� symbol displaying with a Unicode U+00C2 character in front of it.

Question

Using the same java application code and the same j2sdk_1.4.1_02fcs java package I get a display difference between Redhat 7.3 and Redhat EL4 AS.
nf = NumberFormat.getCurrencyInstance(new Locale(lan, con));
String res = nf.format(money);
This results in a single strange character preceeding the monetary symbol for UK pound symbol.
Instead of just displaying a � character in front of monetary values I am getting a Unicode U+00C2 character in from of is as shown :
Good = �1.23
Bad = ��1.23
I used the following simple test program to show this :
>>
import java.text.NumberFormat;
import java.util.Locale;
public class Test555
public static void main (String[] args)
NumberFormat nf = NumberFormat.getCurrencyInstance(new Locale("en", "GB"));
System.out.println(nf.format(1.23));
<<
I compiled this and ran the class on both machines...
on a Redhat 7.3 machine: �1.23
on a Redhat EL4 AS machine : ��1.23
The /bin/unicode_start program only works on the console in a VT or xwindows with a TERM type of xterm, but allows the console to properly display the characters.

Answer 1

The upgrade to Red Hat 8.0 and beyond changed the default character
encoding from ISO-8859-15 to UTF-8. The UTF-8 translation scheme will translate the Unicode representation of the Pound Sterling to a 16-bit UTF-8 representation by prepending an 0xC2 to the 0xA3 (the Pound). It is this 0xC2 that we see represented as the capital A circumflex or the "T" symbol we noticed earlier.
Is there a way to remove the prepending 0xC2 that was added by the 16bit UTF-8 representation ?