Character literals for Unicode (control) characters

[Lars Ingebrigtsen]

I was implementing support for the <bdo> HTML tag the other day.  (It's
for overriding bidi directionality in text.)  This is what I ended up
with:

(defun shr-tag-bdo (dom)
  (let* ((direction (dom-attr dom 'dir))
         (char (cond
                ((equal direction "ltr")
                 #x202d)                ; LRO
                ((equal direction "rtl")
                 #x202e))))             ; RLO
    (when char
      (insert char))
    (shr-generic dom)
    (when char
      (insert #x202c))))                ; PDF

And it just struck me that it would be kinda nice if Emacs had a literal
character syntax for these things.  I mean, we have such a syntax for
some "problematic" ASCII characters already: We recommend writing ?\s
instead of ? , and we recommend writing ?\n instead of ?
, because that's just very confusing.

And then I thought -- well, if we should have a literal syntax for
Unicode control characters, why not for all of them?  We do have the
mapping already in Emacs, so it wouldn't be very difficult to
implement...

So. Three options:

1) Add a new syntax, perhaps something like ?\ucRIGHT-TO-LEFT-OVERRIDE
for the Unicode control characters we care about.

2) Add a syntax for all Unicode characters, like ?\ucPILE-OF-POO.  We
can just write ?💩, so this isn't totally necessary, but perhaps it's
nice?

c) Do nothing, and continue writing code like the code above.  Or start
using the Unicode control characters directly in the code,
but ‮there lies madness‬.  (Note Unicode control characters around the
last part of the previous sentence.)

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no