C++23 Characters

C++ is definitely a language that has Lots of Ways to do It – kind of like Perl’s TIMTOWTSAC. UTF-8 is a suitable source encoding for C++ source, so that you can write string literals in UTF-8. Convenient for adding a poop emoji to an error message, or if your application is very local and doesn’t need translations.

A tiny example (which I have not bothered compiling, YMMV):

QString getSheep() {
  return QStringLiteral("🐑");
}

There’s a sheep right there in the source code! But now you’re rather dependent on font support for reading the code – a KDE Plasma desktop comes with sufficient fonts for all your farm animals, but maybe a co-worker doesn’t have a sheep-enabled font and can’t read your code (or review that a sheep is a sheep indeed). There’s a universally-readable solution, though, available with C++23.

\N escapes.

You can write \N{NAME} in C++23 source and it compiles down to a single Unicode character, which becomes however many bytes it needs to be. This is well-described at CppReference, which then refers you to the Unicode list of character names.

KDE Plasma users can readily install KCharSelect which knows these names already. Type a search term into the box and you get matching Unicode characters. Off in the details box on the right there is a Name field which you can use.

KCharSelect with search term 'sheep' displays CJK and Kangxi radicals for 'sheep' and a sheep symbol

QString getSheep() {
  return QStringLiteral("\N{SHEEP}");
}

It might not be quite as pretty, but it is universally readable. Enjoy!

Note that this is not only C++-standard-dependent, but also compiler-dependent. Clang 15 seems to support \N regardless of standard, while Clang 13 does not support it at all.