BigDecimal（java）中用于指数表示法的缩放问题

Question

In our application, we need to check whether the given number is in range or not and for this we are using below code -

 boolean isValidRangeNumber(Double no,int precision, int scale){
            BigDecimal bigDecimal = new BigDecimal(no.toString());
            int intPartLength = precision - scale;
            int size = (bigDecimal.longValue() + "").toString().length();

            if(bigDecimal.precision() <= precision && bigDecimal.scale() <= scale && size <= intPartLength)
                return true;

            return false;
        }

But when we are calling isValidRangeNumber(0.0000009, 10,7) than its failing as bigDecimal.scale() is giving 8 which can be max 7 whereas its a valid number. Please suggest, what is wrong with this BigDecimal usage.

Answer 1

If you inspect following :

bigDecimal.toPlainString() // This will show 0.00000090 -> scale of 8, appends trailing 0

Because of an extra trailing zero, following condition evaluates to false.

bigDecimal.scale() <= scale // 8 <= 7

You can change into following :

boolean isValidRangeNumber(Double no, int precision, int scale) {
    BigDecimal bigDecimal = new BigDecimal(no.toString()).stripTrailingZeros(); //Get rid of trailing zeros
    bigDecimal.toPlainString();
    int intPartLength = precision - scale;

    int size = (bigDecimal.longValue() + "").toString().length();

    if (bigDecimal.precision() <= precision && bigDecimal.scale() <= scale
            && size <= intPartLength)
        return true;

    return false;
}