Jan-Benedict Glaw <[EMAIL PROTECTED]> writes: >> - if (! (0 <= *p && *p <= 127)) >> + if (*p & ~0x7f)
> The `char' type, in difference to all other integer type, has no > default signedness. On some compilers, "char" is "signed char", on > others, it's a "unsigned char". Yes. The above code should work with either kind of compiler. > To really keep readability, I'd > rather have a local unsigned char variable and operate on it with > if (*foo >= 127) return false. Would that be any more readable? (It'd certainly be longer. :-)
