protected XMLStreamReader2 constructNonNsStreamReader(String content, boolean coal)
throws XMLStreamException {
XMLInputFactory f = getInputFactory();
setNamespaceAware(f, false);
setCoalescing(f, coal);
return (XMLStreamReader2) f.createXMLStreamReader(new StringReader(content));
}
protected XMLStreamReader2 constructNsStreamReader(InputStream in, boolean coal)
throws XMLStreamException {
XMLInputFactory f = getInputFactory();
setNamespaceAware(f, true);
setCoalescing(f, coal);
return (XMLStreamReader2) f.createXMLStreamReader(in);
}
public static Element parseWithCRC(InputStream inputStream)
throws XMLStreamException, CRCMismatchException {
XMLInputFactory xif = XMLInputFactory.newFactory();
Element element = parse(xif.createXMLStreamReader(inputStream));
checkCRC(element);
element.removeChildren(CRC_ELEMENT);
return element;
}
public void testXMLStreamWriter() throws Exception {
FlushRoot control = getControlObject();
ByteArrayOutputStream baos = new ByteArrayOutputStream();
XMLOutputFactory xof = XMLOutputFactory.newFactory();
XMLStreamWriter xsw = xof.createXMLStreamWriter(baos);
marshaller.marshal(control, xsw);
XMLInputFactory xif = XMLInputFactory.newFactory();
XMLStreamReader xsr = xif.createXMLStreamReader(new ByteArrayInputStream(baos.toByteArray()));
Object test = unmarshaller.unmarshal(xsr);
assertEquals(control, test);
}
protected static XMLStreamReader2 constructStreamReaderForFile(XMLInputFactory f, String filename)
throws IOException, XMLStreamException {
File inf = new File(filename);
XMLStreamReader sr = f.createXMLStreamReader(inf.toURL().toString(), new FileReader(inf));
assertEquals(sr.getEventType(), START_DOCUMENT);
return (XMLStreamReader2) sr;
}
@Override
public Scenario parse(InputStream inputStream)
throws IOException, SAXException, ParserConfigurationException {
OutputStream output = new ByteArrayOutputStream();
JsonXMLConfig config = new JsonXMLConfigBuilder().autoArray(true).prettyPrint(false).build();
try {
/* Create source (XML). */
XMLEventReader reader = XMLInputFactory.newInstance().createXMLEventReader(inputStream);
// XMLStreamReader reader = XMLInputFactory.newInstance().createXMLStreamReader(inputStream);
// Source source = new StAXSource(reader);
/* Create result (JSON). */
XMLEventWriter writer = new JsonXMLOutputFactory(config).createXMLEventWriter(output);
// XMLStreamWriter writer = new JsonXMLOutputFactory(config).createXMLStreamWriter(output);
// Result result = new StAXResult(writer);
/*
* Copy events from reader to writer.
*/
writer.add(reader);
/* Copy source to result via "identity transform". */
// TransformerFactory.newInstance().newTransformer().transform(source, result);
} catch (XMLStreamException e) {
e.printStackTrace();
} finally {
/* As per StAX specification, XMLStreamReader/Writer.close()
doesn't close
the underlying stream. */
output.close();
inputStream.close();
}
/*
try {
json = xmlToJson(inputStream);
} catch (JSONException e) {
e.printStackTrace();
}
String jsonString = json.toString();
System.out.println(jsonString);
GenericDocument genericDocument = new GenericDocument();
genericDocument.setJson(jsonString);
*/
GenericDocument genericDocument = new GenericDocument();
String json = output.toString();
genericDocument.setJson(json);
return genericDocument;
}
protected static boolean setSupportDTD(XMLInputFactory f, boolean state)
throws XMLStreamException {
try {
f.setProperty(XMLInputFactory.SUPPORT_DTD, Boolean.valueOf(state));
return (willSupportDTD(f) == state);
} catch (IllegalArgumentException e) {
// Let's assume that the property (or specific value) is NOT supported...
return false;
}
}
public static void main(String[] args) throws Exception {
String urlString;
if (args.length == 0) {
urlString = "http://www.w3c.org";
System.out.println("Using " + urlString);
} else urlString = args[0];
URL url = new URL(urlString);
InputStream in = url.openStream();
XMLInputFactory factory = XMLInputFactory.newInstance();
XMLStreamReader parser = factory.createXMLStreamReader(in);
while (parser.hasNext()) {
int event = parser.next();
if (event == XMLStreamConstants.START_ELEMENT) {
if (parser.getLocalName().equals("a")) {
String href = parser.getAttributeValue(null, "href");
if (href != null) System.out.println(href);
}
}
}
}
public void testOutputStreamUTF8() throws Exception {
FlushRoot control = getControlObject();
ByteArrayOutputStream baos = new ByteArrayOutputStream();
marshaller.marshal(control, baos);
XMLInputFactory xif = XMLInputFactory.newFactory();
Object test = unmarshaller.unmarshal(new ByteArrayInputStream(baos.toByteArray()));
assertEquals(control, test);
}
public void testJSONOutputStreamUTF8() throws Exception {
marshaller.setProperty(MarshallerProperties.MEDIA_TYPE, "application/json");
unmarshaller.setProperty(UnmarshallerProperties.MEDIA_TYPE, "application/json");
FlushRoot control = getControlObject();
ByteArrayOutputStream baos = new ByteArrayOutputStream();
marshaller.marshal(control, baos);
XMLInputFactory xif = XMLInputFactory.newFactory();
Object test = unmarshaller.unmarshal(new ByteArrayInputStream(baos.toByteArray()));
assertEquals(control, test);
}
protected static boolean setNamespaceAware(XMLInputFactory f, boolean state)
throws XMLStreamException {
/* Let's not assert, but see if it sticks. Some implementations
* might choose to silently ignore setting, at least for 'false'?
*/
try {
f.setProperty(XMLInputFactory.IS_NAMESPACE_AWARE, state ? Boolean.TRUE : Boolean.FALSE);
return (isNamespaceAware(f) == state);
} catch (IllegalArgumentException e) {
/* Let's assume, then, that the property (or specific value for it)
* is NOT supported...
*/
return false;
}
}
/**
* ns4 is declared on Envelope and Body and is used in faultcode. So making sure the correct ns4
* is picked up for payload source
*/
public void testPayloadSource1() throws Exception {
String soap18Msg =
"<S:Envelope xmlns:S='http://schemas.xmlsoap.org/soap/envelope/' xmlns:ns4='A'>"
+ "<S:Body xmlns:ns4='http://schemas.xmlsoap.org/soap/envelope/'>"
+ "<S:Fault>"
+ "<faultcode>ns4:Server</faultcode>"
+ "<faultstring>com.sun.istack.XMLStreamException2</faultstring>"
+ "</S:Fault>"
+ "</S:Body>"
+ "</S:Envelope>";
Message message = useStreamCodec(soap18Msg);
Source source = message.readPayloadAsSource();
InputStream is = getInputStream(source);
XMLStreamReader xsr = XMLInputFactory.newInstance().createXMLStreamReader(is);
xsr.next();
xsr.next();
assertEquals("http://schemas.xmlsoap.org/soap/envelope/", xsr.getNamespaceURI("ns4"));
}
@Override
public DbData readData(InputStream in) throws XMLStreamException {
XMLStreamReader sr = _staxInFactory.createXMLStreamReader(in);
DbData result = new DbData();
sr.nextTag();
expectTag(FIELD_TABLE, sr);
try {
while (sr.nextTag() == XMLStreamReader.START_ELEMENT) {
result.addRow(readRow(sr));
}
} catch (IllegalArgumentException iae) {
throw new XMLStreamException("Data problem: " + iae.getMessage(), sr.getLocation());
}
sr.close();
return result;
}
@Override
protected void map(LongWritable key, Text value, Mapper.Context context)
throws IOException, InterruptedException {
String document = value.toString();
System.out.println("'" + document + "'");
try {
XMLStreamReader reader =
XMLInputFactory.newInstance()
.createXMLStreamReader(new ByteArrayInputStream(document.getBytes()));
String propertyName = "";
String propertyValue = "";
String currentElement = "";
while (reader.hasNext()) {
int code = reader.next();
switch (code) {
case XMLStreamConstants.START_ELEMENT: // START_ELEMENT:
currentElement = reader.getLocalName();
break;
case XMLStreamConstants.CHARACTERS: // CHARACTERS:
if (currentElement.equalsIgnoreCase("uid")) {
propertyName += reader.getText().trim();
System.out.println(propertyName);
} else if (currentElement.equalsIgnoreCase("location")) {
propertyValue += reader.getText().trim();
System.out.println(propertyValue);
} else if (currentElement.equalsIgnoreCase("age")) {
propertyValue += ("," + reader.getText().trim());
System.out.println(propertyValue);
}
break;
}
}
reader.close();
context.write(new Text(propertyName.trim()), new Text(propertyValue.trim()));
} catch (Exception e) {
throw new IOException(e);
}
}
/**
* @param args
* @throws Exception
*/
public static void main(String[] args) throws Exception {
if (args.length != 1) {
printUsage();
}
filename = args[0];
XMLInputFactory xmlif = null;
try {
xmlif = XMLInputFactory.newInstance();
xmlif.setProperty(XMLInputFactory.IS_REPLACING_ENTITY_REFERENCES, Boolean.TRUE);
xmlif.setProperty(XMLInputFactory.IS_SUPPORTING_EXTERNAL_ENTITIES, Boolean.FALSE);
xmlif.setProperty(XMLInputFactory.IS_NAMESPACE_AWARE, Boolean.TRUE);
xmlif.setProperty(XMLInputFactory.IS_COALESCING, Boolean.TRUE);
} catch (Exception ex) {
ex.printStackTrace();
}
System.out.println("XMLInputFactory: " + xmlif);
System.out.println("filename = " + filename);
XMLStreamReader xmlr = null;
try {
FileInputStream fis = new FileInputStream(filename);
xmlr = xmlif.createXMLStreamReader(fis);
} catch (Exception ex) {
ex.printStackTrace();
}
for (int event = xmlr.next(); event != XMLStreamConstants.END_DOCUMENT; event = xmlr.next()) {
if (event == XMLStreamConstants.START_ELEMENT) {
String element = xmlr.getLocalName();
}
}
}
public abstract class StaxMateTestBase extends junit.framework.TestCase {
protected static final String BASE64_ENCODED =
"TWFuIGlzIGRpc3Rpbmd1aXNoZWQsIG5vdCBvbmx5IGJ5IGhpcyByZWFzb24sIGJ1dCBieSB0aGlz"
+ "IHNpbmd1bGFyIHBhc3Npb24gZnJvbSBvdGhlciBhbmltYWxzLCB3aGljaCBpcyBhIGx1c3Qgb2Yg"
+ "dGhlIG1pbmQsIHRoYXQgYnkgYSBwZXJzZXZlcmFuY2Ugb2YgZGVsaWdodCBpbiB0aGUgY29udGlu"
+ "dWVkIGFuZCBpbmRlZmF0aWdhYmxlIGdlbmVyYXRpb24gb2Yga25vd2xlZGdlLCBleGNlZWRzIHRo"
+ "ZSBzaG9ydCB2ZWhlbWVuY2Ugb2YgYW55IGNhcm5hbCBwbGVhc3VyZS4=";
protected static final String BASE64_DECODED_STRING =
"Man is distinguished, not only by his reason, but by this singular passion"
+ " from other animals, which is a lust of the mind, that by a perseverance of delight"
+ " in the continued and indefatigable generation of knowledge, exceeds the short vehemence"
+ " of any carnal pleasure.";
protected static final byte[] BASE64_DECODED_BYTES;
static {
byte[] data = null;
try {
data = BASE64_DECODED_STRING.getBytes("UTF-8");
} catch (Exception e) {
}
BASE64_DECODED_BYTES = data;
}
protected static XMLInputFactory _staxInputFactory = XMLInputFactory.newInstance();
protected SMInputFactory _inputFactory;
/*
/**********************************************************************
/* Factory methods
/**********************************************************************
*/
protected SMInputFactory getInputFactory() {
if (_inputFactory == null) {
_inputFactory = new SMInputFactory(getStaxInputFactory());
}
return _inputFactory;
}
protected XMLInputFactory getStaxInputFactory() {
return _staxInputFactory;
}
protected XMLStreamReader getCoalescingReader(String content) throws XMLStreamException {
XMLInputFactory f = XMLInputFactory.newInstance();
f.setProperty(XMLInputFactory.IS_COALESCING, Boolean.TRUE);
return f.createXMLStreamReader(new StringReader(content));
}
protected XMLStreamWriter getSimpleWriter(Writer w) throws XMLStreamException {
XMLOutputFactory f = XMLOutputFactory.newInstance();
return f.createXMLStreamWriter(w);
}
/*
/**********************************************************************
/* Assertion support
/**********************************************************************
*/
protected void assertTokenType(int expType, int actType) throws XMLStreamException {
assertEquals(expType, actType);
}
protected void assertTokenType(int expType, XMLStreamReader sr) throws XMLStreamException {
assertTokenType(expType, sr.getEventType());
}
protected void assertTokenType(int expType, SMInputCursor crsr) throws XMLStreamException {
assertTokenType(expType, crsr.getCurrEventCode());
}
protected void assertToken(SMEvent expEvent, SMEvent actEvent) throws XMLStreamException {
assertTokenType(expEvent.getEventCode(), actEvent.getEventCode());
}
protected void assertElem(XMLStreamReader sr, String expURI, String expLN)
throws XMLStreamException {
assertEquals(expLN, sr.getLocalName());
String actURI = sr.getNamespaceURI();
if (expURI == null || expURI.length() == 0) {
if (actURI != null && actURI.length() > 0) {
fail("Expected no namespace, got URI '" + actURI + "'");
}
} else {
assertEquals(expURI, sr.getNamespaceURI());
}
}
protected void assertException(Throwable e, String match) {
String msg = e.getMessage();
String lmsg = msg.toLowerCase();
String lmatch = match.toLowerCase();
if (lmsg.indexOf(lmatch) < 0) {
fail(
"Expected an exception with sub-string \""
+ match
+ "\": got one with message \""
+ msg
+ "\"");
}
}
/*
/**********************************************************************
/* Other accessors
/**********************************************************************
*/
/** Note: calling this method will move stream to the next non-textual event. */
protected String collectAllText(XMLStreamReader sr) throws XMLStreamException {
StringBuilder sb = new StringBuilder(100);
while (true) {
int type = sr.getEventType();
if (type == CHARACTERS || type == SPACE || type == CDATA) {
sb.append(sr.getText());
sr.next();
} else {
break;
}
}
return sb.toString();
}
}
protected XMLStreamReader getCoalescingReader(String content) throws XMLStreamException {
XMLInputFactory f = XMLInputFactory.newInstance();
f.setProperty(XMLInputFactory.IS_COALESCING, Boolean.TRUE);
return f.createXMLStreamReader(new StringReader(content));
}
protected static void setLazyParsing(XMLInputFactory f, boolean state) throws XMLStreamException {
f.setProperty(XMLInputFactory2.P_LAZY_PARSING, state ? Boolean.TRUE : Boolean.FALSE);
}
protected static void setSupportExternalEntities(XMLInputFactory f, boolean state)
throws XMLStreamException {
f.setProperty(
XMLInputFactory.IS_SUPPORTING_EXTERNAL_ENTITIES, state ? Boolean.TRUE : Boolean.FALSE);
}
protected static void setReplaceEntities(XMLInputFactory f, boolean state)
throws XMLStreamException {
f.setProperty(
XMLInputFactory.IS_REPLACING_ENTITY_REFERENCES, state ? Boolean.TRUE : Boolean.FALSE);
}
protected static XMLEventReader2 constructEventReader(XMLInputFactory f, String content)
throws XMLStreamException {
return (XMLEventReader2) f.createXMLEventReader(new StringReader(content));
}
/**
* Can annotate SVG heatmap.plus charts made by R. Reads and writes using StAX, adding row and col
* attributes to <path> elements corresponding to data points in the heatmap. All indexes can be
* calculated using nRow, nCol, nRowAnnotations and nColAnnotations.
*
* @param chart
* @throws FactoryConfigurationError
* @throws XMLStreamException
* @throws FileNotFoundException
*/
public void annotateHeatMap(HeatMapChart chart)
throws XMLStreamException, FactoryConfigurationError, FileNotFoundException {
XMLEventReader reader =
XMLInputFactory.newInstance().createXMLEventReader(new FileInputStream(inFile));
OutputStream os = new FileOutputStream(outFile);
XMLEventWriter writer = XMLOutputFactory.newInstance().createXMLEventWriter(os);
// get values from HeatMapChart
int nRow = chart.getData().getRowTargets().size();
int nCol = chart.getData().getColumnTargets().size();
// TODO get from HeatMapChart:
int nRowAnnotations = 0;
int nColAnnotations = 0;
// skip the headers and <def> bit until we reach <g id="">
while (true) {
XMLEvent event = (XMLEvent) reader.next();
if (event.isStartElement()) {
StartElement se = event.asStartElement();
if (se.getName().getLocalPart().equals(G) && se.getAttributeByName(ID) != null) {
LOG.info("<g id=\"\"> reached");
writer.add(event);
break;
}
}
writer.add(event);
}
// annotation begins here
// ROW ANNOTATIONS
if (nRowAnnotations > 0) {
LOG.info("parsing " + nRowAnnotations + " row annotations");
annotateHeatMapBlock(nRow, nRowAnnotations, "rowAnnotation", writer, reader);
}
// COLUMN ANNOTATIONS
if (nColAnnotations > 0) {
LOG.info("parsing " + nColAnnotations + " col annotations");
annotateHeatMapBlock(nColAnnotations, nCol, "colAnnotatation", writer, reader);
}
// MATRIX ANNOTATIONS
LOG.info("parsing " + (nRow * nCol) + " matrix values");
annotateHeatMapBlock(nRow, nCol, "matrix", writer, reader);
// COLUMN NAMES
LOG.info("parsing " + nCol + " column names");
int counter = 0;
while (counter < nCol) {
XMLEvent event = (XMLEvent) reader.next();
if (event.isStartElement() && event.asStartElement().getName().getLocalPart().equals(G)) {
@SuppressWarnings("unchecked")
Iterator<Attribute> attributes = event.asStartElement().getAttributes();
StartElement newSe = eventFactory.createStartElement(new QName(G), attributes, null);
writer.add(newSe);
writer.add(eventFactory.createAttribute(ID, "colName"));
writer.add(eventFactory.createAttribute(new QName("col"), Integer.toString(counter + 1)));
counter++;
} else {
writer.add(event);
}
}
// ROW NAMES
LOG.info("parsing " + nRow + " row names");
counter = 0;
while (counter < nRow) {
XMLEvent event = (XMLEvent) reader.next();
if (event.isStartElement() && event.asStartElement().getName().getLocalPart().equals(G)) {
@SuppressWarnings("unchecked")
Iterator<Attribute> attributes = event.asStartElement().getAttributes();
StartElement newSe = eventFactory.createStartElement(new QName(G), attributes, null);
writer.add(newSe);
writer.add(eventFactory.createAttribute(ID, "rowName"));
writer.add(
eventFactory.createAttribute(new QName("row"), Integer.toString(nRow - counter)));
counter++;
} else {
writer.add(event);
}
}
// finish rest of file
while (reader.hasNext()) {
XMLEvent event = (XMLEvent) reader.next();
if (event.isEndElement()) {
// close the <g id=""> tag, right before the </svg> end element
if (event.asEndElement().getName().getLocalPart().equals("svg")) {
EndElement newEe = eventFactory.createEndElement(new QName(G), null);
writer.add(newEe);
}
}
writer.add(event);
}
writer.close();
}
protected static void setValidating(XMLInputFactory f, boolean state) throws XMLStreamException {
f.setProperty(XMLInputFactory.IS_VALIDATING, Boolean.valueOf(state));
}
protected static void setCoalescing(XMLInputFactory f, boolean state) throws XMLStreamException {
f.setProperty(XMLInputFactory.IS_COALESCING, Boolean.valueOf(state));
}
protected static boolean isNamespaceAware(XMLInputFactory f) throws XMLStreamException {
return ((Boolean) f.getProperty(XMLInputFactory.IS_NAMESPACE_AWARE)).booleanValue();
}
protected static XMLInputFactory2 getNewInputFactory() {
return (XMLInputFactory2) XMLInputFactory.newInstance();
}
public static Element parse(InputStream inputStream) throws XMLStreamException {
XMLInputFactory xif = XMLInputFactory.newFactory();
return parse(xif.createXMLStreamReader(inputStream));
}
protected static boolean willSupportDTD(XMLInputFactory f) throws XMLStreamException {
return ((Boolean) f.getProperty(XMLInputFactory.SUPPORT_DTD)).booleanValue();
}
public static Element parse(InputStreamReader inputStreamReader)
throws FileNotFoundException, XMLStreamException {
XMLInputFactory xif = XMLInputFactory.newFactory();
return parse(xif.createXMLStreamReader(inputStreamReader));
}
protected static XMLStreamReader2 constructStreamReader(XMLInputFactory f, byte[] data)
throws XMLStreamException {
return (XMLStreamReader2) f.createXMLStreamReader(new ByteArrayInputStream(data));
}
