/**
* Serialize a CAS to a file in XCAS format
*
* @param aCas CAS to serialize
* @param name output file
* @throws IOException if an I/O failure occurs
* @throws SAXException if an error occurs generating the XML text
*/
private void writeXCas(CAS aCas, File name) throws IOException, SAXException {
FileOutputStream out = null;
try {
out = new FileOutputStream(name);
XCASSerializer ser = new XCASSerializer(aCas.getTypeSystem());
XMLSerializer xmlSer = new XMLSerializer(out, true);
ser.serialize(aCas, xmlSer.getContentHandler());
} finally {
if (out != null) {
out.close();
}
}
}
/**
* Serialize a CAS to a file in XMI format
*
* @param aCas CAS to serialize
* @param name output file
* @throws SAXException -
* @throws Exception -
* @throws ResourceProcessException -
*/
private void writeXmi(CAS aCas, File name, String modelFileName)
throws IOException, SAXException {
FileOutputStream out = null;
try {
// write XMI
out = new FileOutputStream(name);
XmiCasSerializer ser = new XmiCasSerializer(aCas.getTypeSystem());
XMLSerializer xmlSer = new XMLSerializer(out, false);
ser.serialize(aCas, xmlSer.getContentHandler());
} finally {
if (out != null) {
out.close();
}
}
}
/** Copied and modified from {@link org.apache.uima.util.CasToInlineXml} */
private static String toXML(CAS cas, AnnotationsToElements converter) throws SAXException {
ByteArrayOutputStream byteArrayOutputStream = new ByteArrayOutputStream();
XMLSerializer sax2xml = new XMLSerializer(byteArrayOutputStream, false);
// get document text
String docText = cas.getDocumentText();
char[] docCharArray = docText.toCharArray();
// get iterator over annotations sorted by increasing start position and
// decreasing end position
FSIterator<AnnotationFS> iterator = cas.getAnnotationIndex().iterator();
// This is basically a recursive algorithm that has had the recursion
// removed through the use of an explicit Stack. We iterate over the
// annotations, and if an annotation contains other annotations, we
// push the parent annotation on the stack, process the children, and
// then come back to the parent later.
List<AnnotationFS> stack = new ArrayList<AnnotationFS>();
int pos = 0;
ContentHandler handler = sax2xml.getContentHandler();
handler.startDocument();
// write the start tag
converter.startRootElement(handler);
// now use null is a placeholder for this artificial Document annotation
AnnotationFS curAnnot = null;
while (iterator.isValid()) {
AnnotationFS nextAnnot = iterator.get();
if (curAnnot == null || nextAnnot.getBegin() < curAnnot.getEnd()) {
// nextAnnot's start point is within the span of curAnnot
if (curAnnot == null || nextAnnot.getEnd() <= curAnnot.getEnd()) // crossover span check
{
// nextAnnot is contained within curAnnot
// write text between current pos and beginning of nextAnnot
try {
handler.characters(docCharArray, pos, nextAnnot.getBegin() - pos);
pos = nextAnnot.getBegin();
converter.startAnnotationElement(nextAnnot, handler);
// push parent annotation on stack
stack.add(curAnnot);
// move on to next annotation
curAnnot = nextAnnot;
} catch (StringIndexOutOfBoundsException e) {
System.err.println(
"Invalid annotation range: "
+ nextAnnot.getBegin()
+ ","
+ nextAnnot.getEnd()
+ " in document of length "
+ docText.length());
}
}
iterator.moveToNext();
} else {
// nextAnnot begins after curAnnot ends
// write text between current pos and end of curAnnot
try {
handler.characters(docCharArray, pos, curAnnot.getEnd() - pos);
pos = curAnnot.getEnd();
} catch (StringIndexOutOfBoundsException e) {
System.err.println(
"Invalid annotation range: "
+ curAnnot.getBegin()
+ ","
+ curAnnot.getEnd()
+ " in document of length "
+ docText.length());
}
converter.endAnnotationElement(curAnnot, handler);
// pop next containing annotation off stack
curAnnot = stack.remove(stack.size() - 1);
}
}
// finished writing all start tags, now finish up
if (curAnnot != null) {
try {
handler.characters(docCharArray, pos, curAnnot.getEnd() - pos);
pos = curAnnot.getEnd();
} catch (StringIndexOutOfBoundsException e) {
System.err.println(
"Invalid annotation range: "
+ curAnnot.getBegin()
+ ","
+ curAnnot.getEnd()
+ "in document of length "
+ docText.length());
}
converter.endAnnotationElement(curAnnot, handler);
while (!stack.isEmpty()) {
curAnnot = stack.remove(stack.size() - 1); // pop
if (curAnnot == null) {
break;
}
try {
handler.characters(docCharArray, pos, curAnnot.getEnd() - pos);
pos = curAnnot.getEnd();
} catch (StringIndexOutOfBoundsException e) {
System.err.println(
"Invalid annotation range: "
+ curAnnot.getBegin()
+ ","
+ curAnnot.getEnd()
+ "in document of length "
+ docText.length());
}
converter.endAnnotationElement(curAnnot, handler);
}
}
if (pos < docCharArray.length) {
handler.characters(docCharArray, pos, docCharArray.length - pos);
}
converter.endRootElement(handler);
handler.endDocument();
// return XML string
return new String(byteArrayOutputStream.toByteArray());
}