More meaningful error message if commenting fails, and perhaps retry

[Piedone]

Recently, running the workflow failed in our repo, see the workflow output, with the following error:

Error: error adding "This pull request has merge conflicts. Please resolve those before requesting a review.": HttpError: invalid json response body at https://api.github.com/repos/Lombiq/Open-Source-Orchard-Core-Extensions/issues/291/comments reason: Unexpected end of JSON input
    at /home/runner/work/_actions/eps1lon/actions-label-merge-conflict/9[29](https://github.com/Lombiq/Open-Source-Orchard-Core-Extensions/actions/runs/3554453936/jobs/5970562928#step:3:31)24ba33a60e436034b0ac3838de523bf7df071/dist/index.js:7594:23
    at Generator.throw (<anonymous>)
    at rejected (/home/runner/work/_actions/eps1lon/actions-label-merge-conflict/92924ba33a60e436034b0ac3838de523bf7df071/dist/index.js:7364:65)
    at processTicksAndRejections (node:internal/process/task_queues:96:5)

I suspect this was due to a temporary issue with the GitHub web API, because re-running the workflow succeeded, and the action perfectly updated only those PRs that needed it.

My request would be to please consider adding a more meaningful error message in such cases, e.g. (if this is indeed a GH API error) something along the lines of "Error: error adding "This pull request has merge conflicts. Please resolve those before requesting a review." due to an invalid response from the GitHub API. This is most possibly just a temporary issue; try re-running the action. Don't worry, it won't duplicate labels or comments."

Ideally, there would also be some retry logic, so the action would back off for like a second, then try again, then back off for 3 second and try again, and if that still fails, then fail with an error message like "Error: error adding "This pull request has merge conflicts. Please resolve those before requesting a review." due to an invalid response from the GitHub API, even after 3 attempts. This is most possibly just a temporary issue; try re-running the action in a few minutes. Don't worry, it won't duplicate labels or comments."

[eps1lon]

This is most possibly just a temporary issue;

How would I know that?

Implementing back-off logic for rare, one-off issues sounds like a lot of work for little gain. Still, would happily review a PR.

[Piedone]

Since on a second try the same comment succeeded, it was indeed a temporary error. I only see the above-mentioned error message so I don't know what exactly the response was, but the action could check the HTTP response code (before trying to parse the response) and if it's a 5xx only then display this message.

GitHub doesn't have a stellar incident history lately, so I'd argue this is not necessarily a one-off issue (as it is a useful practice to treat remote APIs as potentially unreliable generally) but I don't want to imply it's a huge one either. Such retry would be more like a nice-to-have thing.