Suppose you have a Unicode code unit wchar_t and you want to know whether it represents a numeric digit. If you have the ICU library, you can check if its code point’s u_charType is U_DECIMAL_DIGIT_NUMBER. But what about plain Win32?

For Win32, you can use the Get­String­TypeW function to obtain properties for each code unit.

bool IsUnicodeDigit(wchar_t ch)
{
    WORD type;
    return GetStringTypeW(CT_CTYPE1, &ch, 1, &type) &&
           (type & C1_DIGIT);
}

We ask the Get­String­TypeW function for the CT_CTYPE1 value for one character, passing an output buffer of size 1. We then check whether the result says that it is a digit.

The Get­String­TypeW function produces a 16-bit value for each provided code unit. There are more than 16 things you can ask about, so they are broken into groups, and you specify which group you want. Group 1 contains the basic classifications that support POSIX functions like isdigit and isalnum.

Here’s one way it could be done. (I’m not saying this is how it actually is done.)

Bonus reading: The difference between C1_SPACE-ing out and drawing a C1_BLANK.