Package edu.cmu.lemurproject

Examples of edu.cmu.lemurproject.WarcRecord$WarcHeader

152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
        if (whichStream == null) {
            return false;
        }

        WarcRecord newRecord = WarcRecord.readNextWarcRecord(whichStream);
        if (newRecord == null) {
            // try advancing the file
            if (openNextFile()) {
                newRecord = WarcRecord.readNextWarcRecord(whichStream);
            }

            if (newRecord == null) {
                return false;
            }
        }

        totalNumBytesRead += (long) newRecord.getTotalRecordLength();
        newRecord.setWarcFilePath(filePathList[currentFilePath].toString());

        // now, set our output variables
        value.setRecord(newRecord);
        key.set(recordNumber);
View Full Code Here
43
44
45
46
47
48
49
public class WritableWarcRecord implements Writable {

    WarcRecord record = null;

    public WritableWarcRecord() {
        record = new WarcRecord();
    }
View Full Code Here
47
48
49
50
51
52
53
    public WritableWarcRecord() {
        record = new WarcRecord();
    }

    public WritableWarcRecord(WarcRecord o) {
        record = new WarcRecord(o);
    }
View Full Code Here
55
56
57
58
59
60
61
    public WarcRecord getRecord() {
        return record;
    }

    public void setRecord(WarcRecord rec) {
        record = new WarcRecord(rec);
    }
View Full Code Here
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
  public void map(LongWritable key, WritableWarcRecord record,
      OutputCollector<Text, BehemothDocument> output, Reporter reporter)
      throws IOException {

    WarcRecord wr = record.getRecord();

    if (wr.getHeaderRecordType().equals("response") == false)
      return;

    byte[] binarycontent = wr.getContent();

    String uri = wr.getHeaderMetadataItem("WARC-Target-URI");

    // skip non http documents
    if (uri.startsWith("http") == false)
      return;

    String ip = wr.getHeaderMetadataItem("WARC-IP-Address");

    HttpResponse response;
    try {
      response = new HttpResponse(binarycontent);
    } catch (ProtocolException e) {
View Full Code Here
TOP

Related Classes of edu.cmu.lemurproject.WarcRecord$WarcHeader

  • com.digitalpebble.behemoth.io.warc.WARCConverterJob
  • edu.cmu.lemurproject.WarcFileRecordReader
  • edu.cmu.lemurproject.WritableWarcRecord
    • 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.