Name: dfC67450 Date: 11/04/97



java.text.DecimalFormat.format(double n, StringBuffer str, FieldPosition fp) incorrectly
uses internal representation of double.
See similar bug #4054318.
 
Here is the test demonstrating the bug:

-----------------Test.java------------------------
import java.text.*;
import java.math.*;
public class Test {
    public static void main (String args[]){
        DecimalFormat df = new DecimalFormat();
        df.setMinimumFractionDigits(10);
        df.setGroupingUsed(false);
        double d = 1.000000000000001E7;
        BigDecimal bd = new BigDecimal(d);
        StringBuffer sb = new StringBuffer("");
        FieldPosition fp = new FieldPosition(0);
        System.out.println("d = " + d);
        System.out.println("BigDecimal.toString(): " + bd.toString());
        System.out.println("DecimalFormat.format(): " + df.format(d, sb, fp));
    }
}

---------Output from the test---------------------
d = 1.000000000000001E7
BigDecimal.toString(): 10000000.00000000931322574615478515625
DecimalFormat.format(): 10000000.0000000096
--------------------------------------------------

======================================================================