Create analyzer class

hiraoyogi.gemspec

@@ -18,6 +18,7 @@ Gem::Specification.new do |spec|
   spec.test_files    = spec.files.grep(%r{^(test|spec|features)/})
   spec.require_paths = ["lib"]
 
+  spec.add_dependency "natto"
   spec.add_dependency "nokogiri"
   spec.add_dependency "redis"
 

lib/hiraoyogi.rb

@@ -1,3 +1,4 @@
+require "hiraoyogi/analyzer"
 require "hiraoyogi/crawler"
 require "hiraoyogi/database"
 require "hiraoyogi/version"

lib/hiraoyogi/analyzer.rb

@@ -0,0 +1,31 @@
+# -*- coding: utf-8 -*-
+
+require "natto"
+
+module Hiraoyogi
+  class Analyzer
+    DISALLOW_TYPE = %w(助詞 助動詞)
+
+    def initialize
+      @natto = Natto::MeCab.new
+    end
+
+    def analyze_text(text)
+      result = {}
+
+      @natto.parse(text) do |line|
+        next if DISALLOW_TYPE.include?(type(line.feature))
+        result[line.surface] ||= 0
+        result[line.surface] += 1
+      end
+
+      result
+    end
+
+    private
+
+    def type(feature)
+      feature.split(",")[0]
+    end
+  end
+end

spec/hiraoyogi/analyzer_spec.rb

@@ -0,0 +1,21 @@
+# -*- coding: utf-8 -*-
+
+require "spec_helper"
+
+module Hiraoyogi
+  describe Analyzer do
+    let(:analyzer) do
+      described_class.new
+    end
+
+    let(:text) do
+      "すもももももももものうち"
+    end
+
+    describe "#analyze_text" do
+      it "should create index table" do
+        expect(analyzer.analyze_text(text)).to include "すもも" => 1, "もも" => 2, "うち" => 1
+      end
+    end
+  end
+end