Startseite › Tilmans Blog

Borland Real48 in Java

24. Januar 2009 - 9:47 – Tilman

When writing a converter tool for a binary file format, I had problems reading some of the floating point numbers. After a little bit of research I found out that this very special format was used in programs written in Borland Pascal/Delphi. Since I couldn't find any Java class that's able to read this kind of number I wrote it myself.

As far as I remember these are basically reverse IEEE floating point numbers of six bytes. (It's been a while since I wrote this. See the PDF file by Richard Biffl for a detailed description of the format.)

References:

And here is the code:

package de.tilman.util;

import java.math.BigDecimal;

/**
 * A class for representation and conversion of 6-byte/48-bit floating
 * point numbers used in Borland Pascal/Delphi.
 *
 * @author Tilman Walther
 */
public class Real48 {
       
        char[] bytes;
        double value;
       
        /**
         * Represents a number in Real48 format.
         *
         * @param bytes the byte values of the number
         */
        public Real48(char[] bytes) {
                this.bytes = bytes;
                this.value = Real48.getDoubleFromBytes(bytes);
        }
       
        /**
         * Returns the value of the number as double.
         *
         * @return the number as double
         */
        public double getDoubleValue() {
                return value;
        }
       
        /**
         * Returns the byte values of the Real48 as char[].
         *
         * @return the byte values
         */
        public char[] getBytes() {
                return bytes;
        }
       
        /**
         * Converts a Real48 in a double.
         *
         * @param bytes the number in Real48 format
         * @return the number as double
         */
        public static double getDoubleFromBytes(char[] bytes) {
                long[] real48 = new long[6];
                real48[0] = bytes[0];
                real48[1] = bytes[1];
                real48[2] = bytes[2];
                real48[3] = bytes[3];
                real48[4] = bytes[4];
                real48[5] = bytes[5];
               
                // get sign
                long sign = (real48[0] & 0x80) >> 7;
               
                // get significand
                long significand = ((real48[0] % 0x80) << 32)
                                                        + (real48[1] << 24)
                                                        + (real48[2] << 16)
                                                        + (real48[3] << 8)
                                                        + (real48[4]);
               
                // get exponent
                long exponent = bytes[5];
               
                // if the exponent is zero the real represents zero, too
                if (exponent == 0) {
                        return 0.0;
                }
               
                // convert exponent (bias for Real48 is 129, bias for IEEE 754-double is 1023)
                exponent += 894;
               
                long bits = (sign << 63) + (exponent << 52) + (significand << 13);
               
                double value = Double.longBitsToDouble(bits);
               
                // since Real48 has only a precision of 11 we need to set all digits > 11 to zero
                BigDecimal bd = new BigDecimal(value);
                bd = bd.setScale(11, BigDecimal.ROUND_DOWN);
               
                return bd.doubleValue();
        }

}