In the output XML file, what is the actual hexadecimal encoding of the relevant characters, as shown by a hex editor?
What software are you using to display the output file on screen?
The most common reason for characters appearing on screen as question marks or empty boxes is that the software you are using to display the file is configured to assume a different character encoding from the one that the file actually uses.

Michael Kay
Software AG

-----Original Message-----
From: [] On Behalf Of Sonali J Kanaujia
Sent: 26 November 2002 21:31
To: Saxon List
Subject: [saxon] Scandinavian Character Encoding


I'm seeing encoding problems while using Saxon 6.4.2 and Xerces 2.0 on Solaris 2.8 (Sun 1.3 JVM).
The input xml contains Scandinavian characters.

Both the xml and the stylesheet contain:
<?xml version="1.0" encoding="iso-8859-1"?>

Additionally I've set the xs:output encoding attribute too.
<xsl:output encoding="iso-8859-1" method="xml" indent="yes"/>

But the Scandinavian chars in the output xml are corrupt and are seen as question marks or empty boxes.

Interestingly, on Windows 2000 (IBM 1.3 JVM), the characters are displayed correctly using the same stylesheet and input xml.

I'm using  transformer.transform(saxSrc, new StreamResult(outputStream)); to transform the SAXSource and get the result in a ByteArrayOutputStream.

Any help would be appreaciated.