Created
October 21, 2012 18:22
-
-
Save clicube/3927974 to your computer and use it in GitHub Desktop.
seibenize string
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# coding: utf-8 | |
require 'uri' | |
require 'open-uri' | |
require 'rexml/document' | |
require 'json' | |
require_relative 'romkan' | |
# http://0xcc.net/ruby-romkan/ | |
YAHOO_APP_ID = 'YOUR YAHOO APPLICATION ID' | |
INVERSION_RATE = 0.03 | |
TYPO_RATE = 0.03 | |
TYPO_MAP = { | |
'a' => %w[q w s z], | |
'b' => %w[v g h n], | |
'c' => %w[x d f v], | |
'd' => %w[s e r f c x], | |
'e' => %w[w r d s], | |
'f' => %w[d r t g v c], | |
'g' => %w[f t y h b v], | |
'h' => %w[g y u j n b], | |
'i' => %w[u o k j], | |
'j' => %w[h u i k m n], | |
'k' => %w[j i o l m], | |
'l' => %w[k o p], | |
'm' => %w[n j k], | |
'n' => %w[b h j m], | |
'o' => %w[i p l k], | |
'p' => %w[o l], | |
'q' => %w[w a], | |
'r' => %w[e t f d], | |
's' => %w[a w e d x z], | |
't' => %w[r y g f], | |
'u' => %w[y i j h], | |
'v' => %w[c f g b], | |
'w' => %w[q e s a], | |
'x' => %w[z s d c], | |
'y' => %w[t u h g], | |
'z' => %w[a s x] | |
} | |
class String | |
def to_seibe | |
# puts "input: #{self}" | |
kana = '' | |
uri = 'http://jlp.yahooapis.jp/MAService/V1/parse?' | |
params = { | |
'appid' => YAHOO_APP_ID, | |
'sentence' => self, | |
'results' => 'ma', | |
'response' => 'reading' | |
} | |
uri << params.map{|k,v| "#{k}=#{URI.encode(v)}" }.join('&') | |
response = open(uri).read | |
xmldoc = REXML::Document.new response | |
xmldoc.elements.each('ResultSet/ma_result/word_list/word/reading') do |item| | |
kana << item.text | |
end | |
# puts "kana: #{kana}" | |
roma = kana.to_roma | |
# puts "roma: #{roma}" | |
roma = roma.split(//).inject do |res,c| | |
if rand > INVERSION_RATE | |
res << c | |
else | |
res[res.length-1,0] = c | |
end | |
res | |
end | |
roma = roma.split(//).map do |c| | |
if rand < TYPO_RATE && TYPO_MAP.has_key?(c) | |
TYPO_MAP[c].sample | |
else | |
c | |
end | |
end.join | |
seibed_roma = roma | |
# puts "seibed_roma: #{seibed_roma}" | |
seibed_kana = seibed_roma.to_kana | |
# puts "seibed_kana: #{seibed_kana}" | |
seibed_text = '' | |
uri = 'http://www.google.com/transliterate?' | |
params = { | |
'langpair' => 'ja-Hira|ja', | |
'text' => seibed_kana | |
} | |
uri << params.map{|k,v| "#{k}=#{URI.encode(v)}" }.join('&') | |
response = open(uri).read | |
translated_obj = JSON.parse(response) | |
translated_obj.each do |elm| | |
seibed_text << elm[1][0] | |
end | |
# puts "seibed_text: #{seibed_text}" | |
seibed_text | |
end | |
end | |
if $0 == __FILE__ | |
ARGV.each do |arg| | |
puts arg.to_seibe | |
end | |
end |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment