docs: catch invalid target references #4235
Conversation
taimoorzaeem
force-pushed
the
docs/invalid-refs
branch
2 times, most recently
from
a2717fd to
908f594
Compare
The scheduled linkcheck job fails without notification. To catch the failure, a github workflow has been added which comments on issue 4106 so that the maintainers get notified. Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
taimoorzaeem
force-pushed
the
docs/invalid-refs
branch
from
908f594 to
a225b01
Compare
|
The peter-evans/create-or-update-comment uses github API for this. I think we need a |
|
Hm, why does Linkcheck / linkcheck (pull_request) fail? I thought we only ran it once a week and not per PR? |
Yes, you are right. I added a fake link to make it fail. I changed linkcheck temporarily to just run for this PR, to test the commenting workflow. Once this works, I'll remove the fake link before it could be merged. |
| pull_request: # Temporarily test using PR, remove this before merge | ||
| branches: | ||
| - main |
There was a problem hiding this comment.
With the pull_request event you won't have any permissions for the github token, thus you can't create comments. At least not when the PR comes from a fork.
We'd need to create this PR's branch in the postgrest repo itself to be able to test it that way.
There was a problem hiding this comment.
We'd need to create this PR's branch in the postgrest repo itself to be able to test it that way.
If we are testing it that way, then I guess I should leave this for maintainers with write access on this repo.
| on-linkcheck-fail: | ||
| runs-on: ubuntu-latest | ||
| needs: linkcheck | ||
| if: ${{ always() && needs.linkcheck.result == 'failure' }} |
There was a problem hiding this comment.
We should not run a second job, but just a second step in the same job. Otherwise that first job will still fail causing another notification to whomever touched the scheduled workflow file last.
The scheduled linkcheck job fails without notification. To catch the failure, a github workflow has been added which comments on issue #4106 so that the maintainers get notified.