Package edu.stanford.nlp.international.morph

Examples of edu.stanford.nlp.international.morph.MorphoFeatures

70
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
120
121
122
123
124
125
126
127
128
129
130
131
132
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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
  /**
   * Hand-written rules to convert SAMA analyses to feature structures.
   */
  @Override
  public MorphoFeatures strToFeatures(String spec) {
    MorphoFeatures features = new ArabicMorphoFeatures();

    // Check for the boundary symbol
    if(spec == null || spec.equals("")) {
      return features;
    }
    //Possessiveness
    if(isActive(MorphoFeatureType.POSS) && spec.contains("POSS")) {
      features.addFeature(MorphoFeatureType.POSS,possVals[0]);
    }

    //Nominals and pronominals. Mona ignores Pronominals in ERTS, but they seem to help...
    // NSUFF -- declinable nominals
    // VSUFF -- enclitic pronominals
    // PRON -- ordinary pronominals
    if(spec.contains("NSUFF") || spec.contains("NOUN") || spec.contains("ADJ")) {
      // Nominal phi feature indicators are different than the indicators
      // that we process with processInflectionalFeatures()
      if(isActive(MorphoFeatureType.NGEN)) {
        if(spec.contains("FEM")) {
          features.addFeature(MorphoFeatureType.NGEN, genVals[1]);
        } else if(spec.contains("MASC") || !pNounNoMorph.matcher(spec).find()) {
          features.addFeature(MorphoFeatureType.NGEN, genVals[0]);
        }
      }

      // WSGDEBUG -- Number for nominals only
      if(isActive(MorphoFeatureType.NNUM)) {
        if(spec.contains("DU")) {
          features.addFeature(MorphoFeatureType.NNUM, numVals[1]);
        } else if(spec.contains("PL")) {
          features.addFeature(MorphoFeatureType.NNUM, numVals[2]);
        } else if (!pNounNoMorph.matcher(spec).find()){ // (spec.contains("SG"))
          features.addFeature(MorphoFeatureType.NNUM, numVals[0]);
        }
      }

      //Definiteness
      if(isActive(MorphoFeatureType.DEF)) {
        if (spec.contains("DET")) {
          features.addFeature(MorphoFeatureType.DEF, defVals[1]);
        } else if (!pNounNoMorph.matcher(spec).find()){
          features.addFeature(MorphoFeatureType.DEF, defVals[0]);
        }
      }

      // Proper nouns (probably a stupid feature)
      if (isActive(MorphoFeatureType.PROP)) {
        if (spec.contains("PROP")) {
          features.addFeature(MorphoFeatureType.PROP,"");
        }
      }

    } else if(spec.contains("PRON") || (spec.contains("VSUFF_DO") && !pVerbMood.matcher(spec).find())) {
      if(spec.contains("DEM_PRON")) {
        features.addFeature(MorphoFeatureType.DEF, defVals[0]);
        Matcher m = pDemPronounFeatures.matcher(spec);
        if (m.find()) {
          spec = m.group(1);
          processInflectionalFeaturesHelper(features, spec);
        }

      } else {
        processInflectionalFeatures(features, spec);
      }

    // Verbs (marked for tense)
    } else if(pVerbTenseMarker.matcher(spec).find()) {

      // Tense feature
      if(isActive(MorphoFeatureType.TENSE)) {
        if(spec.contains("PV"))
          features.addFeature(MorphoFeatureType.TENSE, tenseVals[0]);
        else if(spec.contains("IV"))
          features.addFeature(MorphoFeatureType.TENSE, tenseVals[1]);
        else if(spec.contains("CV"))
          features.addFeature(MorphoFeatureType.TENSE, tenseVals[2]);
      }

      // Inflectional features
      processInflectionalFeatures(features, spec);

      if(isActive(MorphoFeatureType.MOOD)) {
        Matcher moodMatcher = pMood.matcher(spec);
        if(moodMatcher.find()) {
          String moodStr = moodMatcher.group(1);
          switch (moodStr) {
            case "I":
              features.addFeature(MorphoFeatureType.MOOD, moodVals[0]);
              break;
            case "S":
              features.addFeature(MorphoFeatureType.MOOD, moodVals[1]);
              break;
            case "J":
              features.addFeature(MorphoFeatureType.MOOD, moodVals[2]);
              break;
          }
        }
      }

      if(isActive(MorphoFeatureType.VOICE)) {
        if(spec.contains("PASS")) {
          features.addFeature(MorphoFeatureType.VOICE, voiceVals[1]);
        } else {
          features.addFeature(MorphoFeatureType.VOICE, voiceVals[0]);
        }
      }
    }
    return features;
  }
View Full Code Here
287
288
289
290
291
292
293
294
295
296
297
298
    try {
      BufferedReader br = new BufferedReader(new InputStreamReader(new FileInputStream(fName)));

      int nLine = 0;
      for(String line;(line = br.readLine()) != null; nLine++) {
        MorphoFeatures mFeats = fSpec.strToFeatures(line.trim());
        System.out.printf("%s\t%s%n", line.trim(), mFeats.toString());
      }
      br.close();
      System.out.printf("%nRead %d lines%n",nLine);

    } catch (FileNotFoundException e) {
View Full Code Here
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
    private static final long serialVersionUID = -4611776415583633186L;

    @Override
    public MorphoFeatures fromTagString(String str) {
      String[] feats = str.split("\\-");
      MorphoFeatures mFeats = new ArabicMorphoFeatures();
      // First element is the base POS
//      String baseTag = feats[0];
      for(int i = 1; i < feats.length; i++) {
        String[] keyValue = feats[i].split(KEY_VAL_DELIM);
        if(keyValue.length != 2) continue;
        MorphoFeatureType fName = MorphoFeatureType.valueOf(keyValue[0].trim());
        mFeats.addFeature(fName, keyValue[1].trim());
      }
      return mFeats;
    }
View Full Code Here
245
246
247
248
249
250
251
252
253
254
255
256
    if(t.isPreTerminal() && tagSpec != null) {
      if( !(t.firstChild().label() instanceof CoreLabel) || ((CoreLabel) t.firstChild().label()).originalText() == null )
        throw new RuntimeException(String.format("%s: Term lacks morpho analysis: %s",this.getClass().getName(),t.toString()));

      String morphoStr = ((CoreLabel) t.firstChild().label()).originalText();
      MorphoFeatures feats = tagSpec.strToFeatures(morphoStr);
      baseCat = feats.getTag(baseCat);
    }

    //Update the label(s)
    String newCat = baseCat + newCategory.toString();
    t.setValue(newCat);
View Full Code Here
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
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
120
121
122
123
124
125
126
127
128
129
130
131
132
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
161
162
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
      throw new IllegalArgumentException("French does not support feature type: " + feat.toString());
  }

  @Override
  public MorphoFeatures strToFeatures(String spec) {
    MorphoFeatures feats = new MorphoFeatures();

    //Usually this is the boundary symbol
    if(spec == null || spec.equals(""))
      return feats;

    boolean isOtherActive = isActive(MorphoFeatureType.OTHER);
   
    if(spec.startsWith("ADV")) {
      feats.setAltTag("ADV");
      if(spec.contains("int")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "advint");
        }
        feats.setAltTag("ADVWH");
      }

    } else if(spec.startsWith("A")) {
      feats.setAltTag("ADJ");
      if(spec.contains("int")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "adjint");
        }
        feats.setAltTag("ADJWH");
      }
     
      addPhiFeatures(feats,spec);

    } else if(spec.equals("CC") || spec.equals("C-C")) {
      if (isOtherActive) {
        feats.addFeature(MorphoFeatureType.OTHER, "Cc");
      }
      feats.setAltTag("CC");

    } else if(spec.equals("CS") || spec.equals("C-S")) {
      if (isOtherActive) {
        feats.addFeature(MorphoFeatureType.OTHER, "Cs");
      }
      feats.setAltTag("CS");

    } else if(spec.startsWith("CL")) {
      feats.setAltTag("CL");
      if(spec.contains("suj") || spec.equals("CL-S-3fp")) {//"CL-S-3fp" is equivalent to suj
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER,"Sbj");
        }
        feats.setAltTag("CLS");

      } else if(spec.contains("obj")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Obj");
        }
        feats.setAltTag("CLO");

      } else if(spec.contains("refl")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Rfl");
        }
        feats.setAltTag("CLR");
      }

      addPhiFeatures(feats,spec);

    } else if(spec.startsWith("D")) {
      feats.setAltTag("DET");
      if(spec.contains("int")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "dint");
        }
        feats.setAltTag("DETWH");
      }

      addPhiFeatures(feats,spec);

    } else if(spec.startsWith("N")) {
      feats.setAltTag("N");//TODO These are usually N-card...make these CD?
      if(spec.contains("P")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Np");
        }
        feats.setAltTag("NPP");

      } else if(spec.contains("C")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Nc");
        }
        feats.setAltTag("NC");
      }

      addPhiFeatures(feats,spec);

    } else if(spec.startsWith("PRO")) {
      feats.setAltTag("PRO");
      if(spec.contains("int")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER,"Ni");
        }
        feats.setAltTag("PROWH");

      } else if(spec.contains("rel")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Nr");
        }
        feats.setAltTag("PROREL");
      }

      addPhiFeatures(feats,spec);

    } else if(spec.startsWith("V")) {
      feats.setAltTag("V");
      if(spec.contains("Y")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER,"Vp");
        }
        feats.setAltTag("VIMP");

      } else if(spec.contains("W")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Vf");
        }
        feats.setAltTag("VINF");
       
      } else if(spec.contains("S") || spec.contains("T")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Vs");
        }
        feats.setAltTag("VS");
       
      } else if(spec.contains("K")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Vp");
        }
        feats.setAltTag("VPP");
       
      } else if(spec.contains("G")) {
        if (isOtherActive) {
          feats.addFeature(MorphoFeatureType.OTHER, "Vr");
        }
        feats.setAltTag("VPR");
      }
     
      addPhiFeatures(feats,spec);
   
    } else if(spec.equals("P") || spec.equals("I")) {
      feats.setAltTag(spec);
     
    }
//    else {
//      System.err.println("Could not map spec: " + spec);
//    }
View Full Code Here
93
94
95
96
97
98
99
100
101
102
103
        morphStr += "-" + subCat + "--";
      } else {
        morphStr += "---";
      }
    }
    MorphoFeatures feats = morpho.strToFeatures(morphStr);
    if(feats.getAltTag() != null && !feats.getAltTag().equals("")) {
      label.setValue(feats.getAltTag());
      label.setTag(feats.getAltTag());
    }
  }
View Full Code Here
253
254
255
256
257
258
259
260
261
262
263
264
      mfs.activate(MorphoFeatureType.GEN);
      mfs.activate(MorphoFeatureType.NUM);
      mfs.activate(MorphoFeatureType.PER);

      for(String line; (line = br.readLine()) != null;) {
        MorphoFeatures feats = mfs.strToFeatures(line);
        System.out.printf("%s\t%s%n", line.trim(),feats.toString());
      }

      br.close();

    } catch (FileNotFoundException e) {
View Full Code Here
217
218
219
220
221
222
223
224
225
226
227
228
229
          morphStr += "-" + subCat + "--";
        } else {
          morphStr += "---";
        }
      }
      MorphoFeatures feats = spec.strToFeatures(morphStr);
      if(feats.getAltTag() != null && !feats.getAltTag().equals("")) {
        CoreLabel cl = (CoreLabel) preYield.get(i);
        cl.setValue(feats.getAltTag());
        cl.setTag(feats.getAltTag());
      }
    }
  }
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
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
      MorphoFeatureSpecification featureSpec = new ArabicMorphoFeatureSpecification();
      featureSpec.activate(MorphoFeatureType.NGEN);
      featureSpec.activate(MorphoFeatureType.NNUM);
      featureSpec.activate(MorphoFeatureType.DEF);
      featureSpec.activate(MorphoFeatureType.TENSE);
      MorphoFeatures features = featureSpec.strToFeatures(tag);

      // Rule #1 : ت --> ة
      if (features.getValue(MorphoFeatureType.NGEN).equals("F") &&
          features.getValue(MorphoFeatureType.NNUM).equals("SG") &&
          rawToken.endsWith("ت-") &&
          !stripRewrites) {
        lastLabel = RewriteSymbol;
      } else if (rawToken.endsWith("ة-")) {
        assert token.endsWith("ة");
        token = token.substring(0, token.length() - 1) + "ت";
        lastLabel = RewriteSymbol;
      }

      // Rule #2 : لل --> ل ال
      if (lastToken.equals("ل") &&
          features.getValue(MorphoFeatureType.DEF).equals("D")) {
        if (rawToken.startsWith("-ال")) {
          if (!token.startsWith("ا"))
            System.err.println("Bad REWAL: " + rawToken + " / " + token);
          token = token.substring(1);
          rewritten = rewritten.substring(1);
          if (!stripRewrites)
            firstLabel = RewriteSymbol;
        } else if (rawToken.startsWith("-ل")) {
          if (!token.startsWith("ل"))
            System.err.println("Bad REWAL: " + rawToken + " / " + token);
          if (!stripRewrites)
            firstLabel = RewriteSymbol;
        } else {
          System.err.println("Ignoring REWAL: " + rawToken + " / " + token);
        }
      }
     
      // Rule #3 : ي --> ى
      // Rule #4 : ا --> ى
      if (rawToken.endsWith("ى-")) {
        if (features.getValue(MorphoFeatureType.TENSE) != null) {
          // verb: ى becomes ا
          token = token.substring(0, token.length() - 1) + "ا";
        } else {
          // assume preposition:
          token = token.substring(0, token.length() - 1) + "ي";
View Full Code Here
50
51
52
53
54
55
56
57
58
59
60
    if( ! universalMap.containsKey(shortTag)) {
      System.err.printf("%s: No universal tag for LDC tag %s%n", this.getClass().getName(),shortTag);
      universalTag = shortTag;
    }
  
    MorphoFeatures feats = new MorphoFeatures(morphoSpec.strToFeatures(rawTag));
   
    String functionalTag = feats.getTag(universalTag);
   
    return functionalTag;
  }
View Full Code Here
TOP

Related Classes of edu.stanford.nlp.international.morph.MorphoFeatures

  • edu.stanford.nlp.international.arabic.ArabicMorphoFeatureSpecification
  • edu.stanford.nlp.international.arabic.ArabicMorphoFeatureSpecification$ArabicMorphoFeatures
  • edu.stanford.nlp.international.arabic.pipeline.UniversalPOSMapper
  • edu.stanford.nlp.international.arabic.process.IOBUtils
  • edu.stanford.nlp.international.french.FrenchMorphoFeatureSpecification
  • edu.stanford.nlp.international.french.scripts.SplitCanditoTrees
  • edu.stanford.nlp.international.morph.MorphoFeatureSpecification.MorphoFeatureType
  • edu.stanford.nlp.parser.lexparser.ArabicTreebankParserParams
  • edu.stanford.nlp.parser.lexparser.FrenchTreebankParserParams
  • edu.stanford.nlp.trees.international.french.FrenchTreeNormalizer
    • 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.