bug#30749: 26.0.91; (cl-)case is confused when a nil atom is used instead of a keylist

[Alex Bennée]

I got confused why my case statement was failing as I'd used the second
form assuming that having nil as an atom would work. I'm not sure why
the quoting fixed it but it's certainly a confusing inconsistency. The
following behaviour shows up on emacs -Q:

(require 'cl)

(case nil
  ((nil) "nil")
  ((t) "default"))
"nil"

(case nil
  (nil "nil")
  (t "default"))
"default"

(case nil
  ('nil "nil")
  ('t "default"))
"nil"

(case t
  ('nil "nil")
  ('t "default"))
"default"

I think this is probably fixable with just making the docstring clearer
as the results of handling the duel atom/list of nil might have
unexpected results.

In GNU Emacs 26.0.91 (build 16, x86_64-pc-linux-gnu, X toolkit, Xaw3d scroll 
bars)
 of 2018-03-06 built on zen
Repository revision: 0efe0bd233de20bfb5bd9d06b255fc8ecf04602b
Windowing system distributor 'The X.Org Foundation', version 11.0.11804000
System Description:     Ubuntu 16.04.3 LTS

--
Alex Bennée