private static Element createResults0(Element element) {
Element newElement = null;
String tag = element.getLocalName();
if (ResultsElement.TAG.equals(tag)) {
newElement = new ResultsElement();
} else if (ResultElement.TAG.equals(tag)) {
newElement = new ResultElement();
} else {
LOG.error("Unknown element: " + tag);
}
XMLUtil.copyAttributes(element, newElement);
for (int i = 0; i < element.getChildCount(); i++) {
Node child = element.getChild(i);
if (child instanceof Text) {
child = child.copy();
} else {
child = ResultsElement.createResults0((Element) child);
}
if (newElement != null && child != null) {
newElement.appendChild(child);
}
}
LOG.trace("XML :" + newElement.toXML());
return newElement;
}
private String childrenToString(Element e) {
if (e == null) {
return null;
}
String ret = "";
for (int i = 0; i < e.getChildCount(); i++) {
Node n = e.getChild(i);
if (n instanceof Element) {
ret += getText((Element) n) + ",";
}
}
return ret.substring(0, ret.length() - 1);
}
/**
* Unmarshal the text element into this object.
*
* <p>This unmarshaller only handles plain text content, although it can recognise the three
* different type elements of text, html and xhtml. This is an area that can be improved in a
* future implementation, if necessary.
*
* @param element The text element.
* @param validationProperties
* @throws UnmarshallException If the specified element is not of the correct type, where the
* localname is used to specify the valid name. Also thrown if there is an issue accessing the
* data.
*/
public SwordValidationInfo unmarshall(Element element, Properties validationProperties)
throws UnmarshallException {
if (!isInstanceOf(element, xmlName)) {
return handleIncorrectElement(element, validationProperties);
}
ArrayList<SwordValidationInfo> validationItems = new ArrayList<SwordValidationInfo>();
ArrayList<SwordValidationInfo> attributeItems = new ArrayList<SwordValidationInfo>();
try {
processUnexpectedAttributes(element, attributeItems);
int length = element.getChildCount();
if (length > 0) {
try {
unmarshallContent(element);
} catch (UnmarshallException ume) {
log.error(
"Error accessing the content of the " + xmlName.getQualifiedName() + " element");
if (validationProperties == null) {
throw ume;
} else {
SwordValidationInfo info =
new SwordValidationInfo(
xmlName, SwordValidationInfo.ERROR_WITH_CONTENT, SwordValidationInfoType.ERROR);
info.setContentDescription(element.getValue());
validationItems.add(info);
}
}
}
} catch (Exception ex) {
log.error("Unable to parse an element in " + getQualifiedName() + ": " + ex.getMessage());
if (validationProperties == null) {
throw new UnmarshallException("Unable to parse an element in " + getQualifiedName(), ex);
}
}
SwordValidationInfo result = null;
if (validationProperties != null) {
result = validate(validationItems, attributeItems, validationProperties);
}
return result;
}
protected static void createSubclassedChildren(
Element oldElement, AbstractEditorElement newElement) {
if (oldElement != null) {
for (int i = 0; i < oldElement.getChildCount(); i++) {
Node node = oldElement.getChild(i);
Node newNode = null;
if (node instanceof Text) {
String value = node.getValue();
newNode = new Text(value);
} else if (node instanceof Comment) {
newNode = new Comment(node.getValue());
} else if (node instanceof ProcessingInstruction) {
newNode = new ProcessingInstruction((ProcessingInstruction) node);
} else if (node instanceof Element) {
newNode = createEditorElement((Element) node);
} else {
throw new RuntimeException("Cannot create new node: " + node.getClass());
}
newElement.appendChild(newNode);
}
}
}
/**
* @param text
* @param validate
* @return
* @throws org.purl.sword.base.UnmarshallException
*/
public SwordValidationInfo unmarshall(Element text, Properties validationProperties)
throws UnmarshallException {
if (!isInstanceOf(text, xmlName)) {
return handleIncorrectElement(text, validationProperties);
}
ArrayList<SwordValidationInfo> validationItems = new ArrayList<SwordValidationInfo>();
ArrayList<SwordValidationInfo> attributeItems = new ArrayList<SwordValidationInfo>();
try {
initialise();
// get the attributes
int attributeCount = text.getAttributeCount();
Attribute attribute = null;
for (int i = 0; i < attributeCount; i++) {
attribute = text.getAttribute(i);
if (ATTRIBUTE_TYPE.equals(attribute.getQualifiedName())) {
boolean success = true;
String value = attribute.getValue();
if (ContentType.TEXT.toString().equals(value)) {
type = ContentType.TEXT;
} else if (ContentType.HTML.toString().equals(value)) {
type = ContentType.HTML;
} else if (ContentType.XHTML.toString().equals(value)) {
type = ContentType.XHTML;
} else {
log.error("Unable to parse extract type in " + getQualifiedName());
SwordValidationInfo info =
new SwordValidationInfo(
xmlName,
new XmlName(attribute),
"Invalid content type has been specified",
SwordValidationInfoType.ERROR);
info.setContentDescription(value);
attributeItems.add(info);
success = false;
}
if (success) {
SwordValidationInfo info = new SwordValidationInfo(xmlName, new XmlName(attribute));
info.setContentDescription(type.toString());
attributeItems.add(info);
}
} else {
SwordValidationInfo info =
new SwordValidationInfo(
xmlName,
new XmlName(attribute),
SwordValidationInfo.UNKNOWN_ATTRIBUTE,
SwordValidationInfoType.INFO);
info.setContentDescription(attribute.getValue());
attributeItems.add(info);
}
}
// retrieve all of the sub-elements
int length = text.getChildCount();
if (length > 0) {
content = unmarshallString(text);
}
} catch (Exception ex) {
log.error("Unable to parse an element in " + getQualifiedName() + ": " + ex.getMessage());
throw new UnmarshallException("Unable to parse an element in " + getQualifiedName(), ex);
}
SwordValidationInfo result = null;
if (validationProperties != null) {
result = validate(validationItems, attributeItems, validationProperties);
}
return result;
}
private static List<CoreMap> toTimexCoreMaps(Element docElem, CoreMap originalDocument) {
// --Collect Token Offsets
HashMap<Integer, Integer> beginMap = new HashMap<Integer, Integer>();
HashMap<Integer, Integer> endMap = new HashMap<Integer, Integer>();
boolean haveTokenOffsets = true;
for (CoreMap sent : originalDocument.get(CoreAnnotations.SentencesAnnotation.class)) {
for (CoreLabel token : sent.get(CoreAnnotations.TokensAnnotation.class)) {
Integer tokBegin = token.get(CoreAnnotations.TokenBeginAnnotation.class);
Integer tokEnd = token.get(CoreAnnotations.TokenEndAnnotation.class);
if (tokBegin == null || tokEnd == null) {
haveTokenOffsets = false;
}
int charBegin = token.get(CoreAnnotations.CharacterOffsetBeginAnnotation.class);
int charEnd = token.get(CoreAnnotations.CharacterOffsetEndAnnotation.class);
beginMap.put(charBegin, tokBegin);
endMap.put(charEnd, tokEnd);
}
}
// --Set Timexes
List<CoreMap> timexMaps = new ArrayList<CoreMap>();
int offset = 0;
Element textElem = docElem.getFirstChildElement("text");
for (int i = 0; i < textElem.getChildCount(); i++) {
Node content = textElem.getChild(i);
if (content instanceof Text) {
Text text = (Text) content;
offset += text.getValue().length();
} else if (content instanceof Element) {
Element child = (Element) content;
if (child.getLocalName().equals("TIMEX3")) {
Timex timex = new Timex(child);
if (child.getChildCount() != 1) {
throw new RuntimeException("TIMEX3 should only contain text " + child);
}
String timexText = child.getValue();
CoreMap timexMap = new ArrayCoreMap();
// (timex)
timexMap.set(TimexAnnotation.class, timex);
// (text)
timexMap.set(CoreAnnotations.TextAnnotation.class, timexText);
// (characters)
int charBegin = offset;
timexMap.set(CoreAnnotations.CharacterOffsetBeginAnnotation.class, charBegin);
offset += timexText.length();
int charEnd = offset;
timexMap.set(CoreAnnotations.CharacterOffsetEndAnnotation.class, charEnd);
// (tokens)
if (haveTokenOffsets) {
Integer tokBegin = beginMap.get(charBegin);
int searchStep = 1; // if no exact match, search around the character offset
while (tokBegin == null) {
tokBegin = beginMap.get(charBegin - searchStep);
if (tokBegin == null) {
tokBegin = beginMap.get(charBegin + searchStep);
}
searchStep += 1;
}
searchStep = 1;
Integer tokEnd = endMap.get(charEnd);
while (tokEnd == null) {
tokEnd = endMap.get(charEnd - searchStep);
if (tokEnd == null) {
tokEnd = endMap.get(charEnd + searchStep);
}
searchStep += 1;
}
timexMap.set(CoreAnnotations.TokenBeginAnnotation.class, tokBegin);
timexMap.set(CoreAnnotations.TokenEndAnnotation.class, tokEnd);
}
// (add)
timexMaps.add(timexMap);
} else {
throw new RuntimeException("unexpected element " + child);
}
} else {
throw new RuntimeException("unexpected content " + content);
}
}
return timexMaps;
}