domain_prefix 0.2.2.4 → 0.3.0.0

data/domain_prefix.gemspec

@@ -2,14 +2,16 @@
 # DO NOT EDIT THIS FILE DIRECTLY
 # Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
 # -*- encoding: utf-8 -*-
+# stub: domain_prefix 0.3.0.0 ruby lib
 
 Gem::Specification.new do |s|
   s.name = "domain_prefix"
-  s.version = "0.2.2.4"
+  s.version = "0.3.0.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
+  s.require_paths = ["lib"]
   s.authors = ["tadman"]
-  s.date = "2013-09-16"
+  s.date = "2014-02-28"
   s.description = "A library to extract information about top-level domain and registered name from generic and international domain names"
   s.email = "github@tadman.ca"
   s.extra_rdoc_files = [
@@ -18,6 +20,8 @@ Gem::Specification.new do |s|
   ]
   s.files = [
     ".document",
+    "Gemfile",
+    "Gemfile.lock",
     "LICENSE",
     "README.md",
     "Rakefile",
@@ -25,23 +29,32 @@ Gem::Specification.new do |s|
     "data/effective_tld_names.dat",
     "domain_prefix.gemspec",
     "lib/domain_prefix.rb",
+    "lib/domain_prefix/tree.rb",
     "test/helper.rb",
     "test/sample/README",
     "test/sample/test.txt",
     "test/test_domain_prefix.rb"
   ]
   s.homepage = "http://github.com/twg/domain_prefix"
-  s.require_paths = ["lib"]
-  s.rubygems_version = "1.8.23"
+  s.rubygems_version = "2.2.0"
   s.summary = "Domain Prefix Extraction Library"
 
   if s.respond_to? :specification_version then
-    s.specification_version = 3
+    s.specification_version = 4
 
     if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
+      s.add_runtime_dependency(%q<simpleidn>, [">= 0.0.5"])
+      s.add_development_dependency(%q<turn>, [">= 0.9.0"])
+      s.add_development_dependency(%q<jeweler>, [">= 2.0.0"])
     else
+      s.add_dependency(%q<simpleidn>, [">= 0.0.5"])
+      s.add_dependency(%q<turn>, [">= 0.9.0"])
+      s.add_dependency(%q<jeweler>, [">= 2.0.0"])
     end
   else
+    s.add_dependency(%q<simpleidn>, [">= 0.0.5"])
+    s.add_dependency(%q<turn>, [">= 0.9.0"])
+    s.add_dependency(%q<jeweler>, [">= 2.0.0"])
   end
 end
 

data/lib/domain_prefix.rb

@@ -1,65 +1,11 @@
 # encoding: UTF-8
+require 'simpleidn'
 
 module DomainPrefix
-  SEPARATOR = '.'.freeze
-
-  class Tree < Hash
-    def insert(path)
-      leaf = path.split(SEPARATOR).reverse.inject(self) do |tree, component|
-        # Seeds an element into the tree structure by referencing it
-        tree[component.sub(/^!/, '')] ||= Tree.new
-      end
-
-      if (path.match(/^[\!]/))
-        leaf[:required] = 0
-      else
-        leaf[:required] = 1
-      end
-      
-      self
-    end
-    
-    def follow(path)
-      path = path.to_s.split(SEPARATOR) unless (path.is_a?(Array))
-      path = path.reverse
-
-      index = traverse(path)
-
-      index and index <= path.length and path[0, index].reverse
-    end
-
-  protected
-    def traverse(path, index = 0)
-      component = path[index]
-
-      unless (component)
-        return self[:required] == 0 ? index : nil
-      end
-
-      named_branch = self[component]
+  require 'domain_prefix/tree'
 
-      if (named_branch)
-        result = named_branch.traverse(path, index + 1)
-
-        return result if (result)
-      end
-
-      wildcard_branch = self["*"]
-      
-      if (wildcard_branch)
-        result = wildcard_branch.traverse(path, index + 1)
-
-        return result if (result)
-      end
-
-      if (!named_branch and !wildcard_branch and self[:required])
-        return index + self[:required]
-      end
+  SEPARATOR = '.'.freeze
 
-      return
-    end
-  end
-  
   TLDIFIER_SOURCE_FILE = File.expand_path(File.join('..', 'data', 'effective_tld_names.dat'), File.dirname(__FILE__))
 
   TLD_SET = begin
@@ -82,19 +28,24 @@ module DomainPrefix
     'local' => true
   }.freeze
   
+  # Returns a cleaned up, canonical version of a domain name.
   def rfc3492_canonical_domain(domain)
-    # FIX: Full implementation of http://www.ietf.org/rfc/rfc3492.txt required
+    # FIX: Full implementation of RFC3429 required.
+    # http://www.ietf.org/rfc/rfc3492.txt
     domain and domain.downcase
   end
 
+  # Returns true if the given tld is listed as public, false otherwise.
   def public_tld?(tld)
     !NONPUBLIC_TLD.key?(tld)
   end
   
-  def registered_domain(domain, rules = :strict)
-    return unless (domain)
+  # Returns the registered domain name for a given FQDN or nil if one cannot
+  # be determined.
+  def registered_domain(fqdn, rules = :strict)
+    return unless (fqdn)
     
-    components = rfc3492_canonical_domain(domain).split(SEPARATOR)
+    components = rfc3492_canonical_domain(fqdn).split(SEPARATOR)
     
     return if (components.empty? or components.find(&:empty?))
 
@@ -115,6 +66,8 @@ module DomainPrefix
     suffix.join(SEPARATOR)
   end
 
+  # Returns the public suffix (e.g. "co.uk") for a given domain or nil if one
+  # cannot be determined.
   def public_suffix(domain)
     return unless (domain)
 
@@ -133,15 +86,19 @@ module DomainPrefix
     suffix.join(SEPARATOR)
   end
 
-  def tld(domain)
-    suffix = public_suffix(rfc3492_canonical_domain(domain))
+  # Returns the very top-level domain for a given domain, or nil if one cannot
+  # be determined.
+  def tld(fqdn)
+    suffix = public_suffix(rfc3492_canonical_domain(fqdn))
     
     suffix and suffix.split(SEPARATOR).last
   end
-  
-  def name(domain)
-    if (domain = registered_domain(domain))
-      domain.split(SEPARATOR).first
+
+  # Returns the name component of a given domain or nil if one cannot be
+  # determined.  
+  def name(fqdn)
+    if (fqdn = registered_domain(fqdn))
+      fqdn.split(SEPARATOR).first
     else
       nil
     end

data/lib/domain_prefix/tree.rb

@@ -0,0 +1,61 @@
+class DomainPrefix::Tree < Hash
+  def insert(path)
+    components = path.sub(/^!/, '').split(DomainPrefix::SEPARATOR).reverse
+
+    leaves = components.inject([ self ]) do |trees, part|
+      [ part, SimpleIDN.to_unicode(part), SimpleIDN.to_ascii(part) ].uniq.flat_map do |l|
+        trees.collect do |tree|
+          tree[l] ||= self.class.new
+        end
+      end
+    end
+
+    required = path.match(/^[\!]/) ? 0 : 1
+
+    leaves.each do |leaf|
+      leaf[:required] = required
+    end
+    
+    self
+  end
+  
+  def follow(path)
+    path = path.to_s.split(SEPARATOR) unless (path.is_a?(Array))
+    path = path.reverse
+
+    index = traverse(path)
+
+    index and index <= path.length and path[0, index].reverse
+  end
+
+protected
+  def traverse(path, index = 0)
+    component = path[index]
+
+    unless (component)
+      return self[:required] == 0 ? index : nil
+    end
+
+    named_branch = self[component]
+
+    if (named_branch)
+      result = named_branch.traverse(path, index + 1)
+
+      return result if (result)
+    end
+
+    wildcard_branch = self["*"]
+    
+    if (wildcard_branch)
+      result = wildcard_branch.traverse(path, index + 1)
+
+      return result if (result)
+    end
+
+    if (!named_branch and !wildcard_branch and self[:required])
+      return index + self[:required]
+    end
+
+    return
+  end
+end

data/test/helper.rb

@@ -1,6 +1,8 @@
 require 'rubygems'
 require 'test/unit'
 
+require 'turn'
+
 $LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
 $LOAD_PATH.unshift(File.dirname(__FILE__))
 

data/test/sample/test.txt

@@ -60,12 +60,12 @@ checkPublicSuffix('a.b.c.kobe.jp', 'b.c.kobe.jp');
 checkPublicSuffix('city.kobe.jp', 'city.kobe.jp');
 checkPublicSuffix('www.city.kobe.jp', 'city.kobe.jp');
 // TLD with a wildcard rule and exceptions.
-checkPublicSuffix('om', null);
-checkPublicSuffix('test.om', null);
-checkPublicSuffix('b.test.om', 'b.test.om');
-checkPublicSuffix('a.b.test.om', 'b.test.om');
-checkPublicSuffix('songfest.om', 'songfest.om');
-checkPublicSuffix('www.songfest.om', 'songfest.om');
+checkPublicSuffix('ck', null);
+checkPublicSuffix('test.ck', null);
+checkPublicSuffix('b.test.ck', 'b.test.ck');
+checkPublicSuffix('a.b.test.ck', 'b.test.ck');
+checkPublicSuffix('www.ck', 'www.ck');
+checkPublicSuffix('www.www.ck', 'www.ck');
 // US K12.
 checkPublicSuffix('us', null);
 checkPublicSuffix('test.us', 'test.us');
@@ -75,4 +75,24 @@ checkPublicSuffix('test.ak.us', 'test.ak.us');
 checkPublicSuffix('www.test.ak.us', 'test.ak.us');
 checkPublicSuffix('k12.ak.us', null);
 checkPublicSuffix('test.k12.ak.us', 'test.k12.ak.us');
-checkPublicSuffix('www.test.k12.ak.us', 'test.k12.ak.us');
+checkPublicSuffix('www.test.k12.ak.us', 'test.k12.ak.us');
+// IDN labels.
+checkPublicSuffix('食狮.com.cn', '食狮.com.cn');
+checkPublicSuffix('食狮.公司.cn', '食狮.公司.cn');
+checkPublicSuffix('www.食狮.公司.cn', '食狮.公司.cn');
+checkPublicSuffix('shishi.公司.cn', 'shishi.公司.cn');
+checkPublicSuffix('公司.cn', null);
+checkPublicSuffix('食狮.中国', '食狮.中国');
+checkPublicSuffix('www.食狮.中国', '食狮.中国');
+checkPublicSuffix('shishi.中国', 'shishi.中国');
+checkPublicSuffix('中国', null);
+// Same as above, but punycoded.
+checkPublicSuffix('xn--85x722f.com.cn', 'xn--85x722f.com.cn');
+checkPublicSuffix('xn--85x722f.xn--55qx5d.cn', 'xn--85x722f.xn--55qx5d.cn');
+checkPublicSuffix('www.xn--85x722f.xn--55qx5d.cn', 'xn--85x722f.xn--55qx5d.cn');
+checkPublicSuffix('shishi.xn--55qx5d.cn', 'shishi.xn--55qx5d.cn');
+checkPublicSuffix('xn--55qx5d.cn', null);
+checkPublicSuffix('xn--85x722f.xn--fiqs8s', 'xn--85x722f.xn--fiqs8s');
+checkPublicSuffix('www.xn--85x722f.xn--fiqs8s', 'xn--85x722f.xn--fiqs8s');
+checkPublicSuffix('shishi.xn--fiqs8s', 'shishi.xn--fiqs8s');
+checkPublicSuffix('xn--fiqs8s', null);

data/test/test_domain_prefix.rb

@@ -2,6 +2,7 @@ require_relative 'helper'
 
 class TestDomainPrefix < Test::Unit::TestCase
   def test_initialization
+    assert DomainPrefix::TLD_SET.length > 0
   end
   
   def test_examples
@@ -15,7 +16,7 @@ class TestDomainPrefix < Test::Unit::TestCase
       'example.gc.ca' =>  %w[ example.gc.ca gc.ca ],
       'example.co.uk' =>  %w[ example.co.uk co.uk ],
       'example.com.au' => %w[ example.com.au com.au ],
-      'example.au' => [ nil, nil ],
+      'example.au' => %w[ example.au au ],
       'example.bar.jp' => %w[ bar.jp jp ],
       'example.bar.hokkaido.jp' =>%w[ bar.hokkaido.jp hokkaido.jp ],
       'example.metro.tokyo.jp' => %w[ metro.tokyo.jp tokyo.jp ]

metadata

@@ -1,16 +1,57 @@
 --- !ruby/object:Gem::Specification
 name: domain_prefix
 version: !ruby/object:Gem::Version
-  version: 0.2.2.4
-  prerelease: 
+  version: 0.3.0.0
 platform: ruby
 authors:
 - tadman
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-09-16 00:00:00.000000000 Z
-dependencies: []
+date: 2014-02-28 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: simpleidn
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.0.5
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.0.5
+- !ruby/object:Gem::Dependency
+  name: turn
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.9.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.9.0
+- !ruby/object:Gem::Dependency
+  name: jeweler
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.0.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.0.0
 description: A library to extract information about top-level domain and registered
   name from generic and international domain names
 email: github@tadman.ca
@@ -20,7 +61,9 @@ extra_rdoc_files:
 - LICENSE
 - README.md
 files:
-- .document
+- ".document"
+- Gemfile
+- Gemfile.lock
 - LICENSE
 - README.md
 - Rakefile
@@ -28,32 +71,32 @@ files:
 - data/effective_tld_names.dat
 - domain_prefix.gemspec
 - lib/domain_prefix.rb
+- lib/domain_prefix/tree.rb
 - test/helper.rb
 - test/sample/README
 - test/sample/test.txt
 - test/test_domain_prefix.rb
 homepage: http://github.com/twg/domain_prefix
 licenses: []
+metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.23
+rubygems_version: 2.2.0
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: Domain Prefix Extraction Library
 test_files: []