Package opennlp.tools.util

Examples of opennlp.tools.util.InvalidFormatException

126
127
128
129
130
131
132
133
134
135
136
    for (int i = 0; i < tokenTags.length; i++) {
      int split = tokenTags[i].lastIndexOf("_");

      if (split == -1) {
        throw new InvalidFormatException("Cannot find \"_\" inside token '" + tokenTags[i] + "'!");
      }

      sentence[i] = tokenTags[i].substring(0, split);
      tags[i] = tokenTags[i].substring(split+1);
    }
View Full Code Here
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
  @Override
  public void validateArtifactMap() throws InvalidFormatException {

    if (this.artifactProvider
        .getManifestProperty(USE_ALPHA_NUMERIC_OPTIMIZATION) == null)
      throw new InvalidFormatException(USE_ALPHA_NUMERIC_OPTIMIZATION
          + " is a mandatory property!");

    Object abbreviationsEntry = this.artifactProvider
        .getArtifact(ABBREVIATIONS_ENTRY_NAME);

    if (abbreviationsEntry != null && !(abbreviationsEntry instanceof Dictionary)) {
      throw new InvalidFormatException("Abbreviations dictionary '" + abbreviationsEntry +
              "' has wrong type, needs to be of type Dictionary!");
    }
  }
View Full Code Here
144
145
146
147
148
149
150
151
    } catch (Exception e) {
      String msg = "Could not instantiate the " + subclassName
          + ". The initialization throw an exception.";
      System.err.println(msg);
      e.printStackTrace();
      throw new InvalidFormatException(msg, e);
    }
  }
View Full Code Here
298
299
300
301
302
303
304
305
306
307
308
        String[] tags = tagString.split(" ");

        StringList word = entry.getTokens();

        if (word.size() != 1)
          throw new InvalidFormatException("Each entry must have exactly one token! "+word);

        newPosDict.dictionary.put(word.getToken(0), tags);
      }});

    newPosDict.caseSensitive = isCaseSensitive;
View Full Code Here
93
94
95
96
97
98
99
100
101
102
103
    if (NAME_ELEMENT_NAMES.contains(name)) {

      String nameType = attributes.get("TYPE");

      if (!EXPECTED_TYPES.contains(nameType)) {
        throw new InvalidFormatException("Unknown timex, numex or namex type: "
            + nameType + ", expected one of " + EXPECTED_TYPES);
      }

      incompleteNames.add(new Span(text.size(), text.size(), nameType.toLowerCase(Locale.ENGLISH)));
    }
View Full Code Here
219
220
221
222
223
224
225
226
227
228
      xmlReader = XMLReaderFactory.createXMLReader();
      xmlReader.setContentHandler(profileContentHandler);
      xmlReader.parse(new InputSource(new UncloseableInputStream(in)));
    }
    catch (SAXException e) {
      throw new InvalidFormatException("The profile data stream has " +
            "an invalid format!", e);
    }
    return profileContentHandler.mIsCaseSensitiveDictionary;
  }
View Full Code Here
60
61
62
63
64
65
66
      if (tagChars.charAt(ci) == '>' || StringUtil.isWhitespace(tagChars.charAt(ci))) {
        return tagChars.subSequence(fromOffset, ci).toString();
      }
    }
   
    throw new InvalidFormatException("Failed to extract tag name!");
  }
View Full Code Here
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
    int c;
    while ((c = in.read()) != -1) {
     
      if ('<' == c) {
        if (isInsideTag) {
          throw new InvalidFormatException("Did not expect < char!");
        }
       
        if (buffer.toString().trim().length() > 0) {
          handler.characters(buffer.toString().trim());
        }
       
        buffer.setLength(0);
       
        isInsideTag = true;
        isStartTag = true;
      }
     
      buffer.appendCodePoint(c);
       
      if ('/' == c && lastChar == '<') {
        isStartTag = false;
      }
         
      if ('>' == c) {
       
        if (!isInsideTag) {
          throw new InvalidFormatException("Did not expect > char!");
        }
       
        if (isStartTag) {
          handler.startElement(extractTagName(buffer), getAttributes(buffer));
        }
        else {
          handler.endElement(extractTagName(buffer));
        }
       
        buffer.setLength(0);
       
        isInsideTag = false;
      }
     
      lastChar = c;
    }
   
    if (isInsideTag) {
      throw new InvalidFormatException("Did not find matching > char!");
    }
  }
View Full Code Here
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
  @Override
  protected void validateArtifactMap() throws InvalidFormatException {
    super.validateArtifactMap();

    if (!(artifactMap.get(MAXENT_MODEL_ENTRY_NAME) instanceof AbstractModel)) {
      throw new InvalidFormatException("Unable to find " + MAXENT_MODEL_ENTRY_NAME +
          " maxent model!");
    }

    if (!ModelUtil.validateOutcomes(getMaxentModel(), SentenceDetectorME.SPLIT,
        SentenceDetectorME.NO_SPLIT)) {
      throw new InvalidFormatException("The maxent model is not compatible " +
          "with the sentence detector!");
    }
  }
View Full Code Here
104
105
106
107
108
109
110
111
112
113
    }
    else if ("ORG".equals(type)) {
      type = "organization";
    }
    else {
      throw new InvalidFormatException("Unknown type: " + type);
    }
   
    return new Span(begin, end, type);
  }
View Full Code Here
TOP

Related Classes of opennlp.tools.util.InvalidFormatException

  • opennlp.tools.chunker.ChunkerFactory
  • opennlp.tools.chunker.ChunkerModel
  • opennlp.tools.dictionary.serializer.DictionarySerializer
  • opennlp.tools.doccat.DoccatFactory
  • opennlp.tools.doccat.DoccatModel
  • opennlp.tools.formats.brat.BratAnnotationStream$BratAnnotationParser
  • opennlp.tools.formats.brat.BratAnnotationStream$RelationAnnotationParser
  • opennlp.tools.formats.brat.BratAnnotationStream$SpanAnnotationParser
  • opennlp.tools.formats.Conll02NameSampleStream
  • opennlp.tools.formats.ConllXPOSSampleStream
    • 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.