superslau-pinyinify 1.0.0

data/lib/pinyinify.rb

@@ -0,0 +1,49 @@
+
+module Pinyinify
+
+  # :stopdoc:
+  VERSION = '1.0.0'
+  LIBPATH = ::File.expand_path(::File.dirname(__FILE__)) + ::File::SEPARATOR
+  PATH = ::File.dirname(LIBPATH) + ::File::SEPARATOR
+  # :startdoc:
+
+  # Returns the version string for the library.
+  #
+  def self.version
+    VERSION
+  end
+
+  # Returns the library path for the module. If any arguments are given,
+  # they will be joined to the end of the libray path using
+  # <tt>File.join</tt>.
+  #
+  def self.libpath( *args )
+    args.empty? ? LIBPATH : ::File.join(LIBPATH, args.flatten)
+  end
+
+  # Returns the lpath for the module. If any arguments are given,
+  # they will be joined to the end of the path using
+  # <tt>File.join</tt>.
+  #
+  def self.path( *args )
+    args.empty? ? PATH : ::File.join(PATH, args.flatten)
+  end
+
+  # Utility method used to require all files ending in .rb that lie in the
+  # directory below this file that has the same name as the filename passed
+  # in. Optionally, a specific _directory_ name can be passed in such that
+  # the _filename_ does not have to be equivalent to the directory.
+  #
+  def self.require_all_libs_relative_to( fname, dir = nil )
+    dir ||= ::File.basename(fname, '.*')
+    search_me = ::File.expand_path(
+        ::File.join(::File.dirname(fname), dir, '**', '*.rb'))
+
+    Dir.glob(search_me).sort.each {|rb| require rb}
+  end
+
+end  # module Pinyinify
+
+Pinyinify.require_all_libs_relative_to(__FILE__)
+
+# EOF

data/lib/pinyinify/transliterate/transliterate.rb

@@ -0,0 +1,36 @@
+module Pinyinify
+  
+  require 'yaml'
+  require 'rubygems'  
+  require 'active_support'  
+  require 'active_support/multibyte'
+  
+  class Transliterate
+
+    attr_reader :characters
+    
+    def transliterate(string)
+      result = ""
+      ActiveSupport::Multibyte::Chars.u_unpack(string).each_with_index {|c,i|
+         lookup_result=lookup(c)
+         lookup_result==nil ?  result=result+string.mb_chars[i] : result=result+lookup_result.strip+" " 
+      }
+      result.strip
+    end
+    
+    def initialize
+      
+      hanzi_file_yml = File.join(File.dirname(__FILE__), %w[ .. .. .. data hanzi.yml ])
+      @characters = YAML.load( IO.read(hanzi_file_yml) )
+
+    end
+    
+    private
+      
+      def lookup(character_code) 
+        ucode="U"+character_code.to_s(base=16).upcase      #convert into form U12345
+        res = @characters[ucode]
+        res==nil ? nil : res.split(" ").first.gsub(/[^-A-Z\s]/, "").downcase
+      end      
+  end
+end

data/pinyinify.gemspec

@@ -0,0 +1,37 @@
+# -*- encoding: utf-8 -*-
+
+Gem::Specification.new do |s|
+  s.name = %q{pinyinify}
+  s.version = "1.0.0"
+
+  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
+  s.authors = ["FIXME (who is writing this software)"]
+  s.date = %q{2009-03-15}
+  s.default_executable = %q{pinyinify}
+  s.description = %q{Transliterate from Mandarin to Pinyin}
+  s.email = %q{superslau@gmail.com}
+  s.executables = ["pinyinify"]
+  s.extra_rdoc_files = ["History.txt", "README.txt", "bin/pinyinify"]
+  s.files = ["History.txt", "README.txt", "Rakefile", "bin/pinyinify", "data/hanzi.dat", "data/hanzi.yml", "lib/pinyinify.rb", "lib/pinyinify/transliterate/transliterate.rb", "pinyinify.gemspec", "spec/pinyinify/transliterate/transliterate_spec.rb", "spec/pinyinify_spec.rb", "spec/spec_helper.rb", "test/test_pinyinify.rb"]
+  s.has_rdoc = true
+  s.homepage = %q{http://github.com/superslau/pinyinify/}
+  s.rdoc_options = ["--main", "README.txt"]
+  s.require_paths = ["lib"]
+  s.rubyforge_project = %q{pinyinify}
+  s.rubygems_version = %q{1.3.1}
+  s.summary = %q{Transliterate from Mandarin to Pinyin}
+  s.test_files = ["test/test_pinyinify.rb"]
+
+  if s.respond_to? :specification_version then
+    current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
+    s.specification_version = 2
+
+    if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
+      s.add_development_dependency(%q<bones>, [">= 2.4.2"])
+    else
+      s.add_dependency(%q<bones>, [">= 2.4.2"])
+    end
+  else
+    s.add_dependency(%q<bones>, [">= 2.4.2"])
+  end
+end

data/spec/pinyinify/transliterate/transliterate_spec.rb

@@ -0,0 +1,58 @@
+# spec/pinyinify/transliterate/transliterate_spec.rb
+# $Id$
+
+# Require the spec helper relative to this file
+require File.join(File.dirname(__FILE__), %w[ .. .. spec_helper])
+
+# No need to type Pinyinify:: before each call
+include Pinyinify
+
+describe Transliterate do
+  setup do
+      @t = Transliterate.new
+  end
+      
+  # All of our specs for Transliterate will go in here
+  it "should be instantiated without any arguments" do
+    lambda { Transliterate.new() }.should_not raise_error
+    lambda { Transliterate.new( "moo" )}.should raise_error(ArgumentError)
+  end
+  
+  it "should accept a string on a transliterate method, and return a string" do
+    @t.transliterate("text").should be_instance_of(String)
+  end
+  
+  it "should accept a non chinese string and return the same value back " do
+    @t.transliterate("text test english").should == "text test english"
+  end  
+  
+  it "should accept a chinese character and return the correct transliterated string" do
+    @t.transliterate("周").should be_instance_of(String)
+    @t.transliterate("周").should == "zhou"
+  end
+  
+  it "should accept a chinese phrase and return the correct transliterated string" do
+    @t.transliterate("最長的電影").should be_instance_of(String)
+    @t.transliterate("最長的電影").should == "zui chang de dian ying" #this transliteration came from itunes :)
+    @t.transliterate("东风破").should == "dong feng po" #this transliteration also came from itunes    
+  end
+  
+  it "should accept a mixed chinese and english phrase and return the correct transliterated string" do
+
+    @t.transliterate("最長的電影is a good song").should == "zui chang de dian ying is a good song" 
+  end
+  
+  it "should load the hanzi data in from the data file on instantiation" do
+    IO.should_receive(:read).and_return(
+      "--- \nU742B: BENG3\nU3E61: XUAN2 XUAN4"
+    )
+    
+    t = Transliterate.new
+    t.characters.should be_a_kind_of(Hash)
+    t.characters.keys.size.should == 2    
+  end
+  
+  
+end
+
+# EOF

data/spec/pinyinify_spec.rb

@@ -0,0 +1,7 @@
+
+require File.join(File.dirname(__FILE__), %w[spec_helper])
+
+describe Pinyinify do
+end
+
+# EOF

data/spec/spec_helper.rb

@@ -0,0 +1,16 @@
+
+require File.expand_path(
+    File.join(File.dirname(__FILE__), %w[.. lib pinyinify]))
+
+Spec::Runner.configure do |config|
+  # == Mock Framework
+  #
+  # RSpec uses it's own mocking framework by default. If you prefer to
+  # use mocha, flexmock or RR, uncomment the appropriate line:
+  #
+  # config.mock_with :mocha
+  # config.mock_with :flexmock
+  # config.mock_with :rr
+end
+
+# EOF

metadata

@@ -0,0 +1,77 @@
+--- !ruby/object:Gem::Specification 
+name: superslau-pinyinify
+version: !ruby/object:Gem::Version 
+  version: 1.0.0
+platform: ruby
+authors: 
+- FIXME (who is writing this software)
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-03-15 00:00:00 -07:00
+default_executable: pinyinify
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: bones
+  type: :development
+  version_requirement: 
+  version_requirements: !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: 2.4.2
+    version: 
+description: Transliterate from Mandarin to Pinyin
+email: superslau@gmail.com
+executables: 
+- pinyinify
+extensions: []
+
+extra_rdoc_files: 
+- History.txt
+- README.txt
+- bin/pinyinify
+files: 
+- History.txt
+- README.txt
+- Rakefile
+- bin/pinyinify
+- data/hanzi.dat
+- data/hanzi.yml
+- lib/pinyinify.rb
+- lib/pinyinify/transliterate/transliterate.rb
+- pinyinify.gemspec
+- spec/pinyinify/transliterate/transliterate_spec.rb
+- spec/pinyinify_spec.rb
+- spec/spec_helper.rb
+- test/test_pinyinify.rb
+has_rdoc: true
+homepage: http://github.com/superslau/pinyinify/
+post_install_message: 
+rdoc_options: 
+- --main
+- README.txt
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+requirements: []
+
+rubyforge_project: pinyinify
+rubygems_version: 1.2.0
+signing_key: 
+specification_version: 2
+summary: Transliterate from Mandarin to Pinyin
+test_files: 
+- test/test_pinyinify.rb