Package org.maltparserx.core.io.dataformat

Examples of org.maltparserx.core.io.dataformat.ColumnDescription

71
72
73
74
75
76
77
78
79
80
81
82
83
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
    }
    Iterator<ColumnDescription> columns = dataFormatInstance.iterator();
   
    for (int i : syntaxGraph.getTokenIndices()) {
      try {
        ColumnDescription column = null;
        while (columns.hasNext()) {
          column = columns.next();

          if (column.getCategory() == ColumnDescription.INPUT) { // && column.getType() != ColumnDescription.IGNORE) {
            TokenNode node = syntaxGraph.getTokenNode(i);
            if (!column.getName().equals("ID")) {
              if (node.hasLabel(column.getSymbolTable())) {
                output.append(node.getLabelSymbol(column.getSymbolTable()));
                if (output.length() != 0) {
                  writer.write(output.toString());
                } else {
                  writer.write('_');
                }
              } else {
                writer.write('_');
              }
            } else {
              writer.write(Integer.toString(i));
            }
          } else if (column.getCategory() == ColumnDescription.HEAD /* && column.getType() != ColumnDescription.IGNORE */&& syntaxGraph instanceof DependencyStructure) {
            if (((DependencyStructure)syntaxGraph).getDependencyNode(i).hasHead()) {
              writer.write(Integer.toString(((DependencyStructure)syntaxGraph).getDependencyNode(i).getHead().getIndex()));
            } else {
              writer.write(Integer.toString(0));
            }
           
          } else if (column.getCategory() == ColumnDescription.DEPENDENCY_EDGE_LABEL /* && column.getType() != ColumnDescription.IGNORE */ && syntaxGraph instanceof DependencyStructure) {
            if (((DependencyStructure)syntaxGraph).getDependencyNode(i).hasHead() && ((DependencyStructure)syntaxGraph).getDependencyNode(i).hasHeadEdgeLabel(column.getSymbolTable())) {
              output.append(((DependencyStructure)syntaxGraph).getDependencyNode(i).getHeadEdgeLabelSymbol(column.getSymbolTable()));
            } else {
              output.append(((DependencyStructure)syntaxGraph).getDefaultRootEdgeLabelSymbol(column.getSymbolTable()));
            }
           
            if (output.length() != 0) {
              writer.write(output.toString());
            }
          } else {
            writer.write(column.getDefaultOutput());
          }
          if (columns.hasNext()) {
            writer.write(TAB);
          }
          output.setLength(0);
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
    if (!(arguments[1] instanceof String)) {
      throw new FeatureException("Could not initialize SplitFeature: the second argument is not a string. ");
    }
    setParentFeature((FeatureFunction)arguments[0]);
    setSeparators((String)arguments[1]);
    ColumnDescription parentColumn = dataFormatInstance.getColumnDescriptionByName(parentFeature.getSymbolTable().getName());
    if (parentColumn.getType() != ColumnDescription.STRING) {
      throw new FeatureException("Could not initialize SplitFeature: the first argument must be a string. ");
    }
    setColumn(dataFormatInstance.addInternalColumnDescription("SPLIT_"+parentFeature.getSymbolTable().getName(), parentColumn));
    setSymbolTable(column.getSymbolTable());
//    setSymbolTable(tableHandler.addSymbolTable("SPLIT_"+parentFeature.getSymbolTable().getName(), parentFeature.getSymbolTable()));
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
    if (!(arguments[1] instanceof Integer)) {
      throw new FeatureException("Could not initialize SuffixFeature: the second argument is not a string. ");
    }
    setParentFeature((FeatureFunction)arguments[0]);
    setSuffixLength(((Integer)arguments[1]).intValue());
    ColumnDescription parentColumn = dataFormatInstance.getColumnDescriptionByName(parentFeature.getSymbolTable().getName());
    if (parentColumn.getType() != ColumnDescription.STRING) {
      throw new FeatureException("Could not initialize SuffixFeature: the first argument must be a string. ");
    }
    setColumn(dataFormatInstance.addInternalColumnDescription("SUFFIX_"+suffixLength+"_"+parentFeature.getSymbolTable().getName(), parentColumn));
    setSymbolTable(column.getSymbolTable());
//    setSymbolTable(tableHandler.addSymbolTable("SUFFIX_"+suffixLength+"_"+parentFeature.getSymbolTable().getName(), parentFeature.getSymbolTable()));
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
    if (!(arguments[1] instanceof FeatureFunction)) {
      throw new FeatureException("Could not initialize MergeFeature: the second argument is not a feature. ");
    }
    setFirstFeature((FeatureFunction)arguments[0]);
    setSecondFeature((FeatureFunction)arguments[1]);
    ColumnDescription firstColumn = (firstFeature.getSymbolTable() != null)?dataFormatInstance.getColumnDescriptionByName(firstFeature.getSymbolTable().getName()):null;
    ColumnDescription secondColumn = (secondFeature.getSymbolTable() != null)?dataFormatInstance.getColumnDescriptionByName(secondFeature.getSymbolTable().getName()):null;
//    if (firstColumn.getType() != secondColumn.getType()) {
//      throw new FeatureException("Could not initialize MergeFeature: the first and the second arguments are not of the same type.");
//    }
//    setColumn(dataFormatInstance.addInternalColumnDescription("MERGE2_"+firstFeature.getSymbolTable().getName()+"_"+secondFeature.getSymbolTable().getName(), firstColumn));
View Full Code Here
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
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
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
          }
          sentenceCount++;
        } else if (currentHeaderTable == NegraTables.SENTENCE) {
          if (line.length() >= 2 && line.charAt(0) == '#' && Character.isDigit(line.charAt(1))) { // Non-terminal
            Iterator<ColumnDescription> columns = dataFormatInstance.iterator();
            ColumnDescription column = null;
            currentNonTerminalSize++;
            char[] lineChars = line.toCharArray();
            int start = 0;
            int secedgecounter = 0;
            for (int i = 0, n = lineChars.length; i < n; i++) {
              if (lineChars[i] == '\t' && start == i) {
                start++;
              } else if (lineChars[i] == '\t' || i == n - 1) {
                if (columns.hasNext()) {
                  column = columns.next();
                }
                if (column.getPosition() == 0) {
                  int index = Integer.parseInt((i == n - 1)?line.substring(start+1):line.substring(start+1, i));
                  child = nonterminals.get(index);
                  if (child == null) {
                    if (index != 0) {
                      child = ((PhraseStructure)syntaxGraph).addNonTerminalNode(index-START_ID_OF_NONTERMINALS+1);
                    }
                    nonterminals.put(index,child);
                  }
                } else if (column.getPosition() == 2 && child != null) {
                  syntaxGraph.addLabel(child, "CAT", (i == n - 1)?line.substring(start):line.substring(start, i));
                } else if (column.getCategory() == ColumnDescription.PHRASE_STRUCTURE_EDGE_LABEL) {
                  edgelabelSymbol.setLength(0);
                  edgelabelSymbol.append((i == n - 1)?line.substring(start):line.substring(start, i));
                  edgelabelTableName.setLength(0);
                  edgelabelTableName.append(column.getName());
                } else if (column.getCategory() == ColumnDescription.PHRASE_STRUCTURE_NODE_LABEL && child != null) {
                  int index = Integer.parseInt((i == n - 1)?line.substring(start):line.substring(start, i));
                  parent = nonterminals.get(index);
                  if (parent == null) {
                    if (index == 0) {
                      parent = phraseStructure.getPhraseStructureRoot(); 
                    } else {
                      parent = phraseStructure.addNonTerminalNode(index-START_ID_OF_NONTERMINALS+1);
                    }
                    nonterminals.put(index,parent);
                  }
                  Edge e = phraseStructure.addPhraseStructureEdge(parent, child);
                  syntaxGraph.addLabel(e, edgelabelTableName.toString(), edgelabelSymbol.toString());
                } else if (column.getCategory() == ColumnDescription.SECONDARY_EDGE_LABEL && child != null) {
                  if (secedgecounter % 2 == 0) {
                    edgelabelSymbol.setLength(0);
                    edgelabelSymbol.append((i == n - 1)?line.substring(start):line.substring(start, i));
                    secedgecounter++;
                  } else {
                    int index = Integer.parseInt((i == n - 1)?line.substring(start):line.substring(start, i));
                    if (index == 0) {
                      parent = phraseStructure.getPhraseStructureRoot();
                    } else if (index < START_ID_OF_NONTERMINALS) {
                      parent = phraseStructure.getTokenNode(index);
                    } else {
                      parent = nonterminals.get(index);
                      if (parent == null) {
                        parent = phraseStructure.addNonTerminalNode(index-START_ID_OF_NONTERMINALS+1);
                        nonterminals.put(index,parent);
                      }
                    }
                    Edge e = phraseStructure.addSecondaryEdge(parent, child);
                    e.addLabel(column.getSymbolTable(), edgelabelSymbol.toString());
                    secedgecounter++;
                  }
                }
                start = i + 1;
              }
            }
          } else { // Terminal
            Iterator<ColumnDescription> columns = dataFormatInstance.iterator();
            ColumnDescription column = null;
           
            currentTerminalSize++;
            child = syntaxGraph.addTokenNode(currentTerminalSize);
            char[] lineChars = line.toCharArray();
            int start = 0;
            int secedgecounter = 0;
            for (int i = 0, n = lineChars.length; i < n; i++) {
              if (lineChars[i] == '\t' && start == i) {
                start++;
              } else if (lineChars[i] == '\t' || i == n - 1) {
                if (columns.hasNext()) {
                  column = columns.next();
                }
                if (column.getCategory() == ColumnDescription.INPUT && child != null) {
                  syntaxGraph.addLabel(child, column.getName(), (i == n - 1)?line.substring(start):line.substring(start, i));
                } else if (column.getCategory() == ColumnDescription.PHRASE_STRUCTURE_EDGE_LABEL && child != null) { // && column.getName().equals("EDGELABEL")) {
                  edgelabelSymbol.setLength(0);
                  edgelabelSymbol.append((i == n - 1)?line.substring(start):line.substring(start, i));
                  edgelabelTableName.setLength(0);
                  edgelabelTableName.append(column.getName());
                } else if (column.getCategory() == ColumnDescription.PHRASE_STRUCTURE_NODE_LABEL && child != null) {
                  int index = Integer.parseInt((i == n - 1)?line.substring(start):line.substring(start, i));
                  parent = nonterminals.get(index);
                  if (parent == null) {
                    if (index == 0) {
                      parent = phraseStructure.getPhraseStructureRoot(); 
                    } else {
                      parent = phraseStructure.addNonTerminalNode(index-START_ID_OF_NONTERMINALS+1);
                    }
                    nonterminals.put(index,parent);
                  }

                  Edge e = phraseStructure.addPhraseStructureEdge(parent, child);
                  syntaxGraph.addLabel(e, edgelabelTableName.toString(), edgelabelSymbol.toString());
                } else if (column.getCategory() == ColumnDescription.SECONDARY_EDGE_LABEL && child != null) {
                  if (secedgecounter % 2 == 0) {
                    edgelabelSymbol.setLength(0);
                    edgelabelSymbol.append((i == n - 1)?line.substring(start):line.substring(start, i));
                    secedgecounter++;
                  } else {
                    int index = Integer.parseInt((i == n - 1)?line.substring(start):line.substring(start, i));
                    if (index == 0) {
                      parent = phraseStructure.getPhraseStructureRoot();
                    } else if (index < START_ID_OF_NONTERMINALS) {
                      parent = phraseStructure.getTokenNode(index);
                    } else {
                      parent = nonterminals.get(index);
                      if (parent == null) {
                        parent = phraseStructure.addNonTerminalNode(index-START_ID_OF_NONTERMINALS+1);
                        nonterminals.put(index,parent);
                      }
                    }
                    Edge e = phraseStructure.addSecondaryEdge(parent, child);
                    e.addLabel(column.getSymbolTable(), edgelabelSymbol.toString());
                    secedgecounter++;
                  }
                }
                start = i + 1;
              }
View Full Code Here
31
32
33
34
35
36
37
38
39
40
41
      throw new SyntaxGraphException("Could not initialize InputColumnFeature: the first argument is not a string. ");
    }
    if (!(arguments[1] instanceof AddressFunction)) {
      throw new SyntaxGraphException("Could not initialize InputColumnFeature: the second argument is not an address function. ");
    }
    ColumnDescription column = dataFormatInstance.getColumnDescriptionByName((String)arguments[0]);
    if (column == null) {
      throw new SyntaxGraphException("Could not initialize InputColumnFeature: the input column type '"+(String)arguments[0]+"' could not be found in the data format specification. ' ");
    }
    setColumn(column);
    setAddressFunction((AddressFunction)arguments[1]);
View Full Code Here
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
          if (i == 0) {
            terminalCounter++;
            node = syntaxGraph.addTokenNode(terminalCounter);
          }
          if (columns.hasNext()) {
            ColumnDescription column = columns.next();
            if (column.getCategory() == ColumnDescription.INPUT && node != null) {
              syntaxGraph.addLabel(node, column.getName(), input.toString());
            } else if (column.getCategory() == ColumnDescription.HEAD) {
              if (syntaxGraph instanceof DependencyStructure) {
                if (column.getCategory() != ColumnDescription.IGNORE && !input.toString().equals(IGNORE_COLUMN_SIGN)) {
//                if (column.getType() != ColumnDescription.IGNORE && !input.toString().equals(IGNORE_COLUMN_SIGN)) { // bugfix
                //if (!input.toString().equals(IGNORE_COLUMN_SIGN)) {
                  edge = ((DependencyStructure)syntaxGraph).addDependencyEdge(Integer.parseInt(input.toString()), terminalCounter);
                }
              }
              else {
                close();
                throw new DataFormatException("The input graph is not a dependency graph and therefore it is not possible to add dependncy edges. ");
              }
            } else if (column.getCategory() == ColumnDescription.DEPENDENCY_EDGE_LABEL && edge != null) {
              //if (column.getType() != ColumnDescription.IGNORE && !input.toString().equals(IGNORE_COLUMN_SIGN)) { // bugfix not working for everybody
                syntaxGraph.addLabel(edge, column.getName(), input.toString());
              //} // bugfix
            }
          }
          input.setLength(0);
          nNewLines = 0;
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
55
56
      throw new FeatureException("Could not initialize Merge3Feature: the third argument is not a feature. ");
    }
    setFirstFeature((FeatureFunction)arguments[0]);
    setSecondFeature((FeatureFunction)arguments[1]);
    setThirdFeature((FeatureFunction)arguments[2]);
    ColumnDescription firstColumn = (firstFeature.getSymbolTable() != null)?dataFormatInstance.getColumnDescriptionByName(firstFeature.getSymbolTable().getName()):null;
    ColumnDescription secondColumn = (secondFeature.getSymbolTable() != null)?dataFormatInstance.getColumnDescriptionByName(secondFeature.getSymbolTable().getName()):null;
    ColumnDescription thirdColumn =  (thirdFeature.getSymbolTable() != null)?dataFormatInstance.getColumnDescriptionByName(thirdFeature.getSymbolTable().getName()):null;
    if (firstFeature.getType() != secondFeature.getType() || firstFeature.getType() != thirdFeature.getType()) {
      throw new FeatureException("Could not initialize MergeFeature: the arguments are not of the same type.");
    }
    if (firstColumn != null || secondColumn != null || thirdColumn != null) {
      setColumn(dataFormatInstance.addInternalColumnDescription("MERGE3_"+firstFeature.getMapIdentifier()+"_"+secondFeature.getMapIdentifier()+"_"+thirdFeature.getMapIdentifier(),
View Full Code Here
41
42
43
44
45
46
47
48
49
50
51
52
    if (!(arguments[1] instanceof Integer)) {
      throw new FeatureException("Could not initialize PrefixFeature: the second argument is not a string. ");
    }
    setParentFeature((FeatureFunction)arguments[0]);
    setPrefixLength(((Integer)arguments[1]).intValue());
    ColumnDescription parentColumn = dataFormatInstance.getColumnDescriptionByName(parentFeature.getSymbolTable().getName());
    if (parentColumn.getType() != ColumnDescription.STRING) {
      throw new FeatureException("Could not initialize PrefixFeature: the first argument must be a string. ");
    }
    setColumn(dataFormatInstance.addInternalColumnDescription("PREFIX_"+prefixLength+"_"+parentFeature.getSymbolTable().getName(), parentColumn));
    setSymbolTable(column.getSymbolTable());
//    setSymbolTable(tableHandler.addSymbolTable("PREFIX_"+prefixLength+"_"+parentFeature.getSymbolTable().getName(), parentFeature.getSymbolTable()));
View Full Code Here
32
33
34
35
36
37
38
39
40
41
42
      throw new SyntaxGraphException("Could not initialize OutputColumnFeature: the first argument is not a string. ");
    }
    if (!(arguments[1] instanceof AddressFunction)) {
      throw new SyntaxGraphException("Could not initialize OutputColumnFeature: the second argument is not an address function. ");
    }
    ColumnDescription column = dataFormatInstance.getColumnDescriptionByName((String)arguments[0]);
    if (column == null) {
      throw new SyntaxGraphException("Could not initialize OutputColumnFeature: the output column type '"+(String)arguments[0]+"' could not be found in the data format specification. ' ");
    }
    setColumn(column);
    setAddressFunction((AddressFunction)arguments[1]);
View Full Code Here
TOP

Related Classes of org.maltparserx.core.io.dataformat.ColumnDescription

  • org.maltparserx.core.feature.map.Merge3Feature
  • org.maltparserx.core.feature.map.MergeFeature
  • org.maltparserx.core.feature.map.PrefixFeature
  • org.maltparserx.core.feature.map.SplitFeature
  • org.maltparserx.core.feature.map.SuffixFeature
  • org.maltparserx.core.propagation.Propagation
  • org.maltparserx.core.syntaxgraph.feature.InputColumnFeature
  • org.maltparserx.core.syntaxgraph.feature.OutputColumnFeature
  • org.maltparserx.core.syntaxgraph.headrules.PrioSet
  • org.maltparserx.core.syntaxgraph.reader.NegraReader
    • 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.