Implement sorting using twitter_cldr

lib/naturally.rb

@@ -33,6 +33,10 @@ def self.sort_by(an_array, an_attribute=nil, &block)
     an_array.sort_by { |obj| normalize(obj.send(an_attribute)) }
   end
 
+  def self.sort_with_collator(an_array, collator)
+    an_array.sort_by { |x| normalize(x, collator) }
+  end
+
   # Convert the given number to an array of {Segment}s.
   # This enables it to be sorted against other arrays
   # by the built-in #sort method.
@@ -44,9 +48,9 @@ def self.sort_by(an_array, an_attribute=nil, &block)
   #                 such as 1.2a.3.
   # @return [Array<Segment>] an array of Segments which
   #         can be sorted naturally via a standard #sort.
-  def self.normalize(complex_number)
+  def self.normalize(complex_number, collator = nil)
     tokens = complex_number.to_s.gsub(/\_/,'').scan(/\p{Word}+/)
-    tokens.map { |t| Segment.new(t) }
+    tokens.map { |t| Segment.new(t, collator) }
   end
 
   private

lib/naturally/segment.rb

@@ -6,12 +6,19 @@ module Naturally
   class Segment
     include Comparable
 
-    def initialize(v)
+    def initialize(v, collator = nil)
       @val = v
+      @collator = collator
     end
 
     def <=>(other)
-      to_array <=> other.to_array
+      other_array = other.to_array
+
+      if @collator
+        compare_using_collator_for_strings(to_array, other_array)
+      else
+        to_array <=> other.to_array
+      end
     end
 
     # @return [Array] a representation of myself in array form
@@ -40,5 +47,58 @@ def to_array
         [:str, @val]
       end
     end
+
+    private
+
+    # Compare to arrays according to the rules of Ruby, using a collator to
+    # compare String elements.
+    # https://github.com/ruby/ruby/blob/v3_0_1/array.c#L5173-L5210
+    #
+    # call-seq:
+    #   array <=> other_array -> -1, 0, or 1
+    #
+    # Returns -1, 0, or 1 as +self+ is less than, equal to, or greater than +other_array+.
+    # For each index +i+ in +self+, evaluates <tt>result = self[i] <=> other_array[i]</tt>.
+    #
+    # Returns -1 if any result is -1:
+    #   [0, 1, 2] <=> [0, 1, 3] # => -1
+    #
+    # Returns 1 if any result is 1:
+    #   [0, 1, 2] <=> [0, 1, 1] # => 1
+    #
+    # When all results are zero:
+    # - Returns -1 if +array+ is smaller than +other_array+:
+    #     [0, 1, 2] <=> [0, 1, 2, 3] # => -1
+    # - Returns 1 if +array+ is larger than +other_array+:
+    #     [0, 1, 2] <=> [0, 1] # => 1
+    # - Returns 0 if +array+ and +other_array+ are the same size:
+    #     [0, 1, 2] <=> [0, 1, 2] # => 0
+    #
+    def compare_using_collator_for_strings(array_1, array_2)
+      cmp = 0
+
+      array_1.each_with_index do |element, index|
+        next unless index < array_2.length
+
+        case element
+        when Integer, Symbol
+          cmp = element <=> array_2[index]
+        when String
+          cmp = @collator.compare(element, array_2[index])
+        else
+          raise ArgumentError, "Cannot compare #{e.class} with #{array_2[index].class}"
+        end
+
+        break if cmp != 0
+      end
+
+      return cmp unless cmp == 0
+
+      length_difference = array_1.length - array_2.length
+
+      return 0 if length_difference == 0
+      return 1 if length_difference > 0
+      return -1
+    end
   end
 end

naturally.gemspec

@@ -13,6 +13,8 @@ Gem::Specification.new do |gem|
   gem.homepage      = "http://github.com/dogweather/naturally"
   gem.required_ruby_version = '>= 2.0'
 
+  gem.add_development_dependency 'twitter_cldr'
+
   gem.files         = `git ls-files`.split($/)
   gem.executables   = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
   gem.test_files    = gem.files.grep(%r{^(test|spec|features)/})

spec/naturally_spec.rb

@@ -208,4 +208,38 @@ def it_sorts(opts = {})
       ]
     end
   end
+
+  describe 'using a collator' do
+    require 'twitter_cldr'
+
+    let(:collator) { TwitterCldr::Collation::Collator.new(:de) }
+
+    it 'sorts unicode characters correctly when using a collator' do
+      unicode_strings = %w( Öl10 b öl3 a Öl1 Öl2 A B )
+      actual = Naturally.sort_with_collator(unicode_strings, collator)
+
+      expect(actual).to eq %w( a A b B öl3 Öl1 Öl2 Öl10 )
+    end
+
+    # https://github.com/dogweather/naturally/issues/20#issuecomment-450617803
+    it 'sorts neither like the Duden nor the telephone book for German' do
+      names = [
+        'Müller, Franziska',
+        'Muller, Inge',
+        'Müller, Hansi',
+        'Muller, Erika',
+        'Mueller, Gerd'
+      ]
+
+      actual = Naturally.sort_with_collator(names, collator)
+
+      expect(actual).to eq [
+        'Mueller, Gerd',
+        'Muller, Erika',
+        'Muller, Inge',
+        'Müller, Franziska',
+        'Müller, Hansi'
+      ]
+    end
+  end
 end