Java Trees类代码示例

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
public MemoryTreebank readTreebank() {
  MemoryTreebank treebank = new MemoryTreebank(trf);
  if (!new File(filename).isFile()) {
    return treebank;
  }
  treebank.loadPath(filename);
  // remove Parse failed
  Iterator<Tree> itr = treebank.iterator();
  while (itr.hasNext()) {
    Tree t = itr.next();
    
    Tree firstLeaf = Trees.getLeaf(t, 0);
    if (firstLeaf.value().equals("Parse")) {
      itr.remove();
    }
  }
  return treebank;
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
@Override
public Tree evaluate(Tree t, TregexMatcher m) {
  Tree topNode = children[0].evaluate(t,m);
  Tree bottomNode = children[1].evaluate(t,m);
  if(Tsurgeon.verbose) {
    System.err.println("Excising...original tree:");
    t.pennPrint(System.err);
    System.err.println("top: " + topNode + "\nbottom:" + bottomNode);
  }
  if(topNode==t)
    return null;
  Tree parent = topNode.parent(t);
  if(Tsurgeon.verbose)
    System.err.println("Parent: " + parent);
  int i = Trees.objectEqualityIndexOf(parent,topNode);
  parent.removeChild(i);
  for(Tree child : bottomNode.children()) {
    parent.addChild(i,child);
    i++;
  }
  if(Tsurgeon.verbose)
    t.pennPrint(System.err);
  return t;
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
/** If one were to call any of these other methods to get a parse tree for some input sentence
 * with the -outputFormatOptions flag of "lexicalize", they would receive their parse tree,
 * in the -outputFormat of their choice, with every leaf marked with it's head word.
 * This function does exactly that on an existing parse tree.
 * NOTE that this WILL re-lexicalize a pre-lexicalized tree, so don't pass in a tree that
 * has been lexicalized and expect to get back the same thing as what you passed in.
 */
public String lexicalize_parse_tree(String tree) throws TApplicationException
{
    try
    {
        Tree parseTree = Tree.valueOf(tree);
        Tree lexicalizedTree = Trees.lexicalize(parseTree, tlp.headFinder());
        treePrinter = ParserUtil.setOptions(null, tlp); // use defaults
        Function<Tree, Tree> a = TreeFunctions.getLabeledToDescriptiveCoreLabelTreeFunction();
        lexicalizedTree = a.apply(lexicalizedTree);
        return ParserUtil.TreeObjectToString(lexicalizedTree, treePrinter);
    }
    catch (Exception e)
    {
        // FIXME
        throw new TApplicationException(TApplicationException.INTERNAL_ERROR, e.getMessage());
    }
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
/**
 * Generate a SyntaxTreeNode Annotation corresponding to this Tree. Work
 * recursively so that the annotations are actually generated from the bottom
 * up, in order to build the consists list of annotation IDs.
 * 
 * @param tree
 *          the current subtree
 * @param rootTree
 *          the whole sentence, used to find the span of the current subtree
 * @return a GATE Annotation of type "SyntaxTreeNode"
 */
protected Annotation annotatePhraseStructureRecursively(
    AnnotationSet annotationSet, StanfordSentence stanfordSentence,
    Tree tree, Tree rootTree) {
  Annotation annotation = null;
  Annotation child;
  String label = tree.value();
  List<Tree> children = tree.getChildrenAsList();
  if(children.size() == 0) { return null; }
  /* implied else */
  /*
   * following line generates ClassCastException IntPair span =
   * tree.getSpan(); edu.stanford.nlp.ling.CategoryWordTag at
   * edu.stanford.nlp.trees.Tree.getSpan(Tree.java:393) but I think it's a bug
   * in the parser, so I'm hacking around it as follows.
   */
  int startPos = Trees.leftEdge(tree, rootTree);
  int endPos = Trees.rightEdge(tree, rootTree);
  Long startNode = stanfordSentence.startPos2offset(startPos);
  Long endNode = stanfordSentence.endPos2offset(endPos);
  List<Integer> consists = new ArrayList<Integer>();
  Iterator<Tree> childIter = children.iterator();
  while(childIter.hasNext()) {
    child =
        annotatePhraseStructureRecursively(annotationSet, stanfordSentence,
            childIter.next(), rootTree);
    if((child != null) && (!child.getType().equals(inputTokenType))) {
      consists.add(child.getId());
    }
  }
  annotation =
      annotatePhraseStructureConstituent(annotationSet, startNode, endNode,
          label, consists, tree.depth());
  return annotation;
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
/**
 * Generate a SyntaxTreeNode Annotation corresponding to this Tree.  Work 
 * recursively so that the annotations are actually generated from the 
 * bottom up, in order to build the consists list of annotation IDs.
 * 
 * @param tree  the current subtree
 * @param rootTree  the whole sentence, used to find the span of the current subtree
 * @return a GATE Annotation of type "SyntaxTreeNode"
 */
protected Annotation annotatePhraseStructureRecursively(AnnotationSet annotationSet, StanfordSentence stanfordSentence, Tree tree, Tree rootTree) {
  Annotation annotation = null;
  Annotation child;
  String label   = tree.value();

  List<Tree> children = tree.getChildrenAsList();

  if (children.size() == 0) {
    return null;
  }
  /* implied else */

  /* following line generates ClassCastException
   * 		IntPair span = tree.getSpan();
   * edu.stanford.nlp.ling.CategoryWordTag
   * at edu.stanford.nlp.trees.Tree.getSpan(Tree.java:393)
   * but I think it's a bug in the parser, so I'm hacking 
   * around it as follows. */
  int startPos = Trees.leftEdge(tree, rootTree);
  int endPos   = Trees.rightEdge(tree, rootTree);
  
  Long startNode = stanfordSentence.startPos2offset(startPos);
  Long endNode   = stanfordSentence.endPos2offset(endPos);

  List<Integer> consists = new ArrayList<Integer>();

  Iterator<Tree> childIter = children.iterator();
  while (childIter.hasNext()) {
    child = annotatePhraseStructureRecursively(annotationSet, stanfordSentence, childIter.next(), rootTree);
    if  ( (child != null)  &&
      (! child.getType().equals(inputTokenType) )) {
      consists.add(child.getId());
    }
  }
  annotation = annotatePhraseStructureConstituent(annotationSet, startNode, endNode, label, consists, tree.depth());

  return annotation;
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
private static Tree pruneHelper(Tree root, Tree nodeToPrune) {
  if(nodeToPrune==root)
    return null;
  Tree parent = nodeToPrune.parent(root);
  parent.removeChild(Trees.objectEqualityIndexOf(parent,nodeToPrune));
  if(parent.children().length==0)
    return pruneHelper(root,parent);
  return root;
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
@Override
public Tree evaluate(Tree t, TregexMatcher m) {
  Tree result = t;
  for (TsurgeonPattern child : children) {
    Tree nodeToDelete = child.evaluate(t, m);
    if (nodeToDelete == t) {
      result = null;
    }
    Tree parent = nodeToDelete.parent(t);
    parent.removeChild(Trees.objectEqualityIndexOf(parent,nodeToDelete));
  }
  return result;
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
@Override
public Tree evaluate(Tree t, TregexMatcher m) {
  Tree nodeToMove = children[0].evaluate(t,m);
  Tree oldParent = nodeToMove.parent(t);
  oldParent.removeChild(Trees.objectEqualityIndexOf(oldParent,nodeToMove));
  Pair<Tree,Integer> position = l.evaluate(t,m);
  position.first().insertDtr(nodeToMove,position.second());
  return t;
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
public static Tree firstLeaf(Tree t) {
  return Trees.getLeaf(t, 0);
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
@Override
boolean satisfies(Tree t1, Tree t2, Tree root) {
  return Trees.rightEdge(t1, root) <= Trees.leftEdge(t2, root);
}
 

import edu.stanford.nlp.trees.Trees; //导入依赖的package包/类
private List<ScoredObject<Tree>> getNext()
{
  try {
    String line;
    int parsesExpected = 0;
    int sentenceId = lastSentenceId;
    ScoredObject<Tree> curParse = null;
    List<ScoredObject<Tree>> curParses = null;
    while ((line = br.readLine()) != null) {
      line = line.trim();
      if (line.length() > 0) {
        if (parsesExpected == 0) {
          // Finished processing parses
          String[] fields = wsDelimiter.split(line, 2);
          parsesExpected = Integer.parseInt(fields[0]);
          sentenceId = Integer.parseInt(fields[1]);
          if (expectConsecutiveSentenceIds) {
           if (sentenceId != lastSentenceId+1) {
              if (lastSentenceId < sentenceId) {
                StringBuilder sb = new StringBuilder("Missing sentences");
                for (int i = lastSentenceId+1; i < sentenceId; i++) {
                  sb.append(" ").append(i);
                }
                logger.warning(sb.toString());
              } else {
                logger.warning("sentenceIds are not increasing (last="
                      + lastSentenceId + ", curr=" + sentenceId + ")");
              }
            }
          }
          lastSentenceId = sentenceId;
          curParses = new ArrayList<ScoredObject<Tree>>(parsesExpected);
        } else {
          if (curParse == null) {
            // read score
            double score = Double.parseDouble(line);
            curParses.add(curParse = new ScoredObject<Tree>(null, score));
          } else {
            // Reading a parse
            curParse.setObject(Trees.readTree(line));
            curParse = null;
            parsesExpected--;
            if (parsesExpected == 0) {
              return curParses;
            }
          }
        }
      }
    }
  } catch (IOException ex) {
    throw new RuntimeException(ex);
  }
  return null;
}