hanzi 0.2.0

data/lib/hanzi.rb

@@ -0,0 +1,81 @@
+# encoding: utf-8
+
+class Hanzi
+  class << self
+    attr_accessor :data
+
+    def load_data
+      return if @data
+      @data = []
+
+      file_path = File.expand_path('../../lib/data/cedict_ts.u8', __FILE__)
+      File.open(file_path).each_line do |line|
+        next if line.start_with?('#')
+        line = line.force_encoding('utf-8')
+
+        # CC-CEDICT format:
+        # Traditional Simplified [pin1 yin1] /English equivalent 1/equivalent 2/
+        line_data = {}
+        line_data[:traditional] = line[0, line.index(' ')]
+
+        line = line[line.index(' ') + 1, line.length]
+        line_data[:simplified] = line[0, line.index(' ')]
+
+        line = line[line.index('['), line.length]
+        line_data[:pinyin] = line[1, line.index(']') - 1].downcase
+
+        line = line[line.index('/'), line.rindex('/')]
+        line_data[:english] = line[1, line.rindex('/') - 1]
+
+        @data << line_data
+      end
+
+    end
+
+    def to_pinyin(text, options={})
+      load_data if @data.nil?
+
+      result = ''
+      pos = 0
+
+      loop do
+        char = text[pos]
+        break if !char
+
+        if char.ord < 0x4E00 || char.ord > 0x9FFF
+          # it's not a chinese character.
+          result << char
+          pos += 1
+        else
+          # it's a chinese character. start by trying to find a long word match,
+          # and if it fails, all the way down to a single hanzi.
+          match = nil
+          match_length = 0
+          4.downto(1) do |length|
+            match = find_match(text[pos, length])
+            match_length = length
+            break if match
+          end
+
+          if match
+            result << match[:pinyin].gsub("\s", '')
+            pos += match_length
+          else
+            result << char
+            pos += 1
+          end
+        end
+      end
+
+      result
+    end
+
+    private
+    def find_match(text)
+      entry = @data.find do |word|
+        word[:simplified] == text || word[:traditional] == text
+      end
+    end
+
+  end
+end

data/test/helper.rb

@@ -0,0 +1,17 @@
+require 'rubygems'
+require 'bundler'
+begin
+  Bundler.setup(:default, :development)
+rescue Bundler::BundlerError => e
+  $stderr.puts e.message
+  $stderr.puts "Run `bundle install` to install missing gems"
+  exit e.status_code
+end
+require 'test/unit'
+
+$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+$LOAD_PATH.unshift(File.dirname(__FILE__))
+require 'hanzi'
+
+class Test::Unit::TestCase
+end

data/test/test_hanzi.rb

@@ -0,0 +1,37 @@
+# encoding: utf-8
+
+require 'helper'
+
+class TestHanzi < Test::Unit::TestCase
+
+  def test_should_init_data
+    Hanzi.load_data
+    assert Hanzi.data.count > 0
+  end
+
+  def test_convert_with_tones
+    result = Hanzi.to_pinyin('为什么')
+    assert_equal 'wei4shen2me5', result
+  end
+
+  def test_second_word
+    result = Hanzi.to_pinyin('走红')
+    assert_equal 'zou3hong2', result
+  end
+
+  def test_can_convert_traditional
+    result = Hanzi.to_pinyin('簡單')
+    assert_equal 'jian3dan1', result
+  end
+
+  def test_can_convert_with_surrounding_english
+    result = Hanzi.to_pinyin('no! 为什么！')
+    assert_equal 'no! wei4shen2me5！', result
+  end
+
+  def test_can_convert_sentence_of_hanzi
+    result = Hanzi.to_pinyin('你好， 我是康昱辰。')
+    assert_equal 'ni3hao3， wo3shi4kang1yu4chen2。', result
+  end
+
+end

metadata

@@ -0,0 +1,95 @@
+--- !ruby/object:Gem::Specification
+name: hanzi
+version: !ruby/object:Gem::Version
+  version: 0.2.0
+  prerelease: 
+platform: ruby
+authors:
+- Steve Jackson
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-01-12 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rdoc
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
+- !ruby/object:Gem::Dependency
+  name: jeweler
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.8.4
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.8.4
+description: Convert Hanzi to pinyin. Unlike other similar gems, this includes tones
+  and can accurately translate common words.
+email: steven.j.jackson@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files:
+- LICENSE.txt
+- README.rdoc
+files:
+- .document
+- Gemfile
+- LICENSE.txt
+- README.rdoc
+- Rakefile
+- VERSION
+- hanzi.gemspec
+- lib/data/cedict_ts.u8
+- lib/hanzi.rb
+- test/helper.rb
+- test/test_hanzi.rb
+homepage: http://github.com/stevejackson/hanzi
+licenses:
+- MIT
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+      segments:
+      - 0
+      hash: 2598915566647370745
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.24
+signing_key: 
+specification_version: 3
+summary: Convert Hanzi to pinyin. Unlike other similar gems, this includes tones and
+  can accurately translate common words.
+test_files: []