morphy 0.0.2

data/lib/morphy.rb

@@ -0,0 +1,98 @@
+require "dawg"
+
+class Morphy
+  class Word
+    attr_accessor :para_id
+    
+    def initialize(word,para_id,index)
+      @word = word
+      @para_id = para_id.to_i
+      @index = index.to_i
+      @prefix_id = Morphy.paradigms[@para_id][@index*3]
+      @suffix_id = Morphy.paradigms[@para_id][@index*3+1]
+      @grammeme_id = Morphy.paradigms[@para_id][@index*3+2]
+      
+    end
+    
+    def to_s
+      @word      
+    end
+    
+    def normal_form
+      prefix = Morphy.prefixes[Morphy.paradigms[@para_id][0]]
+      suffix = Morphy.suffixes[Morphy.paradigms[@para_id][1]]      
+      
+      "#{prefix}#{stem}#{suffix}"
+    end
+    def grammemes
+      Morphy.grammemes[@grammeme_id]      
+    end
+    def stem
+      prefix = Morphy.prefixes[Morphy.paradigms[@para_id][0]]
+      suffix = Morphy.suffixes[Morphy.paradigms[@para_id][1]]
+      grammeme = Morphy.grammemes[Morphy.paradigms[@para_id][2]]
+      word = @word.dup
+      word.gsub!(Morphy.prefixes[@prefix_id],"")
+      word.gsub!(Morphy.suffixes[@suffix_id],"")
+      word
+    end
+
+    def same_paradigm?(other)
+      @para_id == other.para_id      
+    end
+
+    def tag
+      Morphy.grammemes[@grammeme_id].join(",") 
+    end
+
+    def lexemme
+      (0..(Morphy.paradigms[@para_id].length / 3)-1).map do |index|        
+        prefix = Morphy.prefixes[Morphy.paradigms[@para_id][index*3]]
+        suffix = Morphy.suffixes[Morphy.paradigms[@para_id][index*3+1]]
+        grammeme = Morphy.grammemes[Morphy.paradigms[@para_id][index*3+2]]
+        Word.new(prefix+stem+suffix,@para_id,index)
+      end      
+    end
+
+    def inflect(grammemes)
+      words = lexemme
+      words.each do |word|
+        return word if word.grammemes.last(grammemes.length) == grammemes
+      end
+      nil
+    end      
+  end
+  def initialize    
+    
+    path = File.dirname(__FILE__)+"/dictionary/"
+
+    @dawg = Dawg.load("#{path}/dawg.dat") # why it's eating so much memory?
+    @@suffixes ||= File.open("#{path}/suffixes.txt", 'r').read.split("\n")
+    @@prefixes ||= File.open("#{path}/prefixes.txt", 'r').read.split("\n")
+    @@grammemes ||= File.open("#{path}/grammemes.txt", 'r').read.split("\n").map{|g| g.split(",")}    
+    @@paradigms ||= Marshal.load(File.read("#{path}/paradigms.dat"))
+  end
+  def self.paradigms
+    @@paradigms    
+  end
+  def self.prefixes
+    @@prefixes    
+  end
+  def self.suffixes
+    @@suffixes
+  end
+  def self.grammemes
+    @@grammemes    
+  end
+  def find_similar(word)
+    results = @dawg.find_similar(word)
+    results = results.map do |result|
+      word,para_id,index = result.split(" ")
+      Word.new(word,para_id,index)
+    end
+    results
+  end
+  def to_s
+    "Morphy"    
+  end
+end

data/morphy.gemspec

@@ -0,0 +1,19 @@
+# coding: utf-8
+
+Gem::Specification.new do |spec|
+  spec.name          = "morphy"
+  spec.version       = "0.0.2"
+  spec.date          = '2015-05-29'
+  spec.authors       = ["Maksatbek Mansurov"]
+  spec.email         = ["maksat.mansurov@gmail.com"]
+  spec.description   = %q{Morphological analyzer (POS tagger + inflection engine) for Russian language in ruby. Inspired by pymorphy2}
+  spec.summary       = %q{Morphological analyzer for Russian language in ruby. Inspired by pymorphy2}
+  spec.homepage      = "https://github.com/baltavay/morphy"
+  spec.license       = "MIT"
+
+  spec.files         = `git ls-files`.split($/)
+  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 'dawg', '~> 0.0', '>= 0.0.2'
+end

metadata

@@ -0,0 +1,76 @@
+--- !ruby/object:Gem::Specification
+name: morphy
+version: !ruby/object:Gem::Version
+  version: 0.0.2
+platform: ruby
+authors:
+- Maksatbek Mansurov
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2015-05-29 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: dawg
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.0'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.0.2
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.0'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.0.2
+description: Morphological analyzer (POS tagger + inflection engine) for Russian language
+  in ruby. Inspired by pymorphy2
+email:
+- maksat.mansurov@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- ".gitignore"
+- Gemfile
+- LICENSE
+- README.md
+- lib/dictionary/dawg.dat
+- lib/dictionary/grammemes.txt
+- lib/dictionary/paradigms.dat
+- lib/dictionary/prefixes.txt
+- lib/dictionary/suffixes.txt
+- lib/morphy.rb
+- morphy.gemspec
+homepage: https://github.com/baltavay/morphy
+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: 2.2.2
+signing_key: 
+specification_version: 4
+summary: Morphological analyzer for Russian language in ruby. Inspired by pymorphy2
+test_files: []