lorem_jp 0.0.1

data/lib/lorem_jp.rb

@@ -0,0 +1,223 @@
+# -*- coding: utf-8 -*-
+
+# Class for Japanese Lorem Ipsum generation
+#
+class LoremJP
+  # Version number
+  VERSION = '0.0.1'
+
+  # Default dictionary directory
+  DICTIONARY_DIR     = File.absolute_path('../../data', __FILE__)
+  # Default dictionary file
+  DEFAULT_DICTIONARY = File.join(DICTIONARY_DIR, 'dict.txt')
+
+  class << self
+    # Singleton interface of {#sentence}
+    # @param [Hash] options
+    # @option options [Integer] :chain number of words considered as
+    #                                  past state in Marcov chain
+    # @option options [String]  :dictionary file name of dictionary
+    # @return [String] Japanese meaningless sentence
+    def sentence(options = {})
+      return singleton_for_dict(options[:dictionary]).sentence(options)
+    end
+
+    private
+
+    def singleton_for_dict(dictionary)
+      @singleton ||= {}
+
+      @singleton[dictionary] ||= self.new(:dictionary => dictionary,
+                                          :lazy => true)
+
+      return @singleton[dictionary]
+    end
+  end
+
+  # @param [Hash] options
+  # @option options [String]  :dictionary file name of dictionary
+  # @option options [Boolean] :lazy       load dictionary file
+  #                                       on first generation if true
+  def initialize(options = {})
+    @dictionary = options[:dictionary]
+    @chain      = options[:chain] || 1
+
+    lazy = options[:lazy] || false
+
+    @dict   = []
+    @tree   = {}
+    @loaded = false
+
+    if ! lazy
+      load_dict(@dictionary)
+    end
+  end
+
+  # @param [Hash] options
+  # @option options [Integer] :chain number of words considered as past state
+  #                                  in Marcov chain
+  # @return [String] Japanese meaningless sentence
+  def sentence(options = {})
+    unless @loaded
+      load_dict(@dictionary)
+    end
+
+    chain = options[:chain] || @chain
+
+    unless chain > 0
+      raise ArgumentError,
+            "invalid chain option value (#{chain})."
+    end
+    unless chain <= @chain
+      raise ArgumentError,
+            "chain option value (#{chain}) exceeds dict's chain (#{@chain})"
+    end
+
+    tokens = []
+    stack = [ 0 ] * chain
+
+    loop do
+      cands = lookup_candidates(stack)
+      cand = cands[rand(cands.length)]
+      break if cand < 0   # EOS
+
+      tokens << @dict[cand]
+
+      stack.shift
+      stack << cand
+    end
+
+    return tokens.join('')
+  end
+
+  private
+
+  def load_dict(dictionary)
+    if dictionary.respond_to?(:close)
+      load_dict_from_stream(dictionary)
+    else
+      load_dict_from_file(dictionary_file(dictionary))
+    end
+  end
+
+  def dictionary_file(filename)
+    pathname = nil
+    if filename
+      [ nil, DICTIONARY_DIR ].each do |basedir|
+        begin
+          pathname = File.realpath(filename, basedir)
+          break
+        rescue Errno::ENOENT
+        end
+      end
+    else
+      filename = DEFAULT_DICTIONARY
+      pathname = File.realpath(filename)
+    end
+
+    unless pathname
+      raise ArgumentError,
+            "dictionary file (#{filename}) not found"
+    end
+
+    return pathname
+  end
+
+  def load_dict_from_file(filename)
+    open(filename, 'r:utf-8:utf-8') { |handle|
+      load_dict_from_stream(handle)
+    }
+  end
+
+  def load_dict_from_stream(stream)
+    step = 0
+
+    stack = []
+
+    stream.each do |line|
+      line.chomp!
+
+      case step
+      when 0
+        # chain
+        @chain = line.to_i
+        step = 1
+      when 1
+        # first word dict entry is '' (empty)
+        @dict << line
+        step = 2
+      when 2
+        # word dictionary
+        if line == ''   # separator
+          step = 3
+        else
+          @dict << line
+        end
+      else
+        # probability tree
+
+        # turn heading spaces into preceding stack
+        new_stack = []
+        tokens = line.split(%r{[ ]}xmo)
+        while tokens.length > 0
+          if tokens[0].empty?
+            tokens.shift    # trim first (empty) token
+            new_stack << stack.shift
+          else
+            new_stack << tokens.join('')
+            tokens = []
+          end
+        end
+        stack = new_stack
+
+        insert_tree_node(stack)
+      end
+    end
+
+    @loaded = true
+  end
+
+  def insert_tree_node(stack)
+    stack = stack.dup
+
+    node = @tree
+    while stack.length > 0
+      token = stack.shift
+
+      if token =~ %r{=}xmo
+        child, cands = token.split(%r{=}, 2)
+
+        word_id = child.to_i
+
+        node[word_id] = cands.split(%r{, \s*}xmo).map { |token| token.to_i }
+
+        break
+      else
+        word_id = token.to_i
+        node[word_id] ||= {}
+        node = node[word_id]
+      end
+    end
+  end
+
+  def lookup_candidates(stack)
+    stack = stack.dup
+
+    node = @tree
+    while stack.length > 0
+      break if node.nil?
+      break if node.is_a?(Array)
+
+      word = stack.shift
+      node = node[word]
+    end
+
+    if node.is_a?(Hash)
+      return node.keys
+    elsif node.is_a?(Array)
+      return node
+    else
+      return [ -1 ]   # EOS
+    end
+  end
+end

data/lib/lorem_jp/cli.rb

@@ -0,0 +1,40 @@
+# -*- coding: utf-8 -*-
+require 'lorem_jp'
+require 'optparse'
+
+class LoremJP
+
+  # Class for command line interface of Japanese Lorem Ipsum
+  #
+  class CLI
+    # Main routine for command line interface of +lorem_jp+
+    # @return [void]
+    def self.main
+      dict  = nil
+      chain = nil
+
+      opt = OptionParser.new
+
+      opt.on('-f DICT',  'dictionary filename') {
+        |v| dict = v
+      }
+      opt.on('-c CHAIN', 'chain of precedences' +
+                         ' (default: set in dictionary)') {
+        |v| chain = v.to_i
+      }
+
+      opt.parse! ARGV
+
+      options = {}
+      options[:dictionary] = dict   if dict
+      options[:chain]      = chain  if chain
+
+      puts LoremJP.sentence(options)
+    end
+  end
+
+end
+
+if __FILE__ == $0
+  LoremJP::CLI.main
+end

data/lorem_jp.gemspec

@@ -0,0 +1,37 @@
+# -*- coding: utf-8 -*-
+
+Gem::Specification.new do |spec|
+  spec.name          = 'lorem_jp'
+  spec.version       = '0.0.1'
+  spec.authors       = ['ITO Nobuaki']
+  spec.email         = ['daydream.trippers@gmail.com']
+  spec.description   = %q{Japanese Lorem Ipsum generator}
+  spec.summary       = %q{Japanese Lorem Ipsum generator}
+  spec.homepage      = 'https://github.com/dayflower/lorem_jp/'
+  spec.license       = 'MIT'
+
+  spec.files         = [
+    'lorem_jp.gemspec',
+    'Gemfile',
+    'Rakefile',
+    'LICENSE.txt',
+    'README.md',
+    'lib/lorem_jp.rb',
+    'lib/lorem_jp/cli.rb',
+    'data/dict.txt',
+    'bin/lorem_jp',
+    'build/make_dict.rb',
+    'build/fetcher.rb',
+    'build/recipes/789.rb',
+    'build/recipes/2363.rb',
+    'build/recipes/52960.rb',
+    'test/lorem_jp_spec.rb',
+  ]
+
+  spec.executables   = spec.files.grep(%r{^bin/}) { |f| File.basename(f) }
+  spec.test_files    = spec.files.grep(%r{^(test|spec|features)/})
+  spec.require_paths = ['lib']
+
+  spec.add_development_dependency 'bundler', '~> 1.3'
+  spec.add_development_dependency 'rake'
+end

data/test/lorem_jp_spec.rb

@@ -0,0 +1,60 @@
+# -*- coding: utf-8 -*-
+require 'minitest/spec'
+require 'minitest/autorun'
+require 'lorem_jp'
+require 'stringio'
+
+describe LoremJP do
+  it 'generate sentence with 1 chain' do
+    dict = StringIO.new(<<-'END_DICT')
+1
+
+A
+B
+C
+D
+
+0=1
+1=2
+2=3,4
+3=-1
+4=-1
+    END_DICT
+
+    lorem = LoremJP.new( :dictionary => dict, :lazy => false )
+
+    5.times do
+      lorem.sentence.must_match %r{^ A B (C | D) $}xmo
+    end
+  end
+
+  it 'generate sentence with 2 chain' do
+    dict = StringIO.new(<<-'END_DICT')
+2
+
+A
+B
+C
+D
+
+0
+ 0=1
+ 1=2,4
+1
+ 2=3,4
+ 4=2,3
+2
+ 3=-1
+ 4=-1
+4
+ 2=-1
+ 3=-1
+    END_DICT
+
+    lorem = LoremJP.new( :dictionary => dict, :lazy => false )
+
+    5.times do
+      lorem.sentence.must_match %r{^ A (B (C|D) | D (B|C) ) $}xmo
+    end
+  end
+end

metadata

@@ -0,0 +1,90 @@
+--- !ruby/object:Gem::Specification
+name: lorem_jp
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+platform: ruby
+authors:
+- ITO Nobuaki
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-06-10 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: bundler
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.3'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.3'
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+description: Japanese Lorem Ipsum generator
+email:
+- daydream.trippers@gmail.com
+executables:
+- lorem_jp
+extensions: []
+extra_rdoc_files: []
+files:
+- lorem_jp.gemspec
+- Gemfile
+- Rakefile
+- LICENSE.txt
+- README.md
+- lib/lorem_jp.rb
+- lib/lorem_jp/cli.rb
+- data/dict.txt
+- bin/lorem_jp
+- build/make_dict.rb
+- build/fetcher.rb
+- build/recipes/789.rb
+- build/recipes/2363.rb
+- build/recipes/52960.rb
+- test/lorem_jp_spec.rb
+homepage: https://github.com/dayflower/lorem_jp/
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.0.2
+signing_key: 
+specification_version: 4
+summary: Japanese Lorem Ipsum generator
+test_files:
+- test/lorem_jp_spec.rb
+has_rdoc: