Package org.apache.nutch.plugin

Examples of org.apache.nutch.plugin.Extension

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
        "org.apache.nutch.parse.Parser").getExtensions();

    String contentType, command, timeoutString;

    for (int i = 0; i < extensions.length; i++) {
      Extension extension = extensions[i];

      // only look for extensions defined by plugin parse-ext
      if (!extension.getDescriptor().getPluginId().equals("parse-ext"))
        continue;

      contentType = extension.getAttribute("contentType");
      if (contentType == null || contentType.equals(""))
        continue;

      command = extension.getAttribute("command");
      if (command == null || command.equals(""))
        continue;

      timeoutString = extension.getAttribute("timeout");
      if (timeoutString == null || timeoutString.equals(""))
        timeoutString = "" + TIMEOUT_DEFAULT;

      TYPE_PARAMS_MAP.put(contentType, new String[] { command, timeoutString });
    }
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
      parserFactory = new ParserFactory(conf);
  }
   
  /** Unit test for <code>getExtensions(String)</code> method. */
  public void testGetExtensions() throws Exception {
    Extension ext = (Extension)parserFactory.getExtensions("text/html").get(0);
    assertEquals("parse-html", ext.getDescriptor().getPluginId());
    ext = (Extension) parserFactory.getExtensions("text/html; charset=ISO-8859-1").get(0);
    assertEquals("parse-html", ext.getDescriptor().getPluginId());
    ext = (Extension)parserFactory.getExtensions("foo/bar").get(0);
    assertEquals("parse-text", ext.getDescriptor().getPluginId());
  }
View Full Code Here
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
      throw new ParserNotFound(url, contentType);
    }

    parsers = new Vector(parserExts.size());
    for (Iterator i=parserExts.iterator(); i.hasNext(); ){
      Extension ext = (Extension) i.next();
      Parser p = null;
      try {
        //check to see if we've cached this parser instance yet
        p = (Parser) this.conf.getObject(ext.getId());
        if (p == null) {
          // go ahead and instantiate it and then cache it
          p = (Parser) ext.getExtensionInstance();
          this.conf.setObject(ext.getId(),p);
        }
        parsers.add(p);
      } catch (PluginRuntimeException e) {
        if (LOG.isWarnEnabled()) {
          e.printStackTrace(LogUtil.getWarnStream(LOG));
          LOG.warn("ParserFactory:PluginRuntimeException when "
                 + "initializing parser plugin "
                 + ext.getDescriptor().getPluginId()
                 + " instance in getParsers "
                 + "function: attempting to continue instantiating parsers");
        }
      }
    }
View Full Code Here
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
   *         {@link PluginRuntimeException} instantiating the {@link Parser}.
   */
  public Parser getParserById(String id) throws ParserNotFound {

    Extension[] extensions = this.extensionPoint.getExtensions();
    Extension parserExt = null;

    if (id != null) {
      parserExt = getExtension(extensions, id);
    }
    if (parserExt == null) {
      parserExt = getExtensionFromAlias(extensions, id);
    }

    if (parserExt == null) {
      throw new ParserNotFound("No Parser Found for id [" + id + "]");
    }
   
    // first check the cache          
    if (this.conf.getObject(parserExt.getId()) != null) {
      return (Parser) this.conf.getObject(parserExt.getId());

    // if not found in cache, instantiate the Parser   
    } else {
      try {
        Parser p = (Parser) parserExt.getExtensionInstance();
        this.conf.setObject(parserExt.getId(), p);
        return p;
      } catch (PluginRuntimeException e) {
        if (LOG.isWarnEnabled()) {
          LOG.warn("Canno initialize parser " +
                   parserExt.getDescriptor().getPluginId() +
                   " (cause: " + e.toString());
        }
        throw new ParserNotFound("Cannot init parser for id [" + id + "]");
      }
    }
View Full Code Here
290
291
292
293
294
295
296
297
298
299
300
    if (plugins != null) {
     
      for (Iterator i = plugins.iterator(); i.hasNext();) {
        String parsePluginId = (String) i.next();
       
        Extension ext = getExtension(extensions, parsePluginId, contentType);
        // the extension returned may be null
        // that means that it was not enabled in the plugin.includes
        // nutch conf property, but it was mapped in the
        // parse-plugins.xml
        // file.
View Full Code Here
50
51
52
53
54
55
56
57
58
59
60
61
      throw new RuntimeException(URLFilter.X_POINT_ID+" not found.");

    Extension[] extensions = point.getExtensions();

    for (int i = 0; i < extensions.length; i++) {
      Extension extension = extensions[i];
      filter = (URLFilter)extension.getExtensionInstance();
      if (filter.getClass().getName().equals(filterName)) {
        break;
      } else {
        filter = null;
      }
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
                    throw new RuntimeException(URLFilter.X_POINT_ID
                            + " not found.");
                Extension[] extensions = point.getExtensions();
                HashMap filterMap = new HashMap();
                for (int i = 0; i < extensions.length; i++) {
                    Extension extension = extensions[i];
                    URLFilter filter = (URLFilter) extension
                            .getExtensionInstance();
                    if (!filterMap.containsKey(filter.getClass().getName())) {
                        filterMap.put(filter.getClass().getName(), filter);
                    }
                }
View Full Code Here
58
59
60
61
62
63
64
65
66
67
68
69
        ExtensionPoint point = PluginRepository.get(conf).getExtensionPoint(ScoringFilter.X_POINT_ID);
        if (point == null) throw new RuntimeException(ScoringFilter.X_POINT_ID + " not found.");
        Extension[] extensions = point.getExtensions();
        HashMap filterMap = new HashMap();
        for (int i = 0; i < extensions.length; i++) {
          Extension extension = extensions[i];
          ScoringFilter filter = (ScoringFilter) extension.getExtensionInstance();
          if (!filterMap.containsKey(filter.getClass().getName())) {
            filterMap.put(filter.getClass().getName(), filter);
          }
        }
        if (orderedFilters == null) {
View Full Code Here
72
73
74
75
76
77
78
79
80
81
82
83
84
85
   * used. If none match, then the {@link NutchDocumentAnalyzer} is used.
   */
  public NutchAnalyzer get(String lang) {

    NutchAnalyzer analyzer = DEFAULT_ANALYZER;
    Extension extension = getExtension(lang);
    if (extension != null) {
        try {
            analyzer = (NutchAnalyzer) extension.getExtensionInstance();
        } catch (PluginRuntimeException pre) {
            analyzer = DEFAULT_ANALYZER;
        }
    }
    return analyzer;
View Full Code Here
86
87
88
89
90
91
92
93
94
95
96
  }

  private Extension getExtension(String lang) {

    if (lang == null) { return null; }
    Extension extension = (Extension) this.conf.getObject(lang);
    if (extension == null) {
      extension = findExtension(lang);
      if (extension != null) {
        this.conf.setObject(lang, extension);
      }
View Full Code Here
TOP

Related Classes of org.apache.nutch.plugin.Extension

  • org.apache.nutch.analysis.AnalyzerFactory
  • org.apache.nutch.indexer.field.FieldFilters
  • org.apache.nutch.indexer.IndexingFilters
  • org.apache.nutch.indexer.IndexWriters
  • org.apache.nutch.net.URLFilterChecker
  • org.apache.nutch.net.URLFilters
  • org.apache.nutch.net.urlnormalizer.host.HostURLNormalizer
  • org.apache.nutch.net.URLNormalizerChecker
  • org.apache.nutch.net.URLNormalizers
  • org.apache.nutch.parse.ext.ExtParser
    • 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.