w3clove 0.7.9 → 0.7.10

data/README.rdoc

@@ -1,4 +1,4 @@
-= w3clove {<img src="http://travis-ci.org/jaimeiniesta/w3clove.png" />}[http://travis-ci.org/jaimeiniesta/w3clove]
+= w3clove {<img src="http://travis-ci.org/jaimeiniesta/w3clove.png" />}[http://travis-ci.org/jaimeiniesta/w3clove] {<img src="https://codeclimate.com/badge.png" />}[https://codeclimate.com/github/jaimeiniesta/w3clove]
 
 Site-wide markup validator. Validate a whole web site on the W3C Markup Validator, from the command line, and generate a comprehensive report of all errors found.
 

data/lib/w3clove/sitemap.rb

@@ -53,7 +53,7 @@ module W3Clove
     # to absolute links, remove anchors from links, include the sitemap url, and exclude links that don't
     # seem to point to HTML (like images, multimedia, text, javascript...)
     def pages_in_sitemap
-      pages = xml_locations.map {|loc| W3Clove::Page.new(loc.text)}
+      pages = xml_locations.select {|loc| looks_like_html?(loc.text)}.map {|loc| W3Clove::Page.new(loc.text)}
       if pages.empty?
         m     = MetaInspector.new(url, timeout)
         links = [m.url]
@@ -75,10 +75,10 @@ module W3Clove
     # That is, it does not look like javascript, image, pdf...
     def looks_like_html?(url)
       u         = URI.parse(URI.encode(url))
-      scheme    = u.scheme
-      extension = u.path.split(".").last
+      scheme    = u.scheme                if u.scheme
+      extension = u.path.split(".").last  if u.path
 
-      (scheme =~ /http[s]?/i) && (extension !~ /gif|jpg|jpeg|png|tiff|bmp|txt|pdf|doc|xls|wav|mp3|ogg/i)
+      (scheme && extension) && (scheme =~ /http[s]?/i) && (extension !~ /gif|jpg|jpeg|png|tiff|bmp|txt|pdf|doc|rtf|xml|xls|csv|wav|mp3|ogg/i)
     end
 
     def xml_locations

data/lib/w3clove/version.rb

@@ -1,5 +1,5 @@
 # -*- encoding: utf-8 -*-
 
 module W3Clove
-  VERSION = "0.7.9"
+  VERSION = "0.7.10"
 end

data/spec/samples/exclusions.html

@@ -30,6 +30,9 @@ Via: 1.1 varnish
     <a href="/file.txt">link</a><br />
     <a href="/file.doc">link</a><br />
     <a href="/file.xls">link</a><br />
+    <a href="/file.rtf">link</a><br />
+    <a href="/file.xml">link</a><br />
+    <a href="/file.csv">link</a><br />
 
     <a href="/file.wav">link</a><br />
     <a href="/file.mp3">link</a><br />

data/spec/samples/exclusions.xml

@@ -0,0 +1,81 @@
+HTTP/1.1 200 OK
+Date: Thu, 29 Dec 2011 23:08:54 GMT
+Server: Apache/2.2.21 (Unix) mod_ssl/2.2.3 OpenSSL/0.9.8e-fips-rhel5
+Transfer-Encoding: chunked
+Content-Type: text/xml
+
+<?xml version="1.0" encoding="UTF-8"?>
+<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
+  <url>
+    <loc>http://example.com/</loc>
+  </url>
+  <url>
+    <loc>http://example.com/faqs</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.gif</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.GIF</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.gif?size=thumb</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.jpg</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.jpeg</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.png</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.tiff</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.bmp</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.pdf</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.txt</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.doc</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.xls</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.rtf</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.xml</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.csv</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.wav</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.mp3</loc>
+  </url>
+  <url>
+    <loc>http://example.com/file.ogg</loc>
+  </url>
+  <url>
+    <loc>mailto:user@example.com</loc>
+  </url>
+  <url>
+    <loc>ftp://user:password@example.com</loc>
+  </url>
+  <url>
+    <loc>telnet://user:password@example.com</loc>
+  </url>
+  <url>
+    <loc>javascript:alert('hey');</loc>
+  </url>
+</urlset>

data/spec/sitemap_spec.rb

@@ -11,6 +11,7 @@ describe W3Clove::Sitemap do
     @sitemap_with_protocol_relative       = W3Clove::Sitemap.new('http://protocol-relative.com')
     @sitemap_with_protocol_relative_https = W3Clove::Sitemap.new('https://protocol-relative.com')
     @sitemap_for_exclusions               = W3Clove::Sitemap.new('http://example.com/exclusions')
+    @sitemap_for_exclusions_xml           = W3Clove::Sitemap.new('http://example.com/exclusions.xml')
     @sitemap_for_absolute_urls            = W3Clove::Sitemap.new('http://w3clove.com/faqs')
     @sitemap_international                = W3Clove::Sitemap.new('http://example.com/international')
 
@@ -98,13 +99,19 @@ describe W3Clove::Sitemap do
       urls.should_not  include 'http://w3clove.com/faqs/'
     end
 
-    it "should exclude non-html pages" do
+    it "should exclude non-html pages from HTML sitemaps" do
       @sitemap_for_exclusions.pages.length.should == 3
       @sitemap_for_exclusions.pages[0].url.should == 'http://example.com/exclusions'
       @sitemap_for_exclusions.pages[1].url.should == 'http://example.com/'
       @sitemap_for_exclusions.pages[2].url.should == 'http://example.com/faqs'
     end
 
+    it "should exclude non-html pages from XML sitemaps" do
+      @sitemap_for_exclusions_xml.pages.length.should == 2
+      @sitemap_for_exclusions_xml.pages[0].url.should == 'http://example.com/'
+      @sitemap_for_exclusions_xml.pages[1].url.should == 'http://example.com/faqs'
+    end
+
     it "should not crash when URLs have international characters" do
       @sitemap_international.pages.length.should == 9
       @sitemap_international.pages[0].url.should == 'http://example.com/international'

data/spec/spec_helper.rb

@@ -14,6 +14,7 @@ FakeWeb.register_uri(:get, "http://zigotica.com", :response => open("#{$samples_
 FakeWeb.register_uri(:get, "http://protocol-relative.com", :response => open("#{$samples_dir}/protocol_relative.html").read)
 FakeWeb.register_uri(:get, "https://protocol-relative.com", :response => open("#{$samples_dir}/protocol_relative.html").read)
 FakeWeb.register_uri(:get, "http://example.com/exclusions", :response => open("#{$samples_dir}/exclusions.html").read)
+FakeWeb.register_uri(:get, "http://example.com/exclusions.xml", :response => open("#{$samples_dir}/exclusions.xml").read)
 FakeWeb.register_uri(:get, "http://w3clove.com/faqs", :response => open("#{$samples_dir}/w3clove_faqs.response").read)
 FakeWeb.register_uri(:get, "http://example.com/international", :response => open("#{$samples_dir}/international.response").read)
 

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: w3clove
 version: !ruby/object:Gem::Version 
-  hash: 17
+  hash: 23
   prerelease: 
   segments: 
   - 0
   - 7
-  - 9
-  version: 0.7.9
+  - 10
+  version: 0.7.10
 platform: ruby
 authors: 
 - Jaime Iniesta
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2012-09-12 00:00:00 Z
+date: 2012-11-02 00:00:00 Z
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: w3c_validators
@@ -162,6 +162,7 @@ files:
 - spec/samples/absolute_links.html
 - spec/samples/eparreno.com.html
 - spec/samples/exclusions.html
+- spec/samples/exclusions.xml
 - spec/samples/guides.rubyonrails.org.html
 - spec/samples/international.response
 - spec/samples/protocol_relative.html