in reply to printf exact field width of floating point number
It is significant overkill, but Number::WithError can be coerced into writing floats to a fixed precision. Have Number::WithError "round" the number, stringize it, and split the string to extract the number.