Package org.apache.nutch.metadata

Examples of org.apache.nutch.metadata.MetaWrapper

191
192
193
194
195
196
197
198
199
200
201
202
203
204
        HashMap sliceWriters = new HashMap();
        String segmentName = job.get("segment.merger.segmentName");
       
        public void write(WritableComparable key, Writable value) throws IOException {
          // unwrap
          MetaWrapper wrapper = (MetaWrapper)value;
          SegmentPart sp = SegmentPart.parse(wrapper.getMeta(SEGMENT_PART_KEY));
          Writable o = (Writable)wrapper.get();
          String slice = wrapper.getMeta(SEGMENT_SLICE_KEY);
          if (o instanceof CrawlDatum) {
            if (sp.partName.equals(CrawlDatum.GENERATE_DIR_NAME)) {
              g_out = ensureSequenceFile(slice, CrawlDatum.GENERATE_DIR_NAME);
              g_out.append(key, o);
            } else if (sp.partName.equals(CrawlDatum.FETCH_DIR_NAME)) {
View Full Code Here
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
    String lastCname = null;
    String lastPDname = null;
    String lastPTname = null;
    TreeMap linked = new TreeMap();
    while (values.hasNext()) {
      MetaWrapper wrapper = (MetaWrapper)values.next();
      Object o = wrapper.get();
      String spString = wrapper.getMeta(SEGMENT_PART_KEY);
      if (spString == null) {
        throw new IOException("Null segment part, key=" + key);       
      }
      SegmentPart sp = SegmentPart.parse(spString);
      if (o instanceof CrawlDatum) {
        CrawlDatum val = (CrawlDatum)o;
        // check which output dir it belongs to
        if (sp.partName.equals(CrawlDatum.GENERATE_DIR_NAME)) {
          if (lastG == null) {
            lastG = val;
            lastGname = sp.segmentName;
          } else {
            // take newer
            if (lastGname.compareTo(sp.segmentName) < 0) {
              lastG = val;
              lastGname = sp.segmentName;
            }
          }
        } else if (sp.partName.equals(CrawlDatum.FETCH_DIR_NAME)) {
          if (lastF == null) {
            lastF = val;
            lastFname = sp.segmentName;
          } else {
            // take newer
            if (lastFname.compareTo(sp.segmentName) < 0) {
              lastF = val;
              lastFname = sp.segmentName;
            }
          }
        } else if (sp.partName.equals(CrawlDatum.PARSE_DIR_NAME)) {
          if (val.getStatus() == CrawlDatum.STATUS_SIGNATURE) {
            if (lastSig == null) {
              lastSig = val;
              lastSigname = sp.segmentName;
            } else {
              // take newer
              if (lastSigname.compareTo(sp.segmentName) < 0) {
                lastSig = val;
                lastSigname = sp.segmentName;
              }
            }
            continue;
          }
          // collect all LINKED values from the latest segment
          ArrayList segLinked = (ArrayList)linked.get(sp.segmentName);
          if (segLinked == null) {
            segLinked = new ArrayList();
            linked.put(sp.segmentName, segLinked);
          }
          segLinked.add(val);
        } else {
          throw new IOException("Cannot determine segment part: " + sp.partName);
        }
      } else if (o instanceof Content) {
        if (lastC == null) {
          lastC = (Content)o;
          lastCname = sp.segmentName;
        } else {
          if (lastCname.compareTo(sp.segmentName) < 0) {
            lastC = (Content)o;
            lastCname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseData) {
        if (lastPD == null) {
          lastPD = (ParseData)o;
          lastPDname = sp.segmentName;
        } else {
          if (lastPDname.compareTo(sp.segmentName) < 0) {
            lastPD = (ParseData)o;
            lastPDname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseText) {
        if (lastPT == null) {
          lastPT = (ParseText)o;
          lastPTname = sp.segmentName;
        } else {
          if (lastPTname.compareTo(sp.segmentName) < 0) {
            lastPT = (ParseText)o;
            lastPTname = sp.segmentName;
          }
        }
      }
    }
    curCount++;
    String sliceName = null;
    MetaWrapper wrapper = new MetaWrapper();
    if (sliceSize > 0) {
      sliceName = String.valueOf(curCount / sliceSize);
      wrapper.setMeta(SEGMENT_SLICE_KEY, sliceName);
    }
    SegmentPart sp = new SegmentPart();
    // now output the latest values
    if (lastG != null) {
      wrapper.set(lastG);
      sp.partName = CrawlDatum.GENERATE_DIR_NAME;
      sp.segmentName = lastGname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastF != null) {
      wrapper.set(lastF);
      sp.partName = CrawlDatum.FETCH_DIR_NAME;
      sp.segmentName = lastFname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastSig != null) {
      wrapper.set(lastSig);
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = lastSigname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastC != null) {
      wrapper.set(lastC);
      sp.partName = Content.DIR_NAME;
      sp.segmentName = lastCname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPD != null) {
      wrapper.set(lastPD);
      sp.partName = ParseData.DIR_NAME;
      sp.segmentName = lastPDname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPT != null) {
      wrapper.set(lastPT);
      sp.partName = ParseText.DIR_NAME;
      sp.segmentName = lastPTname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (linked.size() > 0) {
      String name = (String)linked.lastKey();
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = name;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      ArrayList segLinked = (ArrayList)linked.get(name);
      for (int i = 0; i < segLinked.size(); i++) {
        CrawlDatum link = (CrawlDatum)segLinked.get(i);
        wrapper.set(link);
        output.collect(key, wrapper);
      }
    }
  }
View Full Code Here
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
         
          @Override
          public synchronized boolean next(Writable key, Writable value) throws IOException {
            LOG.debug("Running OIF.next()");
           
            MetaWrapper wrapper = (MetaWrapper) value;
            try {
              wrapper.set(getValueClass().newInstance());
            } catch (Exception e) {
              throw new IOException(e.toString());
            }

            boolean res = super.next(key, (Writable) wrapper.get());
            wrapper.setMeta(SEGMENT_PART_KEY, spString);
            return res;
          }
         
          @Override
          public Writable createValue() {
            return new MetaWrapper();
          }
         
        };
      } catch (IOException e) {
        throw new RuntimeException("Cannot create RecordReader: ", e);
View Full Code Here
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
    String lastPDname = null;
    String lastPTname = null;
    TreeMap<String, ArrayList<CrawlDatum>> linked =
      new TreeMap<String, ArrayList<CrawlDatum>>();
    while (values.hasNext()) {
      MetaWrapper wrapper = values.next();
      Object o = wrapper.get();
      String spString = wrapper.getMeta(SEGMENT_PART_KEY);
      if (spString == null) {
        throw new IOException("Null segment part, key=" + key);       
      }
      SegmentPart sp = SegmentPart.parse(spString);
      if (o instanceof CrawlDatum) {
        CrawlDatum val = (CrawlDatum)o;
        // check which output dir it belongs to
        if (sp.partName.equals(CrawlDatum.GENERATE_DIR_NAME)) {
          if (lastG == null) {
            lastG = val;
            lastGname = sp.segmentName;
          } else {
            // take newer
            if (lastGname.compareTo(sp.segmentName) < 0) {
              lastG = val;
              lastGname = sp.segmentName;
            }
          }
        } else if (sp.partName.equals(CrawlDatum.FETCH_DIR_NAME)) {
          // only consider fetch status and ignore fetch retry status
          // https://issues.apache.org/jira/browse/NUTCH-1520
          // https://issues.apache.org/jira/browse/NUTCH-1113
          if (CrawlDatum.hasFetchStatus(val) &&
            val.getStatus() != CrawlDatum.STATUS_FETCH_RETRY &&
            val.getStatus() != CrawlDatum.STATUS_FETCH_NOTMODIFIED) {
            if (lastF == null) {
              lastF = val;
              lastFname = sp.segmentName;
            } else {
              if (lastFname.compareTo(sp.segmentName) < 0) {
                lastF = val;
                lastFname = sp.segmentName;
              }
            }
          }
        } else if (sp.partName.equals(CrawlDatum.PARSE_DIR_NAME)) {
          if (val.getStatus() == CrawlDatum.STATUS_SIGNATURE) {
            if (lastSig == null) {
              lastSig = val;
              lastSigname = sp.segmentName;
            } else {
              // take newer
              if (lastSigname.compareTo(sp.segmentName) < 0) {
                lastSig = val;
                lastSigname = sp.segmentName;
              }
            }
            continue;
          }
          // collect all LINKED values from the latest segment
          ArrayList<CrawlDatum> segLinked = linked.get(sp.segmentName);
          if (segLinked == null) {
            segLinked = new ArrayList<CrawlDatum>();
            linked.put(sp.segmentName, segLinked);
          }
          segLinked.add(val);
        } else {
          throw new IOException("Cannot determine segment part: " + sp.partName);
        }
      } else if (o instanceof Content) {
        if (lastC == null) {
          lastC = (Content)o;
          lastCname = sp.segmentName;
        } else {
          if (lastCname.compareTo(sp.segmentName) < 0) {
            lastC = (Content)o;
            lastCname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseData) {
        if (lastPD == null) {
          lastPD = (ParseData)o;
          lastPDname = sp.segmentName;
        } else {
          if (lastPDname.compareTo(sp.segmentName) < 0) {
            lastPD = (ParseData)o;
            lastPDname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseText) {
        if (lastPT == null) {
          lastPT = (ParseText)o;
          lastPTname = sp.segmentName;
        } else {
          if (lastPTname.compareTo(sp.segmentName) < 0) {
            lastPT = (ParseText)o;
            lastPTname = sp.segmentName;
          }
        }
      }
    }
  // perform filtering based on full merge record
    if (mergeFilters != null &&
       !mergeFilters.filter(key, lastG, lastF, lastSig, lastC, lastPD, lastPT,
                    linked.isEmpty() ? null : linked.lastEntry().getValue())){
      return;
    }

    curCount++;
    String sliceName = null;
    MetaWrapper wrapper = new MetaWrapper();
    if (sliceSize > 0) {
      sliceName = String.valueOf(curCount / sliceSize);
      wrapper.setMeta(SEGMENT_SLICE_KEY, sliceName);
    }
    SegmentPart sp = new SegmentPart();
    // now output the latest values
    if (lastG != null) {
      wrapper.set(lastG);
      sp.partName = CrawlDatum.GENERATE_DIR_NAME;
      sp.segmentName = lastGname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastF != null) {
      wrapper.set(lastF);
      sp.partName = CrawlDatum.FETCH_DIR_NAME;
      sp.segmentName = lastFname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastSig != null) {
      wrapper.set(lastSig);
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = lastSigname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastC != null) {
      wrapper.set(lastC);
      sp.partName = Content.DIR_NAME;
      sp.segmentName = lastCname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPD != null) {
      wrapper.set(lastPD);
      sp.partName = ParseData.DIR_NAME;
      sp.segmentName = lastPDname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPT != null) {
      wrapper.set(lastPT);
      sp.partName = ParseText.DIR_NAME;
      sp.segmentName = lastPTname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (linked.size() > 0) {
      String name = linked.lastKey();
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = name;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      ArrayList<CrawlDatum> segLinked = linked.get(name);
      for (int i = 0; i < segLinked.size(); i++) {
        CrawlDatum link = segLinked.get(i);
        wrapper.set(link);
        output.collect(key, wrapper);
      }
    }
  }
View Full Code Here
184
185
186
187
188
189
190
191
192
193
194
            splitReader.close();
          }
         
          @Override
          public MetaWrapper createValue() {
            return new MetaWrapper();
          }
         
        };
      } catch (IOException e) {
        throw new RuntimeException("Cannot create RecordReader: ", e);
View Full Code Here
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
    String lastCname = null;
    String lastPDname = null;
    String lastPTname = null;
    TreeMap linked = new TreeMap();
    while (values.hasNext()) {
      MetaWrapper wrapper = (MetaWrapper)values.next();
      Object o = wrapper.get();
      String spString = wrapper.getMeta(SEGMENT_PART_KEY);
      if (spString == null) {
        throw new IOException("Null segment part, key=" + key);       
      }
      SegmentPart sp = SegmentPart.parse(spString);
      if (o instanceof CrawlDatum) {
        CrawlDatum val = (CrawlDatum)o;
        // check which output dir it belongs to
        if (sp.partName.equals(CrawlDatum.GENERATE_DIR_NAME)) {
          if (lastG == null) {
            lastG = val;
            lastGname = sp.segmentName;
          } else {
            // take newer
            if (lastGname.compareTo(sp.segmentName) < 0) {
              lastG = val;
              lastGname = sp.segmentName;
            }
          }
        } else if (sp.partName.equals(CrawlDatum.FETCH_DIR_NAME)) {
          if (lastF == null) {
            lastF = val;
            lastFname = sp.segmentName;
          } else {
            // take newer
            if (lastFname.compareTo(sp.segmentName) < 0) {
              lastF = val;
              lastFname = sp.segmentName;
            }
          }
        } else if (sp.partName.equals(CrawlDatum.PARSE_DIR_NAME)) {
          if (val.getStatus() == CrawlDatum.STATUS_SIGNATURE) {
            if (lastSig == null) {
              lastSig = val;
              lastSigname = sp.segmentName;
            } else {
              // take newer
              if (lastSigname.compareTo(sp.segmentName) < 0) {
                lastSig = val;
                lastSigname = sp.segmentName;
              }
            }
            continue;
          }
          // collect all LINKED values from the latest segment
          ArrayList segLinked = (ArrayList)linked.get(sp.segmentName);
          if (segLinked == null) {
            segLinked = new ArrayList();
            linked.put(sp.segmentName, segLinked);
          }
          segLinked.add(val);
        } else {
          throw new IOException("Cannot determine segment part: " + sp.partName);
        }
      } else if (o instanceof Content) {
        if (lastC == null) {
          lastC = (Content)o;
          lastCname = sp.segmentName;
        } else {
          if (lastCname.compareTo(sp.segmentName) < 0) {
            lastC = (Content)o;
            lastCname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseData) {
        if (lastPD == null) {
          lastPD = (ParseData)o;
          lastPDname = sp.segmentName;
        } else {
          if (lastPDname.compareTo(sp.segmentName) < 0) {
            lastPD = (ParseData)o;
            lastPDname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseText) {
        if (lastPT == null) {
          lastPT = (ParseText)o;
          lastPTname = sp.segmentName;
        } else {
          if (lastPTname.compareTo(sp.segmentName) < 0) {
            lastPT = (ParseText)o;
            lastPTname = sp.segmentName;
          }
        }
      }
    }
    curCount++;
    String sliceName = null;
    MetaWrapper wrapper = new MetaWrapper();
    if (sliceSize > 0) {
      sliceName = String.valueOf(curCount / sliceSize);
      wrapper.setMeta(SEGMENT_SLICE_KEY, sliceName);
    }
    SegmentPart sp = new SegmentPart();
    // now output the latest values
    if (lastG != null) {
      wrapper.set(lastG);
      sp.partName = CrawlDatum.GENERATE_DIR_NAME;
      sp.segmentName = lastGname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastF != null) {
      wrapper.set(lastF);
      sp.partName = CrawlDatum.FETCH_DIR_NAME;
      sp.segmentName = lastFname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastSig != null) {
      wrapper.set(lastSig);
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = lastSigname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastC != null) {
      wrapper.set(lastC);
      sp.partName = Content.DIR_NAME;
      sp.segmentName = lastCname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPD != null) {
      wrapper.set(lastPD);
      sp.partName = ParseData.DIR_NAME;
      sp.segmentName = lastPDname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPT != null) {
      wrapper.set(lastPT);
      sp.partName = ParseText.DIR_NAME;
      sp.segmentName = lastPTname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (linked.size() > 0) {
      String name = (String)linked.lastKey();
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = name;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      ArrayList segLinked = (ArrayList)linked.get(name);
      for (int i = 0; i < segLinked.size(); i++) {
        CrawlDatum link = (CrawlDatum)segLinked.get(i);
        wrapper.set(link);
        output.collect(key, wrapper);
      }
    }
  }
View Full Code Here
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
         
          @Override
          public synchronized boolean next(Writable key, Writable value) throws IOException {
            LOG.debug("Running OIF.next()");
           
            MetaWrapper wrapper = (MetaWrapper) value;
            try {
              wrapper.set(getValueClass().newInstance());
            } catch (Exception e) {
              throw new IOException(e.toString());
            }

            boolean res = super.next(key, (Writable) wrapper.get());
            wrapper.setMeta(SEGMENT_PART_KEY, spString);
            return res;
          }
         
          @Override
          public Writable createValue() {
            return new MetaWrapper();
          }
         
        };
      } catch (IOException e) {
        throw new RuntimeException("Cannot create RecordReader: ", e);
View Full Code Here
189
190
191
192
193
194
195
196
197
198
199
200
201
202
        HashMap sliceWriters = new HashMap();
        String segmentName = job.get("segment.merger.segmentName");
       
        public void write(WritableComparable key, Writable value) throws IOException {
          // unwrap
          MetaWrapper wrapper = (MetaWrapper)value;
          SegmentPart sp = SegmentPart.parse(wrapper.getMeta(SEGMENT_PART_KEY));
          Writable o = (Writable)wrapper.get();
          String slice = wrapper.getMeta(SEGMENT_SLICE_KEY);
          if (o instanceof CrawlDatum) {
            if (sp.partName.equals(CrawlDatum.GENERATE_DIR_NAME)) {
              g_out = ensureSequenceFile(slice, CrawlDatum.GENERATE_DIR_NAME);
              g_out.append(key, o);
            } else if (sp.partName.equals(CrawlDatum.FETCH_DIR_NAME)) {
View Full Code Here
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
    String lastPDname = null;
    String lastPTname = null;
    TreeMap<String, ArrayList<CrawlDatum>> linked =
      new TreeMap<String, ArrayList<CrawlDatum>>();
    while (values.hasNext()) {
      MetaWrapper wrapper = values.next();
      Object o = wrapper.get();
      String spString = wrapper.getMeta(SEGMENT_PART_KEY);
      if (spString == null) {
        throw new IOException("Null segment part, key=" + key);       
      }
      SegmentPart sp = SegmentPart.parse(spString);
      if (o instanceof CrawlDatum) {
        CrawlDatum val = (CrawlDatum)o;
        // check which output dir it belongs to
        if (sp.partName.equals(CrawlDatum.GENERATE_DIR_NAME)) {
          if (lastG == null) {
            lastG = val;
            lastGname = sp.segmentName;
          } else {
            // take newer
            if (lastGname.compareTo(sp.segmentName) < 0) {
              lastG = val;
              lastGname = sp.segmentName;
            }
          }
        } else if (sp.partName.equals(CrawlDatum.FETCH_DIR_NAME)) {
          if (lastF == null) {
            lastF = val;
            lastFname = sp.segmentName;
          } else {
            // take newer
            if (lastFname.compareTo(sp.segmentName) < 0) {
              lastF = val;
              lastFname = sp.segmentName;
            }
          }
        } else if (sp.partName.equals(CrawlDatum.PARSE_DIR_NAME)) {
          if (val.getStatus() == CrawlDatum.STATUS_SIGNATURE) {
            if (lastSig == null) {
              lastSig = val;
              lastSigname = sp.segmentName;
            } else {
              // take newer
              if (lastSigname.compareTo(sp.segmentName) < 0) {
                lastSig = val;
                lastSigname = sp.segmentName;
              }
            }
            continue;
          }
          // collect all LINKED values from the latest segment
          ArrayList<CrawlDatum> segLinked = linked.get(sp.segmentName);
          if (segLinked == null) {
            segLinked = new ArrayList<CrawlDatum>();
            linked.put(sp.segmentName, segLinked);
          }
          segLinked.add(val);
        } else {
          throw new IOException("Cannot determine segment part: " + sp.partName);
        }
      } else if (o instanceof Content) {
        if (lastC == null) {
          lastC = (Content)o;
          lastCname = sp.segmentName;
        } else {
          if (lastCname.compareTo(sp.segmentName) < 0) {
            lastC = (Content)o;
            lastCname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseData) {
        if (lastPD == null) {
          lastPD = (ParseData)o;
          lastPDname = sp.segmentName;
        } else {
          if (lastPDname.compareTo(sp.segmentName) < 0) {
            lastPD = (ParseData)o;
            lastPDname = sp.segmentName;
          }
        }
      } else if (o instanceof ParseText) {
        if (lastPT == null) {
          lastPT = (ParseText)o;
          lastPTname = sp.segmentName;
        } else {
          if (lastPTname.compareTo(sp.segmentName) < 0) {
            lastPT = (ParseText)o;
            lastPTname = sp.segmentName;
          }
        }
      }
    }
    curCount++;
    String sliceName = null;
    MetaWrapper wrapper = new MetaWrapper();
    if (sliceSize > 0) {
      sliceName = String.valueOf(curCount / sliceSize);
      wrapper.setMeta(SEGMENT_SLICE_KEY, sliceName);
    }
    SegmentPart sp = new SegmentPart();
    // now output the latest values
    if (lastG != null) {
      wrapper.set(lastG);
      sp.partName = CrawlDatum.GENERATE_DIR_NAME;
      sp.segmentName = lastGname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastF != null) {
      wrapper.set(lastF);
      sp.partName = CrawlDatum.FETCH_DIR_NAME;
      sp.segmentName = lastFname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastSig != null) {
      wrapper.set(lastSig);
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = lastSigname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastC != null) {
      wrapper.set(lastC);
      sp.partName = Content.DIR_NAME;
      sp.segmentName = lastCname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPD != null) {
      wrapper.set(lastPD);
      sp.partName = ParseData.DIR_NAME;
      sp.segmentName = lastPDname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (lastPT != null) {
      wrapper.set(lastPT);
      sp.partName = ParseText.DIR_NAME;
      sp.segmentName = lastPTname;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      output.collect(key, wrapper);
    }
    if (linked.size() > 0) {
      String name = linked.lastKey();
      sp.partName = CrawlDatum.PARSE_DIR_NAME;
      sp.segmentName = name;
      wrapper.setMeta(SEGMENT_PART_KEY, sp.toString());
      ArrayList<CrawlDatum> segLinked = linked.get(name);
      for (int i = 0; i < segLinked.size(); i++) {
        CrawlDatum link = segLinked.get(i);
        wrapper.set(link);
        output.collect(key, wrapper);
      }
    }
  }
View Full Code Here
174
175
176
177
178
179
180
181
182
183
184
            reader.close();
          }
         
          @Override
          public MetaWrapper createValue() {
            return new MetaWrapper();
          }
         
        };
      } catch (IOException e) {
        throw new RuntimeException("Cannot create RecordReader: ", e);
View Full Code Here
TOP

Related Classes of org.apache.nutch.metadata.MetaWrapper

  • org.apache.nutch.segment.SegmentMerger
  • org.apache.nutch.segment.SegmentMerger$ObjectInputFormat
  • org.apache.nutch.segment.SegmentMerger$SegmentOutputFormat
    • 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.