SOAP and Numeric Character References

[Anna Krajewska <a....@wasko.pl>]

Hi

I managed to find the reason why polish characters (with tails) looks so strange in the SOAPEnvelope. The thing is that they are replaced by numeric character references (NCR) which are defined as &# <hexadecimal value of the character> ; . When I create my DOM Element and put it on my screen I can see polish characters, they are encoded in utf-8 and each takes only one byte. When I do something like that:

SOAPBodyElement input = new SOAPBodyElement( myDomElement);

and then put the input on the screen - polish characters are gone, they are replaced by NCRs. So the implementation of the SOAPBodyElement is the place where source of the problem starts. Now, is it possible to force SOAPBodyElement not to replace my polish characters with NCRs? DOM Element is able to serialize Element which contain polish characters into byte stream in "utf-8" with no problem. So why SOAP can't?

Regards

Ania