w3clove 0.4.5 → 0.5

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 (5) hide show
  1. data/lib/w3clove/sitemap.rb +4 -2
  2. data/lib/w3clove/version.rb +1 -1
  3. data/spec/samples/zigotica.com.html +37 -0
  4. data/spec/sitemap_spec.rb +10 -2
  5. metadata +4 -4
data/lib/w3clove/sitemap.rb CHANGED
@@ -43,15 +43,17 @@ module W3Clove
43
43
  private
44
44
 
45
45
  # Scrapes the url in search of links.
46
+ #
46
47
  # It first assumes it's an XML sitemap; if no locations found, it will try to
47
48
  # scrape the links from HTML.
49
+ #
48
50
  # For HTML sources, it will only get the links that start with the sitemap url, convert relative links
49
- # to absolute links, and remove anchors from links
51
+ # to absolute links, remove anchors from links, and include the sitemap url
50
52
  def pages_in_sitemap
51
53
  pages = xml_locations.map {|loc| W3Clove::Page.new(loc.text)}
52
54
  if pages.empty?
53
55
  m = MetaInspector.new(url)
54
- links = m.absolute_links.select {|l| l.start_with?(m.url)}.map {|l| l.split('#')[0]}.uniq
56
+ links = ([m.url] + m.absolute_links.select {|l| l.start_with?(m.url)}.map {|l| l.split('#')[0]}).uniq
55
57
  pages = links.map {|link| W3Clove::Page.new(link)}
56
58
  end
57
59
  pages
data/lib/w3clove/version.rb CHANGED
@@ -1,5 +1,5 @@
1
1
  # -*- encoding: utf-8 -*-
2
2
 
3
3
  module W3Clove
4
- VERSION = "0.4.5"
4
+ VERSION = "0.5"
5
5
  end
data/spec/samples/zigotica.com.html ADDED
@@ -0,0 +1,37 @@
1
+ <!DOCTYPE html>
2
+ <html>
3
+ <head>
4
+ <meta charset="utf-8">
5
+ <meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=0" />
6
+ <title>Sergi Meseguer aka zigotica, home</title>
7
+ <link rel="stylesheet" href="css/style.css">
8
+ </head>
9
+
10
+ <body>
11
+ <div id="wrapper">
12
+ <img src="img/avatar.jpg" alt="avatar">
13
+ <section>
14
+ <h1>Sergi Meseguer</h1>
15
+ <p>I am an enthusiast developer who codes front and back end solutions.</p>
16
+ <p class="summary">Follow me:</p>
17
+
18
+ <ul>
19
+ <li><a href="https://github.com/zigotica"><img src="img/github-32x32.png" alt="github"></a></li>
20
+ <li><a href="http://twitter.com/zigotica"><img src="img/twitter-32x32.png" alt="twitter"></a></li>
21
+ <li><a href="http://zigotica.tumblr.com/"><img src="img/tumblr-32x32.png" alt="tumblr"></a></li>
22
+ <li><a href="https://plus.google.com/113605874723194545353"><img src="img/gplus-32x32.png" alt="Google+"></a></li>
23
+ <li><a href="http://www.linkedin.com/in/sergimeseguer"><img src="img/linkedin-32x32.png" alt="linkedin"></a></li>
24
+ </ul>
25
+ </section>
26
+ <footer><small>Contents and image © 2011, Sergi Meseguer. All rights reserved.</small></footer>
27
+ </div>
28
+ <script type="text/javascript">
29
+ var gaJsHost = (("https:" == document.location.protocol) ? "https://ssl." : "http://www.");
30
+ document.write(unescape("%3Cscript src='" + gaJsHost + "google-analytics.com/ga.js' type='text/javascript'%3E%3C/script%3E"));
31
+ </script>
32
+ <script type="text/javascript">
33
+ var pageTracker = _gat._getTracker("UA-3717420-1");
34
+ pageTracker._trackPageview();
35
+ </script>
36
+ </body>
37
+ </html>
data/spec/sitemap_spec.rb CHANGED
@@ -10,6 +10,9 @@ describe W3Clove::Sitemap do
10
10
  @sitemap_html = W3Clove::Sitemap.new('http://guides.rubyonrails.org')
11
11
  @sitemap_html.stub!(:doc).and_return(open("#{$samples_dir}/guides.rubyonrails.org.html"))
12
12
 
13
+ @sitemap_no_links = W3Clove::Sitemap.new('http://zigotica.com')
14
+ @sitemap_no_links.stub!(:doc).and_return(open("#{$samples_dir}/zigotica.com.html"))
15
+
13
16
  MarkupValidator.any_instance.stubs(:validate_uri).returns(stubbed_validator_results)
14
17
  end
15
18
 
@@ -26,8 +29,13 @@ describe W3Clove::Sitemap do
26
29
  end
27
30
 
28
31
  it "should get pages from the sample guides.rubyonrails.org site" do
29
- @sitemap_html.pages.length.should == 34
30
- @sitemap_html.pages.map {|p| p.url}.should == ["http://guides.rubyonrails.org/index.html", "http://guides.rubyonrails.org/", "http://guides.rubyonrails.org/getting_started.html", "http://guides.rubyonrails.org/migrations.html", "http://guides.rubyonrails.org/active_record_validations_callbacks.html", "http://guides.rubyonrails.org/association_basics.html", "http://guides.rubyonrails.org/active_record_querying.html", "http://guides.rubyonrails.org/layouts_and_rendering.html", "http://guides.rubyonrails.org/form_helpers.html", "http://guides.rubyonrails.org/action_controller_overview.html", "http://guides.rubyonrails.org/routing.html", "http://guides.rubyonrails.org/active_support_core_extensions.html", "http://guides.rubyonrails.org/i18n.html", "http://guides.rubyonrails.org/action_mailer_basics.html", "http://guides.rubyonrails.org/testing.html", "http://guides.rubyonrails.org/security.html", "http://guides.rubyonrails.org/debugging_rails_applications.html", "http://guides.rubyonrails.org/performance_testing.html", "http://guides.rubyonrails.org/configuring.html", "http://guides.rubyonrails.org/command_line.html", "http://guides.rubyonrails.org/caching_with_rails.html", "http://guides.rubyonrails.org/asset_pipeline.html", "http://guides.rubyonrails.org/plugins.html", "http://guides.rubyonrails.org/rails_on_rack.html", "http://guides.rubyonrails.org/generators.html", "http://guides.rubyonrails.org/contributing_to_ruby_on_rails.html", "http://guides.rubyonrails.org/api_documentation_guidelines.html", "http://guides.rubyonrails.org/ruby_on_rails_guides_guidelines.html", "http://guides.rubyonrails.org/3_1_release_notes.html", "http://guides.rubyonrails.org/3_0_release_notes.html", "http://guides.rubyonrails.org/2_3_release_notes.html", "http://guides.rubyonrails.org/2_2_release_notes.html", "http://guides.rubyonrails.org/credits.html", "http://guides.rubyonrails.org/v2.3.11/"]
32
+ @sitemap_html.pages.length.should == 35
33
+ @sitemap_html.pages.map {|p| p.url}.should == ["http://guides.rubyonrails.org", "http://guides.rubyonrails.org/index.html", "http://guides.rubyonrails.org/", "http://guides.rubyonrails.org/getting_started.html", "http://guides.rubyonrails.org/migrations.html", "http://guides.rubyonrails.org/active_record_validations_callbacks.html", "http://guides.rubyonrails.org/association_basics.html", "http://guides.rubyonrails.org/active_record_querying.html", "http://guides.rubyonrails.org/layouts_and_rendering.html", "http://guides.rubyonrails.org/form_helpers.html", "http://guides.rubyonrails.org/action_controller_overview.html", "http://guides.rubyonrails.org/routing.html", "http://guides.rubyonrails.org/active_support_core_extensions.html", "http://guides.rubyonrails.org/i18n.html", "http://guides.rubyonrails.org/action_mailer_basics.html", "http://guides.rubyonrails.org/testing.html", "http://guides.rubyonrails.org/security.html", "http://guides.rubyonrails.org/debugging_rails_applications.html", "http://guides.rubyonrails.org/performance_testing.html", "http://guides.rubyonrails.org/configuring.html", "http://guides.rubyonrails.org/command_line.html", "http://guides.rubyonrails.org/caching_with_rails.html", "http://guides.rubyonrails.org/asset_pipeline.html", "http://guides.rubyonrails.org/plugins.html", "http://guides.rubyonrails.org/rails_on_rack.html", "http://guides.rubyonrails.org/generators.html", "http://guides.rubyonrails.org/contributing_to_ruby_on_rails.html", "http://guides.rubyonrails.org/api_documentation_guidelines.html", "http://guides.rubyonrails.org/ruby_on_rails_guides_guidelines.html", "http://guides.rubyonrails.org/3_1_release_notes.html", "http://guides.rubyonrails.org/3_0_release_notes.html", "http://guides.rubyonrails.org/2_3_release_notes.html", "http://guides.rubyonrails.org/2_2_release_notes.html", "http://guides.rubyonrails.org/credits.html", "http://guides.rubyonrails.org/v2.3.11/"]
34
+ end
35
+
36
+ it "should include sitemap url at least, even if no links were found" do
37
+ @sitemap_no_links.pages.length.should == 1
38
+ @sitemap_no_links.pages[0].url.should == 'http://zigotica.com'
31
39
  end
32
40
  end
33
41
 
metadata CHANGED
@@ -1,13 +1,12 @@
1
1
  --- !ruby/object:Gem::Specification
2
2
  name: w3clove
3
3
  version: !ruby/object:Gem::Version
4
- hash: 5
4
+ hash: 1
5
5
  prerelease:
6
6
  segments:
7
7
  - 0
8
- - 4
9
8
  - 5
10
- version: 0.4.5
9
+ version: "0.5"
11
10
  platform: ruby
12
11
  authors:
13
12
  - Jaime Iniesta
@@ -15,7 +14,7 @@ autorequire:
15
14
  bindir: bin
16
15
  cert_chain: []
17
16
 
18
- date: 2011-11-10 00:00:00 Z
17
+ date: 2011-11-22 00:00:00 Z
19
18
  dependencies:
20
19
  - !ruby/object:Gem::Dependency
21
20
  name: w3c_validators
@@ -147,6 +146,7 @@ files:
147
146
  - spec/samples/absolute_links.html
148
147
  - spec/samples/guides.rubyonrails.org.html
149
148
  - spec/samples/sitemap.xml
149
+ - spec/samples/zigotica.com.html
150
150
  - spec/sitemap_spec.rb
151
151
  - spec/spec_helper.rb
152
152
  - w3clove.gemspec