labeler
labeler copied to clipboard
actions
Fix labeler
Labeler has a bunch of more or less related problems.
- it doesn't properly annotate its
action.yml - it confuses booleans and strings #404
- it doesn't handle permissions errors #136
- it's actively hostile to forks (#136)
The general implementation to address this fits the outlines in https://github.com/actions/labeler/issues/136#issuecomment-1169411989
https://github.com/check-spelling/use-labeler/actions/runs/2729552526
Annotations
1 error and 1 warning
❌ label
Resource not accessible by integration
⚠️ jsoref/labeler running under 'pull_request' is misconfigured
It needs permissions: pull-requests: write to work. Update the workflow. See https://github.com/actions/labeler/blob/6a315d4ea58951035b498eef56668feaba24489f/README.md#create-workflow
https://github.com/check-spelling/use-labeler/runs/7492560360?check_suite_focus=true
Run jsoref/labeler@main Workflow is not configured to label forks, exiting
This could include an info item, but, I don't really see the point.
@thboop? I understand you're busy, but it'd be really helpful if someone would look (I can, of course, resolve conflicts, but it isn't worth it if no one will look at the PR.)
Is it possible to split the fixes into separate PRs? I don't think it'll be reviewed otherwise.
@ofek: it isn't worth my time to do such a thing until a GitHub employee shows that they're willing to actually review things. This project appears to be entirely orphaned. @ethomson prove me wrong.
To your question: the commits are distinct and anyone w/ write permissions could trivially cherry-pick them individually.
It is considered a very serious error to unintentionally delete a label in a tool for automating labeling. I think that PR should be reflected quickly.
👋 From AWS! We rely on this labeling solution and would love to see this bug worked out. Thanks!
@ford-at-aws I've enabled github actions on my fork: https://github.com/jsoref/labeler/actions -- and I've merged the changes into it.
Actions details for the relevant commits
✅ Merge pull request #1 from jsoref/fix-labeler Build & Test #2: Commit 3cc0b12 pushed by jsoref main
✅ Fix labeler Build & Test #1: Pull request https://github.com/jsoref/labeler/pull/1 opened by jsoref fix-labeler
Switching from actions/[email protected] to jsoref/[email protected]
In theory where you would have used actions/[email protected], you could use jsoref/[email protected].
Or where you could have used actions/labeler@e54e5b338fbd6e6cdb5d60f51c22335fc57c401e, you could use actions/labeler@6b5f93ed5fc722246dd157acf1984bb1e3065b4d.
I will not delete the above tag. I can't delete the above commit sha -- do note that the sha doesn't (at the present time) refer to a commit in this repository, it's just https://github.com/jsoref/labeler/commit/6b5f93ed5fc722246dd157acf1984bb1e3065b4d, but by referencing this repository, you should be less sensitive to me renaming my repository (which I don't intend to ever do).
Updated 2022-12-20
Switching from actions/[email protected] to jsoref/[email protected]
In theory where you would have used actions/[email protected], you could use jsoref/[email protected].
Or where you could have used actions/labeler@5c7539237e04b714afd8ad9b4aed733815b9fab4, you could use actions/labeler@c6732c9ca7d4d449e8d513bbd5be7900a2461cff.
I will not delete the above tag. I can't delete the above commit sha -- do note that the sha doesn't (at the present time) refer to a commit in this repository, it's just https://github.com/jsoref/labeler/commit/c6732c9ca7d4d449e8d513bbd5be7900a2461cff, but by referencing this repository, you should be less sensitive to me renaming my repository (which I don't intend to ever do).
Hi @jsoref do you think my issue https://github.com/actions/labeler/issues/442 is related?
@efb4f5ff-1298-471a-8973-3d47447115dc: dunno. I don't suppose you'd be willing to try using the labeler as versioned in https://github.com/actions/labeler/pull/405#issuecomment-1264073531 for a bit? I imagine you'd be able to confirm its behavior pretty quickly... You could even create a second copy (not really a fork, just a copy) of your repository w/ the alternate labeler and see how it behaves...
@MaksimZhukov can you look at this PR as well? At least one linked repository has been using this for months.
