fromgit-url2mhtml 0.0.2
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.
- data/ChangeLog +7 -0
- data/README +50 -0
- data/Rakefile +114 -0
- data/bin/url2mhtml +7 -0
- data/doc/classes/Url2mhtml.html +260 -0
- data/doc/classes/Url2mhtml.src/M000001.html +21 -0
- data/doc/classes/Url2mhtml.src/M000002.html +25 -0
- data/doc/classes/Url2mhtml.src/M000003.html +23 -0
- data/doc/classes/Url2mhtml.src/M000004.html +41 -0
- data/doc/classes/Url2mhtml.src/M000005.html +18 -0
- data/doc/classes/Url2mhtml.src/M000006.html +30 -0
- data/doc/classes/Url2mhtml.src/M000007.html +30 -0
- data/doc/classes/Url2mhtml.src/M000008.html +22 -0
- data/doc/classes/Url2mhtmlTest.html +167 -0
- data/doc/classes/Url2mhtmlTest.src/M000009.html +17 -0
- data/doc/classes/Url2mhtmlTest.src/M000010.html +17 -0
- data/doc/classes/Url2mhtmlTest.src/M000011.html +19 -0
- data/doc/created.rid +1 -0
- data/doc/files/lib/url2mhtml_rb.html +110 -0
- data/doc/files/test/test_helper_rb.html +108 -0
- data/doc/files/test/url2mhtml_test_rb.html +108 -0
- data/doc/fr_class_index.html +28 -0
- data/doc/fr_file_index.html +29 -0
- data/doc/fr_method_index.html +37 -0
- data/doc/index.html +24 -0
- data/doc/rdoc-style.css +208 -0
- data/lib/url2mhtml.rb +106 -0
- data/test/test_helper.rb +3 -0
- data/test/url2mhtml_test.rb +16 -0
- metadata +105 -0
data/ChangeLog
ADDED
data/README
ADDED
@@ -0,0 +1,50 @@
|
|
1
|
+
|
2
|
+
= url2mhtml
|
3
|
+
|
4
|
+
|
5
|
+
== Description
|
6
|
+
|
7
|
+
|
8
|
+
== Installation
|
9
|
+
|
10
|
+
=== Archive Installation
|
11
|
+
|
12
|
+
rake install
|
13
|
+
|
14
|
+
=== Gem Installation
|
15
|
+
|
16
|
+
gem sources -a http://gems.github.com
|
17
|
+
gem install fromgit-url2mhtml
|
18
|
+
|
19
|
+
|
20
|
+
== Features/Problems
|
21
|
+
|
22
|
+
|
23
|
+
== Synopsis
|
24
|
+
|
25
|
+
|
26
|
+
== Copyright
|
27
|
+
|
28
|
+
Author:: forgithubid <forgithubid@gmail.com>
|
29
|
+
Copyright:: Copyright (c) 2008 forgithubid
|
30
|
+
License:: The MIT License
|
31
|
+
|
32
|
+
Copyright (c) 2008 forgithubid
|
33
|
+
|
34
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
35
|
+
of this software and associated documentation files (the "Software"), to deal
|
36
|
+
in the Software without restriction, including without limitation the rights
|
37
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
38
|
+
copies of the Software, and to permit persons to whom the Software is
|
39
|
+
furnished to do so, subject to the following conditions:
|
40
|
+
|
41
|
+
The above copyright notice and this permission notice shall be included in
|
42
|
+
all copies or substantial portions of the Software.
|
43
|
+
|
44
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
45
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
46
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
47
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
48
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
49
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
50
|
+
THE SOFTWARE.
|
data/Rakefile
ADDED
@@ -0,0 +1,114 @@
|
|
1
|
+
require 'rubygems'
|
2
|
+
require 'rake'
|
3
|
+
require 'rake/clean'
|
4
|
+
require 'rake/testtask'
|
5
|
+
require 'rake/packagetask'
|
6
|
+
require 'rake/gempackagetask'
|
7
|
+
require 'rake/rdoctask'
|
8
|
+
require 'rake/contrib/rubyforgepublisher'
|
9
|
+
require 'rake/contrib/sshpublisher'
|
10
|
+
require 'fileutils'
|
11
|
+
require 'lib/url2mhtml'
|
12
|
+
include FileUtils
|
13
|
+
|
14
|
+
NAME = "url2mhtml"
|
15
|
+
AUTHOR = "forgithubid"
|
16
|
+
EMAIL = "forgithubid@gmail.com"
|
17
|
+
DESCRIPTION = "generate to MHTML from URL."
|
18
|
+
#RUBYFORGE_PROJECT = "url2mhtml"
|
19
|
+
HOMEPATH = "http://github.com/fromgit/url2mhtml/tree/master"
|
20
|
+
BIN_FILES = %w( url2mhtml )
|
21
|
+
|
22
|
+
VERS = Url2mhtml::VERSION
|
23
|
+
REV = File.read(".svn/entries")[/committed-rev="(d+)"/, 1] rescue nil
|
24
|
+
CLEAN.include ['**/.*.sw?', '*.gem', '.config']
|
25
|
+
RDOC_OPTS = [
|
26
|
+
'--title', "#{NAME} documentation",
|
27
|
+
"--charset", "utf-8",
|
28
|
+
"--opname", "index.html",
|
29
|
+
"--line-numbers",
|
30
|
+
"--main", "README",
|
31
|
+
"--inline-source",
|
32
|
+
]
|
33
|
+
|
34
|
+
task :default => [:test]
|
35
|
+
task :package => [:clean]
|
36
|
+
|
37
|
+
Rake::TestTask.new("test") do |t|
|
38
|
+
t.libs << "test"
|
39
|
+
t.pattern = "test/**/*_test.rb"
|
40
|
+
t.verbose = true
|
41
|
+
end
|
42
|
+
|
43
|
+
spec = Gem::Specification.new do |s|
|
44
|
+
s.name = NAME
|
45
|
+
s.version = VERS
|
46
|
+
s.platform = Gem::Platform::RUBY
|
47
|
+
s.has_rdoc = true
|
48
|
+
s.extra_rdoc_files = ["README", "ChangeLog"]
|
49
|
+
s.rdoc_options += RDOC_OPTS + ['--exclude', '^(examples|extras)/']
|
50
|
+
s.summary = DESCRIPTION
|
51
|
+
s.description = DESCRIPTION
|
52
|
+
s.author = AUTHOR
|
53
|
+
s.email = EMAIL
|
54
|
+
s.homepage = HOMEPATH
|
55
|
+
s.executables = BIN_FILES
|
56
|
+
# s.rubyforge_project = RUBYFORGE_PROJECT
|
57
|
+
s.bindir = "bin"
|
58
|
+
s.require_path = "lib"
|
59
|
+
#s.autorequire = ""
|
60
|
+
s.test_files = Dir["test/*_test.rb"]
|
61
|
+
|
62
|
+
s.add_dependency('mechanize', '>=0.7.8')
|
63
|
+
s.add_dependency('tmail', '>=1.2.3.1')
|
64
|
+
|
65
|
+
s.files = %w(README ChangeLog Rakefile) +
|
66
|
+
Dir.glob("{bin,doc,test,lib,templates,generator,extras,website,script}/**/*") +
|
67
|
+
Dir.glob("ext/**/*.{h,c,rb}") +
|
68
|
+
Dir.glob("examples/**/*.rb") +
|
69
|
+
Dir.glob("tools/*.rb") +
|
70
|
+
Dir.glob("rails/*.rb")
|
71
|
+
|
72
|
+
s.extensions = FileList["ext/**/extconf.rb"].to_a
|
73
|
+
end
|
74
|
+
|
75
|
+
Rake::GemPackageTask.new(spec) do |p|
|
76
|
+
p.need_tar = true
|
77
|
+
p.gem_spec = spec
|
78
|
+
end
|
79
|
+
|
80
|
+
task :install do
|
81
|
+
name = "#{NAME}-#{VERS}.gem"
|
82
|
+
sh %{rake package}
|
83
|
+
sh %{sudo gem install pkg/#{name}}
|
84
|
+
end
|
85
|
+
|
86
|
+
task :uninstall => [:clean] do
|
87
|
+
sh %{sudo gem uninstall #{NAME}}
|
88
|
+
end
|
89
|
+
|
90
|
+
|
91
|
+
Rake::RDocTask.new do |rdoc|
|
92
|
+
rdoc.rdoc_dir = 'html'
|
93
|
+
rdoc.options += RDOC_OPTS
|
94
|
+
rdoc.template = "resh"
|
95
|
+
#rdoc.template = "#{ENV['template']}.rb" if ENV['template']
|
96
|
+
if ENV['DOC_FILES']
|
97
|
+
rdoc.rdoc_files.include(ENV['DOC_FILES'].split(/,\s*/))
|
98
|
+
else
|
99
|
+
rdoc.rdoc_files.include('README', 'ChangeLog')
|
100
|
+
rdoc.rdoc_files.include('lib/**/*.rb')
|
101
|
+
rdoc.rdoc_files.include('ext/**/*.c')
|
102
|
+
end
|
103
|
+
end
|
104
|
+
|
105
|
+
|
106
|
+
desc 'Show information about the gem.'
|
107
|
+
task :debug_gem do
|
108
|
+
puts spec.to_ruby
|
109
|
+
end
|
110
|
+
|
111
|
+
desc 'Update gem spec'
|
112
|
+
task :gemspec do
|
113
|
+
open("#{NAME}.gemspec", 'w').write spec.to_ruby
|
114
|
+
end
|
data/bin/url2mhtml
ADDED
@@ -0,0 +1,260 @@
|
|
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: Url2mhtml</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">Url2mhtml</td>
|
54
|
+
</tr>
|
55
|
+
<tr class="top-aligned-row">
|
56
|
+
<td><strong>In:</strong></td>
|
57
|
+
<td>
|
58
|
+
<a href="../files/lib/url2mhtml_rb.html">
|
59
|
+
lib/url2mhtml.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
|
+
|
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="#M000004">append_relative_contents</a>
|
90
|
+
<a href="#M000008">capture</a>
|
91
|
+
<a href="#M000007">create_mail</a>
|
92
|
+
<a href="#M000006">create_mail_part</a>
|
93
|
+
<a href="#M000001">get_agent</a>
|
94
|
+
<a href="#M000002">get_content</a>
|
95
|
+
<a href="#M000003">get_contents</a>
|
96
|
+
<a href="#M000005">resolve_relative_uri</a>
|
97
|
+
</div>
|
98
|
+
</div>
|
99
|
+
|
100
|
+
</div>
|
101
|
+
|
102
|
+
|
103
|
+
<!-- if includes -->
|
104
|
+
|
105
|
+
<div id="section">
|
106
|
+
|
107
|
+
|
108
|
+
<div id="constants-list">
|
109
|
+
<h3 class="section-bar">Constants</h3>
|
110
|
+
|
111
|
+
<div class="name-list">
|
112
|
+
<table summary="Constants">
|
113
|
+
<tr class="top-aligned-row context-row">
|
114
|
+
<td class="context-item-name">VERSION</td>
|
115
|
+
<td>=</td>
|
116
|
+
<td class="context-item-value">'0.0.2'</td>
|
117
|
+
</tr>
|
118
|
+
<tr class="top-aligned-row context-row">
|
119
|
+
<td class="context-item-name">ContentInfo</td>
|
120
|
+
<td>=</td>
|
121
|
+
<td class="context-item-value">Struct.new(:uri,:type,:body,:is_root,:title)</td>
|
122
|
+
</tr>
|
123
|
+
</table>
|
124
|
+
</div>
|
125
|
+
</div>
|
126
|
+
|
127
|
+
|
128
|
+
|
129
|
+
|
130
|
+
|
131
|
+
|
132
|
+
<!-- if method_list -->
|
133
|
+
<div id="methods">
|
134
|
+
<h3 class="section-bar">Public Class methods</h3>
|
135
|
+
|
136
|
+
<div id="method-M000004" class="method-detail">
|
137
|
+
<a name="M000004"></a>
|
138
|
+
|
139
|
+
<div class="method-heading">
|
140
|
+
<a href="Url2mhtml.src/M000004.html" target="Code" class="method-signature"
|
141
|
+
onclick="popupCode('Url2mhtml.src/M000004.html');return false;">
|
142
|
+
<span class="method-name">append_relative_contents</span><span class="method-args">(page,content_info_list)</span>
|
143
|
+
</a>
|
144
|
+
</div>
|
145
|
+
|
146
|
+
<div class="method-description">
|
147
|
+
</div>
|
148
|
+
</div>
|
149
|
+
|
150
|
+
<div id="method-M000008" class="method-detail">
|
151
|
+
<a name="M000008"></a>
|
152
|
+
|
153
|
+
<div class="method-heading">
|
154
|
+
<a href="Url2mhtml.src/M000008.html" target="Code" class="method-signature"
|
155
|
+
onclick="popupCode('Url2mhtml.src/M000008.html');return false;">
|
156
|
+
<span class="method-name">capture</span><span class="method-args">(uri)</span>
|
157
|
+
</a>
|
158
|
+
</div>
|
159
|
+
|
160
|
+
<div class="method-description">
|
161
|
+
</div>
|
162
|
+
</div>
|
163
|
+
|
164
|
+
<div id="method-M000007" class="method-detail">
|
165
|
+
<a name="M000007"></a>
|
166
|
+
|
167
|
+
<div class="method-heading">
|
168
|
+
<a href="Url2mhtml.src/M000007.html" target="Code" class="method-signature"
|
169
|
+
onclick="popupCode('Url2mhtml.src/M000007.html');return false;">
|
170
|
+
<span class="method-name">create_mail</span><span class="method-args">(title,parts)</span>
|
171
|
+
</a>
|
172
|
+
</div>
|
173
|
+
|
174
|
+
<div class="method-description">
|
175
|
+
</div>
|
176
|
+
</div>
|
177
|
+
|
178
|
+
<div id="method-M000006" class="method-detail">
|
179
|
+
<a name="M000006"></a>
|
180
|
+
|
181
|
+
<div class="method-heading">
|
182
|
+
<a href="Url2mhtml.src/M000006.html" target="Code" class="method-signature"
|
183
|
+
onclick="popupCode('Url2mhtml.src/M000006.html');return false;">
|
184
|
+
<span class="method-name">create_mail_part</span><span class="method-args">(content)</span>
|
185
|
+
</a>
|
186
|
+
</div>
|
187
|
+
|
188
|
+
<div class="method-description">
|
189
|
+
</div>
|
190
|
+
</div>
|
191
|
+
|
192
|
+
<div id="method-M000001" class="method-detail">
|
193
|
+
<a name="M000001"></a>
|
194
|
+
|
195
|
+
<div class="method-heading">
|
196
|
+
<a href="Url2mhtml.src/M000001.html" target="Code" class="method-signature"
|
197
|
+
onclick="popupCode('Url2mhtml.src/M000001.html');return false;">
|
198
|
+
<span class="method-name">get_agent</span><span class="method-args">()</span>
|
199
|
+
</a>
|
200
|
+
</div>
|
201
|
+
|
202
|
+
<div class="method-description">
|
203
|
+
</div>
|
204
|
+
</div>
|
205
|
+
|
206
|
+
<div id="method-M000002" class="method-detail">
|
207
|
+
<a name="M000002"></a>
|
208
|
+
|
209
|
+
<div class="method-heading">
|
210
|
+
<a href="Url2mhtml.src/M000002.html" target="Code" class="method-signature"
|
211
|
+
onclick="popupCode('Url2mhtml.src/M000002.html');return false;">
|
212
|
+
<span class="method-name">get_content</span><span class="method-args">(uri,is_root)</span>
|
213
|
+
</a>
|
214
|
+
</div>
|
215
|
+
|
216
|
+
<div class="method-description">
|
217
|
+
</div>
|
218
|
+
</div>
|
219
|
+
|
220
|
+
<div id="method-M000003" class="method-detail">
|
221
|
+
<a name="M000003"></a>
|
222
|
+
|
223
|
+
<div class="method-heading">
|
224
|
+
<a href="Url2mhtml.src/M000003.html" target="Code" class="method-signature"
|
225
|
+
onclick="popupCode('Url2mhtml.src/M000003.html');return false;">
|
226
|
+
<span class="method-name">get_contents</span><span class="method-args">(uri,is_root,content_info_list)</span>
|
227
|
+
</a>
|
228
|
+
</div>
|
229
|
+
|
230
|
+
<div class="method-description">
|
231
|
+
</div>
|
232
|
+
</div>
|
233
|
+
|
234
|
+
<div id="method-M000005" class="method-detail">
|
235
|
+
<a name="M000005"></a>
|
236
|
+
|
237
|
+
<div class="method-heading">
|
238
|
+
<a href="Url2mhtml.src/M000005.html" target="Code" class="method-signature"
|
239
|
+
onclick="popupCode('Url2mhtml.src/M000005.html');return false;">
|
240
|
+
<span class="method-name">resolve_relative_uri</span><span class="method-args">(base_uri,target_uri)</span>
|
241
|
+
</a>
|
242
|
+
</div>
|
243
|
+
|
244
|
+
<div class="method-description">
|
245
|
+
</div>
|
246
|
+
</div>
|
247
|
+
|
248
|
+
|
249
|
+
</div>
|
250
|
+
|
251
|
+
|
252
|
+
</div>
|
253
|
+
|
254
|
+
|
255
|
+
<div id="validator-badges">
|
256
|
+
<p><small><a href="http://validator.w3.org/check/referer">[Validate]</a></small></p>
|
257
|
+
</div>
|
258
|
+
|
259
|
+
</body>
|
260
|
+
</html>
|
@@ -0,0 +1,21 @@
|
|
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>get_agent (Url2mhtml)</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/url2mhtml.rb, line 9</span>
|
14
|
+
<span class="ruby-keyword kw">def</span> <span class="ruby-keyword kw">self</span>.<span class="ruby-identifier">get_agent</span>
|
15
|
+
<span class="ruby-identifier">user_agent_alias</span> =<span class="ruby-value str">'Windows IE 6'</span>
|
16
|
+
<span class="ruby-identifier">agent</span> = <span class="ruby-constant">WWW</span><span class="ruby-operator">::</span><span class="ruby-constant">Mechanize</span>.<span class="ruby-identifier">new</span>
|
17
|
+
<span class="ruby-identifier">agent</span>.<span class="ruby-identifier">user_agent_alias</span> =<span class="ruby-identifier">user_agent_alias</span>
|
18
|
+
<span class="ruby-identifier">agent</span>
|
19
|
+
<span class="ruby-keyword kw">end</span></pre>
|
20
|
+
</body>
|
21
|
+
</html>
|
@@ -0,0 +1,25 @@
|
|
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>get_content (Url2mhtml)</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/url2mhtml.rb, line 16</span>
|
14
|
+
<span class="ruby-keyword kw">def</span> <span class="ruby-keyword kw">self</span>.<span class="ruby-identifier">get_content</span>(<span class="ruby-identifier">uri</span>,<span class="ruby-identifier">is_root</span>)
|
15
|
+
<span class="ruby-identifier">got_content</span> = <span class="ruby-identifier">get_agent</span>.<span class="ruby-identifier">get</span>(<span class="ruby-identifier">uri</span>)
|
16
|
+
<span class="ruby-identifier">type</span>=<span class="ruby-identifier">got_content</span>.<span class="ruby-identifier">response</span>[<span class="ruby-value str">'content-type'</span>]
|
17
|
+
<span class="ruby-identifier">body</span>=<span class="ruby-identifier">got_content</span>.<span class="ruby-identifier">body</span>
|
18
|
+
|
19
|
+
<span class="ruby-identifier">title</span>= ( <span class="ruby-identifier">is_root</span> <span class="ruby-operator">&&</span> <span class="ruby-regexp re">/html/</span>.<span class="ruby-identifier">match</span>(<span class="ruby-identifier">type</span>) ) <span class="ruby-operator">?</span> <span class="ruby-identifier">got_content</span>.<span class="ruby-identifier">title</span> <span class="ruby-operator">:</span> <span class="ruby-value str">'no title'</span>
|
20
|
+
|
21
|
+
<span class="ruby-identifier">content_info</span>=<span class="ruby-constant">ContentInfo</span>.<span class="ruby-identifier">new</span>(<span class="ruby-identifier">uri</span>,<span class="ruby-identifier">type</span>,<span class="ruby-identifier">body</span>,<span class="ruby-identifier">is_root</span>,<span class="ruby-identifier">title</span>)
|
22
|
+
<span class="ruby-keyword kw">return</span> <span class="ruby-identifier">content_info</span>,<span class="ruby-identifier">got_content</span>
|
23
|
+
<span class="ruby-keyword kw">end</span></pre>
|
24
|
+
</body>
|
25
|
+
</html>
|
@@ -0,0 +1,23 @@
|
|
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>get_contents (Url2mhtml)</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/url2mhtml.rb, line 27</span>
|
14
|
+
<span class="ruby-keyword kw">def</span> <span class="ruby-keyword kw">self</span>.<span class="ruby-identifier">get_contents</span>(<span class="ruby-identifier">uri</span>,<span class="ruby-identifier">is_root</span>,<span class="ruby-identifier">content_info_list</span>)
|
15
|
+
|
16
|
+
<span class="ruby-identifier">content_info</span>,<span class="ruby-identifier">got_content</span>=<span class="ruby-identifier">get_content</span>(<span class="ruby-identifier">uri</span>,<span class="ruby-identifier">is_root</span>)
|
17
|
+
<span class="ruby-identifier">content_info_list</span> <span class="ruby-operator"><<</span> <span class="ruby-identifier">content_info</span>
|
18
|
+
|
19
|
+
<span class="ruby-identifier">append_relative_contents</span>(<span class="ruby-identifier">got_content</span>,<span class="ruby-identifier">content_info_list</span>) <span class="ruby-keyword kw">if</span> <span class="ruby-regexp re">/html/</span>.<span class="ruby-identifier">match</span>(<span class="ruby-identifier">content_info</span>.<span class="ruby-identifier">type</span>)
|
20
|
+
<span class="ruby-identifier">content_info_list</span>
|
21
|
+
<span class="ruby-keyword kw">end</span></pre>
|
22
|
+
</body>
|
23
|
+
</html>
|
@@ -0,0 +1,41 @@
|
|
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>append_relative_contents (Url2mhtml)</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/url2mhtml.rb, line 36</span>
|
14
|
+
<span class="ruby-keyword kw">def</span> <span class="ruby-keyword kw">self</span>.<span class="ruby-identifier">append_relative_contents</span>(<span class="ruby-identifier">page</span>,<span class="ruby-identifier">content_info_list</span>)
|
15
|
+
<span class="ruby-identifier">base_uri</span>=<span class="ruby-identifier">page</span>.<span class="ruby-identifier">uri</span>
|
16
|
+
<span class="ruby-identifier">raw_image_uris</span>=<span class="ruby-identifier">page</span>.<span class="ruby-identifier">search</span>(<span class="ruby-value str">'//img'</span>).<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">i</span>[<span class="ruby-value str">'src'</span>]}
|
17
|
+
<span class="ruby-identifier">raw_image_uris</span>.<span class="ruby-identifier">push</span>(<span class="ruby-operator">*</span>(<span class="ruby-identifier">page</span>.<span class="ruby-identifier">search</span>(<span class="ruby-value str">'//body'</span>).<span class="ruby-identifier">find_all</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">i</span>[<span class="ruby-value str">'background'</span>]}.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">i</span>[<span class="ruby-value str">'background'</span>]}))
|
18
|
+
<span class="ruby-identifier">raw_image_uris</span>.<span class="ruby-identifier">push</span>(<span class="ruby-operator">*</span>(<span class="ruby-identifier">page</span>.<span class="ruby-identifier">search</span>(<span class="ruby-value str">'//th'</span>).<span class="ruby-identifier">find_all</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">i</span>[<span class="ruby-value str">'background'</span>]}.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">i</span>[<span class="ruby-value str">'background'</span>]}))
|
19
|
+
<span class="ruby-identifier">raw_image_uris</span>.<span class="ruby-identifier">push</span>(<span class="ruby-operator">*</span>(<span class="ruby-identifier">page</span>.<span class="ruby-identifier">search</span>(<span class="ruby-value str">'//td'</span>).<span class="ruby-identifier">find_all</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">i</span>[<span class="ruby-value str">'background'</span>]}.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">i</span>[<span class="ruby-value str">'background'</span>]}))
|
20
|
+
<span class="ruby-identifier">image_uris</span>=<span class="ruby-identifier">raw_image_uris</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">resolve_relative_uri</span>(<span class="ruby-identifier">base_uri</span>,<span class="ruby-identifier">i</span>)}.<span class="ruby-identifier">grep</span>(<span class="ruby-regexp re">/^(http|ftp)/</span>)
|
21
|
+
|
22
|
+
<span class="ruby-identifier">raw_frame_uris</span>=<span class="ruby-identifier">page</span>.<span class="ruby-identifier">frames</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">f</span><span class="ruby-operator">|</span> <span class="ruby-identifier">f</span>.<span class="ruby-identifier">uri</span>}
|
23
|
+
<span class="ruby-identifier">frame_uris</span>=<span class="ruby-identifier">raw_frame_uris</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">resolve_relative_uri</span>(<span class="ruby-identifier">base_uri</span>,<span class="ruby-identifier">i</span>)}.<span class="ruby-identifier">grep</span>(<span class="ruby-regexp re">/^(http|ftp)/</span>)
|
24
|
+
|
25
|
+
<span class="ruby-identifier">raw_iframe_uris</span>=<span class="ruby-identifier">page</span>.<span class="ruby-identifier">iframes</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">f</span><span class="ruby-operator">|</span> <span class="ruby-identifier">f</span>.<span class="ruby-identifier">uri</span>}
|
26
|
+
<span class="ruby-identifier">iframe_uris</span>=<span class="ruby-identifier">raw_iframe_uris</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">resolve_relative_uri</span>(<span class="ruby-identifier">base_uri</span>,<span class="ruby-identifier">i</span>)}.<span class="ruby-identifier">grep</span>(<span class="ruby-regexp re">/^(http|ftp)/</span>)
|
27
|
+
|
28
|
+
<span class="ruby-identifier">raw_css_uris</span>=<span class="ruby-identifier">page</span>.<span class="ruby-identifier">search</span>(<span class="ruby-value str">'link[@rel="stylesheet"]'</span>).<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">l</span><span class="ruby-operator">|</span> <span class="ruby-identifier">l</span>[<span class="ruby-value str">'href'</span>]}
|
29
|
+
<span class="ruby-identifier">css_uris</span>=<span class="ruby-identifier">raw_css_uris</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">resolve_relative_uri</span>(<span class="ruby-identifier">base_uri</span>,<span class="ruby-identifier">i</span>)}.<span class="ruby-identifier">grep</span>(<span class="ruby-regexp re">/^(http|ftp)/</span>)
|
30
|
+
|
31
|
+
<span class="ruby-identifier">raw_script_uris</span>=<span class="ruby-identifier">page</span>.<span class="ruby-identifier">search</span>(<span class="ruby-value str">'script'</span>).<span class="ruby-identifier">find_all</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">s</span><span class="ruby-operator">|</span> <span class="ruby-identifier">s</span>[<span class="ruby-value str">'src'</span>]}.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">s</span><span class="ruby-operator">|</span> <span class="ruby-identifier">s</span>[<span class="ruby-value str">'src'</span>]}
|
32
|
+
<span class="ruby-identifier">script_uris</span>=<span class="ruby-identifier">raw_script_uris</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">i</span><span class="ruby-operator">|</span> <span class="ruby-identifier">resolve_relative_uri</span>(<span class="ruby-identifier">base_uri</span>,<span class="ruby-identifier">i</span>)}.<span class="ruby-identifier">grep</span>(<span class="ruby-regexp re">/^(http|ftp)/</span>)
|
33
|
+
|
34
|
+
|
35
|
+
<span class="ruby-identifier">raw_urls</span> = ( <span class="ruby-identifier">image_uris</span> <span class="ruby-operator">+</span> <span class="ruby-identifier">frame_uris</span> <span class="ruby-operator">+</span> <span class="ruby-identifier">iframe_uris</span> <span class="ruby-operator">+</span> <span class="ruby-identifier">css_uris</span> <span class="ruby-operator">+</span> <span class="ruby-identifier">script_uris</span> )
|
36
|
+
<span class="ruby-identifier">target_content_urls</span> = <span class="ruby-identifier">raw_urls</span>.<span class="ruby-identifier">map</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">u</span><span class="ruby-operator">|</span> <span class="ruby-identifier">u</span>.<span class="ruby-identifier">gsub</span>(<span class="ruby-regexp re">/#.*/</span>,<span class="ruby-value str">''</span>)}.<span class="ruby-identifier">uniq</span>.<span class="ruby-identifier">find_all</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">u</span><span class="ruby-operator">|</span> <span class="ruby-identifier">content_info_list</span>.<span class="ruby-identifier">any?</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">content</span><span class="ruby-operator">|</span> <span class="ruby-identifier">u</span> <span class="ruby-operator">!=</span> <span class="ruby-identifier">content</span>.<span class="ruby-identifier">uri</span>}}
|
37
|
+
|
38
|
+
<span class="ruby-identifier">target_content_urls</span>.<span class="ruby-identifier">each</span>{<span class="ruby-operator">|</span><span class="ruby-identifier">uri</span><span class="ruby-operator">|</span> <span class="ruby-identifier">get_contents</span>(<span class="ruby-identifier">uri</span>,<span class="ruby-keyword kw">false</span>,<span class="ruby-identifier">content_info_list</span>)}
|
39
|
+
<span class="ruby-keyword kw">end</span></pre>
|
40
|
+
</body>
|
41
|
+
</html>
|
@@ -0,0 +1,18 @@
|
|
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>resolve_relative_uri (Url2mhtml)</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/url2mhtml.rb, line 63</span>
|
14
|
+
<span class="ruby-keyword kw">def</span> <span class="ruby-keyword kw">self</span>.<span class="ruby-identifier">resolve_relative_uri</span>(<span class="ruby-identifier">base_uri</span>,<span class="ruby-identifier">target_uri</span>)
|
15
|
+
<span class="ruby-constant">URI</span>.<span class="ruby-identifier">join</span>(<span class="ruby-identifier">base_uri</span>.<span class="ruby-identifier">to_s</span>,<span class="ruby-identifier">target_uri</span>).<span class="ruby-identifier">to_s</span>
|
16
|
+
<span class="ruby-keyword kw">end</span></pre>
|
17
|
+
</body>
|
18
|
+
</html>
|
@@ -0,0 +1,30 @@
|
|
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>create_mail_part (Url2mhtml)</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/url2mhtml.rb, line 67</span>
|
14
|
+
<span class="ruby-keyword kw">def</span> <span class="ruby-keyword kw">self</span>.<span class="ruby-identifier">create_mail_part</span>(<span class="ruby-identifier">content</span>)
|
15
|
+
<span class="ruby-identifier">part</span>=<span class="ruby-constant">TMail</span><span class="ruby-operator">::</span><span class="ruby-constant">Mail</span>.<span class="ruby-identifier">new</span>
|
16
|
+
|
17
|
+
<span class="ruby-identifier">part</span>[<span class="ruby-value str">'content-location'</span>]=<span class="ruby-identifier">content</span>.<span class="ruby-identifier">uri</span>
|
18
|
+
<span class="ruby-identifier">part</span>.<span class="ruby-identifier">content_type</span> = <span class="ruby-identifier">content</span>.<span class="ruby-identifier">type</span>
|
19
|
+
<span class="ruby-keyword kw">if</span> <span class="ruby-regexp re">/html/</span>.<span class="ruby-identifier">match</span>(<span class="ruby-identifier">content</span>.<span class="ruby-identifier">type</span>)
|
20
|
+
<span class="ruby-identifier">part</span>.<span class="ruby-identifier">transfer_encoding</span> = <span class="ruby-value str">'8bit'</span>
|
21
|
+
<span class="ruby-identifier">part</span>.<span class="ruby-identifier">body</span> = <span class="ruby-identifier">content</span>.<span class="ruby-identifier">body</span>
|
22
|
+
<span class="ruby-keyword kw">else</span>
|
23
|
+
<span class="ruby-identifier">part</span>.<span class="ruby-identifier">transfer_encoding</span> = <span class="ruby-value str">'base64'</span>
|
24
|
+
<span class="ruby-identifier">b64encoded_body</span> = [<span class="ruby-identifier">content</span>.<span class="ruby-identifier">body</span>].<span class="ruby-identifier">pack</span>(<span class="ruby-value str">'m'</span>).<span class="ruby-identifier">chomp</span>.<span class="ruby-identifier">gsub</span>(<span class="ruby-regexp re">/.{76}/</span>, <span class="ruby-value str">"\\1\n"</span>)
|
25
|
+
<span class="ruby-identifier">part</span>.<span class="ruby-identifier">body</span> = <span class="ruby-identifier">b64encoded_body</span>
|
26
|
+
<span class="ruby-keyword kw">end</span>
|
27
|
+
<span class="ruby-identifier">part</span>
|
28
|
+
<span class="ruby-keyword kw">end</span></pre>
|
29
|
+
</body>
|
30
|
+
</html>
|