twingly-url 5.1.1 → 6.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 5a3c908215d3999313c0230890446e979995175f5ae34f05f778a8ff51ecf7de
-  data.tar.gz: f3546864fab23385d42f25f41506a74c6309716ce34b2cf00c5ce52bbf497d77
+  metadata.gz: 56c971217509cb49a880e44835cd28fb7cbd0993ed3c770e916dbc5692aa3a4a
+  data.tar.gz: d09caeb8ea2a01b7523565fa8088852922afbcd042ce650bb1de41a13010bed4
 SHA512:
-  metadata.gz: 2c8232dda0e4d15d8a22c99692ca76ae47a27ad5986222481aa2c6f2d6aec5f762ac26ac26a77a3653436c8a97647f8f00a2a4dd89086e16b7f13b1a06112147
-  data.tar.gz: 89e956970ea3b14ce5fccd527d4eedbc7c910749a0f6f60fc8b280316dab5de6e7c9a5a26ad724d41d487d9814547eba4184123c03df0d43a944238ad42558ed
+  metadata.gz: 4b960358767ae47df3d234adcb16242bce1de669b27af753691c206b5d8c0899ed2cf1272a1468fe74faa61e008647b015b028277d63878e1054301e01b43585
+  data.tar.gz: cb82dbf7140f0cae7ddc8068a6fe892ddb1bab5afde458d21f0c34e5fd6e36bfe189c916bcdc27f5f4f3db88130bcf288f75fe5955413c2102dde0d3d9d7b299

data/README.md

@@ -11,7 +11,6 @@ Twingly URL tools.
     * `Twingly::URL::Hasher.blogstream_hash(url)` - MD5 hexdigest
     * `Twingly::URL::Hasher.documentdb_hash(url)` - SHA256 unsigned long, native endian digest
     * `Twingly::URL::Hasher.autopingdb_hash(url)` - SHA256 64-bit signed, native endian digest
-    * `Twingly::URL::Hasher.pingloggerdb_hash(url)` - SHA256 64-bit unsigned, native endian digest
 * `twingly/url/utilities` - Utilities to work with URLs
     * `Twingly::URL::Utilities.extract_valid_urls` - Returns Array of valid `Twingly::URL`
 

data/lib/twingly/public_suffix_list.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 require "addressable/idna"
 require "public_suffix"
 

data/lib/twingly/url.rb

@@ -1,4 +1,6 @@
-require "addressable/idna"
+# frozen_string_literal: true
+
+require "addressable/idna/pure"
 require "addressable/uri"
 require "public_suffix"
 
@@ -11,19 +13,33 @@ module Twingly
   class URL
     include Comparable
 
-    ACCEPTED_SCHEMES = /\Ahttps?\z/i
+    ACCEPTED_SCHEMES = /\Ahttps?\z/i.freeze
     CUSTOM_PSL = PublicSuffixList.with_punycoded_names
-    ENDS_WITH_SLASH = /\/+$/
+    ENDS_WITH_SLASH = /\/+$/.freeze
+    STARTS_WITH_WWW = /\Awww\./i.freeze
     ERRORS_TO_EXTEND = [
       Addressable::IDNA::PunycodeBigOutput,
       Addressable::URI::InvalidURIError,
       PublicSuffix::DomainInvalid,
-    ]
+    ].freeze
+    NBSP = "\u00A0"
+    SPACE = "\u0020"
+    WHITESPACE_CHARS = [
+      NBSP,
+      SPACE,
+    ].join.freeze
+    LEADING_AND_TRAILING_WHITESPACE =
+      /\A[#{WHITESPACE_CHARS}]+|[#{WHITESPACE_CHARS}]+\z/.freeze
 
     private_constant :ACCEPTED_SCHEMES
     private_constant :CUSTOM_PSL
+    private_constant :STARTS_WITH_WWW
     private_constant :ENDS_WITH_SLASH
     private_constant :ERRORS_TO_EXTEND
+    private_constant :NBSP
+    private_constant :SPACE
+    private_constant :WHITESPACE_CHARS
+    private_constant :LEADING_AND_TRAILING_WHITESPACE
 
     class << self
       def parse(potential_url)
@@ -35,8 +51,9 @@ module Twingly
         raise
       end
 
-      def internal_parse(potential_url)
-        addressable_uri = to_addressable_uri(potential_url)
+      def internal_parse(input)
+        potential_url   = clean_input(input)
+        addressable_uri = Addressable::URI.heuristic_parse(potential_url)
         raise Twingly::URL::Error::ParseError if addressable_uri.nil?
 
         scheme = addressable_uri.scheme
@@ -58,15 +75,16 @@ module Twingly
         raise
       end
 
-      def to_addressable_uri(potential_url)
-       if potential_url.is_a?(Addressable::URI)
-          potential_url
-        else
-          potential_url = String(potential_url)
-          potential_url = potential_url.scrub
+      def clean_input(input)
+        input = String(input)
+        input = input.scrub
+        input = strip_whitespace(input)
+      end
 
-          Addressable::URI.heuristic_parse(potential_url)
-        end
+      def strip_whitespace(input)
+        return input unless input.encoding == Encoding::UTF_8
+
+        input.gsub(LEADING_AND_TRAILING_WHITESPACE, "")
       end
 
       # Workaround for the following bug in addressable:
@@ -83,7 +101,8 @@ module Twingly
 
       private :new
       private :internal_parse
-      private :to_addressable_uri
+      private :clean_input
+      private :strip_whitespace
       private :try_addressable_normalize
     end
 
@@ -189,6 +208,16 @@ module Twingly
       self.to_s <=> other.to_s
     end
 
+    def eql?(other)
+      return false unless other.is_a?(self.class)
+
+      self.hash == other.hash
+    end
+
+    def hash
+      self.to_s.hash
+    end
+
     def to_s
       addressable_uri.to_s
     end
@@ -203,7 +232,7 @@ module Twingly
 
     def normalize_blogspot(host, domain)
       if domain.sld.downcase == "blogspot"
-        host.sub(/\Awww\./i, "").sub(/#{domain.tld}\z/i, "com")
+        host.sub(STARTS_WITH_WWW, "").sub(/#{domain.tld}\z/i, "com")
       else
         host
       end

data/lib/twingly/url/error.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 module Twingly
   class URL
     module Error

data/lib/twingly/url/hasher.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 require 'digest'
 
 require "twingly/url"
@@ -28,10 +30,6 @@ module Twingly
       def autopingdb_hash(url)
         SHA256_DIGEST.digest(url).unpack("q")[0]
       end
-
-      def pingloggerdb_hash(url)
-        SHA256_DIGEST.digest(url).unpack("Q")[0]
-      end
     end
   end
 end

data/lib/twingly/url/null_url.rb

@@ -1,4 +1,4 @@
-require "twingly/url"
+# frozen_string_literal: true
 
 module Twingly
   class URL
@@ -24,6 +24,16 @@ module Twingly
         self.to_s <=> other.to_s
       end
 
+      def eql?(other)
+        return false unless other.is_a?(self.class)
+
+        self.hash == other.hash
+      end
+
+      def hash
+        self.to_s.hash
+      end
+
       def to_s
         ""
       end

data/lib/twingly/url/utilities.rb

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 require "twingly/url"
 
 module Twingly

data/lib/twingly/version.rb

@@ -1,5 +1,7 @@
+# frozen_string_literal: true
+
 module Twingly
   class URL
-    VERSION = "5.1.1"
+    VERSION = "6.0.0"
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: twingly-url
 version: !ruby/object:Gem::Version
-  version: 5.1.1
+  version: 6.0.0
 platform: ruby
 authors:
 - Twingly AB
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-02-14 00:00:00.000000000 Z
+date: 2019-02-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: addressable
@@ -107,15 +107,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - "~>"
     - !ruby/object:Gem::Version
-      version: '2.2'
+      version: '2.4'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.7.4
+rubygems_version: 3.0.1
 signing_key: 
 specification_version: 4
 summary: Ruby library for URL handling