/**
* Method that not only gets currently available text from the reader, but also checks that its
* consistenly accessible using different (basic) StAX methods.
*/
protected static String getAndVerifyText(XMLStreamReader sr) throws XMLStreamException {
/* 05-Apr-2006, TSa: Although getText() is available for DTD
* and ENTITY_REFERENCE, getTextXxx() are not. Thus, can not
* do more checks for those types.
*/
int type = sr.getEventType();
if (type == ENTITY_REFERENCE || type == DTD) {
return sr.getText();
}
int expLen = sr.getTextLength();
/* Hmmh. It's only ok to return empty text for DTD event... well,
* maybe also for CDATA, since empty CDATA blocks are legal?
*/
/* !!! 01-Sep-2004, TSa:
* note: theoretically, in coalescing mode, it could be possible
* to have empty CDATA section(s) get converted to CHARACTERS,
* which would be empty... may need to enhance this to check that
* mode is not coalescing? Or something
*/
if (type == CHARACTERS) {
assertTrue("Stream reader should never return empty Strings.", (expLen > 0));
}
String text = sr.getText();
assertNotNull("getText() should never return null.", text);
assertEquals(
"Expected text length of " + expLen + ", got " + text.length(), expLen, text.length());
char[] textChars = sr.getTextCharacters();
int start = sr.getTextStart();
String text2 = new String(textChars, start, expLen);
assertEquals(text, text2);
return text;
}
private static Element parseElement(XMLStreamReader xsr) throws XMLStreamException {
// xsr points to a START_ELEMENT event. Create the element and read all its attributes
// Then read all its children events
Element element = new Element(xsr.getLocalName());
// text that will be added to the element. Text can come in different events, so we add it here
// and add it to the element at the end
StringBuilder elementText = new StringBuilder();
int attributeCount = xsr.getAttributeCount();
for (int i = 0; i < attributeCount; i++) {
element.putAttribute(xsr.getAttributeLocalName(i), xsr.getAttributeValue(i));
}
while (xsr.hasNext()) {
xsr.next();
if (xsr.getEventType() == XMLStreamConstants.END_ELEMENT) {
// element is closed. Move the cursor and return it
// check if there is some text to add before (empty text is not added, but added text is not
// trimmed)
// we set empty text also if the element has no children
if (!elementText.toString().trim().isEmpty() || !element.hasChildren()) {
element.setText(elementText.toString());
}
// xsr.next();
return element;
} else if (xsr.getEventType() == XMLStreamConstants.CHARACTERS) {
// an attribute of the current element
elementText.append(xsr.getText());
} else if (xsr.getEventType() == XMLStreamConstants.START_ELEMENT) {
// new element begins -> read it recursively and add it to the current element
element.addChild(parseElement(xsr));
}
}
// we reached the end of the document without the tag end -> error parsing
throw new XMLStreamException(
"End of the document unexpectedly reached. Element " + element.getName() + " not closed");
}
@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);
}
}
/** 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();
}
