How to Preserve the Input's Declared Encoding in the Output of javax.xml.transform.Transformer.transform? (e.g. avoid UTF-16 changing to UTF-8)

Question

Assuming this input XML

<?xml version="1.0" encoding="UTF-16"?>
<test></test>

Writing these lines of code :

StreamSource source = new StreamSource(new StringReader(/* the above XML*/));
StringWriter stringWriter = new StringWriter();
StreamResult streamResult = new StreamResult(stringWriter);
TransformerFactory.newInstance().newTransformer().transform(source, streamResult);
return stringWriter.getBuffer().toString();

Outputs for me this XML:

<?xml version="1.0" encoding="UTF-8"?>
<test></test>

(the declared encoding of UTF-16 is converted to the default UTF-8)

I know I can explicitly ask for UTF-16 output

transformer.setOutputProperty(OutputKeys.ENCODING, "UTF-16");

But the question is, how to make the output encoding automatically be the same as the input?

I don't think it's possible without manually peeking into the input stream first. By the way, if you're using the encoding of the XML header you mustn't use strings, only byte arrays. — biziclop, Jan 31 '11 at 20:31
Out of curiosity: why does that requirement exist? The resulting XML should be just as valid if it uses UTF-8 encoding. — Joachim Sauer, Feb 01 '11 at 12:10

score 5 · Accepted Answer · answered Jan 31 '11 at 20:35

To do this, you'll have to use something more sophisticated than a StreamSource. For example, a StAXSource takes an XMLStreamReader, which has the getCharacterEncodingScheme() method that tells you which encoding the input document used - you can the set that as output enocding.

score 3 · Answer 2 · answered Feb 02 '11 at 14:53

try this:

// Create an XML Stream Reader
XMLStreamReader xmlSR = XMLInputFactory.newInstance()
        .createXMLStreamReader(new StringReader(/* the above XML*/));
// Wrap the XML Stream Reader in a StAXSource
StAXSource source = new StAXSource(xmlSR);
// Create a String Writer
StringWriter stringWriter = new StringWriter();
// Create a Stream Result
StreamResult streamResult = new StreamResult(stringWriter);
// Create a transformer
Transformer transformer = TransformerFactory.newInstance().newTransformer();
// Set STANDALONE based on the source stream
transformer.setOutputProperty(OutputKeys.STANDALONE,
        xmlSR.isStandalone() ? "yes" : "no");
// Set ENCODING based on the source stream
transformer.setOutputProperty(OutputKeys.ENCODING,
        xmlSR.getCharacterEncodingScheme());
// Set VERSION based on the source stream
transformer.setOutputProperty(OutputKeys.VERSION, xmlSR.getVersion());
// Transform the source stream to the out stream
transformer.transform(source, streamResult);
// Print the results
return stringWriter.getBuffer().toString();

I am getting: Can't transform a Source of type javax.xml.transform.stax.StAXSource — honzajde, Jul 23 '14 at 12:10

score 1 · Answer 3 · answered Jan 31 '11 at 20:35

1

You need to peek into the stream first. Section F of the XML specification gives you an idea how to auto-detect the encoding.

answered Jan 31 '11 at 20:35

Jochen Bedersdorfer

4,093
24
26

1

please don't write your own code to determine the encoding. it _will_ be broken. – jtahlborn Jan 31 '11 at 21:26

score 1 · Answer 4 · answered Feb 01 '11 at 11:40

1

The XSLT processor doesn't actually know what the input encoding is (the XML parser doesn't tell it, because it doesn't need to know). You can set the output encoding using xsl:output, but to make this the same as the input encoding you're going to have to discover the input encoding first, for example by peeking at the source file before parsing it.

answered Feb 01 '11 at 11:40

Michael Kay

156,231
11
92
164

+1 Correct answer. After parsing and Infoset building, there is no need for this data, and that is why is not kept as part of the Infoset. – Feb 01 '11 at 22:05

How to Preserve the Input's Declared Encoding in the Output of javax.xml.transform.Transformer.transform? (e.g. avoid UTF-16 changing to UTF-8)

4 Answers4

Linked