Package org.apache.nutch.protocol

Examples of org.apache.nutch.protocol.Content

45
46
47
48
49
50
51
52
53
54
55
56
57
  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) {
      e.printStackTrace();
      Assert.fail(e.toString());
    }
View Full Code Here
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
      ftp.setMaxContentLength(maxContentLength);

    // set log level
    //LOG.setLevel(Level.parse((new String(logLevel)).toUpperCase()));

    Content content = ftp.getProtocolOutput(new Text(urlString), new CrawlDatum()).getContent();

    System.err.println("Content-Type: " + content.getContentType());
    System.err.println("Content-Length: " +
                       content.getMetadata().get(Response.CONTENT_LENGTH));
    System.err.println("Last-Modified: " +
                      content.getMetadata().get(Response.LAST_MODIFIED));
    if (dumpContent) {
      System.out.print(new String(content.getContent()));
    }

    ftp = null;
  }
View Full Code Here
68
69
70
71
72
73
74
75
76
  }

  public byte[] getContent() { return content; }

  public Content toContent() {
    return new Content(orig, base, (content != null ? content : EMPTY_CONTENT),
                       getHeader(Response.CONTENT_TYPE),
                       headers, this.conf);
  }
View Full Code Here
60
61
62
63
64
65
66
67
68
  }

  public String getTextContent(String fileName) throws ProtocolException, ParseException {
    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).parseByExtensionId("parse-tika", content).get(content.getUrl());
    return parse.getText();
  }
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
  @Test
  public void testIt() throws ProtocolException, ParseException {
    String urlString;
    Protocol protocol;
    Content content;
    Parse parse;

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

      Configuration conf = NutchConfiguration.create();
      protocol = new ProtocolFactory(conf).getProtocol(urlString);
      content = protocol.getProtocolOutput(new Text(urlString), new CrawlDatum()).getContent();
      parse = new ParseUtil(conf).parseByExtensionId("parse-tika", content).get(content.getUrl());

      Assert.assertEquals("121", parse.getData().getMeta("width"));
      Assert.assertEquals("48", parse.getData().getMeta("height"));
    }
  }
View Full Code Here
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
  @Test
  public void testIt() throws ProtocolException, ParseException {

    String urlString;
    Protocol protocol;
    Content content;
    Parse parse;

    Configuration conf = NutchConfiguration.create();
    urlString = "file:" + sampleDir + fileSeparator + rtfFile;
    protocol = new ProtocolFactory(conf).getProtocol(urlString);
    content = protocol.getProtocolOutput(new Text(urlString),
        new CrawlDatum()).getContent();
    parse = new ParseUtil(conf).parseByExtensionId("parse-tika", content)
        .get(content.getUrl());
    String text = parse.getText();
    Assert.assertEquals("The quick brown fox jumps over the lazy dog", text.trim());

    String title = parse.getData().getTitle();
    Metadata meta = parse.getData().getParseMeta();
View Full Code Here
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
  @Test
  public void testIt() throws ProtocolException, ParseException {
    String urlString;
    Protocol protocol;
    Content content;
    Parse parse;

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

      Configuration conf = NutchConfiguration.create();
      protocol = new ProtocolFactory(conf).getProtocol(urlString);
      content = protocol.getProtocolOutput(new Text(urlString), new CrawlDatum()).getContent();
      parse = new ParseUtil(conf).parseByExtensionId("parse-tika", content).get(content.getUrl());

      int index = parse.getText().indexOf(expectedText);
      Assert.assertTrue(index > 0);
    }
  }
View Full Code Here
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
   */
  @Test
  public void testIt() throws ProtocolException, ParseException {
    String urlString;
    Protocol protocol;
    Content content;
    Parse parse;

    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).parseByExtensionId("parse-tika",
          content).get(content.getUrl());

      // check that there are 2 outlinks:
      // unlike the original parse-rss
      // tika ignores the URL and description of the channel

View Full Code Here
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
    if (!output.getStatus().isSuccess()) {
      System.err.println("Fetch failed with protocol status: " + output.getStatus());
      return (-1);
    }
   
    Content content = output.getContent();

    if (content == null) {
      LOG.error("No content for " + url);
      return (-1);
    }

    if (force) {
      content.setContentType(contentType);
    } else {
      contentType = content.getContentType();
    }

    if (contentType == null) {
      LOG.error("Failed to determine content type!");
      return (-1);
View Full Code Here
118
119
120
121
122
123
124
125
126
127
128
    byte[] buf = new byte[in.available()];
    in.read(buf);
    in.close();
    SWFParser parser = new SWFParser();
    ParseResult parseResult = parser.getParse(new Content("file:" + args[0], "file:" + args[0],
                                          buf, "application/x-shockwave-flash",
                                          new Metadata(),
                                          NutchConfiguration.create()));
    Parse p = parseResult.get("file:" + args[0]);
    System.out.println("Parse Text:");
View Full Code Here
TOP

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

  • com.atlantbh.nutch.filter.xpath.XPathHtmlParserFilterTest
  • com.flaptor.hounder.crawler.Nutch9Fetcher$NutchSegment$SegmentIterator
  • nise.ExtractNutch
  • org.apache.commons.cli.Options
  • org.apache.hadoop.conf.Configuration
  • org.apache.hadoop.fs.FileSystem
  • org.apache.hadoop.fs.Path
  • org.apache.hadoop.util.GenericOptionsParser
  • org.apache.nutch.analysis.lang.LanguageIdentifier
  • org.apache.nutch.analysis.lang.TestHTMLLanguageParser
    • 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.