jRuby Tests for twitter-cldr-rb Normalizer

Gemfile

source :rubygems
gem "twitter_cldr", :path => "~/dev/repos/twitter-cldr-rb"
gem "rspec"

test_java_normalizer.rb

# encoding: UTF-8

include Java

require 'rubygems'
require 'bundler/setup'
require 'lucene-test-framework-3.6.0.jar'
require 'lucene-core-3.6.0'
require 'rspec'
require 'twitter_cldr'

java_import(['java.util.Random', 'java.text.Normalizer', 'org.apache.lucene.util._TestUtil'])

def random_unicode_string
  JavaUtilities.get_proxy_class('org.apache.lucene.util._TestUtil').randomRealisticUnicodeString(Random.new)
end

def java_normalize(string)
  Normalizer.normalize(string, Normalizer::Form::NFD)
end

def ruby_normalize(string)
  TwitterCldr::Normalizers::NFD.normalize(string)
end

include TwitterCldr::Normalizers
describe NFD do
  it "normalizes any unicode string exactly like java.text.Normalizer" do
    50.times do
      p string = random_unicode_string
      
      string_code_point = string.split('').map { |char| Base.char_to_code_point(char) }
      ruby_code_point = ruby_normalize(string).split('').map { |char| Base.char_to_code_point(char) }
      java_code_point = java_normalize(string).split('').map { |char| Base.char_to_code_point(char) }
      
      p "String is #{string_code_point.inspect}"
      p "Ruby is #{ruby_code_point.inspect}"
      p "Java is #{java_code_point.inspect}"
    
      ruby_normalize(string).should == java_normalize(string)
    end
  end
end