如何在 C（Linux utf8 终端）中打印“盒子抽屉”Unicode 字符？

Question

I'm trying to display Unicode characters from (Box Drawing Range: 2500–257F). It's supposed to be standard utf8 (The Unicode Standard, Version 6.2). I'm simply unable to do it.

I first tried to use the good old ASCII characters but the Linux terminal displays in utf8 and there is no conversion (symbol ?) displayed in place.

Could anyone answer these questions:

• How to encode a unicode character in a C variable ( style wchar_t )?
• How to use the escape sequence such as 0x or 0o (hex, oct) for Unicode?

I know U+ but it seems it didn't work.

setlocale(LC_ALL,"");
short a = 0x2500, b = 0x2501;
wchar_t ac = a;
wchar_t bc = b;
wprintf(L"%c%c\n", ac, bc);
exit(0);

I know that the results are related to the font used, but I use a utf8 font ( http://www.unicode.org/charts/fonts.html ) and codes from 2500 to 257F must be displayed... Actually they aren't.

Thanks for your help in advance...

[EDIT LATELY] The issue is solved since... and I found how to use wprintf() with %lc instead of %c... and deeper. Now those bow drawers are part of my student "tools" library to make the console programming learning a little more coloured.

Answer 1

Use a Cstring containing the bytes for the utf-8 versions of those characters. If You print that Cstring, it will print that character.

example for Your two characters:

#include <stdio.h>

int main (int argc, char *argv[])
{
    char block1[] = { 0xe2, 0x94, 0x80, '\0' };
    char block2[] = { 0xe2, 0x94, 0x81, '\0' };
    printf("%s%s\n", block1, block2);
    return 0;
}

prints ─━ for me.

Also, if You'd print a Cstring containing uft-8 character bytes somewhere in it, it would print those characters without problems. /* assuming You use gcc */ And IIRC gcc uses utf-8 internally anyway.

EDIT: Your question changed a bit while I was writing this. And my answer is less relevant now. But from Your symptoms - if You see one ? for each character You expect, I'd say Your terminal font might be missing the glyphs required for those characters.

Answer 2

That depends on what you call "terminal".

The linux console uses various hacks to display unicode but in reality its font is limited to 512 symbols IIRC so it can't really display the whole unicode range and what it can display depends on the font loaded (this may change in the future).

Windows terminals used to access Linux are usually braindamaged one way or another unicode-wise.

Physical terminals are usually worse and only operate in ascii-land

Linux GUI terminals (such as gnome-terminal) can pretty much display everything as long as you have the corresponding fonts.

Are you sure you don't want to use ncurses instead of writing your own terminal widgets?