Examples of LanguageProfile

Examples of org.apache.tika.language.LanguageProfile

39
40
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
  private static void addProfile(Language language) {
    final String PROFILE_SUFFIX = ".ngp";
    final String PROFILE_ENCODING = "UTF-8";

    try {
      final LanguageProfile profile = new LanguageProfile();

      final String languageCode = language.getShortName();
      final String detectionFile = "/" + languageCode + "/" + languageCode + PROFILE_SUFFIX;
      if (!JLanguageTool.getDataBroker().resourceExists(detectionFile)) {
        // that's okay, not every language comes with its own detection file,
        // as Tika supports most languages out of the box.
        return;
      }
      final InputStream stream = JLanguageTool.getDataBroker().getFromResourceDirAsStream(detectionFile);
      try {
        final InputStreamReader in = new InputStreamReader(stream, PROFILE_ENCODING);
        final BufferedReader reader =
                new BufferedReader(in);
        String line = reader.readLine();
        while (line != null) {
          if (line.length() > 0 && !line.startsWith("#")) {
            final int space = line.indexOf(' ');
            profile.add(
                    line.substring(0, space),
                    Long.parseLong(line.substring(space + 1)));
          }
          line = reader.readLine();
        }
View Full Code Here

Examples of org.apache.tika.language.LanguageProfile

197
198
199
200
201
202
203
204
205
206
207
  }

  private void extractLanguage(JCas plainTextView) {
    try {
      LanguageIdentifier li = new LanguageIdentifier(new LanguageProfile(plainTextView.getDocumentText()));
      if (li.getLanguage() != null && !"".equals(li.getLanguage()))
        plainTextView.setDocumentLanguage(li.getLanguage());
    }
    catch (Exception e) {
      this.getContext().getLogger().log(Level.WARNING, new StringBuffer("Could not extract language due to ")
View Full Code Here

Examples of org.apache.tika.language.LanguageProfile

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
    }
  }

  private static void addProfile(Language language) {
    try {
      final LanguageProfile profile = new LanguageProfile();

      final String languageCode = language.getShortName();
      final String detectionFile = "/" + languageCode + "/" + languageCode + PROFILE_SUFFIX;
      if (!JLanguageTool.getDataBroker().resourceExists(detectionFile)) {
        // that's okay, not every language comes with its own detection file,
        // as Tika supports most languages out of the box.
        return;
      }
      try (InputStream stream = JLanguageTool.getDataBroker().getFromResourceDirAsStream(detectionFile)) {
        final InputStreamReader in = new InputStreamReader(stream, PROFILE_ENCODING);
        final BufferedReader reader =
                new BufferedReader(in);
        String line = reader.readLine();
        while (line != null) {
          if (line.length() > 0 && !line.startsWith("#")) {
            final int space = line.indexOf(' ');
            profile.add(
                    line.substring(0, space),
                    Long.parseLong(line.substring(space + 1)));
          }
          line = reader.readLine();
        }
View Full Code Here

Examples of org.apache.tika.language.LanguageProfile

91
92
93
94
95
96
97
98
99
     * @throws Exception
     */
    @Override
    public String translate(String text, String targetLanguage) throws Exception {
        LanguageIdentifier language = new LanguageIdentifier(
                new LanguageProfile(text));
        String sourceLanguage = language.getLanguage();
        return translate(text, sourceLanguage, targetLanguage);
    }
View Full Code Here

Examples of org.apache.tika.language.LanguageProfile

58
59
60
61
62
63
64
65
66
    }

    @Override
    public String translate(String text, String targetLanguage) throws Exception {
        LanguageIdentifier language = new LanguageIdentifier(
                new LanguageProfile(text));
        String sourceLanguage = language.getLanguage();
        return translate(text, sourceLanguage, targetLanguage);
    }
View Full Code Here

Examples of org.apache.tika.language.LanguageProfile

118
119
120
121
122
123
124
125
126
     * @param targetLanguage The target language of translation.
     * @return true if the cache contains a translation of the text, false otherwise.
     */
    public boolean contains(String text, String targetLanguage) {
        LanguageIdentifier language = new LanguageIdentifier(
                new LanguageProfile(text));
        String sourceLanguage = language.getLanguage();
        return contains(text, sourceLanguage, targetLanguage);
    }
View Full Code Here

Examples of org.apache.tika.language.LanguageProfile

101
102
103
104
105
106
107
108
109
  public String translate(String text, String targetLanguage)
      throws Exception {
    if (!this.isAvailable)
      return text;
    LanguageIdentifier language = new LanguageIdentifier(
        new LanguageProfile(text));
    String sourceLanguage = language.getLanguage();
    return translate(text, sourceLanguage, targetLanguage);
  }
View Full Code Here

Examples of org.apache.tika.language.LanguageProfile

96
97
98
99
100
101
102
103
104
    public String translate(String text, String targetLanguage)
            throws Exception {
        if (!this.isAvailable)
            return text;
        LanguageIdentifier language = new LanguageIdentifier(
                new LanguageProfile(text));
        String sourceLanguage = language.getLanguage();
        return translate(text, sourceLanguage, targetLanguage);
    }
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. 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.