XOM API for Processing XML with Java

["Michael Kay" <mike AT saxonica.com>]

You can of course create an OutputStreamWriter with a nominated
CharsetEncoder which in effect provides a callback for handling
non-encodable characters; and if you are asked to write to an
OutputStreamWriter, then you can interrogate it to find out what encoding it
is using. The behaviour you describe is only a default.

Michael Kay
http://www.saxonica.com/


> That Writer's behaviour with unsupported characters is 
> undefined and that OutputStreamWriter seems to write 
> substitution sequences for them are both flaws in the design 
> of java.io, in my humble opinion.  I think it would be safer 
> if the defined behaviour was to throw an exception, at least 
> by default, to avoid silent data loss.
> 
> Ed.
> 
> _______________________________________________
> XOM-interest mailing list
> XOM-interest AT lists.ibiblio.org
> http://lists.ibiblio.org/mailman/listinfo/xom-interest