Package org.apache.nutch.plugin

Examples of org.apache.nutch.plugin.Extension

112
113
114
115
116
117
118
119
120
121
122
123
124
    // get the extensions for domain urlfilter
    String pluginName = "urlfilter-domainblacklist";
    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
93
94
95
96
97
98
99
100
101
102
103
104
105
    // get the extensions for domain urlfilter
    String pluginName = "urlnormalizer-host";
    Extension[] extensions = PluginRepository.get(conf).getExtensionPoint(
      URLNormalizer.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
64
65
66
67
68
69
70
71
72
73
74
75
        if (point == null) throw new RuntimeException(ScoringFilter.X_POINT_ID + " not found.");
        Extension[] extensions = point.getExtensions();
        HashMap<String, ScoringFilter> filterMap =
          new HashMap<String, ScoringFilter>();
        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
51
52
53
54
55
56
57
58
59
60
61
62
      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
51
52
53
54
55
56
57
58
59
60
61
62
        if (point == null)
          throw new RuntimeException(URLFilter.X_POINT_ID + " not found.");
        Extension[] extensions = point.getExtensions();
        Map<String, URLFilter> filterMap = new HashMap<String, URLFilter>();
        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);
          }
        }
        if (orderedFilters == null) {
View Full Code Here
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
   
    List<URLNormalizer> normalizers = new Vector<URLNormalizer>(extensions.size());

    Iterator<Extension> it = extensions.iterator();
    while (it.hasNext()) {
      Extension ext = it.next();
      URLNormalizer normalizer = null;
      try {
        // check to see if we've cached this URLNormalizer instance yet
        normalizer = (URLNormalizer) objectCache.getObject(ext.getId());
        if (normalizer == null) {
          // go ahead and instantiate it and then cache it
          normalizer = (URLNormalizer) ext.getExtensionInstance();
          objectCache.setObject(ext.getId(), normalizer);
        }
        normalizers.add(normalizer);
      } catch (PluginRuntimeException e) {
        e.printStackTrace();
        LOG.warn("URLNormalizers:PluginRuntimeException when "
                + "initializing url normalizer plugin "
                + ext.getDescriptor().getPluginId()
                + " instance in getURLNormalizers "
                + "function: attempting to continue instantiating plugins");
      }
    }
    return normalizers.toArray(new URLNormalizer[normalizers
View Full Code Here
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
      impls = new HashSet<String>(Arrays.asList(names));
    }
    Extension[] extensions = this.extensionPoint.getExtensions();
    HashMap<String, Extension> normalizerExtensions = new HashMap<String, Extension>();
    for (int i = 0; i < extensions.length; i++) {
      Extension extension = extensions[i];
      if (impls != null && !impls.contains(extension.getClazz()))
        continue;
      normalizerExtensions.put(extension.getClazz(), extension);
    }
    List<Extension> res = new ArrayList<Extension>();
    if (orders == null) {
      res.addAll(normalizerExtensions.values());
    } else {
      // first add those explicitly named in correct order
      for (int i = 0; i < orders.length; i++) {
        Extension e = normalizerExtensions.get(orders[i]);
        if (e != null) {
          res.add(e);
          normalizerExtensions.remove(orders[i]);
        }
      }
View Full Code Here
59
60
61
62
63
64
65
66
67
68
69
70
        if (point == null) throw new RuntimeException(ScoringFilter.X_POINT_ID + " not found.");
        Extension[] extensions = point.getExtensions();
        HashMap<String, ScoringFilter> filterMap =
          new HashMap<String, ScoringFilter>();
        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
51
52
53
54
55
56
57
58
59
60
61
62
      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
51
52
53
54
55
56
57
58
59
60
61
62
        if (point == null)
          throw new RuntimeException(URLFilter.X_POINT_ID + " not found.");
        Extension[] extensions = point.getExtensions();
        Map<String, URLFilter> filterMap = new HashMap<String, URLFilter>();
        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);
          }
        }
        if (orderedFilters == 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.