fn float_to_decimal_common_exact<T>( fmt: &mut Formatter<'_>, num: &T, sign: Sign, precision: u16, ) -> Resultwhere T: DecodableFloat,