Bruno,
encoding issues in my experience can be very tricky to debug. Never
rely on any characters you see on the screen or in print to figure out
the actual encoding used. The tool which displays / prints may have
weaved its own encoding/decoding 'magic' and may show you distorted
results. Look at the actual byte stream and check how the characters
are actually encoded in it, i.e. how does the message appear 'on the
wire'.
Manuel
Manuel,
I already tried my tests without tcpmon in the middle and the problems persisted anyway.
But I'll check there on the server side if at least the words were correctly stored as iso-8859-1.
But this does not dismiss my observation about response.getXMLStreamReader().getCharacterEncodingScheme()
That method should definitely return "iso-8859-1" instead of "utf-8". I believe this is a proof that axis2 is not doing what it was supposed to.
I'd ask you to carefully read my previous posting and analyze those evidences I showed.
Thank you,
bruno.
