/**
* Remove traces and non-terminal decorations (e.g., "-SUBJ" in "NP-SUBJ") from a Penn
* Treebank-style tree.
*
* @param inputTree
*/
public void normalizeTree(Tree inputTree) {
inputTree.label().setFromString("ROOT");
List<Pair<TregexPattern, TsurgeonPattern>> ops =
new ArrayList<Pair<TregexPattern, TsurgeonPattern>>();
List<TsurgeonPattern> ps = new ArrayList<TsurgeonPattern>();
String tregexOpStr;
TregexPattern matchPattern;
TsurgeonPattern p;
TregexMatcher matcher;
tregexOpStr = "/\\-NONE\\-/=emptynode";
matchPattern = TregexPatternFactory.getPattern(tregexOpStr);
matcher = matchPattern.matcher(inputTree);
ps.add(Tsurgeon.parseOperation("prune emptynode"));
matchPattern = TregexPatternFactory.getPattern(tregexOpStr);
p = Tsurgeon.collectOperations(ps);
ops.add(new Pair<TregexPattern, TsurgeonPattern>(matchPattern, p));
Tsurgeon.processPatternsOnTree(ops, inputTree);
Label nonterminalLabel;
tregexOpStr = "/.+\\-.+/=nonterminal < __";
matchPattern = TregexPatternFactory.getPattern(tregexOpStr);
matcher = matchPattern.matcher(inputTree);
while (matcher.find()) {
nonterminalLabel = matcher.getNode("nonterminal");
if (nonterminalLabel == null) continue;
nonterminalLabel.setFromString(tlp.basicCategory(nonterminalLabel.value()));
}
}
/**
* Tries to match a pattern against a tree. If it succeeds, apply the surgical operations
* contained in a {@link TsurgeonPattern}.
*
* @param matchPattern A {@link TregexPattern} to be matched against a {@link Tree}.
* @param p A {@link TsurgeonPattern} to apply.
* @param t the {@link Tree} to match against and perform surgery on.
* @return t, which has been surgically modified.
*/
public static Tree processPattern(TregexPattern matchPattern, TsurgeonPattern p, Tree t) {
TregexMatcher m = matchPattern.matcher(t);
while (m.find()) {
t = p.evaluate(t, m);
if (t == null) {
break;
}
m = matchPattern.matcher(t);
}
return t;
}
private void prettyPrint(PrintWriter pw, int indent) {
for (int i = 0; i < indent; i++) {
pw.print(" ");
}
if (neg) {
pw.print('!');
}
if (opt) {
pw.print('?');
}
pw.println(localString());
for (TregexPattern child : getChildren()) {
child.prettyPrint(pw, indent + 1);
}
}
private static List<Pair<TregexPattern, TsurgeonPattern>> loadOps() {
List<Pair<TregexPattern, TsurgeonPattern>> ops = new ArrayList<>();
String line = null;
try {
BufferedReader br = new BufferedReader(new StringReader(editStr));
List<TsurgeonPattern> tsp = new ArrayList<>();
while ((line = br.readLine()) != null) {
if (DEBUG) System.err.print("Pattern is " + line);
TregexPattern matchPattern = TregexPattern.compile(line);
if (DEBUG) System.err.println(" [" + matchPattern + ']');
tsp.clear();
while (continuing(line = br.readLine())) {
TsurgeonPattern p = Tsurgeon.parseOperation(line);
if (DEBUG) System.err.println("Operation is " + line + " [" + p + ']');
tsp.add(p);
}
if (!tsp.isEmpty()) {
TsurgeonPattern tp = Tsurgeon.collectOperations(tsp);
ops.add(new Pair<>(matchPattern, tp));
}
} // while not at end of file
} catch (IOException ioe) {
ioe.printStackTrace();
}
return ops;
}
public static boolean filter(PhraseInfo phrase) {
if (phrase == null || phrase.getTree() == null) return false;
Tree phraseTree = phrase.getTree();
// __ < ( NP < (PRP !< it|them) ) | < ( NP < ( NP < (PRP !< it|them) ) )
String invalidPronounPattern =
" ( NP < (PRP=prp !< " + Utils.wordsConjuction(Rules.VALID_PRONOUNS) + ") ) ";
String filterPattern =
"__ < " + invalidPronounPattern + " | < ( NP <" + invalidPronounPattern + ")";
TregexPattern tregexPattern = TregexPattern.compile(filterPattern);
TregexMatcher matcher = tregexPattern.matcher(phraseTree);
if (matcher.matches()) {
Proof proof = new Proof(ProofType.FAIL_PERSONAL_PRONOUN);
Tree evdTree = matcher.getNode("prp");
proof.setEvidenceTree(evdTree);
phrase.addProof(proof);
return false;
}
return true;
}
public static TregexPattern getPattern(String tregex) {
if (instance == null) {
instance = new TregexPatternFactory();
}
Map<String, TregexPattern> myMap = instance.getMap();
TregexPattern pattern = myMap.get(tregex);
if (pattern == null) {
try {
pattern = TregexPattern.compile(tregex);
myMap.put(tregex, pattern);
} catch (Exception e) {
e.printStackTrace();
}
}
return pattern;
}
// todo: add an option to only print each tree once, regardless. Most useful in conjunction
// with -w
public void visitTree(Tree t) {
treeNumber++;
if (printTree) {
pw.print(treeNumber + ":");
pw.println("Next tree read:");
tp.printTree(t, pw);
}
TregexMatcher match = p.matcher(t);
if (printNonMatchingTrees) {
if (match.find()) numMatches++;
else tp.printTree(t, pw);
return;
}
Tree lastMatchingRootNode = null;
while (match.find()) {
if (oneMatchPerRootNode) {
if (lastMatchingRootNode == match.getMatch()) continue;
else lastMatchingRootNode = match.getMatch();
}
numMatches++;
if (printFilename && treebank instanceof DiskTreebank) {
DiskTreebank dtb = (DiskTreebank) treebank;
pw.print("# ");
pw.println(dtb.getCurrentFilename());
}
if (printSubtreeCode) {
pw.print(treeNumber);
pw.print(':');
pw.println(match.getMatch().nodeNumber(t));
}
if (printMatches) {
if (reportTreeNumbers) {
pw.print(treeNumber);
pw.print(": ");
}
if (printTree) {
pw.println("Found a full match:");
}
if (printWholeTree) {
tp.printTree(t, pw);
} else if (handles != null) {
if (printTree) {
pw.println("Here's the node you were interested in:");
}
for (String handle : handles) {
Tree labeledNode = match.getNode(handle);
if (labeledNode == null) {
System.err.println(
"Error!! There is no matched node \""
+ handle
+ "\"! Did you specify such a label in the pattern?");
} else {
tp.printTree(labeledNode, pw);
}
}
} else {
tp.printTree(match.getMatch(), pw);
}
// pw.println(); // TreePrint already puts a blank line in
} // end if (printMatches)
} // end while match.find()
} // end visitTree
/**
* Prints out all matches of a tree pattern on each tree in the path. Usage: <br>
* <br>
* <code>
* java edu.stanford.nlp.trees.tregex.TregexPattern [[-TCwfosnu] [-filter] [-h <node-name>]]* pattern
* filepath </code>
*
* <p>Arguments:<br>
*
* <ul>
* <li><code>pattern</code>: the tree pattern which optionally names some set of nodes (i.e.,
* gives it the "handle") <code>=name</code> (for some arbitrary string "name")
* <li><code>filepath</code>: the path to files with trees. If this is a directory, there will
* be recursive descent and the pattern will be run on all files beneath the specified
* directory.
* </ul>
*
* <p>Options:<br>
* <li><code>-C</code> suppresses printing of matches, so only the number of matches is printed.
* <li><code>-w</code> causes the whole of a tree that matches to be printed.
* <li><code>-f</code> causes the filename to be printed.
* <li><code>-i <filename></code> causes the pattern to be matched to be read from <code>
* <filename></code> rather than the command line. Don't specify a pattern when this
* option is used.
* <li><code>-o</code> Specifies that each tree node can be reported only once as the root of a
* match (by default a node will be printed once for every <em>way</em> the pattern matches).
* <li><code>-s</code> causes trees to be printed all on one line (by default they are pretty
* printed).
* <li><code>-n</code> causes the number of the tree in which the match was found to be printed
* before every match.
* <li><code>-u</code> causes only the label of each matching node to be printed, not complete
* subtrees.
* <li><code>-t</code> causes only the yield (terminal words) of the selected node to be printed
* (or the yield of the whole tree, if the <code>-w</code> option is used).
* <li><code>-encoding <charset_encoding></code> option allows specification of character
* encoding of trees..
* <li><code>-h <node-handle></code> If a <code>-h</code> option is given, the root tree
* node will not be printed. Instead, for each <code>node-handle</code> specified, the node
* matched and given that handle will be printed. Multiple nodes can be printed by using the
* <code>-h</code> option multiple times on a single command line.
* <li><code>-hf <headfinder-class-name></code> use the specified {@link HeadFinder} class
* to determine headship relations.
* <li><code>-hfArg <string></code> pass a string argument in to the {@link HeadFinder}
* class's constructor. <code>-hfArg</code> can be used multiple times to pass in multiple
* arguments.
* <li><code>-trf <TreeReaderFactory-class-name></code> use the specified {@link
* TreeReaderFactory} class to read trees from files.
* <li><code>-v</code> print every tree that contains no matches of the specified pattern, but
* print no matches to the pattern.
* <li><code>-x</code> Instead of the matched subtree, print the matched subtree's identifying
* number as defined in <tt>tgrep2</tt>:a unique identifier for the subtree and is in the form
* s:n, where s is an integer specifying the sentence number in the corpus (starting with 1),
* and n is an integer giving the order in which the node is encountered in a depth-first
* search starting with 1 at top node in the sentence tree.
* <li><code>-extract <code> <tree-file></code> extracts the subtree s:n specified by
* <tt>code</tt> from the specified <tt>tree-file</tt>. Overrides all other behavior of
* tregex. Can't specify multiple encodings etc. yet.
* <li><code>-extractFile <code-file> <tree-file></code> extracts every subtree
* specified by the subtree codes in <tt>code-file</tt>, which must appear exactly one per
* line, from the specified <tt>tree-file</tt>. Overrides all other behavior of tregex. Can't
* specify multiple encodings etc. yet.
* <li><code>-filter</code> causes this to act as a filter, reading tree input from stdin
* <li><code>-T</code> causes all trees to be printed as processed (for debugging purposes).
* Otherwise only matching nodes are printed.
* <li><code>-macros <filename></code> filename with macro substitutions to use. file with
* tab separated lines original-tab-replacement
* </ul>
*/
public static void main(String[] args) throws IOException {
Timing.startTime();
StringBuilder treePrintFormats = new StringBuilder();
String printNonMatchingTreesOption = "-v";
String subtreeCodeOption = "-x";
String extractSubtreesOption = "-extract";
String extractSubtreesFileOption = "-extractFile";
String inputFileOption = "-i";
String headFinderOption = "-hf";
String headFinderArgOption = "-hfArg";
String trfOption = "-trf";
String headFinderClassName = null;
String[] headFinderArgs = StringUtils.EMPTY_STRING_ARRAY;
String treeReaderFactoryClassName = null;
String printHandleOption = "-h";
String markHandleOption = "-k";
String encodingOption = "-encoding";
String encoding = "UTF-8";
String macroOption = "-macros";
String macroFilename = "";
String yieldOnly = "-t";
String printAllTrees = "-T";
String quietMode = "-C";
String wholeTreeMode = "-w";
String filenameOption = "-f";
String oneMatchPerRootNodeMode = "-o";
String reportTreeNumbers = "-n";
String rootLabelOnly = "-u";
String oneLine = "-s";
Map<String, Integer> flagMap = Generics.newHashMap();
flagMap.put(extractSubtreesOption, 2);
flagMap.put(extractSubtreesFileOption, 2);
flagMap.put(subtreeCodeOption, 0);
flagMap.put(printNonMatchingTreesOption, 0);
flagMap.put(encodingOption, 1);
flagMap.put(inputFileOption, 1);
flagMap.put(printHandleOption, 1);
flagMap.put(markHandleOption, 2);
flagMap.put(headFinderOption, 1);
flagMap.put(headFinderArgOption, 1);
flagMap.put(trfOption, 1);
flagMap.put(macroOption, 1);
flagMap.put(yieldOnly, 0);
flagMap.put(quietMode, 0);
flagMap.put(wholeTreeMode, 0);
flagMap.put(printAllTrees, 0);
flagMap.put(filenameOption, 0);
flagMap.put(oneMatchPerRootNodeMode, 0);
flagMap.put(reportTreeNumbers, 0);
flagMap.put(rootLabelOnly, 0);
flagMap.put(oneLine, 0);
Map<String, String[]> argsMap = StringUtils.argsToMap(args, flagMap);
args = argsMap.get(null);
if (argsMap.containsKey(encodingOption)) {
encoding = argsMap.get(encodingOption)[0];
System.err.println("Encoding set to " + encoding);
}
PrintWriter errPW = new PrintWriter(new OutputStreamWriter(System.err, encoding), true);
if (argsMap.containsKey(extractSubtreesOption)) {
List<String> subTreeStrings =
Collections.singletonList(argsMap.get(extractSubtreesOption)[0]);
extractSubtrees(subTreeStrings, argsMap.get(extractSubtreesOption)[1]);
return;
}
if (argsMap.containsKey(extractSubtreesFileOption)) {
List<String> subTreeStrings =
Arrays.asList(
IOUtils.slurpFile(argsMap.get(extractSubtreesFileOption)[0]).split("\n|\r|\n\r"));
extractSubtrees(subTreeStrings, argsMap.get(extractSubtreesFileOption)[0]);
return;
}
if (args.length < 1) {
errPW.println(
"Usage: java edu.stanford.nlp.trees.tregex.TregexPattern [-T] [-C] [-w] [-f] [-o] [-n] [-s] [-filter] [-hf class] [-trf class] [-h handle]* pattern [filepath]");
return;
}
String matchString = args[0];
if (argsMap.containsKey(macroOption)) {
macroFilename = argsMap.get(macroOption)[0];
}
if (argsMap.containsKey(headFinderOption)) {
headFinderClassName = argsMap.get(headFinderOption)[0];
errPW.println("Using head finder " + headFinderClassName + "...");
}
if (argsMap.containsKey(headFinderArgOption)) {
headFinderArgs = argsMap.get(headFinderArgOption);
}
if (argsMap.containsKey(trfOption)) {
treeReaderFactoryClassName = argsMap.get(trfOption)[0];
errPW.println("Using tree reader factory " + treeReaderFactoryClassName + "...");
}
if (argsMap.containsKey(printAllTrees)) {
TRegexTreeVisitor.printTree = true;
}
if (argsMap.containsKey(inputFileOption)) {
String inputFile = argsMap.get(inputFileOption)[0];
matchString = IOUtils.slurpFile(inputFile, encoding);
String[] newArgs = new String[args.length + 1];
System.arraycopy(args, 0, newArgs, 1, args.length);
args = newArgs;
}
if (argsMap.containsKey(quietMode)) {
TRegexTreeVisitor.printMatches = false;
TRegexTreeVisitor.printNumMatchesToStdOut = true;
}
if (argsMap.containsKey(printNonMatchingTreesOption)) {
TRegexTreeVisitor.printNonMatchingTrees = true;
}
if (argsMap.containsKey(subtreeCodeOption)) {
TRegexTreeVisitor.printSubtreeCode = true;
TRegexTreeVisitor.printMatches = false;
}
if (argsMap.containsKey(wholeTreeMode)) {
TRegexTreeVisitor.printWholeTree = true;
}
if (argsMap.containsKey(filenameOption)) {
TRegexTreeVisitor.printFilename = true;
}
if (argsMap.containsKey(oneMatchPerRootNodeMode)) TRegexTreeVisitor.oneMatchPerRootNode = true;
if (argsMap.containsKey(reportTreeNumbers)) TRegexTreeVisitor.reportTreeNumbers = true;
if (argsMap.containsKey(rootLabelOnly)) {
treePrintFormats.append(TreePrint.rootLabelOnlyFormat).append(',');
} else if (argsMap.containsKey(oneLine)) { // display short form
treePrintFormats.append("oneline,");
} else if (argsMap.containsKey(yieldOnly)) {
treePrintFormats.append("words,");
} else {
treePrintFormats.append("penn,");
}
HeadFinder hf = new CollinsHeadFinder();
if (headFinderClassName != null) {
Class[] hfArgClasses = new Class[headFinderArgs.length];
for (int i = 0; i < hfArgClasses.length; i++) hfArgClasses[i] = String.class;
try {
hf =
(HeadFinder)
Class.forName(headFinderClassName)
.getConstructor(hfArgClasses)
.newInstance(
(Object[])
headFinderArgs); // cast to Object[] necessary to avoid varargs-related
// warning.
} catch (Exception e) {
throw new RuntimeException("Error occurred while constructing HeadFinder: " + e);
}
}
TRegexTreeVisitor.tp =
new TreePrint(treePrintFormats.toString(), new PennTreebankLanguagePack());
try {
// TreePattern p = TreePattern.compile("/^S/ > S=dt $++ '' $-- ``");
TregexPatternCompiler tpc = new TregexPatternCompiler(hf);
Macros.addAllMacros(tpc, macroFilename, encoding);
TregexPattern p = tpc.compile(matchString);
errPW.println("Pattern string:\n" + p.pattern());
errPW.println("Parsed representation:");
p.prettyPrint(errPW);
String[] handles = argsMap.get(printHandleOption);
if (argsMap.containsKey("-filter")) {
TreeReaderFactory trf = getTreeReaderFactory(treeReaderFactoryClassName);
treebank =
new MemoryTreebank(
trf, encoding); // has to be in memory since we're not storing it on disk
// read from stdin
Reader reader = new BufferedReader(new InputStreamReader(System.in, encoding));
((MemoryTreebank) treebank).load(reader);
reader.close();
} else if (args.length == 1) {
errPW.println("using default tree");
TreeReader r =
new PennTreeReader(
new StringReader(
"(VP (VP (VBZ Try) (NP (NP (DT this) (NN wine)) (CC and) (NP (DT these) (NNS snails)))) (PUNCT .))"),
new LabeledScoredTreeFactory(new StringLabelFactory()));
Tree t = r.readTree();
treebank = new MemoryTreebank();
treebank.add(t);
} else {
int last = args.length - 1;
errPW.println("Reading trees from file(s) " + args[last]);
TreeReaderFactory trf = getTreeReaderFactory(treeReaderFactoryClassName);
treebank = new DiskTreebank(trf, encoding);
treebank.loadPath(args[last], null, true);
}
TRegexTreeVisitor vis = new TRegexTreeVisitor(p, handles, encoding);
treebank.apply(vis);
Timing.endTime();
if (TRegexTreeVisitor.printMatches) {
errPW.println("There were " + vis.numMatches() + " matches in total.");
}
if (TRegexTreeVisitor.printNumMatchesToStdOut) {
System.out.println(vis.numMatches());
}
} catch (IOException e) {
e.printStackTrace();
} catch (TregexParseException e) {
errPW.println("Error parsing expression: " + args[0]);
errPW.println("Parse exception: " + e.toString());
}
}
public class ShiftReduceParserQuery implements ParserQuery {
Debinarizer debinarizer = new Debinarizer(false);
List<? extends HasWord> originalSentence;
private State initialState, finalState;
Tree debinarized;
boolean success;
boolean unparsable;
final ShiftReduceParser parser;
List<ParserConstraint> constraints = null;
public ShiftReduceParserQuery(ShiftReduceParser parser) {
this.parser = parser;
}
@Override
public boolean parse(List<? extends HasWord> sentence) {
this.originalSentence = sentence;
initialState = ShiftReduceParser.initialStateFromTaggedSentence(sentence);
return parseInternal();
}
public boolean parse(Tree tree) {
this.originalSentence = tree.yieldHasWord();
initialState = ShiftReduceParser.initialStateFromGoldTagTree(tree);
return parseInternal();
}
// TODO: we are assuming that sentence final punctuation always has
// either . or PU as the tag.
private static TregexPattern rearrangeFinalPunctuationTregex =
TregexPattern.compile(
"__ !> __ <- (__=top <- (__ <<- (/[.]|PU/=punc < /[.!?。！？]/ ?> (__=single <: =punc))))");
private static TsurgeonPattern rearrangeFinalPunctuationTsurgeon =
Tsurgeon.parseOperation("[move punc >-1 top] [if exists single prune single]");
private boolean parseInternal() {
final int maxBeamSize = Math.max(parser.op.testOptions().beamSize, 1);
success = true;
unparsable = false;
PriorityQueue<State> beam =
new PriorityQueue<>(maxBeamSize + 1, ScoredComparator.ASCENDING_COMPARATOR);
beam.add(initialState);
// TODO: don't construct as many PriorityQueues
while (beam.size() > 0) {
// System.err.println("================================================");
// System.err.println("Current beam:");
// System.err.println(beam);
PriorityQueue<State> oldBeam = beam;
beam = new PriorityQueue<>(maxBeamSize + 1, ScoredComparator.ASCENDING_COMPARATOR);
State bestState = null;
for (State state : oldBeam) {
Collection<ScoredObject<Integer>> predictedTransitions =
parser.model.findHighestScoringTransitions(state, true, maxBeamSize, constraints);
// System.err.println("Examining state: " + state);
for (ScoredObject<Integer> predictedTransition : predictedTransitions) {
Transition transition = parser.model.transitionIndex.get(predictedTransition.object());
State newState = transition.apply(state, predictedTransition.score());
// System.err.println(" Transition: " + transition + " (" + predictedTransition.score() +
// ")");
if (bestState == null || bestState.score() < newState.score()) {
bestState = newState;
}
beam.add(newState);
if (beam.size() > maxBeamSize) {
beam.poll();
}
}
}
if (beam.size() == 0) {
// Oops, time for some fallback plan
// This can happen with the set of constraints given by the original paper
// For example, one particular French model had a situation where it would reach
// @Ssub @Ssub .
// without a left(Ssub) transition, so finishing the parse was impossible.
// This will probably result in a bad parse, but at least it
// will result in some sort of parse.
for (State state : oldBeam) {
Transition transition = parser.model.findEmergencyTransition(state, constraints);
if (transition != null) {
State newState = transition.apply(state);
if (bestState == null || bestState.score() < newState.score()) {
bestState = newState;
}
beam.add(newState);
}
}
}
// bestState == null only happens when we have failed to make progress, so quit
// If the bestState is finished, we are done
if (bestState == null || bestState.isFinished()) {
break;
}
}
List<State> bestParses;
if (beam.size() == 0) {
success = false;
unparsable = true;
debinarized = null;
finalState = null;
bestParses = Collections.emptyList();
} else {
// TODO: filter out beam elements that aren't finished
bestParses = Generics.newArrayList(beam);
Collections.sort(bestParses, beam.comparator());
Collections.reverse(bestParses);
finalState = bestParses.get(0);
debinarized = debinarizer.transformTree(finalState.stack.peek());
debinarized =
Tsurgeon.processPattern(
rearrangeFinalPunctuationTregex, rearrangeFinalPunctuationTsurgeon, debinarized);
}
return success;
}
/** TODO: if we add anything interesting to report, we should report it here */
@Override
public boolean parseAndReport(List<? extends HasWord> sentence, PrintWriter pwErr) {
boolean success = parse(sentence);
// System.err.println(getBestTransitionSequence());
// System.err.println(getBestBinarizedParse());
return success;
}
public Tree getBestBinarizedParse() {
return finalState.stack.peek();
}
public List<Transition> getBestTransitionSequence() {
return finalState.transitions.asList();
}
@Override
public double getPCFGScore() {
return finalState.score;
}
@Override
public Tree getBestParse() {
return debinarized;
}
/** TODO: can we get away with not calling this PCFG? */
@Override
public Tree getBestPCFGParse() {
return debinarized;
}
@Override
public Tree getBestDependencyParse(boolean debinarize) {
return null;
}
@Override
public Tree getBestFactoredParse() {
return null;
}
/** TODO: if this is a beam, return all equal parses */
@Override
public List<ScoredObject<Tree>> getBestPCFGParses() {
ScoredObject<Tree> parse = new ScoredObject<>(debinarized, finalState.score);
return Collections.singletonList(parse);
}
@Override
public boolean hasFactoredParse() {
return false;
}
/** TODO: return more if this used a beam */
@Override
public List<ScoredObject<Tree>> getKBestPCFGParses(int kbestPCFG) {
ScoredObject<Tree> parse = new ScoredObject<>(debinarized, finalState.score);
return Collections.singletonList(parse);
}
@Override
public List<ScoredObject<Tree>> getKGoodFactoredParses(int kbest) {
throw new UnsupportedOperationException();
}
@Override
public KBestViterbiParser getPCFGParser() {
// TODO: find some way to treat this as a KBestViterbiParser?
return null;
}
@Override
public KBestViterbiParser getDependencyParser() {
return null;
}
@Override
public KBestViterbiParser getFactoredParser() {
return null;
}
@Override
public void setConstraints(List<ParserConstraint> constraints) {
this.constraints = constraints;
}
@Override
public boolean saidMemMessage() {
return false;
}
@Override
public boolean parseSucceeded() {
return success;
}
/** TODO: skip sentences which are too long */
@Override
public boolean parseSkipped() {
return false;
}
@Override
public boolean parseFallback() {
return false;
}
/** TODO: add memory handling? */
@Override
public boolean parseNoMemory() {
return false;
}
@Override
public boolean parseUnparsable() {
return unparsable;
}
@Override
public List<? extends HasWord> originalSentence() {
return originalSentence;
}
/** TODO: clearly this should be a default method in ParserQuery once Java 8 comes out */
@Override
public void restoreOriginalWords(Tree tree) {
if (originalSentence == null || tree == null) {
return;
}
List<Tree> leaves = tree.getLeaves();
if (leaves.size() != originalSentence.size()) {
throw new IllegalStateException(
"originalWords and sentence of different sizes: "
+ originalSentence.size()
+ " vs. "
+ leaves.size()
+ "\n Orig: "
+ Sentence.listToString(originalSentence)
+ "\n Pars: "
+ Sentence.listToString(leaves));
}
// TODO: get rid of this cast
Iterator<? extends Label> wordsIterator =
(Iterator<? extends Label>) originalSentence.iterator();
for (Tree leaf : leaves) {
leaf.setLabel(wordsIterator.next());
}
}
}