Examples of Outlink

Examples of org.apache.nutch.parse.Outlink

331
332
333
334
335
336
337
338
339
340
341
            }
          }
          if (target != null && !noFollow && !post)
            try {
              URL url = new URL(base, target);
              outlinks.add(new Outlink(url.toString(),
                                       linkText.toString().trim(), conf));
            } catch (MalformedURLException e) {
              // don't care
            }
        }
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

99
100
101
102
103
104
105
106
107
108
109
            Parse parse = new ParseUtil(this.conf).parse(content);
            ParseData theParseData = parse.getData();
            Outlink[] theOutlinks = theParseData.getOutlinks();
           
            for(int count = 0; count < theOutlinks.length; count++) {
              outLinksList.add(new Outlink(theOutlinks[count].getToUrl(), theOutlinks[count].getAnchor(), this.conf));
            }
           
            resultText += entry.getName() + " " + parse.getText() + " ";
          } catch (ParseException e) {
            if (LOG.isInfoEnabled()) {
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

199
200
201
202
203
204
205
206
207
208
209
        } else url = new URL(baseURL, url).toString();
        url = url.replaceAll("&amp;", "&");
        if (LOG.isTraceEnabled()) {
          LOG.trace(" - outlink from JS: '" + url + "'");
        }
        outlinks.add(new Outlink(url, anchor, getConf()));
      }
    } catch (Exception ex) {
      // if it is a malformed URL we just throw it away and continue with
      // extraction.
      if (LOG.isErrorEnabled()) { LOG.error("getJSLinks", ex); }
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

48
49
50
51
52
53
54
55
56
57
58
      setAlbum(value);
    if (name.equals("TPE1-Text"))
      setArtist(value);

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

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

Examples of org.apache.nutch.parse.Outlink

218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
      testDOMs[i]= node;
    }
    try {
     answerOutlinks = new Outlink[][]{
         {
           new Outlink("http://www.nutch.org", "anchor", conf),
         },
         {
           new Outlink("http://www.nutch.org/", "home", conf),
           new Outlink("http://www.nutch.org/docs/bot.html", "bots", conf),
         },
         {
           new Outlink("http://www.nutch.org/", "separate this", conf),
           new Outlink("http://www.nutch.org/docs/ok", "from this", conf),
         },
         {
           new Outlink("http://www.nutch.org/", "home", conf),
           new Outlink("http://www.nutch.org/docs/1", "1", conf),
           new Outlink("http://www.nutch.org/docs/2", "2", conf),
         },
         {
           new Outlink("http://www.nutch.org/frames/top.html", "", conf),
           new Outlink("http://www.nutch.org/frames/left.html", "", conf),
           new Outlink("http://www.nutch.org/frames/invalid.html", "", conf),
           new Outlink("http://www.nutch.org/frames/right.html", "", conf),
         },
         {
           new Outlink("http://www.nutch.org/maps/logo.gif", "", conf),
           new Outlink("http://www.nutch.org/index.html", "", conf),
           new Outlink("http://www.nutch.org/maps/#bottom", "", conf),
           new Outlink("http://www.nutch.org/bot.html", "", conf),
           new Outlink("http://www.nutch.org/docs/index.html", "", conf),
         },
         {
             new Outlink("http://www.nutch.org/index.html", "whitespace test", conf),
         },
         {
         },
         {
           new Outlink("http://www.nutch.org/dummy.jsp", "test2", conf),
         },
         {
         }
      };
  
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

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
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
      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")
         },
         {
           new Outlink("http://www.nutch.org/g", ""),
           new Outlink("http://www.nutch.org/g1", ""),
           new Outlink("http://www.nutch.org/g2", "bla bla"),
           new Outlink("http://www.nutch.org/test.gif", "bla bla"),
         }
      };
  
    } catch (MalformedURLException e) {
       
View Full Code Here

Examples of org.apache.nutch.parse.Outlink

429
430
431
432
433
434
435
436
437
438
439
            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

102
103
104
105
106
107
108
109
110
111
112
            Parse parse = new ParseUtil(this.conf).parse(content).get(content.getUrl());
            ParseData theParseData = parse.getData();
            Outlink[] theOutlinks = theParseData.getOutlinks();
           
            for(int count = 0; count < theOutlinks.length; count++) {
              outLinksList.add(new Outlink(theOutlinks[count].getToUrl(), theOutlinks[count].getAnchor()));
            }
           
            resultText += entry.getName() + " " + parse.getText() + " ";
          } catch (ParseException e) {
            if (LOG.isInfoEnabled()) {
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

218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
        Map<String, String> outlinkMap = new LinkedHashMap<String, String>();

        // normalize urls and put into map
        if (outlinkAr != null && outlinkAr.length > 0) {
          for (int i = 0; i < outlinkAr.length; i++) {
            Outlink outlink = outlinkAr[i];
            String toUrl = normalizeUrl(outlink.getToUrl());

            // only put into map if the url doesn't already exist in the map or
            // if it does and the anchor for that link is null, will replace if
            // url is existing
            boolean existingUrl = outlinkMap.containsKey(toUrl);
            if (toUrl != null
              && (!existingUrl || (existingUrl && outlinkMap.get(toUrl) == null))) {
              outlinkMap.put(toUrl, outlink.getAnchor());
            }
          }
        }

        // collect the outlinks under the fetch time
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.