UTF-8 string literals

What's the best way of embedding UTF-8 (specifically UTF-8, not Unicode) literals in strings in C? You can, of course, do them as byte (char) sequences, but that makes it impossible to write

Code:

---

char * str = "The first letter of the Greek alphabet is " ALPHA ".";

---

Basically, I want to be able to write

Code:

---

#define ALPHA "<something>"

---

I should probably add, since this is platform-dependent, that I'm using GCC on OS X and FreeBSD.

I can see two ways:
1. declare utf-8 string as a byte array:

Code:

---

unsigned char utf8[] = { 'a', 'b', 0 };

---

2. declare string as wide char string, then convert to multibyte.

That doesn't work. As I said, representing them isn't the problem. Reresenting them as a string literal is. Can they be written as "<something>"?

Venturing into new territory myself, but isn't it something like "\u0139"?

Using '\U' or '\u' is C99's support for Universal Character Names (UCN).

The character set used is the Universal Character Set (UCS), as described by ISO/IEC 10646.

You can also use 'u' or 'U' to begin a string literal (much like 'L').
When you use 'u', you get a UTF-16 encoded character or string literal.
When you use 'U', you get a UTF-32 encoded character or string literal.
When you use 'L', you get a "wide" character or string. The character set used in implementation defined.

You use wchar_t for all of those.

If you want portable, UTF-8 encoding in a string literal, you'll have to encode it yourself with escape sequences "\xhhh".

http://www.cl.cam.ac.uk/~mgk25/unicode.html

Having said all that, here's gcc's "defined implementation"
http://gcc.gnu.org/onlinedocs/gcc-4....fined-behavior

If you lookup the "-fexec-charset", you'll find that UTF-8 is already the default. So if you have a Greek "Alpha" on your keyboard, make your source files UTF-8 and put the real character in there (gcc supports UTF-8 source files). If you're source files are in ASCII, you'll have to use escape sequences or load your strings from an external source.

gg

That's great, exactly what I was looking for, thanks!