonade / OpenScPCA-analysis

An open, collaborative project to analyze data from the Single-cell Pediatric Cancer Atlas (ScPCA) Portal
Other
1 stars 8 forks source link

Bump mkdocs and update links action #546

Closed sjspielman closed 4 days ago

sjspielman commented 5 days ago

I recently updated my local version of material-mkdocs, and found that this updated version flags bad anchor links which had not been flagged previously. After playing around with this a bit in the now-closed #545, I realized that in addition to bumping mkdocs-material, I also needed to update the check-links.yml GHA itself to also grab anchor link warning messages. This GHA indeed failed as expected: https://github.com/AlexsLemonade/OpenScPCA-analysis/actions/runs/9683208495/job/26718098940

This PR takes care of all these items:

Edit: One additional question I have is whether we actually don't want the links action to be required to pass before merging into main. During docs development, we may actually expect some #STUB_LINKS. We don't necessarily want that to hold up merging into a feature branch like reproducibility-docs, but we definitely don't want bad links going into main. I don't have the permissions to make this change though, so if we want to make it we can tag in someone else who can add that protection rule.

allyhawkins commented 5 days ago

Edit: One additional question I have is whether we actually don't want the links action to be required to pass before merging into main. During docs development, we may actually expect some #STUB_LINKS. We don't necessarily want that to hold up merging into a feature branch like reproducibility-docs, but we definitely don't want bad links going into main. I don't have the permissions to make this change though, so if we want to make it we can tag in someone else who can add that protection rule.

I would agree that we don't want things that are broken being merged into main, so I would keep the check there. I think if you want to make the check run but don't require it to pass on feature branches then we do need to tag in someone to update branch protection. I assume it's turned on for reproducibility-docs given merging being blocked here, so maybe we adjust to run the CI but it doesn't have to pass CI until it is merged into main?

sjspielman commented 5 days ago

I assume it's turned on for reproducibility-docs given merging being blocked here, so maybe we adjust to run the CI but it doesn't have to pass CI until it is merged into main?

Yes, this is what I was thinking!

sjspielman commented 4 days ago

Noting that the failure in https://github.com/AlexsLemonade/OpenScPCA-analysis/pull/546/commits/7192b2396482be684014558363fb3724ec48f279 is for an expected stub link which can be updated once #533 goes in.

allyhawkins commented 4 days ago

This looks good to me, but we just need to update branch protection rules or wait for #533

Actually it looks like checks aren't required to pass so we are good to go.

sjspielman commented 4 days ago

Actually it looks like checks aren't required to pass so we are good to go.

Ah, indeed! I was just on autopilot assuming they were 😬