Package net.matuschek.spider

Examples of net.matuschek.spider.WebRobot

22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
  public static void main(String[] args)
    throws Exception
  {
    System.out.println("URLs will be logged to urls.txt\n\n");

    WebRobot robby = new WebRobot();
    robby.setStartURL(new URL("http://www.matuschek.net"));
    robby.setMaxDepth(1);
    robby.setSleepTime(0);

    FileWriter logfile = new FileWriter("urls.txt");
    URLLogger log = new URLLogger(logfile);
    robby.setDocManager(log);

    robby.run();
    logfile.close();
  }
View Full Code Here
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
  }



  public void run() throws Exception {
    WebRobot robby = new WebRobot();
    robby.setStartURL(new URL("http://www.matuschek.net"));
    robby.setMaxDepth(1);
    robby.setSleepTime(0);

    // download only the first 5 documents
    DownloadStopper stopit = new DownloadStopper(5,robby);
    robby.setWebRobotCallback(stopit);

    robby.run();
  }
View Full Code Here
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
    // get command line options
    GetOpt opt = new GetOpt(argv);
    String option = null;

    JoBoBase jobobase = JoBoBase.createFromXML();
    WebRobot robby = jobobase.getRobot();

    // referer
    option=opt.getOptionString("r");
    if (option != null) {
      robby.setStartReferer(option);
    }
   
    // maximal depth
    option=opt.getOptionString("d");
    if (option != null) {
  try {
    int maxDepth=Integer.parseInt(option);
    robby.setMaxDepth(maxDepth);
  } catch (NumberFormatException e) {
    System.out.println("Wrong number for maxDepth: "+option);
  }
    }
   
    // walk to other hosts ?
    if (opt.getOptionBoolean("o")) {
      robby.setWalkToOtherHosts(true);
    }
   
    // store directory
    option=opt.getOptionString("s");
    if (option != null) {
      basedir=option;
    }
   
    // minimal file size
    option=opt.getOptionString("m");
    if (option != null) {
      try {
  minSize=Integer.parseInt(option);
      } catch (NumberFormatException e) {}
    }
   
    // agent name
    option=opt.getOptionString("a");
    if (option != null) {
      robby.setAgentName(option);
    }
   
    // ignore robots.txt
    if (opt.getOptionBoolean("i")) {
      robby.setIgnoreRobotsTxt(true);
    }
   
    // wait time
    option=opt.getOptionString("w");
    if (option != null) {
      try {
  int waitTime=Integer.parseInt(option);
  robby.setSleepTime(waitTime*1000);
      } catch (NumberFormatException e) {}
    }
   
    // print usage
    if (opt.getOptionBoolean("?")) {
      printUsage();
      return;
    }
   
    URL u = new URL(argv[argv.length-1]);

    HttpDocToFile docStore=new HttpDocToFile(basedir);
    docStore.setMinFileSize(minSize);

    SystemOutHttpToolCallback statusInfo = new SystemOutHttpToolCallback();

    robby.setStartURL(u);
    robby.setDocManager(docStore);
    robby.setHttpToolCallback(statusInfo);
   
    robby.run();
   
  }
View Full Code Here
67
68
69
70
71
72
73
74
    throws ClassNotFoundException
  {
    log = Category.getInstance(this.getClass());
    docstore = new HttpDocToFile(storageDirectory);
    initializeFilters();
    robot = new WebRobot();
    robot.setFilters(filters);
  }
View Full Code Here
TOP

Related Classes of net.matuschek.spider.WebRobot

  • net.matuschek.examples.LogURL
  • net.matuschek.examples.StopDownload
  • net.matuschek.html.HtmlDocument
  • net.matuschek.http.DownloadRuleSet
  • net.matuschek.http.ExtendedURL
  • net.matuschek.http.HttpDoc
  • net.matuschek.http.HttpHeader
  • net.matuschek.jobo.JoBo
  • net.matuschek.jobo.JoBoBase
  • java.util.StringTokenizer
    • 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.