chenillen-pinyin 0.1.0

data/lib/pinyin.rb

@@ -0,0 +1,75 @@
+require 'singleton'
+
+class Pinyin
+  
+  # Initialize using Pinyin.new
+  # py = Pinyin.new
+  def initialize
+		ma = File.join(File.dirname(File.expand_path(__FILE__)),'../dict/Mandarin.dat')
+		@codes = {}
+		File.readlines(ma).each do |line|
+			nv = line.split(/\s/)
+			@codes[nv[0]] = nv[1]
+		end
+	end
+  
+	# Generating permlink using '-' split
+	# Show the permlink: Interesting-Ruby-Tidbits-That-Dont-Need-Separate-Posts-17
+	# You can also pass a split symbol to_permlink(str, '~')
+	def to_permlink(str)
+		str_to_pinyin(str,'-')
+	end
+
+	# Generating shorter permlink with first letter of each charater
+	# Like: zgyh stands for "中国银行"
+	def to_pinyin_abbr(str)
+		str_to_pinyin(str,'',false,true)
+	end
+
+	#第一个字取全部，后面首字母.名称缩写。eg. liudh 刘德华
+	def to_pinyin_abbr_else(str)
+		str_to_pinyin(str,'',true,nil) 
+	end
+
+	#通用情况 tone为取第几声的标识。eg. ni3hao3zhong1guo2
+	def to_pinyin(str,separator='',tone=false)
+		str_to_pinyin(str,separator,false,false,tone)
+	end  	
+  	
+  	
+  # Private Methods Started...
+  private
+
+  def get_value(code)
+  	@codes[code]
+  end
+
+  def str_to_pinyin(str,separator='',abbr_else=false,abbr=false,tone=false)
+  		res = []
+  		str.unpack('U*').each_with_index do |t,idx|
+  			code = sprintf('%x',t).upcase
+  			val = get_value(code)
+  			#是否找到拼音？
+  			if val
+  				unless tone
+  					val = val.gsub(/\d/,'')
+  				end
+  				if (abbr and !abbr_else) or (abbr_else and idx!=0)
+  					val = val[0..0]
+  				end
+  				res << val.downcase+separator
+  			else
+  				tmp = [t].pack('U*')
+  				res << tmp if tmp =~ /^[_0-9a-zA-Z\s]*$/ #复原，去除特殊字符,如全角符号等。
+  			end
+  		end
+  		unless separator==''
+  			re = Regexp.new("\\#{separator}+")
+  			re2 = Regexp.new("\\#{separator}$")
+  			return res.join('').gsub(/\s+/,separator).gsub(re,separator).gsub(re2,'')
+  		else
+  			return res.join('')
+  		end
+    end
+end
+	

data/pinyin.gemspec

@@ -0,0 +1,31 @@
+# -*- encoding: utf-8 -*-
+
+Gem::Specification.new do |s|
+  s.name = %q{pinyin}
+  s.version = "0.1.0"
+
+  s.required_rubygems_version = Gem::Requirement.new(">= 1.2") if s.respond_to? :required_rubygems_version=
+  s.authors = ["Allen Chan"]
+  s.date = %q{2009-03-11}
+  s.description = %q{Convert Chinese Charaters to Pinyin letters.}
+  s.email = %q{chenillen@gmail.com}
+  s.extra_rdoc_files = ["lib/pinyin.rb", "README.rdoc"]
+  s.files = ["dict/Mandarin.dat", "lib/pinyin.rb", "Manifest", "Rakefile", "README.rdoc", "pinyin.gemspec"]
+  s.has_rdoc = true
+  s.homepage = %q{http://github.com/chenillen/uniquify}
+  s.rdoc_options = ["--line-numbers", "--inline-source", "--title", "Pinyin", "--main", "README.rdoc"]
+  s.require_paths = ["lib"]
+  s.rubyforge_project = %q{pinyin}
+  s.rubygems_version = %q{1.3.1}
+  s.summary = %q{Convert Chinese Charaters to Pinyin letters.}
+
+  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
+    else
+    end
+  else
+  end
+end

metadata

@@ -0,0 +1,64 @@
+--- !ruby/object:Gem::Specification 
+name: chenillen-pinyin
+version: !ruby/object:Gem::Version 
+  version: 0.1.0
+platform: ruby
+authors: 
+- Allen Chan
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-03-11 00:00:00 -07:00
+default_executable: 
+dependencies: []
+
+description: Convert Chinese Charaters to Pinyin letters.
+email: chenillen@gmail.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: 
+- lib/pinyin.rb
+- README.rdoc
+files: 
+- dict/Mandarin.dat
+- lib/pinyin.rb
+- Manifest
+- Rakefile
+- README.rdoc
+- pinyin.gemspec
+has_rdoc: true
+homepage: http://github.com/chenillen/uniquify
+post_install_message: 
+rdoc_options: 
+- --line-numbers
+- --inline-source
+- --title
+- Pinyin
+- --main
+- README.rdoc
+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: "1.2"
+  version: 
+requirements: []
+
+rubyforge_project: pinyin
+rubygems_version: 1.2.0
+signing_key: 
+specification_version: 2
+summary: Convert Chinese Charaters to Pinyin letters.
+test_files: []
+