Encoding error with xerces/JDOM

Question

hej,
We have XML files that must be encoded in ISO-8859-1(i.e.encoding attribute in the document's XML declaration) to handle swedish characters. When trying to read in the XML document via JDOM (default parse is xerces) the following error is thrown:
Error message : Unable to build document (url). Invalid byte 2 of 3-byte UTF-8 sequence.
java.lang.RuntimeException: Unable to build document (url). Invalid byte 2 of 3-byte UTF-8 sequence.
at se.sfa.rfv.sat.installation.XMLFacade.build(XMLFacade.java:398)
at se.sfa.rfv.sat.installation.XMLFacade.main(XMLFacade.java:458)
If we transform the URL into an InputStream then transform the character encoding to UTF-8 likely the swedish characters (�,�, and �) will be removed. Not happening.
Is there are way to tell the underlying parser that the incoming data stream is encoded as ISO-8859-1 and build the document? Must be.
/ Matthew Young

Answer 1

Trying to use the transformer class to translate to ISO-8859-1 encoding but the following code hangs at the transform stage:
         InputStream inputstream = url.openConnection().getInputStream();
         StreamSource source = new StreamSource(inputstream);
         PipedWriter pipedwriter = new PipedWriter();
         PipedReader pipereader = new PipedReader(pipedwriter);
         PrintWriter printwriter = new PrintWriter(pipedwriter);
         StreamResult result = new StreamResult(printwriter);
         TransformerFactory factory = TransformerFactory.newInstance();
         Transformer transformer = factory.newTransformer();
         transformer.setOutputProperty(OutputKeys.ENCODING, "ISO-8859-1");
         transformer.transform(source, result);
         this.document = builder.build(pipereader);The URL passed in is the url to the input XML file. Assume the problem is related to the piping IO streams. Any ideas? / Matthew