-
Notifications
You must be signed in to change notification settings - Fork 192
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
Pipeline template: Fix download test CI #2727
Conversation
6096941
to
b12f086
Compare
81e24ae
to
6280bc9
Compare
nf_core/pipeline-template/.github/workflows/download_pipeline.yml
Outdated
Show resolved
Hide resolved
Downgrading this to draft again. The branch issue seems to be resolved by setting However, there is another issue with Jinja templating that needs fixing as well:
When investigating the cause of this error message in a Testpipeline CI run, I stumbled over two code fragments ( Having no knowledge about their origin, I presumed that those would be odd hallucinations from Copilot and kicked them out in the course of this PR again. Upon a closer look, it turned out that @mashehu added them on purpose, but I am not aware of the rationale. In either case, we must find a solution that does not interfere with the commands, but also settled the issue that prompted their inclusion in the first place. |
The |
1925e80
to
5e26891
Compare
I see. Then everything was just a big misunderstanding from my side. The escape statements did interfere, but only because I lazily copied the file across from my local tools repo to the testpipeline folder for testing instead of using tools, so the Jinja escape statements were not removed in the process. Github Copilot is also exonerated. I was totally surprised that those escape sequences suddenly showed up in the file, even though I could not remember seeing them shortly before, and an automatic Copilot change seemed the only plausible explanation. In reality, that change popped up because I rebased my feature branch onto I have now dropped the two commits in which I deleted them from this branch, so the history is clean again. |
… option for manual dispatch.
5e26891
to
de73d39
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files☔ View full report in Codecov by Sentry. |
When writing the CI test that tested nf-core download on the current pipeline, I intended to run the test for any push on a branch underlying an active PR to dev. This required that the CI action would dynamically determine the branch name, for which I naively queried
/refs/heads/
.However, I now learned that resolving the branch name is a bit more complex than I initially thought, since there may be a
HEAD
,FETCH_HEAD
,ORIG_HEAD
,MERGE_HEAD
andCHERRY_PICK_HEAD
and sadly branch determination already failed in practice (cc @sofstam ) because aMERGE_HEAD
existed in addition toHEAD
.Since it is quite difficult to determine the current branch reliably (except maybe by running
git
?) and we are exclusively running the CI test for merge-PRs to master, I decided to hardcode thedev
branch instead. This works on PRs to master, but it is still possible to manually dispatch the action on any other branch if needed.PR checklist
CHANGELOG.md
is updateddocs
is updated