Support keyword splatting nil

[jeremyevans]

nil is treated similar to the empty hash in this case, passing no keywords and not calling any conversion methods.

Fixes [Bug #20064]

[k0kubun]

jeremyevans requested a review from k0kubun 4 minutes ago

I'm not sure how I'm relevant to this change. Are you asking me to add YJIT/RJIT support or something?

[jeremyevans]

jeremyevans requested a review from k0kubun 4 minutes ago

I'm not sure how I'm relevant to this change. Are you asking me to add YJIT/RJIT support or something?

No, you just came up first in the GitHub list. :) You can ignore.

[k0kubun]

Thanks. Maybe GitHub thinks I did many contributions to the file because I expanded tabs 🙂 Nobu or Koichi would be a better reviewer for this change.

[jeremyevans]

Only failure is TravisCI on ppc64le Linux, where it passed test-spec and test-all, but not test: https://app.travis-ci.com/github/ruby/ruby/jobs/616119319#L2006 (with no output)

[jeremyevans]

This needs more work to handle {**nil}, f(a: 1, **nil), and other cases. I'll work on that later today.