Name: dbT83986 Date: 02/10/99


I was under the impression that the Variant data member of the java.util.Locale SHOULD be used for specifying a
code page. However, when you create a locale and specify the variant, the string is uppercased. This makes it
incompatible with InputStreamReader. For example:

Locale l = new Locale( "en", "GB", "Cp850" );

String s = l.getVariant(); // s == "CP850"

InputStreamReader r = new InputStreamReader( aStream, s );
// throws UnsupportedEncodingException

Note, I had mistakenly entered this as a bug against System.getProperties("file.encoding") which does, in fact,
return a valid string. Apologies.

Please direct us to how and whether the variant can be used.
(Review ID: 49035)
======================================================================