spider 0.4.3 → 0.4.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (66) hide show
  1. data/CHANGES +6 -0
  2. data/README +3 -3
  3. data/doc/classes/BeStaticServerPages.html +197 -0
  4. data/doc/classes/BeStaticServerPages.src/M000030.html +19 -0
  5. data/doc/classes/BeStaticServerPages.src/M000031.html +19 -0
  6. data/doc/classes/BeStaticServerPages.src/M000032.html +18 -0
  7. data/doc/classes/BeStaticServerPages.src/M000033.html +18 -0
  8. data/doc/classes/IncludedInMemcached.html +18 -45
  9. data/doc/classes/IncludedInMemcached.src/M000015.html +18 -0
  10. data/doc/classes/IncludedInMemcached.src/M000016.html +18 -0
  11. data/doc/classes/IncludedInMemcached.src/M000017.html +18 -0
  12. data/doc/classes/LoopingServlet.html +137 -0
  13. data/doc/classes/LoopingServlet.src/M000037.html +23 -0
  14. data/doc/classes/NextUrlsInSQS.html +204 -0
  15. data/doc/classes/NextUrlsInSQS.src/M000018.html +19 -0
  16. data/doc/classes/NextUrlsInSQS.src/M000019.html +22 -0
  17. data/doc/classes/NextUrlsInSQS.src/M000020.html +19 -0
  18. data/doc/classes/QueryServlet.html +137 -0
  19. data/doc/classes/QueryServlet.src/M000038.html +19 -0
  20. data/doc/classes/RobotRules.html +175 -0
  21. data/doc/classes/RobotRules.src/M000034.html +19 -0
  22. data/doc/classes/RobotRules.src/M000035.html +67 -0
  23. data/doc/classes/RobotRules.src/M000036.html +24 -0
  24. data/doc/classes/Spider.html +5 -17
  25. data/doc/classes/Spider.src/M000029.html +21 -0
  26. data/doc/classes/SpiderInstance.html +72 -108
  27. data/doc/classes/SpiderInstance.src/M000021.html +18 -0
  28. data/doc/classes/SpiderInstance.src/M000022.html +22 -0
  29. data/doc/classes/SpiderInstance.src/M000023.html +22 -0
  30. data/doc/classes/SpiderInstance.src/M000024.html +24 -0
  31. data/doc/classes/SpiderInstance.src/M000025.html +18 -0
  32. data/doc/classes/SpiderInstance.src/M000026.html +18 -0
  33. data/doc/classes/SpiderInstance.src/M000027.html +18 -0
  34. data/doc/classes/SpiderInstance.src/M000028.html +18 -0
  35. data/doc/created.rid +1 -1
  36. data/doc/files/lib/spider/included_in_memcached_rb.html +29 -1
  37. data/doc/files/lib/spider/next_urls_in_sqs_rb.html +144 -0
  38. data/doc/files/lib/spider/robot_rules_rb.html +114 -0
  39. data/doc/files/lib/spider/spider_instance_rb.html +1 -2
  40. data/doc/files/lib/spider_rb.html +40 -9
  41. data/doc/files/spec/spec_helper_rb.html +196 -0
  42. data/doc/files/spec/spec_helper_rb.src/M000001.html +20 -0
  43. data/doc/files/spec/spec_helper_rb.src/M000002.html +26 -0
  44. data/doc/files/spec/spec_helper_rb.src/M000003.html +24 -0
  45. data/doc/files/spec/spec_helper_rb.src/M000004.html +18 -0
  46. data/doc/files/spec/spec_helper_rb.src/M000005.html +23 -0
  47. data/doc/files/spec/spider/included_in_memcached_spec_rb.html +142 -0
  48. data/doc/files/spec/spider/included_in_memcached_spec_rb.src/M000006.html +19 -0
  49. data/doc/files/spec/spider/included_in_memcached_spec_rb.src/M000007.html +18 -0
  50. data/doc/files/spec/spider/spider_instance_spec_rb.html +210 -0
  51. data/doc/files/spec/spider/spider_instance_spec_rb.src/M000008.html +21 -0
  52. data/doc/files/spec/spider/spider_instance_spec_rb.src/M000009.html +19 -0
  53. data/doc/files/spec/spider/spider_instance_spec_rb.src/M000010.html +19 -0
  54. data/doc/files/spec/spider/spider_instance_spec_rb.src/M000011.html +27 -0
  55. data/doc/files/spec/spider/spider_instance_spec_rb.src/M000012.html +26 -0
  56. data/doc/files/spec/spider/spider_instance_spec_rb.src/M000013.html +27 -0
  57. data/doc/files/spec/spider_spec_rb.html +127 -0
  58. data/doc/files/spec/spider_spec_rb.src/M000014.html +23 -0
  59. data/doc/fr_class_index.html +5 -0
  60. data/doc/fr_file_index.html +6 -1
  61. data/doc/fr_method_index.html +38 -11
  62. data/doc/index.html +1 -1
  63. data/lib/spider/spider_instance.rb +15 -7
  64. data/spider.gemspec +1 -1
  65. metadata +84 -22
  66. data/lib/test.rb +0 -27
data/doc/classes/NextUrlsInSQS.src/M000020.html ADDED
@@ -0,0 +1,19 @@
1
+ <?xml version="1.0" encoding="iso-8859-1"?>
2
+ <!DOCTYPE html
3
+ PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
4
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
5
+
6
+ <html>
7
+ <head>
8
+ <title>push (NextUrlsInSQS)</title>
9
+ <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
10
+ <link rel="stylesheet" href="../.././rdoc-style.css" type="text/css" media="screen" />
11
+ </head>
12
+ <body class="standalone-code">
13
+ <pre><span class="ruby-comment cmt"># File lib/spider/next_urls_in_sqs.rb, line 62</span>
14
+ <span class="ruby-keyword kw">def</span> <span class="ruby-identifier">push</span>(<span class="ruby-identifier">a_msg</span>)
15
+ <span class="ruby-identifier">encoded_message</span> = <span class="ruby-constant">YAML</span><span class="ruby-operator">::</span><span class="ruby-identifier">dump</span>(<span class="ruby-identifier">a_msg</span>)
16
+ <span class="ruby-ivar">@queue</span>.<span class="ruby-identifier">push</span>(<span class="ruby-identifier">a_msg</span>)
17
+ <span class="ruby-keyword kw">end</span></pre>
18
+ </body>
19
+ </html>
data/doc/classes/QueryServlet.html ADDED
@@ -0,0 +1,137 @@
1
+ <?xml version="1.0" encoding="iso-8859-1"?>
2
+ <!DOCTYPE html
3
+ PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
4
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
5
+
6
+ <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
7
+ <head>
8
+ <title>Class: QueryServlet</title>
9
+ <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
10
+ <meta http-equiv="Content-Script-Type" content="text/javascript" />
11
+ <link rel="stylesheet" href=".././rdoc-style.css" type="text/css" media="screen" />
12
+ <script type="text/javascript">
13
+ // <![CDATA[
14
+
15
+ function popupCode( url ) {
16
+ window.open(url, "Code", "resizable=yes,scrollbars=yes,toolbar=no,status=no,height=150,width=400")
17
+ }
18
+
19
+ function toggleCode( id ) {
20
+ if ( document.getElementById )
21
+ elem = document.getElementById( id );
22
+ else if ( document.all )
23
+ elem = eval( "document.all." + id );
24
+ else
25
+ return false;
26
+
27
+ elemStyle = elem.style;
28
+
29
+ if ( elemStyle.display != "block" ) {
30
+ elemStyle.display = "block"
31
+ } else {
32
+ elemStyle.display = "none"
33
+ }
34
+
35
+ return true;
36
+ }
37
+
38
+ // Make codeblocks hidden by default
39
+ document.writeln( "<style type=\"text/css\">div.method-source-code { display: none }</style>" )
40
+
41
+ // ]]>
42
+ </script>
43
+
44
+ </head>
45
+ <body>
46
+
47
+
48
+
49
+ <div id="classHeader">
50
+ <table class="header-table">
51
+ <tr class="top-aligned-row">
52
+ <td><strong>Class</strong></td>
53
+ <td class="class-name-in-header">QueryServlet</td>
54
+ </tr>
55
+ <tr class="top-aligned-row">
56
+ <td><strong>In:</strong></td>
57
+ <td>
58
+ <a href="../files/spec/spec_helper_rb.html">
59
+ spec/spec_helper.rb
60
+ </a>
61
+ <br />
62
+ </td>
63
+ </tr>
64
+
65
+ <tr class="top-aligned-row">
66
+ <td><strong>Parent:</strong></td>
67
+ <td>
68
+ WEBrick::HTTPServlet::AbstractServlet
69
+ </td>
70
+ </tr>
71
+ </table>
72
+ </div>
73
+ <!-- banner header -->
74
+
75
+ <div id="bodyContent">
76
+
77
+
78
+
79
+ <div id="contextContent">
80
+
81
+
82
+
83
+ </div>
84
+
85
+ <div id="method-list">
86
+ <h3 class="section-bar">Methods</h3>
87
+
88
+ <div class="name-list">
89
+ <a href="#M000038">do_GET</a>&nbsp;&nbsp;
90
+ </div>
91
+ </div>
92
+
93
+ </div>
94
+
95
+
96
+ <!-- if includes -->
97
+
98
+ <div id="section">
99
+
100
+
101
+
102
+
103
+
104
+
105
+
106
+
107
+ <!-- if method_list -->
108
+ <div id="methods">
109
+ <h3 class="section-bar">Public Instance methods</h3>
110
+
111
+ <div id="method-M000038" class="method-detail">
112
+ <a name="M000038"></a>
113
+
114
+ <div class="method-heading">
115
+ <a href="QueryServlet.src/M000038.html" target="Code" class="method-signature"
116
+ onclick="popupCode('QueryServlet.src/M000038.html');return false;">
117
+ <span class="method-name">do_GET</span><span class="method-args">(req, res)</span>
118
+ </a>
119
+ </div>
120
+
121
+ <div class="method-description">
122
+ </div>
123
+ </div>
124
+
125
+
126
+ </div>
127
+
128
+
129
+ </div>
130
+
131
+
132
+ <div id="validator-badges">
133
+ <p><small><a href="http://validator.w3.org/check/referer">[Validate]</a></small></p>
134
+ </div>
135
+
136
+ </body>
137
+ </html>
data/doc/classes/QueryServlet.src/M000038.html ADDED
@@ -0,0 +1,19 @@
1
+ <?xml version="1.0" encoding="iso-8859-1"?>
2
+ <!DOCTYPE html
3
+ PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
4
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
5
+
6
+ <html>
7
+ <head>
8
+ <title>do_GET (QueryServlet)</title>
9
+ <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
10
+ <link rel="stylesheet" href="../.././rdoc-style.css" type="text/css" media="screen" />
11
+ </head>
12
+ <body class="standalone-code">
13
+ <pre><span class="ruby-comment cmt"># File spec/spec_helper.rb, line 62</span>
14
+ <span class="ruby-keyword kw">def</span> <span class="ruby-identifier">do_GET</span>(<span class="ruby-identifier">req</span>, <span class="ruby-identifier">res</span>)
15
+ <span class="ruby-identifier">res</span>[<span class="ruby-value str">'Content-type'</span>] = <span class="ruby-value str">'text/plain'</span>
16
+ <span class="ruby-identifier">res</span>.<span class="ruby-identifier">body</span> = <span class="ruby-value str">&quot;response\n&quot;</span>
17
+ <span class="ruby-keyword kw">end</span></pre>
18
+ </body>
19
+ </html>
data/doc/classes/RobotRules.html ADDED
@@ -0,0 +1,175 @@
1
+ <?xml version="1.0" encoding="iso-8859-1"?>
2
+ <!DOCTYPE html
3
+ PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
4
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
5
+
6
+ <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
7
+ <head>
8
+ <title>Class: RobotRules</title>
9
+ <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
10
+ <meta http-equiv="Content-Script-Type" content="text/javascript" />
11
+ <link rel="stylesheet" href=".././rdoc-style.css" type="text/css" media="screen" />
12
+ <script type="text/javascript">
13
+ // <![CDATA[
14
+
15
+ function popupCode( url ) {
16
+ window.open(url, "Code", "resizable=yes,scrollbars=yes,toolbar=no,status=no,height=150,width=400")
17
+ }
18
+
19
+ function toggleCode( id ) {
20
+ if ( document.getElementById )
21
+ elem = document.getElementById( id );
22
+ else if ( document.all )
23
+ elem = eval( "document.all." + id );
24
+ else
25
+ return false;
26
+
27
+ elemStyle = elem.style;
28
+
29
+ if ( elemStyle.display != "block" ) {
30
+ elemStyle.display = "block"
31
+ } else {
32
+ elemStyle.display = "none"
33
+ }
34
+
35
+ return true;
36
+ }
37
+
38
+ // Make codeblocks hidden by default
39
+ document.writeln( "<style type=\"text/css\">div.method-source-code { display: none }</style>" )
40
+
41
+ // ]]>
42
+ </script>
43
+
44
+ </head>
45
+ <body>
46
+
47
+
48
+
49
+ <div id="classHeader">
50
+ <table class="header-table">
51
+ <tr class="top-aligned-row">
52
+ <td><strong>Class</strong></td>
53
+ <td class="class-name-in-header">RobotRules</td>
54
+ </tr>
55
+ <tr class="top-aligned-row">
56
+ <td><strong>In:</strong></td>
57
+ <td>
58
+ <a href="../files/lib/spider/robot_rules_rb.html">
59
+ lib/spider/robot_rules.rb
60
+ </a>
61
+ <br />
62
+ </td>
63
+ </tr>
64
+
65
+ <tr class="top-aligned-row">
66
+ <td><strong>Parent:</strong></td>
67
+ <td>
68
+ Object
69
+ </td>
70
+ </tr>
71
+ </table>
72
+ </div>
73
+ <!-- banner header -->
74
+
75
+ <div id="bodyContent">
76
+
77
+
78
+
79
+ <div id="contextContent">
80
+
81
+ <div id="description">
82
+ <p>
83
+ Based on Perl&#8216;s WWW::RobotRules module, by Gisle Aas.
84
+ </p>
85
+
86
+ </div>
87
+
88
+
89
+ </div>
90
+
91
+ <div id="method-list">
92
+ <h3 class="section-bar">Methods</h3>
93
+
94
+ <div class="name-list">
95
+ <a href="#M000036">allowed?</a>&nbsp;&nbsp;
96
+ <a href="#M000034">new</a>&nbsp;&nbsp;
97
+ <a href="#M000035">parse</a>&nbsp;&nbsp;
98
+ </div>
99
+ </div>
100
+
101
+ </div>
102
+
103
+
104
+ <!-- if includes -->
105
+
106
+ <div id="section">
107
+
108
+
109
+
110
+
111
+
112
+
113
+
114
+
115
+ <!-- if method_list -->
116
+ <div id="methods">
117
+ <h3 class="section-bar">Public Class methods</h3>
118
+
119
+ <div id="method-M000034" class="method-detail">
120
+ <a name="M000034"></a>
121
+
122
+ <div class="method-heading">
123
+ <a href="RobotRules.src/M000034.html" target="Code" class="method-signature"
124
+ onclick="popupCode('RobotRules.src/M000034.html');return false;">
125
+ <span class="method-name">new</span><span class="method-args">( user_agent )</span>
126
+ </a>
127
+ </div>
128
+
129
+ <div class="method-description">
130
+ </div>
131
+ </div>
132
+
133
+ <h3 class="section-bar">Public Instance methods</h3>
134
+
135
+ <div id="method-M000036" class="method-detail">
136
+ <a name="M000036"></a>
137
+
138
+ <div class="method-heading">
139
+ <a href="RobotRules.src/M000036.html" target="Code" class="method-signature"
140
+ onclick="popupCode('RobotRules.src/M000036.html');return false;">
141
+ <span class="method-name">allowed?</span><span class="method-args">( text_uri )</span>
142
+ </a>
143
+ </div>
144
+
145
+ <div class="method-description">
146
+ </div>
147
+ </div>
148
+
149
+ <div id="method-M000035" class="method-detail">
150
+ <a name="M000035"></a>
151
+
152
+ <div class="method-heading">
153
+ <a href="RobotRules.src/M000035.html" target="Code" class="method-signature"
154
+ onclick="popupCode('RobotRules.src/M000035.html');return false;">
155
+ <span class="method-name">parse</span><span class="method-args">( text_uri, robots_data )</span>
156
+ </a>
157
+ </div>
158
+
159
+ <div class="method-description">
160
+ </div>
161
+ </div>
162
+
163
+
164
+ </div>
165
+
166
+
167
+ </div>
168
+
169
+
170
+ <div id="validator-badges">
171
+ <p><small><a href="http://validator.w3.org/check/referer">[Validate]</a></small></p>
172
+ </div>
173
+
174
+ </body>
175
+ </html>
data/doc/classes/RobotRules.src/M000034.html ADDED
@@ -0,0 +1,19 @@
1
+ <?xml version="1.0" encoding="iso-8859-1"?>
2
+ <!DOCTYPE html
3
+ PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
4
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
5
+
6
+ <html>
7
+ <head>
8
+ <title>new (RobotRules)</title>
9
+ <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
10
+ <link rel="stylesheet" href="../.././rdoc-style.css" type="text/css" media="screen" />
11
+ </head>
12
+ <body class="standalone-code">
13
+ <pre><span class="ruby-comment cmt"># File lib/spider/robot_rules.rb, line 10</span>
14
+ <span class="ruby-keyword kw">def</span> <span class="ruby-identifier">initialize</span>( <span class="ruby-identifier">user_agent</span> )
15
+ <span class="ruby-ivar">@user_agent</span> = <span class="ruby-identifier">user_agent</span>.<span class="ruby-identifier">scan</span>(<span class="ruby-regexp re">/\S+/</span>).<span class="ruby-identifier">first</span>.<span class="ruby-identifier">sub</span>(<span class="ruby-regexp re">%r{/.*}</span>, <span class="ruby-value str">&quot;&quot;</span>).<span class="ruby-identifier">downcase</span>
16
+ <span class="ruby-ivar">@rules</span> = <span class="ruby-constant">Hash</span>.<span class="ruby-identifier">new</span> { <span class="ruby-operator">|</span><span class="ruby-identifier">rules</span>, <span class="ruby-identifier">rule</span><span class="ruby-operator">|</span> <span class="ruby-identifier">rules</span>[<span class="ruby-identifier">rule</span>] = <span class="ruby-constant">Array</span>.<span class="ruby-identifier">new</span> }
17
+ <span class="ruby-keyword kw">end</span></pre>
18
+ </body>
19
+ </html>
data/doc/classes/RobotRules.src/M000035.html ADDED
@@ -0,0 +1,67 @@
1
+ <?xml version="1.0" encoding="iso-8859-1"?>
2
+ <!DOCTYPE html
3
+ PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
4
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
5
+
6
+ <html>
7
+ <head>
8
+ <title>parse (RobotRules)</title>
9
+ <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
10
+ <link rel="stylesheet" href="../.././rdoc-style.css" type="text/css" media="screen" />
11
+ </head>
12
+ <body class="standalone-code">
13
+ <pre><span class="ruby-comment cmt"># File lib/spider/robot_rules.rb, line 15</span>
14
+ <span class="ruby-keyword kw">def</span> <span class="ruby-identifier">parse</span>( <span class="ruby-identifier">text_uri</span>, <span class="ruby-identifier">robots_data</span> )
15
+ <span class="ruby-identifier">uri</span> = <span class="ruby-constant">URI</span>.<span class="ruby-identifier">parse</span>(<span class="ruby-identifier">text_uri</span>)
16
+ <span class="ruby-identifier">location</span> = <span class="ruby-node">&quot;#{uri.host}:#{uri.port}&quot;</span>
17
+ <span class="ruby-ivar">@rules</span>.<span class="ruby-identifier">delete</span>(<span class="ruby-identifier">location</span>)
18
+
19
+ <span class="ruby-identifier">rules</span> = <span class="ruby-identifier">robots_data</span>.<span class="ruby-identifier">split</span>(<span class="ruby-regexp re">/[\015\012]+/</span>).<span class="ruby-identifier">map</span> <span class="ruby-keyword kw">do</span> <span class="ruby-operator">|</span><span class="ruby-identifier">rule</span><span class="ruby-operator">|</span>
20
+ <span class="ruby-identifier">rule</span>.<span class="ruby-identifier">sub</span>(<span class="ruby-regexp re">/\s*#.*$/</span>, <span class="ruby-value str">&quot;&quot;</span>)
21
+ <span class="ruby-keyword kw">end</span>
22
+ <span class="ruby-identifier">anon_rules</span> = <span class="ruby-constant">Array</span>.<span class="ruby-identifier">new</span>
23
+ <span class="ruby-identifier">my_rules</span> = <span class="ruby-constant">Array</span>.<span class="ruby-identifier">new</span>
24
+ <span class="ruby-identifier">current</span> = <span class="ruby-identifier">anon_rules</span>
25
+ <span class="ruby-identifier">rules</span>.<span class="ruby-identifier">each</span> <span class="ruby-keyword kw">do</span> <span class="ruby-operator">|</span><span class="ruby-identifier">rule</span><span class="ruby-operator">|</span>
26
+ <span class="ruby-keyword kw">case</span> <span class="ruby-identifier">rule</span>
27
+ <span class="ruby-keyword kw">when</span> <span class="ruby-regexp re">/^\s*User-Agent\s*:\s*(.+?)\s*$/i</span>
28
+ <span class="ruby-keyword kw">break</span> <span class="ruby-keyword kw">unless</span> <span class="ruby-identifier">my_rules</span>.<span class="ruby-identifier">empty?</span>
29
+
30
+ <span class="ruby-identifier">current</span> = <span class="ruby-keyword kw">if</span> <span class="ruby-identifier">$1</span> <span class="ruby-operator">==</span> <span class="ruby-value str">&quot;*&quot;</span>
31
+ <span class="ruby-identifier">anon_rules</span>
32
+ <span class="ruby-keyword kw">elsif</span> <span class="ruby-identifier">$1</span>.<span class="ruby-identifier">downcase</span>.<span class="ruby-identifier">index</span>(<span class="ruby-ivar">@user_agent</span>)
33
+ <span class="ruby-identifier">my_rules</span>
34
+ <span class="ruby-keyword kw">else</span>
35
+ <span class="ruby-keyword kw">nil</span>
36
+ <span class="ruby-keyword kw">end</span>
37
+ <span class="ruby-keyword kw">when</span> <span class="ruby-regexp re">/^\s*Disallow\s*:\s*(.*?)\s*$/i</span>
38
+ <span class="ruby-keyword kw">next</span> <span class="ruby-keyword kw">if</span> <span class="ruby-identifier">current</span>.<span class="ruby-identifier">nil?</span>
39
+
40
+ <span class="ruby-keyword kw">if</span> <span class="ruby-identifier">$1</span>.<span class="ruby-identifier">empty?</span>
41
+ <span class="ruby-identifier">current</span> <span class="ruby-operator">&lt;&lt;</span> <span class="ruby-keyword kw">nil</span>
42
+ <span class="ruby-keyword kw">else</span>
43
+ <span class="ruby-identifier">disallow</span> = <span class="ruby-constant">URI</span>.<span class="ruby-identifier">parse</span>(<span class="ruby-identifier">$1</span>)
44
+
45
+ <span class="ruby-keyword kw">next</span> <span class="ruby-keyword kw">unless</span> <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">scheme</span>.<span class="ruby-identifier">nil?</span> <span class="ruby-keyword kw">or</span> <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">scheme</span> <span class="ruby-operator">==</span>
46
+ <span class="ruby-identifier">uri</span>.<span class="ruby-identifier">scheme</span>
47
+ <span class="ruby-keyword kw">next</span> <span class="ruby-keyword kw">unless</span> <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">port</span>.<span class="ruby-identifier">nil?</span> <span class="ruby-keyword kw">or</span> <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">port</span> <span class="ruby-operator">==</span> <span class="ruby-identifier">uri</span>.<span class="ruby-identifier">port</span>
48
+ <span class="ruby-keyword kw">next</span> <span class="ruby-keyword kw">unless</span> <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">host</span>.<span class="ruby-identifier">nil?</span> <span class="ruby-keyword kw">or</span>
49
+ <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">host</span>.<span class="ruby-identifier">downcase</span> <span class="ruby-operator">==</span> <span class="ruby-identifier">uri</span>.<span class="ruby-identifier">host</span>.<span class="ruby-identifier">downcase</span>
50
+
51
+ <span class="ruby-identifier">disallow</span> = <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">path</span>
52
+ <span class="ruby-identifier">disallow</span> = <span class="ruby-value str">&quot;/&quot;</span> <span class="ruby-keyword kw">if</span> <span class="ruby-identifier">disallow</span>.<span class="ruby-identifier">empty?</span>
53
+ <span class="ruby-identifier">disallow</span> = <span class="ruby-node">&quot;/#{disallow}&quot;</span> <span class="ruby-keyword kw">unless</span> <span class="ruby-identifier">disallow</span>[<span class="ruby-value">0</span>] <span class="ruby-operator">==</span> <span class="ruby-value">?/</span>
54
+
55
+ <span class="ruby-identifier">current</span> <span class="ruby-operator">&lt;&lt;</span> <span class="ruby-identifier">disallow</span>
56
+ <span class="ruby-keyword kw">end</span>
57
+ <span class="ruby-keyword kw">end</span>
58
+ <span class="ruby-keyword kw">end</span>
59
+
60
+ <span class="ruby-ivar">@rules</span>[<span class="ruby-identifier">location</span>] = <span class="ruby-keyword kw">if</span> <span class="ruby-identifier">my_rules</span>.<span class="ruby-identifier">empty?</span>
61
+ <span class="ruby-identifier">anon_rules</span>.<span class="ruby-identifier">compact</span>
62
+ <span class="ruby-keyword kw">else</span>
63
+ <span class="ruby-identifier">my_rules</span>.<span class="ruby-identifier">compact</span>
64
+ <span class="ruby-keyword kw">end</span>
65
+ <span class="ruby-keyword kw">end</span></pre>
66
+ </body>
67
+ </html>
data/doc/classes/RobotRules.src/M000036.html ADDED
@@ -0,0 +1,24 @@
1
+ <?xml version="1.0" encoding="iso-8859-1"?>
2
+ <!DOCTYPE html
3
+ PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
4
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
5
+
6
+ <html>
7
+ <head>
8
+ <title>allowed? (RobotRules)</title>
9
+ <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
10
+ <link rel="stylesheet" href="../.././rdoc-style.css" type="text/css" media="screen" />
11
+ </head>
12
+ <body class="standalone-code">
13
+ <pre><span class="ruby-comment cmt"># File lib/spider/robot_rules.rb, line 68</span>
14
+ <span class="ruby-keyword kw">def</span> <span class="ruby-identifier">allowed?</span>( <span class="ruby-identifier">text_uri</span> )
15
+ <span class="ruby-identifier">uri</span> = <span class="ruby-constant">URI</span>.<span class="ruby-identifier">parse</span>(<span class="ruby-identifier">text_uri</span>)
16
+ <span class="ruby-identifier">location</span> = <span class="ruby-node">&quot;#{uri.host}:#{uri.port}&quot;</span>
17
+ <span class="ruby-identifier">path</span> = <span class="ruby-identifier">uri</span>.<span class="ruby-identifier">path</span>
18
+
19
+ <span class="ruby-keyword kw">return</span> <span class="ruby-keyword kw">true</span> <span class="ruby-keyword kw">unless</span> <span class="ruby-node">%w{http https}</span>.<span class="ruby-identifier">include?</span>(<span class="ruby-identifier">uri</span>.<span class="ruby-identifier">scheme</span>)
20
+
21
+ <span class="ruby-keyword kw">not</span> <span class="ruby-ivar">@rules</span>[<span class="ruby-identifier">location</span>].<span class="ruby-identifier">any?</span> { <span class="ruby-operator">|</span><span class="ruby-identifier">rule</span><span class="ruby-operator">|</span> <span class="ruby-identifier">path</span>.<span class="ruby-identifier">index</span>(<span class="ruby-identifier">rule</span>) <span class="ruby-operator">==</span> <span class="ruby-value">0</span> }
22
+ <span class="ruby-keyword kw">end</span></pre>
23
+ </body>
24
+ </html>