Package com.clearnlp.constituent

Examples of com.clearnlp.constituent.CTNode

218
219
220
221
222
223
224
225
226
227
228
  }
 
  // KAIST
  private String getSpecialLabel(CTNode C)
  {
    CTNode d = C.c2d.getDependencyHead();
   
    if (CTLibKaist.isPunctuation(C) || CTLibKaist.isPunctuation(d))
      return DEPLibKr.DEP_PUNCT;
   
    if (CTLibKaist.isOnlyEJX(C, DELIM_PLUS))
View Full Code Here
303
304
305
306
307
308
309
310
311
312
313
  /** Adds dependency heads. */
  private void addDEPHeads(DEPTree dTree, CTTree cTree)
  {
    int currId, headId, size = dTree.size(), rootCount = 0;
    DEPNode dNode;
    CTNode cNode;
    String label;
   
    for (currId=1; currId<size; currId++)
    {
      dNode  = dTree.get(currId);
View Full Code Here
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
  void printTreesForCKY(String[] args)
  {
    CTReader reader = new CTReader(UTInput.createBufferedFileReader(args[0]));
    PrintStream fout = UTOutput.createPrintBufferedFileStream(args[1]);
    CTTree tree;
    CTNode root;
    int count;
   
    while ((tree = reader.nextTree()) != null)
    {
      root = tree.getRoot();
     
      if (root.getChildrenSize() == 1)
      {
        count = stripPunct(tree);
       
        if (root.getChildrenSize() > 0 && tree.getTokens().size()-count >= 4 && !containsEmptyCategories(tree) && isCKYTree(root.getChild(0)))
          fout.println(tree+"\n");
      }
    }
   
    reader.close();
View Full Code Here
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
  {
    if (instances == null)  return;
    initPBArgs(cTree.getRoot());
    int    predTokenId;
    String label;
    CTNode cNode;
   
    for (PBInstance instance : instances)
    {
      if (isPBSkip(instance, cTree))  continue;
      predTokenId = cTree.getTerminal(instance.predId).getTokenId() + 1;
     
      for (PBArg arg : instance.getArgs())
      {
        if (arg.label.startsWith(PBLib.PB_LINK))
          continue;
       
        if (arg.label.endsWith("UNDEF"))
          continue;
       
        label = arg.isLabel(PBLib.PB_REL) ? PBLib.PB_C_V : "A"+arg.label.substring(3);
       
        for (PBLoc loc : arg.getLocs())
        {
          if (arg.isLabel(PBLib.PB_REL) && loc.terminalId == instance.predId)
            continue;
         
          cNode = cTree.getNode(loc);
         
          if (!cNode.isEmptyCategoryRec())
            cNode.pbArgs.add(new StringIntPair(label, predTokenId));
        }
      }
    }
  }
View Full Code Here
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
    return null;
  }
 
  private Pair<CTNode,CTNode> getESMPair(CTNode pred)
  {
    CTNode s = pred.getNearestAncestor("+S.*");
   
    if (s != null && s.getParent().isPTag(CTLibEn.PTAG_PRN))
    {
      CTNode next = pred.getNextSibling("+S|SBAR");
     
      if (next != null)
      {
        CTNode ec = getESM(next);
        if (ec != null)  return new Pair<CTNode,CTNode>(s.getParent(), ec);
      }
    }
   
    return null;
View Full Code Here
823
824
825
826
827
828
829
830
831
832
833
834
      return getESMAux(node);
    else if (node.isPTag(CTLibEn.PTAG_SBAR))
    {
      if (node.getChildrenSize() == 2)
      {
        CTNode fst = node.getChild(0);
        CTNode snd = node.getChild(1);
       
        if (fst.isEmptyCategory() && fst.form.equals(CTLibEn.EC_ZERO))
          return getESMAux(snd);
      }
    }
View Full Code Here
838
839
840
841
842
843
844
845
846
847
848
 
  private CTNode getESMAux(CTNode node)
  {
    if (node.isEmptyCategoryRec())
    {
      CTNode ec = node.getFirstTerminal();
     
      if (ec != null && (ec.form.startsWith(CTLibEn.EC_TRACE) || ec.form.startsWith(CTLibEn.EC_ESM)))
        return ec;
    }
   
View Full Code Here
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
    String[] tmp = P_UNDER.split(loc);
    int treeId = Integer.parseInt(tmp[0]);
    int terminalId = Integer.parseInt(tmp[1]);
    CTTree tree = trees.get(treeId);
    List<CTNode> terminals = tree.getTerminals();
    CTNode node = tree.getTerminal(terminalId);
   
    if (isFirst)
    {
      int size = terminals.size();
     
      while (node.isEmptyCategory())
      {
        if (terminalId+1 < size)
          node = terminals.get(++terminalId);
        else
          return null;
      } 
    }
    else
    {
      while (node.isEmptyCategory())
      {
        if (terminalId-1 >= 0)
          node = terminals.get(--terminalId);
        else
          return null;
      }
    }
   
    return new IntIntPair(treeId, node.getTokenId());
  }
View Full Code Here
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
  private void joinConcatenations(PBInstance instance)
  {
    SortedIntArrayList ids = new SortedIntArrayList();
    CTTree tree = instance.getTree();
    int terminalId, height;
    CTNode node, parent;
    List<PBLoc> lNew;
   
    for (PBArg arg : instance.getArgs())
    {
      if (arg.isLabel(PBLib.PB_REL))  continue;
      ids.clear();
     
      for (PBLoc loc : arg.getLocs())
      {
        if (!loc.isType("") && !loc.isType(","))  return;
        if (loc.height > 0)              return;
        ids.add(loc.terminalId);
      }
     
      lNew = new ArrayList<PBLoc>();
     
      while (!ids.isEmpty())
      {
        terminalId = ids.get(0);
        height     = 0;
        node       = tree.getNode(terminalId, height);
       
        while ((parent = node.getParent()) != null && !parent.isPTag(CTLib.PTAG_TOP) && UTHppc.isSubset(ids, parent.getSubTerminalIdSet()))
        {
          node = parent;
          height++;
        }
       
        lNew.add(new PBLoc(terminalId, height, ","));
        ids.removeAll(node.getSubTerminalIdSet());
      }
     
      if (lNew.size() < arg.getLocSize())
      {
        lNew.get(0).type = "";
View Full Code Here
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
  private void fixCyclicLocs(PBInstance instance)
  {
    CTTree  tree  = instance.getTree();
    int    predId = instance.predId;
    boolean isCyc = false;
    CTNode  node, tmp;
   
    StringBuilder build = new StringBuilder();
    build.append(ERR_CYCLIC);
   
    for (PBArg arg : instance.getArgs())
    {
      if (arg.isLabel(PBLib.PB_REL))  continue;
     
      for (PBLoc loc : arg.getLocs())
      {
        if ((node = tree.getNode(loc)).getSubTerminalIdSet().contains(predId))
        {
          if (arg.isLabel(PBLib.PB_ARGM_MOD))
            loc.height = 0;
          else if (arg.isLabel(PBLib.PB_LINK_SLC) && node.isPTag(CTLibEn.PTAG_SBAR) && (tmp = node.getFirstChild("+WH.*")) != null)
            loc.set(tmp.getPBLoc(), loc.type);
          else if (node.isPTag(CTLibEn.PTAG_NP) && (tmp = node.getChild(0)).isPTag(CTLibEn.PTAG_NP) && !tmp.getSubTerminalIdSet().contains(predId))
            loc.height--;
          else
          {
            build.append(":");
            build.append(arg.label);
View Full Code Here
TOP

Related Classes of com.clearnlp.constituent.CTNode

  • com.carrotsearch.hppc.IntArrayList
  • com.carrotsearch.hppc.IntOpenHashSet
  • com.clearnlp.conversion.AbstractC2DConverter
  • com.clearnlp.conversion.EnglishC2DConverter
  • com.clearnlp.conversion.KaistC2DConverter
  • com.clearnlp.experiment.GenerateOntoNotesCoref
  • com.clearnlp.experiment.Tmp
  • com.clearnlp.run.C2DConvertMulti
  • com.clearnlp.run.PBPostProcess
  • java.util.StringTokenizer
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.