Package placeholder.website

Examples of placeholder.website.WebsiteResult

30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
  @Override
  protected boolean __fill() throws WebsiteResultQueueFlushException {
    HashMap<String, String> postFields = new HashMap<String, String>();
    try {
      WebsiteResult result;
      while ((result = this.queue.removeFirst()) != null) {
        // http status
        postFields.put("websites[" + result.getWebsite() + "][httpStatus]", new Integer(result.getHTTPStatus()).toString());

        // check time
        postFields.put("websites[" + result.getWebsite() + "][checkTime]", new Integer(result.getCheckTime()).toString());
       
        // websites
        String[] websites = result.getWebsites();
        for (int i = 0; i < websites.length; i++) {
          postFields.put("websites[" + result.getWebsite() + "][websites][" + i + "]", websites[i]);
        }
       
        // services
        String[] services = result.getServices();
        for (int i = 0; i < services.length; i++) {
          postFields.put("websites[" + result.getWebsite() + "][services][" + i + "]", services[i]);
        }
      }
    } catch (NoSuchElementException e) {
      // do nothing
    }
View Full Code Here
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
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
        // start request
        int responseCode = 0;
        String redirectLocation = null;
        String responseContent = null;
        WebsiteResult result;
        try {
          this.out.titledPrint("\tStarting request...");
         
          String website2 = "http://" + website;
          do {
            HttpRequest request = new HttpRequest(website2);
            request.send();
            responseCode = request.getConnection().getResponseCode();
            redirectLocation = request.getConnection().getHeaderField("Location");
            responseContent = request.getResponseBody();
           
            // follow redirect?
            try {
              URL url = new URL(redirectLocation);
              if (website.equals(url.getHost())) {
                website2 = redirectLocation;
              }
              else {
                website2 = null;
              }
            } catch (MalformedURLException e) {
              website2 = null;
            }
          } while (website2 != null);
          this.out.println("finished!");
        } catch(Exception e) {
          responseCode = -1;
          responseContent = "";
        } finally {
          result = new WebsiteResult(website, responseCode);
          this.resultQueue.add(result);
        }
       
        if (responseCode == 200) {
          // look for other websites adresses
          this.out.titledPrint("\tLooking for other websites...");
          Matcher matcher = pattern.matcher(responseContent);
          while (matcher.find() && matcher.group(1).indexOf(".") != -1) {           
            try {
              String host = new URL(matcher.group(1)).getHost();
             
              // ignore ip addresses
              if (host.matches("\\d{1,3}(\\.\\d{1,3}){3}")) {
                continue;
              }
             
              // ignore sub domains
              String[] parts = host.split("\\.", 5);
              String topLevel = parts[parts.length - 1];
              int maxParts = (topLevel.matches("at|be|biz|ca|cc|ch|com|cz|de|dk|fr|edu|es|eu|in|int|it|info|li|mobi|name|net|nl|org|ru|se|tv|us|ws") ? 3 : 4);

              if (parts.length > maxParts || parts.length == maxParts && !parts[0].equals("www")) {
                // remove lowed namespace
                StringBuilder builder = new StringBuilder();
                for (int i = parts.length - maxParts + 1; i < parts.length; i++) {
                  if (builder.length() > 0) builder.append(".");
                  builder.append(parts[i]);
                }
                host = builder.toString();
              }
             
              result.addWebsite(host);
            } catch (MalformedURLException e) {
              // ignore bad urls
              continue;
            }
          }
          this.out.println(result.getWebsites().length + " found.");
 
          // look for services
          this.out.titledPrint("\tLooking for services...");
          for (String service: CheckServices.check(responseContent)) {
            result.addService(service);
          }
          this.out.println(result.getServices().length + " found.");
        }
        else if (redirectLocation != null) {
          try {
            URL url = new URL(redirectLocation);
            result.addWebsite(url.getHost());
            this.out.titledPrintln("\tFound redirect to website: " + url.getHost());
          } catch (MalformedURLException e) {
            // ignore bad urls
            continue;
          }
View Full Code Here
TOP

Related Classes of placeholder.website.WebsiteResult

  • placeholder.Crawler
  • placeholder.website.queue.http.HttpWebsiteResultQueue
  • java.util.Date
    • 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.