The code needs to use a helper trait to allow iteration on something different than char / char16_t / char32_t. My idea is that such a trait tells which decoder to use (UTF-8 / 16 / 32) and how to convert a type T we're iterating upon to char/char16_t/char32_t. The trait must be customizable by the users.
At a minimum this should be supported:
- QChar -> convert to char16_t via unicode()
- ushort and other unsigned 16 bit integrals -> convert to char16_t
- uint and other unsigned 32 bit integrals -> convert to char32_t
- wchar_t -> 16 / 32 as needed
The code needs to use a helper trait to allow iteration on something different than char / char16_t / char32_t. My idea is that such a trait tells which decoder to use (UTF-8 / 16 / 32) and how to convert a type T we're iterating upon to char/char16_t/char32_t. The trait must be customizable by the users.
At a minimum this should be supported: