Proposal: allow '?' and '!' in identifiers

[Michael Hudson]

jcm <grumble at mailcity.com> writes:

> Erik Max Francis <max at alcyone.com> wrote:
> > Nathaniel Gray wrote:
> 
> >> Why not be hardcore and allow both?
> 
> > Because using a suffixed p to indicate predicates, if that is your
> > liking, is already supported by the language.  Allowing ? in identifiers
> > is not, and would require all sorts of weird parser logic to allow it in
> > identifiers but not break backward compatibility.  It is simply not
> > worth it.
> 
> I understand that you dislike the idea of allowing '?' in identifiers.
> But it would not necessarily require "all sorts of weird parser
> logic".  I expect it's a simple change to the lexer.  

It is.  It takes about a minute if you know where to look (which is in
Parser/tokenizer.c).  I did it once, but seem to have lost the patch.

Cheers,
M.

-- 
  ARTHUR:  Yes.  It was on display in the bottom of a locked filing
           cabinet stuck in a disused lavatory with a sign on the door
           saying "Beware of the Leopard".
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 1