Package com.clearnlp.constituent

Examples of com.clearnlp.constituent.CTReader

742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
    }
  }
 
  void extractPos(String[] args)
  {
    CTReader reader = new CTReader(UTInput.createBufferedFileReader(args[0]));
    Set<String> set = new HashSet<String>();
    Pattern delim = Pattern.compile("\\+");
    CTTree tree;
   
    while ((tree = reader.nextTree()) != null)
    {
      for (CTNode node : tree.getTokens())
        for (String pos : delim.split(node.pTag))
          set.add(pos);
    }
View Full Code Here
896
897
898
899
900
901
902
903
904
905
906
907
908
909
    fnew.close();
  }
 
  void wc(String inputFile)
  {
    CTReader reader = new CTReader(UTInput.createBufferedFileReader(inputFile));
    CTTree tree;
    int sc, wc;
   
    for (sc=0,wc=0; (tree = reader.nextTree()) != null; sc++)
      wc += tree.getTokens().size();
   
    System.out.println(sc+" "+wc);
  }
View Full Code Here
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
    System.out.println(sc+" "+wc);
  }
 
  void stripTrees(String[] args)
  {
    CTReader reader = new CTReader(UTInput.createBufferedFileReader(args[0]));
    PrintStream fout = UTOutput.createPrintBufferedFileStream(args[0]+".strip");
    Set<String> set = new HashSet<String>();
    String forms;
    CTTree tree;
    int i;
   
    for (i=0; (tree = reader.nextTree()) != null; i++)
    {
      forms = tree.toForms();
     
      if (!set.contains(forms))
      {
View Full Code Here
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
    System.out.println(i+" -> "+set.size());
  }
 
  void splitTrees(String[] args)
  {
    CTReader reader = new CTReader(UTInput.createBufferedFileReader(args[0]));
    PrintStream[] fout = new PrintStream[4];
    CTTree tree;
    int i, j;
   
    fout[0] = UTOutput.createPrintBufferedFileStream(args[0]+".trn.parse");
    fout[1] = UTOutput.createPrintBufferedFileStream(args[0]+".trn.raw");
    fout[2] = UTOutput.createPrintBufferedFileStream(args[0]+".tst.parse");
    fout[3] = UTOutput.createPrintBufferedFileStream(args[0]+".tst.raw");
   
    for (i=0; (tree = reader.nextTree()) != null; i++)
    {
      j = (i%6 == 0) ? 2 : 0;
     
      fout[j]  .println(tree.toString()+"\n");
      fout[j+1].println(tree.toForms());
View Full Code Here
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
    for (PrintStream f : fout)  f.close();
  }
 
  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();
    fout.close();
  }
View Full Code Here
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
    return true;
  }
 
  void traverse(String inputFile)
  {
    CTReader reader = new CTReader(UTInput.createBufferedFileReader(inputFile));
    CTTree tree;
   
    while ((tree = reader.nextTree()) != null)
      traverseAux(tree.getRoot());
  }
View Full Code Here
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
   * @param outputDir the directory to save normalized tree files.
   * @param extension the tree file extension (e.g., {@code parse}).
   */
  public void run(String inputDir, String outputDir, String extension)
  {
    CTReader      reader;
    CTTree        tree;
    PrintStream   fout;
    StringBuilder build;
   
    File dir = new File(outputDir);
    if (!dir.exists())  dir.mkdirs();
   
    inputDir  += File.separator;
    outputDir += File.separator;
   
    for (String filename : new File(inputDir).list(new FileExtFilter(extension)))
    {
      reader = new CTReader(UTInput.createBufferedFileReader(inputDir + filename));
      fout   = UTOutput.createPrintBufferedFileStream(outputDir + filename);
     
      while ((tree = reader.nextTree()) != null)
      {
      /*  build = new StringBuilder();
       
        for (CTNode node : tree.getTokens())
        {
          build.append(" ");
          build.append(node.form);
          build.append("/");
          build.append(node.pTag);
        }
       
        fout.println(build.substring(1));*/
       
        for (CTNode node : tree.getTokens())
        {
          build = new StringBuilder();
         
          build.append(node.form);
          build.append(AbstractColumnReader.DELIM_COLUMN);
          build.append(node.pTag);
         
          fout.println(build.toString());
        }
       
        fout.println();
      }
     
      reader.close();
      fout.close();
    }
  }
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
88
89
90
   * @param inputFile the name of the file containing input trees.
   * @param outputFile the name of the file to contain output trees.
   */
  public void print(String inputFile, String outputFile)
  {
    CTReader    reader = new CTReader(UTInput.createBufferedFileReader(inputFile));
    PrintStream fout   = UTOutput.createPrintBufferedFileStream(outputFile);
    CTTree      tree;

    while ((tree = reader.nextTree()) != null)
      fout.println(tree.toStringLine());
   
    fout.close();
  }
View Full Code Here
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
    }
  }
 
  public void normalize(String inputFile, String outputFile)
  {
    CTReader reader = new CTReader(UTInput.createBufferedFileReader(inputFile));
    PrintStream fout = UTOutput.createPrintBufferedFileStream(outputFile);
    CTTree tree;
   
    while ((tree = reader.nextTree()) != null)
    {
      CTLib.normalizeIndices(tree);
      fout.println(tree.toString()+"\n");
    }
   
    reader.close();
    fout.close();
  }
View Full Code Here
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
    Set<String> ids = UTInput.getStringSet(new FileInputStream(args[0]));
    ObjectIntHashMap<String> scm = new ObjectIntHashMap<>();
    ObjectIntHashMap<String> wcm = new ObjectIntHashMap<>();
    Pattern p = Pattern.compile("/");
    String root = args[1];
    CTReader reader;
    String genre;
    CTTree tree;
    int sc, wc;
   
    for (String id : ids)
    {
      reader = new CTReader(UTInput.createBufferedFileReader(root+"/"+id+".parse"));
      genre  = p.split(id)[0];
      sc = wc = 0;
     
      while ((tree = reader.nextTree()) != null)
      {
        wc += tree.getTokens().size();
        sc ++;
      }
     
View Full Code Here
TOP

Related Classes of com.clearnlp.constituent.CTReader

  • com.clearnlp.demo.Demo
  • com.clearnlp.experiment.CompareCTTrees
  • com.clearnlp.experiment.CT2POS
  • com.clearnlp.experiment.GenerateOntoNotesCoref
  • com.clearnlp.experiment.GeneratePOS
  • com.clearnlp.experiment.PrintTreeLine
  • com.clearnlp.experiment.RemoveEdited
  • com.clearnlp.experiment.SemlinkToOntoNotes
  • com.clearnlp.experiment.Tmp
  • com.clearnlp.propbank.PBLib
    • 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.