[#697] Github actions to run all PRC tests

[d-w-moore]

All tests will run from GitHub actions.

Except the interactive PAM test, and that omission will soon be rectified.

[d-w-moore]

All the checks implemented via github actions (except Codacy) are passing . Some cleanup has been done on the code as well. I think we're for a preliminary round of review. Just be aware there are some extra scripts that are not central to the github actions we've implemented here. They are there as examples and tests of the harness functionality.

[trel]

please mark the ones that are 'not central'.

are you suggesting we keep the scripts that are 'not central'? if so, why?

[d-w-moore]

please mark the ones that are 'not central'.

are you suggesting we keep the scripts that are 'not central'? if so, why?

They're illustrative but not necessary, so I'll remove them.

[trel]

They're illustrative but not necessary, so I'll remove them.

Ah, okay - yes please remove them. And consider putting any helpful words / examples in README.md files at the appropriate locations.

[d-w-moore]

We are now down to 6 basic commits.

[d-w-moore]

We're now down to 8 basic commits. If that's too many for the convenience of those reviewing, let me know and I can maybe digest it down further.

[korydraughn]

8 commits isn't bad at all, but let's go ahead and squash commits into their final shape if we know what that is.

[d-w-moore]

Commits now minimized to 2.

[trel]

please squash so this is a candidate for merging. i still see two extra commits i think.

and consider renaming irods/test/setupssl.py to irods/test/setup_ssl.py

[d-w-moore]

Should f59678c and ff808aa be squashed together?

My own editorial on it, keep them separate for proper attribution. And, because there is fundamentally different work being done:

• the first commit is basically the development of the github infrastructure and the multiple-node harness.
• the second commit is the single-node harness and related changes to the setup scripts for the tests themselves.

[alanking]

My top suggestion at this point is to revert / delete that last commit. Why rename setupssl.py? I think we should leave that alone since it's publicly available and documented as setupssl.py. If you want to rename it, let's make a new issue and discuss how to do that at another time... unless you need to rename it in order for these changes to work.

I think the "remove Jenkins test framework support" commit can be linked to #778, yes?

Finally, looking at "[_TENTATIVE_SQUASH] mostly review based changes ..."...

This one should be split up and squashed into the commit(s) with which each change is best suited. For example, it includes a change to get rid of all symlinks. I see one symlink that was introduced in "test harness and container-based tests", so that should be squashed into there.

If no reasonable (I will let you decide what that means) split / squash can be had for these changes, maybe just keep it in its own commit and link it to all of the relevant issues. It's not ideal, but if it can't be done in a clean manner, maybe that's all we can do.

I think the rest of the commits look fine.

[trel]

the rename ssl thing was me - i noticed all the other files had underscores in their naming... and it stood out.

[d-w-moore]

Finally, looking at "[_TENTATIVE_SQUASH] mostly review based changes ..."...

This one should be split up and squashed into the commit(s) with which each change is best suited. For example, it includes a change to get rid of all symlinks. I see one symlink that was introduced in "test harness and container-based tests", so that should be squashed into there.

If no reasonable (I will let you decide what that means) split / squash can be had for these changes, maybe just keep it in its own commit and link it to all of the relevant issues.

I can split it up pretty easily, as it still exists on another branch in unsplit form.

[d-w-moore]

the rename ssl thing was me - i noticed all the other files had underscores in their naming... and it stood out.

I'll work on the merge-compatible squashing tonight and defer to consensus on the renaming or not of setupssl.py

[korydraughn]

Almost there!

[alanking]

Given the new changes and the way things are arranged now, I would make a best-effort attempt at squashing the commits without issue numbers into the commits with issue numbers. If it proves too messy/difficult to sort through, I (personally) am not opposed to "formatting" commits (which exclusively or primarily deal with whitespace) that come at the end of a line of commits. Leaving that up to you, @d-w-moore.

the rename ssl thing was me - i noticed all the other files had underscores in their naming... and it stood out.

Looking into this further, I think it's okay to rename that script. It appears to be used internally for testing only. Please squash that into the commits with issue numbers as well.

[d-w-moore]

... I would make a best-effort attempt at squashing the commits without issue numbers into the commits with issue numbers. If it proves too messy/difficult to sort through, I (personally) am not opposed to "formatting" commits (which exclusively or primarily deal with whitespace) that come at the end of a line of commits. Leaving that up to you, @d-w-moore.

the rename ssl thing was me - i noticed all the other files had underscores in their naming... and it stood out.

Looking into this further, I think it's okay to rename that script. It appears to be used internally for testing only. Please squash that into the commits with issue numbers as well.

No problems, will do the squashing next.

[korydraughn]

Pound it!

[alanking]

The documentation introduced in this PR for running the tests also addresses #637. We will not be linking any specific commit to the issue, so I am doing it here in this comment.

Merging.