url_trimmer 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: d26f5d3528cba18f8d691a3cabd30320d4aca627
-  data.tar.gz: bdefacd3a2b56bde497cdcca064807c1a58728ba
+  metadata.gz: 208db8066a5128f36d5d2a23435569e278346eb4
+  data.tar.gz: 353abe12c38767112e252f3d709fc78ce1a0fa2e
 SHA512:
-  metadata.gz: fdf00065dd4cad05963b8b35e1e0f9c7c38ae452c693411f841e0848e8108bf69390a135855718d7c45ac6938fe541e45cacf34a4a7909a7d8693a2737f051ee
-  data.tar.gz: 9b2f4c34f990a71e1ce250f3f5aa00e76034cd202ed7d1005719edace747aebce59490eae3e08418ad38d0b7bbedb3f5c9e8df8360e5eff68ae14303b3a143ca
+  metadata.gz: 626f3ad8f74050eb434382f8f40cdc2830dfac10812044db58ba99d0e9a0662eed8e78f705c25d5a32ce2b1ee6e405e6accc37e5d36df59a38f7ec71291a668c
+  data.tar.gz: b20cafc045eb1cf0cad0b8b2cb9e6fe60bb6c36de85efb51d0571959b9c41f008a1d49a044c1f087d30e7d35c5492a90646590de40c400b98f2a621c91f5a349

data/lib/url_trimmer.rb

@@ -3,6 +3,8 @@ require "domain_name"
 
 module URLTrimmer
   class Worker
+    URL_REGEXP = %r(\Ahttps?://([^/]+))
+
     def self.uniq_by_domain(urls)
       urls.map! do |url|
         begin
@@ -11,9 +13,9 @@ module URLTrimmer
           url.encode("UTF-8", invalid: :replace, undef: :replace, replace: "").downcase
         end
       end
-      urls.map! { |url| url[%r(\Ahttps?://[^/]+), 0] }
-      urls.compact!
-      urls.uniq! { |url| DomainName(url).domain }
+      urls.select! { |url| url =~ URL_REGEXP }
+      urls.uniq! { |url| DomainName(url[URL_REGEXP, 1]).domain }
+      urls.sort!
       urls
     end
   end

data/lib/url_trimmer/version.rb

@@ -1,3 +1,3 @@
 module URLTrimmer
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

data/spec/lib/url_trimmer_spec.rb

@@ -5,16 +5,16 @@ module URLTrimmer
   describe Worker do
     let(:urls) do
       ["http://www.google.com.ar/blah1", "https://www.google.com.ar/blah2", "https://www.google.com.br/blah3",
-       "http://www.google.com/blah4" "https://plus.google.com/blah5"]
+       "http://www.google.com/blah4", "https://plus.google.com/blah5"]
     end
 
     it "returns a list of unique URLs by domain" do
       unique_urls = Worker.uniq_by_domain(urls)
 
       unique_urls.size.must_equal 3
-      unique_urls.must_include("http://www.google.com.ar")
-      unique_urls.must_include("https://www.google.com.br")
-      unique_urls.must_include("http://www.google.com")
+      unique_urls.must_include("http://www.google.com.ar/blah1")
+      unique_urls.must_include("https://www.google.com.br/blah3")
+      unique_urls.must_include("http://www.google.com/blah4")
     end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: url_trimmer
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Cristian Rasch
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-06-28 00:00:00.000000000 Z
+date: 2014-06-30 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: domain_name