Package org.apache.nutch.protocol

Examples of org.apache.nutch.protocol.ProtocolFactory

441
442
443
444
445
446
447
448
449
450
451
      this.setDaemon(true);                       // don't hang JVM on exit
      this.setName("FetcherThread" + num);        // use an informative name
      this.context = context;
      Configuration conf = context.getConfiguration();
      this.urlFilters = new URLFilters(conf);
      this.protocolFactory = new ProtocolFactory(conf);
      this.normalizers = new URLNormalizers(conf, URLNormalizers.SCOPE_FETCHER);
      this.maxCrawlDelay = conf.getInt("fetcher.max.crawl.delay", 30) * 1000;
      // backward-compatible default setting
      this.byIP = conf.getBoolean("fetcher.threads.per.host.by.ip", true);
      this.ignoreExternalLinks = conf.getBoolean("db.ignore.external.links", false);
View Full Code Here
139
140
141
142
143
144
145
146
147
148
    Collection<WebPage.Field> fields = new HashSet<WebPage.Field>(FIELDS);
    if (job.getConfiguration().getBoolean(PARSE_KEY, false)) {
      ParserJob parserJob = new ParserJob();
      fields.addAll(parserJob.getFields(job));
    }
    ProtocolFactory protocolFactory = new ProtocolFactory(job.getConfiguration());
    fields.addAll(protocolFactory.getFields());

    return fields;
  }
View Full Code Here
103
104
105
106
107
108
109
110
111
112
113
114
    if (LOG.isInfoEnabled()) {
      LOG.info("fetching: " + url);
    }

    ProtocolFactory factory = new ProtocolFactory(conf);
    Protocol protocol = factory.getProtocol(url);
    WebPage page = new WebPage();
   
    ProtocolOutput protocolOutput = protocol.getProtocolOutput(url, page);
   
    if(!protocolOutput.getStatus().isSuccess()) {
View Full Code Here
48
49
50
51
52
53
54
55
56
57
58
 
  public Metadata parseMeta(String fileName, Configuration conf) {
    Metadata metadata = null;
    try {
      String urlString = "file:" + sampleDir + fileSeparator + fileName;    
      Protocol protocol = new ProtocolFactory(conf).getProtocol(urlString);
      Content content = protocol.getProtocolOutput(new Text(urlString),
          new CrawlDatum()).getContent();
      Parse parse = new ParseUtil(conf).parse(content).get(content.getUrl());
      metadata = parse.getData().getParseMeta();
    } catch (Exception e) {
View Full Code Here
96
97
98
99
100
101
102
103
104
105
106
107
    if (LOG.isInfoEnabled()) {
      LOG.info("fetching: " + url);
    }

    ProtocolFactory factory = new ProtocolFactory(conf);
    Protocol protocol = factory.getProtocol(url);
    ProtocolOutput output = protocol.getProtocolOutput(new Text(url), new CrawlDatum());
   
    if (!output.getStatus().isSuccess()) {
      System.err.println("Fetch failed with protocol status: " + output.getStatus());
      return (-1);
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
88
      LOG.info("fetching: " + url);
    }

    IndexingFilters indexers = new IndexingFilters(conf);

    ProtocolFactory factory = new ProtocolFactory(conf);
    Protocol protocol = factory.getProtocol(url);
    CrawlDatum datum = new CrawlDatum();

    ProtocolOutput output = protocol.getProtocolOutput(new Text(url), datum);
   
    if (!output.getStatus().isSuccess()) {
View Full Code Here
80
81
82
83
84
85
86
87
88
89
90
    Configuration conf = NutchConfiguration.create();

    for (int i = 0; i < sampleFiles.length; i++) {
      urlString = "file:" + sampleDir + fileSeparator + sampleFiles[i];

      protocol = new ProtocolFactory(conf).getProtocol(urlString);
      content = protocol.getProtocolOutput(new Text(urlString), new CrawlDatum()).getContent();

      parse = new ParseUtil(conf).parse(content).get(content.getUrl());

      String text = parse.getText().replaceAll("[ \t\r\n]+", " ").trim();
View Full Code Here
69
70
71
72
73
74
75
76
77
78
79
    Content content;
    Parse parse;

    Configuration conf = NutchConfiguration.create();
    urlString = "file:" + sampleDir + fileSeparator + rtfFile;
    protocol = new ProtocolFactory(conf).getProtocol(urlString);
    content = protocol.getProtocolOutput(new UTF8(urlString), new CrawlDatum())
                      .getContent();
    parse = new ParseUtil(conf).parseByExtensionId("parse-rtf", content);
    String text = parse.getText();
    assertEquals("The quick brown fox jumps over the lazy dog", text.trim());
View Full Code Here
65
66
67
68
69
70
71
72
73
74
75
    Configuration conf = NutchConfiguration.create();
    for (int i=0; i<sampleFiles.length; i++) {
      urlString = "file:" + sampleDir + fileSeparator + sampleFiles[i];

      protocol = new ProtocolFactory(conf).getProtocol(urlString);
      content = protocol.getProtocolOutput(new UTF8(urlString), new CrawlDatum()).getContent();
      parse = new ParseUtil(conf).parseByExtensionId("parse-msword", content);

      assertTrue(parse.getText().startsWith(expectedText));
    }
View Full Code Here
104
105
106
107
108
109
110
111
    super.setUp();

    this.urlString = createUrl(this.testFile.getName());

    System.out.println("Testing file: " + this.urlString + "...");
    this.protocol =new ProtocolFactory(NutchConfiguration.create()).getProtocol(this.urlString);
    this.content = this.protocol.getProtocolOutput(new UTF8(this.urlString), new CrawlDatum()).getContent();
  }
View Full Code Here
TOP

Related Classes of org.apache.nutch.protocol.ProtocolFactory

  • org.apache.nutch.analysis.lang.LanguageIdentifier
  • org.apache.nutch.fetcher.FetcherJob
  • org.apache.nutch.fetcher.FetcherReducer$FetcherThread
  • org.apache.nutch.indexer.IndexingFiltersChecker
  • org.apache.nutch.parse.ext.TestExtParser
  • org.apache.nutch.parse.feed.TestFeedParser
  • org.apache.nutch.parse.html.TestMetatagParser
  • org.apache.nutch.parse.metatags.TestMetatagParser
  • org.apache.nutch.parse.mp3.TestMP3Parser
  • org.apache.nutch.parse.msexcel.TestMSExcelParser
    • 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.