-
Notifications
You must be signed in to change notification settings - Fork 4
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
Enable usage of github runner file commands #90
Conversation
This should (hopefully) allow the use of things like `>> $GITHUB_OUTPUT` (when such variables are passed through via `export:`)
Hi @JasonGross, thanks!
A few suggestions:
|
Use DIR_GITHUB_RUNNER_FILE_COMMANDS to avoid reserved GITHUB_ prefix
Done, thanks for the suggestion
I haven't really tested this, so I extended the env test in the Coq demo to write to GitHub step summary (this should be the easiest one to observe the effect of), let's see if it works here
I think all the runner file commands should be based in the same directory. Currently this seems to be the case, and if this changes, we can either add a couple more |
Hm, I guess by default we have ownership issues:
Should we chmod the relevant files? |
Current test at https://github.com/coq-community/docker-coq-action/actions/runs/9036006653/job/24831950493 fails with |
I am thoroughly confused by https://github.com/coq-community/docker-coq-action/actions/runs/9036191567/job/24832574488?pr=90#step:4:614 https://github.com/coq-community/docker-coq-action/actions/runs/9036191567/job/24832574488?pr=90#step:4:391 So the file seems to exist and is writable... Any ideas @erikmd |
I see
|
Hi Jason, thanks for your experiments! I'll take a look tonight and let you know... |
Latest results: the directory exists but looks empty from inside docker (according to |
A little bit more debugging:
and
Docker is invoked with
(why is the total 16 if there are only four things?) and
(why 8 if there is nothing but |
Hi @JasonGross, sorry for replying in-depth just now. I think you are facing an issue directly related to a double-embedding of Docker. I had already fixed something similar in e3df63e. To be more precise:
Details from this job
But, the That's it for the logical explanation. Now I'll take a look at your latest commits. |
Ah, I see, yes, that would fully explain the issue. So I guess just that path needs fixing somehow? And then one of us can just strip off all of my debugging commits / drop the extraneous stuff in the test file |
Yes, exactly! I'm going to push soonish |
…a bit of debugging
Good news! the latest commit succeeded in this log: https://github.com/coq-community/docker-coq-action/actions/runs/9039122047/job/24841442923?pr=90 I propose to keep the PR as-is for now, polish the test ASAP and do a release on Monday or so… Thanks again @JasonGross for suggesting this feature! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
New version looks good to me! Is there anything else to do before merging?
I'll just polish the test, add some documentation (README.md), and squash-merge. |
@JasonGross Also contrarily to |
…t+docs * Merge branch 'master' into pr/90
@JasonGross: the 2 PRs now go live in |
…4xSPC (#93) * emacs: Add file-local-variable `indent-tabs-mode := nil` to avoid inserting further TABs
This should (hopefully) allow the use of things like
>> $GITHUB_OUTPUT
(when such variables are passed through viaexport:
)