Examples of Outlink

Examples of bixo.datum.Outlink

58
59
60
61
62
63
64
65
66
67
68
    }

    @Override
    public void operate(FlowProcess process, FunctionCall<NullContext> funcCall) {
        ParsedDatum datum = new ParsedDatum(funcCall.getArguments());
        Outlink outlinks[] = datum.getOutlinks();

        // Bump the crawl depth value only on a successful parse
        int crawlDepth = (Integer) datum.getPayloadValue(CrawlDbDatum.CRAWL_DEPTH);
        datum.setPayloadValue(CrawlDbDatum.CRAWL_DEPTH, crawlDepth + 1);
View Full Code Here

Examples of bixo.datum.Outlink

60
61
62
63
64
65
66
67
68
69
70
    @SuppressWarnings("rawtypes")
    @Override
    public void operate(FlowProcess process, FunctionCall<NullContext> funcCall) {
       
        AnalyzedDatum datum = new AnalyzedDatum(funcCall.getArguments().getTuple());
        Outlink outlinks[] = datum.getOutlinks();

        TupleEntryCollector collector = funcCall.getOutputCollector();

        if (outlinks.length > 0) {
            float pageScore = datum.getPageScore();
View Full Code Here

Examples of bixo.datum.Outlink

120
121
122
123
124
125
126
127
128
129
130
       
        for (Node node : aNodes) {
            String url = getAttributeFromNode(node, "href");
            String anchor = getAttributeFromNode(node, "name");
            String rel = getAttributeFromNode(node, "rel");
            Outlink link = new Outlink(url, anchor, rel);
            outlinkList.add(link);
        }
   
        return outlinkList.toArray(new Outlink[outlinkList.size()]);
    }
View Full Code Here

Examples of bixo.datum.Outlink

143
144
145
146
147
148
149
150
151
    @Override
    public void endElement(String uri, String localName, String name) throws SAXException {
        super.endElement(uri, localName, name);

        if (localName.equalsIgnoreCase(_inAnchorTag)) {
            addLink(new Outlink(_curUrl, _curAnchor.toString(), _curRelAttributes));
            _inAnchorTag = null;
        }
    }
View Full Code Here

Examples of net.nutch.parse.Outlink

26
27
28
29
30
31
32
33
34
35
36
      setAlbum(value);
    if (name.equals("TPE1-Text"))
      setArtist(value);

    if (name.indexOf("URL Link") > -1) {
      links.add(new Outlink(value, ""));
    } else if (name.indexOf("Text") > -1) {
      text += value + "\n";
    }

    metadata.setProperty(name, value);
View Full Code Here

Examples of net.nutch.parse.Outlink

290
291
292
293
294
295
296
297
298
299
300
            }
          }
          if (target != null)
            try {
              URL url = new URL(base, target);
              outlinks.add(new Outlink(url.toString(),
                                       linkText.toString().trim()));
            } catch (MalformedURLException e) {
              // don't care
            }
        }
View Full Code Here

Examples of net.nutch.parse.Outlink

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
202
203
204
205
206
      testDOMs[i]= node;
    }
    try {
     answerOutlinks = new Outlink[][]{
         {
           new Outlink("http://www.nutch.org", "anchor"),
         },
         {
           new Outlink("http://www.nutch.org/", "home"),
           new Outlink("http://www.nutch.org/docs/bot.html", "bots"),
         },
         {
           new Outlink("http://www.nutch.org/", "separate this"),
           new Outlink("http://www.nutch.org/docs/ok", "from this"),
         },
         {
           new Outlink("http://www.nutch.org/", "home"),
           new Outlink("http://www.nutch.org/docs/1", "1"),
           new Outlink("http://www.nutch.org/docs/2", "2"),
         },
         {
           new Outlink("http://www.nutch.org/frames/top.html", ""),
           new Outlink("http://www.nutch.org/frames/left.html", ""),
           new Outlink("http://www.nutch.org/frames/invalid.html", ""),
           new Outlink("http://www.nutch.org/frames/right.html", ""),
         },
         {
           new Outlink("http://www.nutch.org/index.html", ""),
           new Outlink("http://www.nutch.org/maps/#bottom", ""),
           new Outlink("http://www.nutch.org/bot.html", ""),
           new Outlink("http://www.nutch.org/docs/index.html", ""),
         },
         {
             new Outlink("http://www.nutch.org/index.html", "whitespace test"),
         },
      };
  
    } catch (MalformedURLException e) {
       
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

400
401
402
403
404
405
406
407
408
409
410
            if (target != null && !noFollow && !post)
              try {
               
                URL url = (base.toString().indexOf(';') > 0) ?
                  fixEmbeddedParams(base, target) :  new URL(base, target);
                outlinks.add(new Outlink(url.toString(),
                                         linkText.toString().trim()));
              } catch (MalformedURLException e) {
                // don't care
              }
          }
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
      testDOMs[i]= node;
    }
    try {
     answerOutlinks = new Outlink[][]{
         {
           new Outlink("http://www.nutch.org", "anchor"),
         },
         {
           new Outlink("http://www.nutch.org/", "home"),
           new Outlink("http://www.nutch.org/docs/bot.html", "bots"),
         },
         {
           new Outlink("http://www.nutch.org/", "separate this"),
           new Outlink("http://www.nutch.org/docs/ok", "from this"),
         },
         {
           new Outlink("http://www.nutch.org/", "home"),
           new Outlink("http://www.nutch.org/docs/1", "1"),
           new Outlink("http://www.nutch.org/docs/2", "2"),
         },
         {
           new Outlink("http://www.nutch.org/frames/top.html", ""),
           new Outlink("http://www.nutch.org/frames/left.html", ""),
           new Outlink("http://www.nutch.org/frames/invalid.html", ""),
           new Outlink("http://www.nutch.org/frames/right.html", ""),
         },
         {
           new Outlink("http://www.nutch.org/maps/logo.gif", ""),
           new Outlink("http://www.nutch.org/index.html", ""),
           new Outlink("http://www.nutch.org/maps/#bottom", ""),
           new Outlink("http://www.nutch.org/bot.html", ""),
           new Outlink("http://www.nutch.org/docs/index.html", ""),
         },
         {
             new Outlink("http://www.nutch.org/index.html", "whitespace test"),
         },
         {
         },
         {
           new Outlink("http://www.nutch.org/dummy.jsp", "test2"),
         },
         {
         },
         {
           new Outlink("http://www.nutch.org/;x", "anchor1"),
           new Outlink("http://www.nutch.org/g;x", "anchor2"),
           new Outlink("http://www.nutch.org/g;x?y#s", "anchor3")
         },
         {
           new Outlink("http://www.nutch.org/g;something", "anchor1"),
           new Outlink("http://www.nutch.org/g;something?y#s", "anchor2"),
           new Outlink("http://www.nutch.org/;something?y=1", "anchor3"),
           new Outlink("http://www.nutch.org/;something?y=1#s", "anchor4"),
           new Outlink("http://www.nutch.org/?y=1;somethingelse", "anchor5")
         }
      };
  
    } catch (MalformedURLException e) {
       
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

157
158
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
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
      fromUrlCriginalColectionName=SqlSearcher.getCollectionNameOriginal(collection);     
      fromUrlTimestamp=SqlSearcher.getTimestampOriginal(collection);     
    }
   
    for (int i = 0; i < outlinks.length; i++) {
      Outlink outlink = outlinks[i];
      String toUrl = outlink.getToUrl();
          
      if (this.nwIgnoreInternalLinks)
      {
        String toHost = getHost(toUrl);
       
        if (toHost == null || toHost.equals(fromHost)) { // internal link             
          continue;                               // skip it
        }
      }

      if (this.nwUrlNormalizers != null)
      {
        try {         
          toUrl = this.nwUrlNormalizers. normalize(toUrl, URLNormalizers.SCOPE_LINKDB);
        }
        catch (Exception e) {
          LOG.warn("Skipping " + toUrl + ":" + e);
          toUrl = null;
        }
      }
     
      if (toUrl != null && this.nwUrlFilters != null) {
        try {
          toUrl = this.nwUrlFilters.filter(toUrl); // filter the url
          if (toUrl==null) { 
            LOG.info("LINKDB URL FILTERED"); 
          }
        }
        catch (Exception e) {
          LOG.warn("Skipping " + toUrl + ":" + e);
          toUrl = null;
        }
      }
   
      if (toUrl == null) {
        continue;
      }

      inlinks.clear();
   
      String anchor = outlink.getAnchor();        // truncate long anchors
      
      if (anchor.length() > this.nwMaxAnchorLength) {
        anchor = anchor.substring(0, this.nwMaxAnchorLength);
      }
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.