Package org.apache.nutch.plugin

Examples of org.apache.nutch.plugin.Extension

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
      throw new ParserNotFound(url, contentType);
    }

    parsers = new Vector<Parser>(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) objectCache.getObject(ext.getId());
        if (p == null) {
          // go ahead and instantiate it and then cache it
          p = (Parser) ext.getExtensionInstance();
          objectCache.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
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
   *         {@link PluginRuntimeException} instantiating the {@link Parser}.
   */
  public Parser getParserById(String id) throws ParserNotFound {

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

    ObjectCache objectCache = ObjectCache.get(conf);
   
    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 (objectCache.getObject(parserExt.getId()) != null) {
      return (Parser) objectCache.getObject(parserExt.getId());

    // if not found in cache, instantiate the Parser   
    } else {
      try {
        Parser p = (Parser) parserExt.getExtensionInstance();
        objectCache.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
291
292
293
294
295
296
297
298
299
300
301
    List<Extension> extList = new ArrayList<Extension>();
    if (plugins != null) {
     
      for (String parsePluginId : plugins) {
       
        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
121
122
123
124
125
126
127
128
129
130
131
132
133
    String pluginName = "urlfilter-prefix";
    Extension[] extensions = PluginRepository.get(conf).getExtensionPoint(
        URLFilter.class.getName()).getExtensions();
    for (int i = 0; i < extensions.length; i++) {
      Extension extension = extensions[i];
      if (extension.getDescriptor().getPluginId().equals(pluginName)) {
        attributeFile = extension.getAttribute("file");
        break;
      }
    }
    if (attributeFile != null && attributeFile.trim().equals(""))
      attributeFile = null;
View Full Code Here
112
113
114
115
116
117
118
119
120
121
122
123
124
    // get the extensions for domain urlfilter
    String pluginName = "urlfilter-domain";
    Extension[] extensions = PluginRepository.get(conf).getExtensionPoint(
      URLFilter.class.getName()).getExtensions();
    for (int i = 0; i < extensions.length; i++) {
      Extension extension = extensions[i];
      if (extension.getDescriptor().getPluginId().equals(pluginName)) {
        attributeFile = extension.getAttribute("file");
        break;
      }
    }
   
    // handle blank non empty input
View Full Code Here
63
64
65
66
67
68
69
70
71
72
73
74
          throw new RuntimeException(IndexingFilter.X_POINT_ID + " not found.");
        Extension[] extensions = point.getExtensions();
        HashMap<String, IndexingFilter> filterMap =
          new HashMap<String, IndexingFilter>();
        for (int i = 0; i < extensions.length; i++) {
          Extension extension = extensions[i];
          IndexingFilter filter = (IndexingFilter) extension
              .getExtensionInstance();
          LOG.info("Adding " + filter.getClass().getName());
          if (!filterMap.containsKey(filter.getClass().getName())) {
            filterMap.put(filter.getClass().getName(), filter);
          }
View Full Code Here
47
48
49
50
51
52
53
54
55
56
57
58
      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-tika", ext.getDescriptor().getPluginId());
  }
View Full Code Here
47
48
49
50
51
52
53
54
55
56
57
58
      parserFactory = new ParserFactory(conf);
  }
   
  /** Unit test for <code>getExtensions(String)</code> method. */
  public void testGetExtensions() throws Exception {
    Extension ext = parserFactory.getExtensions("text/html").get(0);
    assertEquals("parse-tika", ext.getDescriptor().getPluginId());
    ext = parserFactory.getExtensions("text/html; charset=ISO-8859-1").get(0);
    assertEquals("parse-tika", ext.getDescriptor().getPluginId());
    ext = parserFactory.getExtensions("foo/bar").get(0);
    assertEquals("parse-tika", ext.getDescriptor().getPluginId());
  }
View Full Code Here
232
233
234
235
236
237
238
239
240
241
242
243
244
    this.conf = conf;

    String pluginName = "urlfilter-suffix";
    Extension[] extensions = PluginRepository.get(conf).getExtensionPoint(URLFilter.class.getName()).getExtensions();
    for (int i = 0; i < extensions.length; i++) {
      Extension extension = extensions[i];
      if (extension.getDescriptor().getPluginId().equals(pluginName)) {
        attributeFile = extension.getAttribute("file");
        break;
      }
    }
    if (attributeFile != null && attributeFile.trim().equals("")) attributeFile = null;
    if (attributeFile != null) {
View Full Code Here
120
121
122
123
124
125
126
127
128
129
130
131
132
    String pluginName = "urlfilter-prefix";
    Extension[] extensions = PluginRepository.get(conf).getExtensionPoint(
        URLFilter.class.getName()).getExtensions();
    for (int i = 0; i < extensions.length; i++) {
      Extension extension = extensions[i];
      if (extension.getDescriptor().getPluginId().equals(pluginName)) {
        attributeFile = extension.getAttribute("file");
        break;
      }
    }
    if (attributeFile != null && attributeFile.trim().equals(""))
      attributeFile = null;
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.