mecab-light 0.1.1 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 9bc1034d15fd3f35b2b1ad7d0ef37c2cc73cc827
-  data.tar.gz: cfbdc262474d9d30a56e9963e9f8b586f319dd33
+  metadata.gz: 8a5511af0c985b3d07f4d529ca04c7138657867c
+  data.tar.gz: 89da28098265ec53e39c01bd3b6fedc75ff614f5
 SHA512:
-  metadata.gz: 1880e49178c264d6c9fb4e702d022d41742f2f6d9d0f9c46fd25367b155744b55c61518063d03e743dcd9734f10b306070c06bf6ff678f2d29ddcefc9203b614
-  data.tar.gz: aa31d0b6a51a66a097815dc30ca973a20fb23d468a8f1ef8cf68ff562f23cf409f4087fe2deaaeea13b23a84a2ed9acba32706ddd781bca038c9f77ed20bfdcf
+  metadata.gz: f1a8cbd1738fcab2d3b3d508cd45440c3755791fe8a061704ba91dad37aef5254719c42f60820ad3fdbf533de9101170e159b57c4cf7bed7529e19a4cec5d52e
+  data.tar.gz: 29244fa47f165c0e26ddffc1e8a733fa97945250988a82ff088eea7da13dde505c4877c7acf51fe9d89b54119d3a8b839e09f2244515f7b46330e7a7b1572040

data/README.md

@@ -1,10 +1,13 @@
-# mecab-light [![Build Status](https://travis-ci.org/hadzimme/mecab-light.png)](https://travis-ci.org/hadzimme/mecab-light)
+# MeCab::Light [![Build Status](https://travis-ci.org/hadzimme/mecab-light.png)](https://travis-ci.org/hadzimme/mecab-light)
 
 Use a sequence of results as an Enumerable object.
 
 ## Installation
 
-At first, install the official gem "mecab-ruby" manually.
+This gem works without official binding.
+MeCab::Light is a 'light' tool.
+This supports only the 'parse' method for now,
+and the method's feature is totally different from its original.
 
 Add this line to your application's Gemfile:
 

data/lib/mecab/light/binding.rb

@@ -0,0 +1,10 @@
+module MeCab
+  module Light
+    module Binding
+      extend FFI::Library
+      ffi_lib 'mecab'
+      attach_function :mecab_new2, [:string], :pointer
+      attach_function :mecab_sparse_tostr, [:pointer, :string], :string
+    end
+  end
+end

data/lib/mecab/light/tagger.rb

@@ -1,12 +1,22 @@
 module MeCab
   module Light
     class Tagger
-      def initialize(*args)
-        @core_tagger = MeCab::Tagger.new(*args)
+      def initialize
+        @mecab = Binding.mecab_new2('')
       end
 
       def parse(string)
-        Result.new(@core_tagger.parse_to_enum(string))
+        Result.new(parse_to_enum(string))
+      end
+
+      private
+      def parse_to_enum(string)
+        parse_to_str(string).sub(/EOS\n$/, '').each_line
+      end
+
+      def parse_to_str(string)
+        encoding = Encoding.default_external
+        Binding.mecab_sparse_tostr(@mecab, string).force_encoding(encoding)
       end
     end
   end

data/lib/mecab/light/version.rb

@@ -1,5 +1,5 @@
 module MeCab
   module Light
-    VERSION = "0.1.1"
+    VERSION = "0.2.0"
   end
 end

data/lib/mecab/light.rb

@@ -1,6 +1,6 @@
-require 'MeCab'
-require 'mecab/tagger'
+require 'ffi'
 require 'mecab/light/version'
 require 'mecab/light/morpheme'
 require 'mecab/light/result'
 require 'mecab/light/tagger'
+require 'mecab/light/binding'

data/mecab-light.gemspec

@@ -18,6 +18,8 @@ Gem::Specification.new do |gem|
   gem.require_paths = ["lib"]
   gem.required_ruby_version = '>= 1.9'
 
+  gem.add_runtime_dependency 'ffi'
+
   gem.add_development_dependency 'rake'
   gem.add_development_dependency 'rspec'
 end

data/spec/mecab_spec.rb

@@ -1,9 +1,27 @@
 # coding: utf-8
 
-$:.unshift(File.expand_path(File.dirname(__FILE__)) + '/lib')
-
 require 'rspec'
-require 'mecab/light'
+begin
+  require 'mecab/light'
+rescue LoadError
+  puts 'loaded mecab-light but MeCab does not exist.'
+  puts 'dummy binding will be loaded.'
+end
+
+module MeCab
+  module Light
+    module Binding
+      def self.mecab_new2(arg)
+        :dummy_mecab
+      end
+
+      def self.mecab_sparse_tostr(mecab, string)
+        result = "見る\t動詞,自立,*,*,一段,基本形,見る,ミル,ミル\nEOS\n"
+        result.force_encoding('ASCII-8BIT')
+      end
+    end
+  end
+end
 
 describe MeCab::Light::Morpheme do
   context 'when initialized with the result line of the word "見る"' do
@@ -14,11 +32,11 @@ describe MeCab::Light::Morpheme do
     subject { @morpheme }
     its(:surface){ should eq '見る' }
     its(:feature){ should eq '動詞,自立,*,*,一段,基本形,見る,ミル,ミル' }
-    describe '#surface' do
+    describe 'surface' do
       subject { @morpheme.surface }
       its(:encoding){ should be Encoding::UTF_8 }
     end
-    describe '#feature' do
+    describe 'feature' do
       subject { @morpheme.feature }
       its(:encoding){ should be Encoding::UTF_8 }
     end
@@ -34,7 +52,7 @@ describe MeCab::Light::Result do
     it { should respond_to :each }
     it { should be_an Enumerable }
     its(:count){ should be 1 }
-    describe '#each' do
+    describe 'each' do
       context 'with block' do
         subject { @result.each{} }
         it 'should return self' do
@@ -48,13 +66,13 @@ describe MeCab::Light::Result do
         end
       end
     end
-    describe '#[]' do
+    describe '[]' do
       context 'when argument 0' do
         subject { @result[0] }
         it { should be_an_instance_of MeCab::Light::Morpheme }
       end
     end
-    describe '#at' do
+    describe 'at' do
       context 'when argument 0' do
         subject { @result.at(0) }
         it { should be_an_instance_of MeCab::Light::Morpheme }
@@ -62,25 +80,11 @@ describe MeCab::Light::Result do
     end
   end
 end
-describe MeCab::Tagger do
-  before { @tagger = MeCab::Tagger.new }
-  subject { @tagger }
-  it { should respond_to :parse_to_enum }
-  describe '#parse_to_enum' do
-    context 'when argument "見る"' do
-      subject { @tagger.parse_to_enum('見る') }
-      it 'should return Enumerator' do
-        should be_an_instance_of Enumerator
-      end
-      its(:count){ should be 1 }
-    end
-  end
-end
 describe MeCab::Light::Tagger do
   before { @tagger = MeCab::Light::Tagger.new }
   subject { @tagger }
   it { should respond_to :parse }
-  describe '#parse' do
+  describe 'parse' do
     context 'when argument "見る"' do
       before { @result = @tagger.parse('見る') }
       subject { @result }
@@ -88,15 +92,15 @@ describe MeCab::Light::Tagger do
         should be_an_instance_of MeCab::Light::Result
       end
       its(:count){ should be 1 }
-      describe 'returned MeCab::Light::Result object' do
-        context '#[0]' do
+      describe 'MeCab::Light::Result (returned)' do
+        context '[0]' do
           before { @morpheme = @result[0] }
           subject { @morpheme }
           it 'should return MeCab::Light::Morpheme object' do
             should be_an_instance_of MeCab::Light::Morpheme
           end
-          describe 'returned MeCab::Light::Morpheme object' do
-            context '#surface' do
+          describe 'MeCab::Light::Morpheme (returned)' do
+            context 'surface' do
               subject { @morpheme.surface }
               it { should eq '見る' }
             end

metadata

@@ -1,15 +1,29 @@
 --- !ruby/object:Gem::Specification
 name: mecab-light
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.2.0
 platform: ruby
 authors:
 - Hajime WAKAHARA
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-03-19 00:00:00.000000000 Z
+date: 2013-03-26 00:00:00.000000000 Z
 dependencies:
+- !ruby/object:Gem::Dependency
+  name: ffi
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
@@ -51,13 +65,12 @@ files:
 - README.md
 - Rakefile
 - lib/mecab/light.rb
+- lib/mecab/light/binding.rb
 - lib/mecab/light/morpheme.rb
 - lib/mecab/light/result.rb
 - lib/mecab/light/tagger.rb
 - lib/mecab/light/version.rb
-- lib/mecab/tagger.rb
 - mecab-light.gemspec
-- spec/lib/MeCab.rb
 - spec/mecab_spec.rb
 homepage: https://github.com/hadzimme/mecab-light
 licenses: []
@@ -83,5 +96,4 @@ signing_key:
 specification_version: 4
 summary: A Thin Wrapper for mecab-ruby
 test_files:
-- spec/lib/MeCab.rb
 - spec/mecab_spec.rb

data/lib/mecab/tagger.rb

@@ -1,7 +0,0 @@
-module MeCab
-  class Tagger
-    def parse_to_enum(string)
-      self.parse(string).sub(/EOS\n$/, '').each_line
-    end
-  end
-end

data/spec/lib/MeCab.rb

@@ -1,9 +0,0 @@
-# coding: utf-8
-
-module MeCab
-  class Tagger
-    def parse(string)
-      "見る\t動詞,自立,*,*,一段,基本形,見る,ミル,ミル\nEOS\n"
-    end
-  end
-end