UTF-8 multilanguage console printout (and input)

[Michele]

Somewhere on this forum there was a discussion about how to print multilanguage strings on console output.
The small attached sample shows the code that use the console codepage set functions using the UTF-8 encoding that is, at least for me, a very powerful option because while allows the multilanguage text let you use the simple and efficient ansi text functions (strcpy, ctrcat, strcmp, ...).

[Vortex]

Hi Michele,

Thanks for your code. Testing your executable on my Windows 7 Sp1 64-bit, I get a lot of garbled characters, probably a limitation of Windows 7.

[Michele]

Hello Vortex, I imagine that the problem is the OS version.
I will not add mine, just report what Gemini says:

The functions SetConsoleOutputCP and SetConsoleCP along with the CP_UTF8 code page constant (code page 65001) have been supported since Windows 2000.

However, the minimal OS version depends on what you are trying to achieve:

Basic Function Support (API availability): Windows 2000 or later. Calling SetConsoleOutputCP(65001) is supported on all Windows NT-based OS versions starting from 2000, as well as Windows XP, Vista, 7, 8, and 10.

Functional UTF-8 Console Rendering: Windows 10 Version 1903 (Build 18362) or later.

Why the distinction matters:
While the functions exist on older OS versions, UTF-8 support inside cmd.exe on legacy releases (Windows 7/8/early Win 10) was notoriously buggy:

Legacy Bugs: On older versions like Windows 7, setting the console to code page 65001 frequently broke C runtime output routines (printf), caused character truncation, or crashed batch scripts.

Font Rendering: Legacy consoles couldn't render multi-byte UTF-8 sequences properly unless using specific TrueType fonts.

Full ConPTY Overhaul: Microsoft fully overhauled the Windows Console infrastructure in Windows 10 version 1903 (Build 18362)—the exact same release that introduced manifest-based UTF-8 ACP support—making UTF-8 output reliable and stable in the terminal.

[TimoVJL]

Windows 7 x64 EN console.

Less fonts than Windows 1x versions.

[Michele]

Timovjl you successfully changed the codepage to UTF-8, in fact many characters are correctly displayed.
I agree that now the problem is the console font that doesn't contains the full set.
A known font known to render all charset is 'Consolas', try changing the face font with:

Code Select Expand

BOOL SetConsoleFontToConsolas(void)
{
    HANDLE hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
    if (hConsole == INVALID_HANDLE_VALUE) return FALSE;

    CONSOLE_FONT_INFOEX cfi;
    cfi.cbSize = sizeof(CONSOLE_FONT_INFOEX);

    // Get the current font info first to preserve defaults
    if (!GetCurrentConsoleFontEx(hConsole, FALSE, &cfi))
    {
        return FALSE;
    }

    // Set font properties
    cfi.dwFontSize.X = 0;   // Width (0 lets Windows calculate based on height)
    cfi.dwFontSize.Y = 16;  // Height (Font size in points/pixels)
    cfi.FontFamily = FF_DONTCARE;
    cfi.FontWeight = FW_NORMAL;

    // Specify a TrueType font that supports UTF-8 Unicode characters
    // Common choices: L"Cascadia Code", L"Consolas", L"Lucida Console"
    wcscpy_s(cfi.FaceName, LF_FACESIZE, L"Consolas");

    // Apply the new font
    return SetCurrentConsoleFontEx(hConsole, FALSE, &cfi);
}
If still doesn't it confirms that the fonts in your OS version are limited.
You said that on a Win10/11 system it works?

[TimoVJL]

Windows 7 have Consolas font, but version 5.? and Windows 1x have version 7 font family.
Less interest to console, albeit file / folder names can have many kind of chars.