Adding pypy github action

[Sangarshanan]

Closes #235

Opening a cleaner PR with some changes suggested by @pradyunsg

[brettcannon]

@Sangarshanan do you mind editing a Python file to trigger the workflow so we can see how it works out? Like fix a spelling mistake in a docstring or just tweak a .py file slightly which is easy to revert before we merge?

[brettcannon]

Just holding up until we can see a run to make sure there isn't anything like YAML errors (which has happened to me when working on action files 😁).

[Sangarshanan]

Not sure as to why there are no logs for these failures

[pradyunsg]

FYI -- there's a button on the UI for GitHub checks ("Rerun all checks") that I just pressed.

[brettcannon]

@Sangarshanan Maybe try working backwards by scaling back the pypy job run-by-run and see if that changes anything?

@pradyunsg the re-run still leads to no output unfortunately :/

[Sangarshanan]

Just figured out that it was an issue with the pytest faulthandler pytest-dev/pytest#5807

Downgraded the Pytest version and the tests have run successfully : )

[brettcannon]

Great job figuring out the problem! I just want to make sure it's recorded as to why an old version of pytest is needed (for now) so people don't unpin it until a fix is out.

[Sangarshanan]

Thanks for being awesome and helping me out @brettcannon < 3

[Sangarshanan]

Do I need to make any more changes @brettcannon @pradyunsg ?

[brettcannon]

@Sangarshanan nope, just had to wait until one of had time to review it one last time.

Thanks for making this happen!

[aixtools]

Hi. While to update #190 I got an error from pypy testing. Not understanding how my patch was breaking things - I opened a new scratch PR that I was going to use to slowly add things to see when it broke. But I see that "master" is not passing for me - which do not understand because this PR #243 did pass all tests.

Anyway - my scratch #244 fails with https://github.com/pypa/packaging/pull/244/checks?check_run_id=339889206 Please consider reverting this addition - as I am stumped (not that my being stumped has any importance - :) - I would be extremely appreciative in any case!)

[brettcannon]

@aixtools (un)fortunately master still passes untouched with this PR so reverting it for another PR doesn't seem prudent to me.