On Tue, Sep 1, 2020 at 8:45 AM Steven D'Aprano <steve@pearwood.info> wrote:
This is slightly revised version of something I sent to Stefano two
weeks ago. I hope he is planning to use this, or something similar, in
the PEP, but for what it's worth here it is for discussion.

Excellent. Could you add some cases that show how `d[1]` differs from `d[1,]`? (And perhaps explain the reason why `d[1, k=2]` follows `d[1]` instead of `d[1,]`. I know the answer, but it's worth being clear about this particular edge case because it has tripped up various attempts.)

--Guido van Rossum (python.org/~guido)