StripUserinfoRule


1   /* StripUserinfoRule
2    * 
3    * Created on Oct 5, 2004
4    *
5    * Copyright (C) 2004 Internet Archive.
6    * 
7    * This file is part of the Heritrix web crawler (crawler.archive.org).
8    * 
9    * Heritrix is free software; you can redistribute it and/or modify
10   * it under the terms of the GNU Lesser Public License as published by
11   * the Free Software Foundation; either version 2.1 of the License, or
12   * any later version.
13   * 
14   * Heritrix is distributed in the hope that it will be useful, 
15   * but WITHOUT ANY WARRANTY; without even the implied warranty of
16   * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
17   * GNU Lesser Public License for more details.
18   * 
19   * You should have received a copy of the GNU Lesser Public License
20   * along with Heritrix; if not, write to the Free Software
21   * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
22   */
23  package org.archive.crawler.url.canonicalize;
24  
25  import java.util.regex.Pattern  ;
26  
27  
28  
29  /**
30   * Strip any 'userinfo' found on http/https URLs.
31   * @author stack
32   * @version $Date: 2007/01/13 01:31:29 $, $Revision: 1.3.18.1 $
33   */
34  public class StripUserinfoRule extends BaseRule {
35  
36      private static final long serialVersionUID = -4271062607638914996L;
37  
38      private static final String   DESCRIPTION = "Strip any 'userinfo' found. " +
39          "Use this rule to equate 'http://stack:psswrd@archive.org/index.htm'" + 
40          " and 'http://archive.org/index.htm'. The resulting canonicalization" +
41          " returns 'http://archive.org/index.htm'. Removes any userinfo" +
42          " found.  Operates on http/https/ftp/ftps schemes only.";
43      
44      /**
45       * Strip userinfo.
46       */
47      private static final Pattern   REGEX =
48          Pattern.compile("^((?:(?:https?)|(?:ftps?))://)(?:[^/]+@)(.*)$",
49              Pattern.CASE_INSENSITIVE);
50  
51      public StripUserinfoRule(String   name) {
52          super(name, DESCRIPTION);
53      }
54  
55      public String   canonicalize(String   url, Object   context) {
56          return doStripRegexMatch(url, REGEX.matcher(url));
57      }
58  }
59
A to Z: JavaDoc & Examples Daily Java News & Articles Open Source Projects Open Source Codes Free Computer Books Remove Frame
Popular Tags