/**
* Parse the
*
* <p>elements
*
* @return a list of tokens from the dictionary's entry
* @throws javax.xml.stream.XMLStreamException
*/
EntryToken procTransduction() throws XMLStreamException {
ArrayList<Integer> lhs = new ArrayList<Integer>();
ArrayList<Integer> rhs = new ArrayList<Integer>();
skipUntil("", COMPILER_LEFT_ELEM);
reader.next();
String name = "";
while (true) {
if (reader.isEndElement() && reader.getLocalName().equals(COMPILER_LEFT_ELEM)) {
break;
}
if (reader.isStartElement()) {
name = reader.getLocalName();
readString(lhs, name);
reader.next();
} else if (reader.isCharacters()) {
readString(lhs, "");
reader.next();
} else if (reader.isEndElement()) {
reader.next();
} else {
throw new RuntimeException(
"Error ("
+ reader.getLocation().getLineNumber()
+ ","
+ reader.getLocation().getColumnNumber()
+ "): unexpected type of event.");
}
}
skipUntil(name, COMPILER_RIGHT_ELEM);
reader.next();
while (true) {
if (reader.isEndElement() && reader.getLocalName().equals(COMPILER_RIGHT_ELEM)) {
break;
}
if (reader.isStartElement()) {
name = reader.getLocalName();
readString(rhs, name);
reader.next();
} else if (reader.isCharacters()) {
readString(rhs, "");
reader.next();
} else if (reader.isEndElement()) {
reader.next();
} else {
throw new RuntimeException(
"Error ("
+ reader.getLocation().getLineNumber()
+ ","
+ reader.getLocation().getColumnNumber()
+ "): unexpected type of event.");
}
}
skipUntil(name, COMPILER_PAIR_ELEM);
EntryToken e = new EntryToken();
e.setSingleTransduction(lhs, rhs);
return e;
}
/**
* Parse the <f> elements
*
* @return a list of tokens from the dictionary's entry
*/
EntryToken procFlag() throws XMLStreamException {
ArrayList<Integer> both_sides = new ArrayList<Integer>();
// String n = attrib(COMPILER_N_ATTR);
// String v = attrib(COMPILER_VALUE_ATTR);
String name = "";
reader.next();
while (true) {
if (reader.isEndElement() && reader.getLocalName().equals(COMPILER_FLAG_ELEM)) {
break;
}
if (reader.isStartElement()) {
name = reader.getLocalName();
readString(both_sides, name);
reader.next();
} else if (reader.isCharacters()) {
readString(both_sides, "");
reader.next();
} else if (reader.isEndElement()) {
reader.next();
} else {
throw new RuntimeException(
"Error ("
+ reader.getLocation().getLineNumber()
+ ","
+ reader.getLocation().getColumnNumber()
+ "): unexpected type of event.");
}
}
/*
* while (true) {
* reader.next();
* int type = reader.getEventType();
* if (type == XMLStreamConstants.END_ELEMENT || type == XMLStreamConstants.START_ELEMENT) {
* name = reader.getLocalName();
*
* System.err.println("name = " + name);
* }
* if (name.equals(COMPILER_FLAG_ELEM)) {
* break;
* }
* readString(both_sides, name);
* }
*/
EntryToken e = new EntryToken();
if (direction.equals(COMPILER_RESTRICTION_LR_VAL)) {
e.setSingleTransduction(new ArrayList<Integer>(), both_sides);
} else {
e.setSingleTransduction(both_sides, new ArrayList<Integer>());
}
return e;
}
/**
* Parse the <par> elements
*
* @return a list of tokens from the dictionary's entry
* @throws javax.xml.stream.XMLStreamException
*/
EntryToken procPar() throws XMLStreamException {
String paradigm_name = attrib(COMPILER_N_ATTR);
if (!paradigms.containsKey(paradigm_name)) {
throw new RuntimeException(
"Error ("
+ reader.getLocation().getLineNumber()
+ "): Undefined paradigm '"
+ paradigm_name
+ "'.");
}
EntryToken e = new EntryToken();
e.setParadigm(paradigm_name);
return e;
}
/**
* Parse the <re> elements
*
* @return a list of tokens from the dictionary's entry
* @throws javax.xml.stream.XMLStreamException
*/
EntryToken procRegexp() throws XMLStreamException {
reader.next();
StringBuffer re = new StringBuffer();
int start = reader.getTextStart();
int length = reader.getTextLength();
while (reader.isCharacters()) {
start = reader.getTextStart();
length = reader.getTextLength();
re.append(reader.getTextCharacters(), start, length);
reader.next();
}
EntryToken et = new EntryToken();
et.setRegexp(re.toString());
return et;
}
/**
* Parse the <i> elements
*
* @return a list of tokens from the dictionary's entry
* @throws javax.xml.stream.XMLStreamException
*/
EntryToken procIdentity() throws XMLStreamException {
ArrayList<Integer> both_sides = new ArrayList<Integer>();
if (!(reader.isStartElement() && reader.isEndElement())) {
// not an emoty node
}
String name = "";
reader.next();
while (true) {
if (reader.isEndElement() && reader.getLocalName().equals(COMPILER_IDENTITY_ELEM)) {
break;
}
if (reader.isStartElement()) {
name = reader.getLocalName();
readString(both_sides, name);
reader.next();
} else if (reader.isCharacters()) {
readString(both_sides, "");
reader.next();
} else if (reader.isEndElement()) {
reader.next();
} else {
throw new RuntimeException(
"Error ("
+ reader.getLocation().getLineNumber()
+ ","
+ reader.getLocation().getColumnNumber()
+ "): unexpected type of event.");
}
}
/*
* while (true) {
* reader.next();
* int type = reader.getEventType();
* if (type == XMLStreamConstants.END_ELEMENT || type == XMLStreamConstants.START_ELEMENT) {
* name = reader.getLocalName();
* }
* if (name.equals(COMPILER_IDENTITY_ELEM)) {
* break;
* }
* readString(both_sides, name);
* }
*/
EntryToken e = new EntryToken();
e.setSingleTransduction(both_sides, both_sides);
return e;
}
/**
* Insert a list of tokens into the paradigm / section being processed
*
* @param elements the list
*/
private void insertEntryTokens(ArrayList<EntryToken> elements) {
if (DEBUG) System.err.println("insertEntryTokens( " + elements);
if (!current_paradigm.equals("")) {
// compilation of paradigms
TransducerComp t = paradigms.get(current_paradigm);
if (t == null) {
t = new TransducerComp();
paradigms.put(current_paradigm, t);
}
Integer e = t.getInitial();
for (int i = 0, limit = elements.size(); i < limit; i++) {
EntryToken entry = elements.get(i);
if (entry.isParadigm()) {
if (!paradigms.containsKey(entry.paradigmName)) {
paradigms.put(entry.paradigmName, new TransducerComp());
}
e = t.insertTransducer(e, paradigms.get(entry.paradigmName));
} else if (entry.isSingleTransduction()) {
e = matchTransduction(entry.leftSide, entry.rightSide, e, t);
} else if (entry.isRegexp()) {
RegexpCompiler analyzer = new RegexpCompiler();
analyzer.initialize(alphabet);
analyzer.compile(entry.regexp);
t.setEpsilon_Tag(alphabet_cast00);
e = t.insertTransducer(e, analyzer.getTransducer());
} else {
throw new RuntimeException(
"Error (" + reader.getLocation().getLineNumber() + "): Invalid entry token.");
}
}
t.setFinal(e);
} else {
// compilation of the dictionary
TransducerComp t;
if (!sections.containsKey(current_section)) {
t = new TransducerComp();
sections.put(current_section, t);
} else {
t = sections.get(current_section);
}
int e = t.getInitial();
for (int i = 0, limit = elements.size(); i < limit; i++) {
EntryToken entry = elements.get(i);
if (entry.isParadigm()) {
final String paradigmName = entry.paradigmName;
if (i == elements.size() - 1) {
// paradigm sufix
if (!suffix_paradigms.containsKey(current_section)) {
suffix_paradigms.put(current_section, new HashMap<String, Integer>());
}
if (suffix_paradigms.get(current_section).containsKey(paradigmName)) {
t.linkStates(e, suffix_paradigms.get(current_section).get(paradigmName), 0);
e = postsuffix_paradigms.get(current_section).get(paradigmName);
} else {
e = t.insertNewSingleTransduction(alphabet_cast00, e);
suffix_paradigms.get(current_section).put(paradigmName, e);
t.setEpsilon_Tag(0);
e = t.insertTransducer(e, paradigms.get(paradigmName));
if (!postsuffix_paradigms.containsKey(current_section)) {
postsuffix_paradigms.put(current_section, new HashMap<String, Integer>());
}
postsuffix_paradigms.get(current_section).put(paradigmName, e);
}
} else if (i == 0) {
// paradigm prefix
if (!prefix_paradigms.containsKey(current_section)) {
prefix_paradigms.put(current_section, new HashMap<String, Integer>());
}
if (prefix_paradigms.get(current_section).containsKey(paradigmName)) {
e = prefix_paradigms.get(current_section).get(paradigmName);
} else {
t.setEpsilon_Tag(0);
e = t.insertTransducer(e, paradigms.get(paradigmName));
prefix_paradigms.get(current_section).put(paradigmName, e);
}
} else {
// paradigm intermediate
if (!paradigms.containsKey(paradigmName)) {
paradigms.put(paradigmName, new TransducerComp());
}
t.setEpsilon_Tag(0);
e = t.insertTransducer(e, paradigms.get(paradigmName));
}
} else if (entry.isRegexp()) {
RegexpCompiler analyzer = new RegexpCompiler();
analyzer.initialize(alphabet);
analyzer.compile(entry.regexp);
t.setEpsilon_Tag(alphabet_cast00);
e = t.insertTransducer(e, analyzer.getTransducer());
} else {
e = matchTransduction(entry.leftSide, entry.rightSide, e, t);
}
}
t.setFinal(e);
}
}