Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

#1654: Use luarocks' GitHub actions #1655

Conversation

MaxVanDeursen
Copy link
Contributor

Fixes #1654

@MaxVanDeursen
Copy link
Contributor Author

Please let me know if an alternative solution is preferred or whether some other factor should be considered prior to merging this PR!

Copy link
Member

@NTBBloodbath NTBBloodbath left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, thanks for addressing this!

Failed CIs can be ignored for merging, they are not related.

@benlubas benlubas merged commit 2f880ad into nvim-neorg:main Feb 23, 2025
3 of 6 checks passed
@benlubas
Copy link
Contributor

benlubas commented Feb 23, 2025

lovely, seems like this deleted half all of the module pages in the wiki

@MaxVanDeursen
Copy link
Contributor Author

MaxVanDeursen commented Feb 24, 2025

@benlubas Sorry to hear that the merge of this PR caused pages to be deleted. Is there something I/we can do to make sure this does not occur in the future or that I could have done differently?

Either way, I do have some time later this week and will have a look at whether I can find the culprit.

@MaxVanDeursen MaxVanDeursen mentioned this pull request Feb 24, 2025
2 tasks
@benlubas
Copy link
Contributor

Nothing you could have done. The doc gen cut is setup in weird way.

We can temporarily disable this ci job and manually restore the wiki. I'll probably do this later today

@MaxVanDeursen
Copy link
Contributor Author

@benlubas I don't mind focusing a bit on the nightly build and see how to solve this. I indeed have now confirmed that the last release of neovim does not have the error, while the nightly build does, locally.

I will submit a PR if I find the fix!

@benlubas
Copy link
Contributor

Oh interesting, I was unable to replicate the error locally. I guess the nightly flake was a little behind.

@vhyrro since you setup the original CI, is there an easy way to switch from nightly to stable?

@MaxVanDeursen
Copy link
Contributor Author

@benlubas See #1662 for a PR which does this.

@MaxVanDeursen
Copy link
Contributor Author

@benlubas I have tracked down the issue: It appears the treesitter's parser is not loaded correctly in the get_first_node nor get_first_node_recursive function of the core/integrations/treesitter module. Calling parser:parse() prior to for_each_tree(...) seems to solve this problem. I do not know whether this is the most efficient solution so am currently looking for a relevant change in neovim that explains this.

@benlubas
Copy link
Contributor

benlubas commented Feb 25, 2025

Recently there was async TS parsing code merged. I think the highlight code now uses async, and the highlight code was previously what performed the full parse.

I think we should only call parse in neovim 0.11, and start to convert these queries to use the async TS API once it's better documented.

Btw I really appreciate that you dug into this. I will merge your fix after work today.

@MaxVanDeursen
Copy link
Contributor Author

@benlubas My pleasure. Regarding your comment about calling parse in 0.11, is that something we want to do instead of #1662, is that something we want to do alongside #1662 at this point in time, or do we want to an issue to track this and do this once 0.11 is released? If you expect or would like any action from me, let me know!

@benlubas
Copy link
Contributor

@MaxVanDeursen

calling parse in 0.11

I think we could/should do this before 0.11 releases. can just use vim.version to check versions. If you want to do this I will absolutely not stop you. I'm going to be really busy next week, but after that I'll have time and I can make the changes.

We also need to at some point either make the wiki generate but not push in PR, or make the wiki generation script fail (ie. don't push) if it doesn't generate the expected files. or both, so that something like this doesn't happen again.

Same deal here. I can do this, or if you're interested, I will not say no :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

The "Automatically Generate Documentation" GitHub Action always fails
3 participants