Re: inputting characters by hexadigit

[David De La Harpe Golden]

Alfred M. Szmidt wrote:
> What about C-\ and having a input method (maybe that already exists?),

The ucs input method _does_ exist. However, especially if you're already
using some other input method, it's a bit of a pain to switch to ucs and
back again just for the odd character. It needs a little expansion (it
only covers the BMP (4-hexit codes) and maybe should handle unicode
names if that's now legally feasible)

Er. Though I guess I should confess that I tend to have the ucs IM a C-\
away - I tend to use system-level input methods (scim/skim) for languages...

M-x ucs-insert (or M-x uc TAB RET) is shorter than switching input
method with C-x RET C-\ ucs RET , though presently doesn't allow C-u for
repetition, which is annoying (though that's presumably easily fixed).

> Inserting unicode in hex cannot be _that_ common that it requires a
> short keybinding...

If you're doing a lot of i18n/m17n/l10n work it can just be handy - and
you're quite likely to be in another input method, too, so the ucs input
method is less handy.

[following is mostly just thread recap...]

☆ The fact remains, there is already an existing short-keybinding
facility in emacs for numeric char code entry ‽ - C-q supports entering
by numeric code. It used to be okay on unibyte emacs , but is a bit
unsuitable for unicode emacs, at least with its default octal (icky
octal) radix.  Despite that, N.B., _it already works for entering
unicode_, except that right now, the existing facility does the wrong
thing for codes between #o240 and #o377.  At the very least, that's a
bug that needs to be resolved, even if nothing else changes as a result
of this thread...

The facility does damage the "purity" of C-q - If the real point of C-q
is to literally insert the next keypress, it obviously blocks that for
0-7 or 0-9 or 0-9a-fA-F depending on your setting of
read-quoted-char-radix.  But if it were to be dropped from C-q, then
it'd be IMO reasonable for people to expect a similarly easy alternative.


http://en.wikipedia.org/wiki/Unicode_input

Emacs _clashes with_ the short keybinding for unicode input typical on
gtk/GNOME desktops - C-s-U (which emacs can't treat as different to
C-u in general due to terminal restrictions).

(... Hmm. Is M-+ taken? would be a bit like the windoze keybinding...)

It's one thing for emacs to provide different short keybindings for
typically-short-keybinding-accessible functionality (as in C-w/M-w/C-y
vs. C-x/C-c/C-v), it's another to not provide a short keybinding at all...

[cheap shot] Vim provides quite a short keybinding - C-v u - for it.

There is some fondness for unicode symbols in some circles, and some
people DO just learn off a few hex codes. I myself tend to do ☘
/ ☠ / ☢ / ☺ / ♬ etc. from memory (or at least remember such symbols
are mostly blocks #x2600 #x2700 ...)