ikku 0.0.4 → 0.0.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: a70c684dd795a3e2b7f41e2e48045ad85ce25771
-  data.tar.gz: 18cf28254326d74bfc6b7594530bbde28b224c53
+  metadata.gz: 4186a5706c4014ef2dc6ac0738e93ef39e548f3d
+  data.tar.gz: 91da30ecd5430d0a3cef04c40fb8df78a6d6c08a
 SHA512:
-  metadata.gz: 70e986f93c25c2b17ee3c9bfea3804491641d408bc8aeb2502a5614f4c1a43d2d994628f85df3574c124c3cd8996701794044f370206da4639d5069a600c79b3
-  data.tar.gz: eb527ec36dcc1570296e3026d34492c26571f4501e757f38a46181bff2165e35decbf5151ea8e4d01242504eef925bdc45f68c69814b53096e2f08e1ff319e87
+  metadata.gz: d8926c58122478c30e3226b888a51494fc9a552d8fdabe49f7f310a9bedeefc16dc275d76e4e1212b752f39f14f056866abcc027be7d11b2aabb474563d67dd0
+  data.tar.gz: 46213e3d2fae5cd31467d01236bb9efbe1bb513d4be587f249f0ae8f41d1ea2cd45a005138c72cee3d152df834015a540bd472beb370dbed0d6788bb49f38094

data/.rspec

@@ -0,0 +1,2 @@
+--color
+--require spec_helper

data/CHANGELOG.md

@@ -1,3 +1,6 @@
+## 0.0.5
+- Improve pattern about 自立・非自立・助動詞
+
 ## 0.0.4
 - Fix bug of Ikku::Scanner#consume
 

data/ikku.gemspec

@@ -20,4 +20,5 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "bundler", "~> 1.7"
   spec.add_development_dependency "pry"
   spec.add_development_dependency "rake", "~> 10.0"
+  spec.add_development_dependency "rspec", "3.2.0"
 end

data/lib/ikku.rb

@@ -1,305 +1,2 @@
-require "csv"
-require "natto"
+require "ikku/reviewer"
 require "ikku/version"
-
-module Ikku
-  class Reviewer
-    def initialize(rule: nil)
-      @rule = rule
-    end
-
-    # Find one available haiku from given text.
-    # @return [Array<Array>]
-    def find(text)
-      nodes = parser.parse(text)
-      nodes.length.times.find do |index|
-        if (phrases = Scanner.new(nodes[index..-1], rule: @rule).scan)
-          break phrases
-        end
-      end
-    end
-
-    # Judge if given text is haiku or not.
-    # @return [true, false]
-    def judge(text)
-      !Scanner.new(parser.parse(text), exactly: true, rule: @rule).scan.nil?
-    end
-
-    # Search all available haikus from given text.
-    # @return [Array<Array>]
-    def search(text)
-      nodes = parser.parse(text)
-      nodes.length.times.map do |index|
-        Scanner.new(nodes[index..-1], rule: @rule).scan
-      end.compact
-    end
-
-    private
-
-    def parser
-      @parser ||= Parser.new
-    end
-  end
-
-  # Find one haiku that starts from the 1st node of given nodes.
-  class Scanner
-    DEFAULT_RULE = [5, 7, 5]
-
-    attr_writer :count
-
-    def initialize(nodes, exactly: false, rule: nil)
-      @exactly = exactly
-      @nodes = nodes
-      @rule = rule
-    end
-
-    # @note Pronounciation count
-    def count
-      @count ||= 0
-    end
-
-    def scan
-      if has_valid_first_node? && has_valid_last_node?
-        @nodes.each_with_index do |node, index|
-          if consume(node)
-            if has_full_count?
-              return phrases unless @exactly
-            end
-          else
-            return
-          end
-        end
-        phrases if has_full_count?
-      end
-    end
-
-    private
-
-    def consume(node)
-      case
-      when node.pronounciation_length > max_consumable_length
-        false
-      when first_of_phrase? && !node.first_of_phrase?
-        false
-      else
-        phrases[phrase_index] ||= []
-        phrases[phrase_index] << node
-        self.count += node.pronounciation_length
-        true
-      end
-    end
-
-    def first_of_phrase?
-      rule.inject([]) do |array, length|
-        array << array.last.to_i + length
-      end.include?(count)
-    end
-
-    def has_full_count?
-      count == rule.inject(0, :+)
-    end
-
-    def has_valid_first_node?
-      @nodes.first.first_of_ikku?
-    end
-
-    def has_valid_last_node?
-      @nodes.last.last_of_ikku?
-    end
-
-    def max_consumable_length
-      rule[0..phrase_index].inject(0, :+) - count
-    end
-
-    def phrase_index
-      rule.length.times.find do |index|
-        count < rule[0..index].inject(0, :+)
-      end || rule.length - 1
-    end
-
-    def phrases
-      @phrases ||= []
-    end
-
-    def rule
-      @rule || DEFAULT_RULE
-    end
-  end
-
-  class Parser
-    def parse(text)
-      mecab.enum_parse(text).map do |mecab_node|
-        Node.new(mecab_node)
-      end.select(&:analyzable?)
-    end
-
-    private
-
-    def mecab
-      @mecab ||= Natto::MeCab.new
-    end
-  end
-
-  class Node
-    STAT_ID_FOR_NORMAL = 0
-    STAT_ID_FOR_UNKNOWN = 1
-    STAT_ID_FOR_BOS = 2
-    STAT_ID_FOR_EOS = 3
-
-    # @param node [Natto::MeCabNode]
-    def initialize(node)
-      @node = node
-    end
-
-    def analyzable?
-      !bos? && !eos?
-    end
-
-    def auxiliary_verb?
-      type == "助動詞"
-    end
-
-    def bos?
-      stat == STAT_ID_FOR_BOS
-    end
-
-    def dependent?
-      subtype1 == "非自立"
-    end
-
-    def element_of_ikku?
-      normal?
-    end
-
-    def last_of_ikku?
-      case
-      when type == "連体詞"
-        false
-      when ["名詞接続", "格助詞", "係助詞", "連体化", "接続助詞", "並立助詞", "副詞化", "数接続"].include?(type)
-        false
-      when auxiliary_verb? && root_form == "だ"
-        false
-      else
-        true
-      end
-    end
-
-    def eos?
-      stat == STAT_ID_FOR_EOS
-    end
-
-    def feature
-      @feature ||= CSV.parse(@node.feature)[0]
-    end
-
-    def filler?
-      type == "フィラー"
-    end
-
-    def first_of_ikku?
-      case
-      when !first_of_phrase?
-        false
-      # when filler?
-      #   false
-      when ["、", "・", " ", "　"].include?(surface)
-        false
-      else
-        true
-      end
-    end
-
-    def first_of_phrase?
-      case
-      when particle?
-        false
-      when auxiliary_verb?
-        false
-      when independent?
-        false
-      when postfix?
-        false
-      when dependent? && ["する", "できる"].include?(root_form)
-        false
-      else
-        true
-      end
-    end
-
-    def independent?
-      subtype1 == "自立"
-    end
-
-    def inspect
-      to_s.inspect
-    end
-
-    def normal?
-      stat == STAT_ID_FOR_NORMAL
-    end
-
-    def particle?
-      type == "助詞"
-    end
-
-    def postfix?
-      subtype1 == "接尾"
-    end
-
-    def pronounciation
-      feature[8]
-    end
-
-    def pronounciation_length
-      @pronounciation_length ||= begin
-        if pronounciation
-          pronounciation_mora.length
-        else
-          0
-        end
-      end
-    end
-
-    def pronounciation_mora
-      if pronounciation
-        pronounciation.tr("ぁ-ゔ","ァ-ヴ").gsub(/[^アイウエオカ-モヤユヨラ-ロワヲンヴー]/, "")
-      end
-    end
-
-    def root_form
-      feature[6]
-    end
-
-    def stat
-      @node.stat
-    end
-
-    def subtype1
-      feature[1]
-    end
-
-    def subtype2
-      feature[2]
-    end
-
-    def subtype3
-      feature[3]
-    end
-
-    def surface
-      @node.surface
-    end
-
-    def symbol?
-      type == "記号"
-    end
-
-    def to_s
-      surface
-    end
-
-    def type
-      feature[0]
-    end
-  end
-end

data/lib/ikku/node.rb

@@ -0,0 +1,119 @@
+require "csv"
+
+module Ikku
+  class Node
+    STAT_ID_FOR_NORMAL = 0
+    STAT_ID_FOR_UNKNOWN = 1
+    STAT_ID_FOR_BOS = 2
+    STAT_ID_FOR_EOS = 3
+
+    # @param node [Natto::MeCabNode]
+    def initialize(node)
+      @node = node
+    end
+
+    def analyzable?
+      !bos? && !eos?
+    end
+
+    def bos?
+      stat == STAT_ID_FOR_BOS
+    end
+
+    def eos?
+      stat == STAT_ID_FOR_EOS
+    end
+
+    def feature
+      @feature ||= CSV.parse(@node.feature)[0]
+    end
+
+    def first_of_ikku?
+      case
+      when !first_of_phrase?
+        false
+      when ["、", "・", " ", "　"].include?(surface)
+        false
+      else
+        true
+      end
+    end
+
+    def first_of_phrase?
+      case
+      when ["助詞", "助動詞"].include?(type)
+        false
+      when ["非自立", "接尾"].include?(subtype1)
+        false
+      when subtype1 == "自立" && ["する", "できる"].include?(root_form)
+        false
+      else
+        true
+      end
+    end
+
+    def inspect
+      to_s.inspect
+    end
+
+    def last_of_ikku?
+      !["名詞接続", "格助詞", "係助詞", "連体化", "接続助詞", "並立助詞", "副詞化", "数接続", "連体詞"].include?(type)
+    end
+
+    def normal?
+      stat == STAT_ID_FOR_NORMAL
+    end
+
+    def pronounciation
+      feature[8]
+    end
+
+    def pronounciation_length
+      @pronounciation_length ||= begin
+        if pronounciation
+          pronounciation_mora.length
+        else
+          0
+        end
+      end
+    end
+
+    def pronounciation_mora
+      if pronounciation
+        pronounciation.tr("ぁ-ゔ","ァ-ヴ").gsub(/[^アイウエオカ-モヤユヨラ-ロワヲンヴー]/, "")
+      end
+    end
+
+    def root_form
+      feature[6]
+    end
+
+    def stat
+      @node.stat
+    end
+
+    def subtype1
+      feature[1]
+    end
+
+    def subtype2
+      feature[2]
+    end
+
+    def subtype3
+      feature[3]
+    end
+
+    def surface
+      @node.surface
+    end
+
+    def to_s
+      surface
+    end
+
+    def type
+      feature[0]
+    end
+  end
+end

data/lib/ikku/parser.rb

@@ -0,0 +1,18 @@
+require "natto"
+require "ikku/node"
+
+module Ikku
+  class Parser
+    def parse(text)
+      mecab.enum_parse(text).map do |mecab_node|
+        Node.new(mecab_node)
+      end.select(&:analyzable?)
+    end
+
+    private
+
+    def mecab
+      @mecab ||= Natto::MeCab.new
+    end
+  end
+end

data/lib/ikku/reviewer.rb

@@ -0,0 +1,42 @@
+require "ikku/parser"
+require "ikku/scanner"
+
+module Ikku
+  class Reviewer
+    def initialize(rule: nil)
+      @rule = rule
+    end
+
+    # Find one available haiku from given text.
+    # @return [Array<Array>]
+    def find(text)
+      nodes = parser.parse(text)
+      nodes.length.times.find do |index|
+        if (phrases = Scanner.new(nodes[index..-1], rule: @rule).scan)
+          break phrases
+        end
+      end
+    end
+
+    # Judge if given text is haiku or not.
+    # @return [true, false]
+    def judge(text)
+      !Scanner.new(parser.parse(text), exactly: true, rule: @rule).scan.nil?
+    end
+
+    # Search all available haikus from given text.
+    # @return [Array<Array>]
+    def search(text)
+      nodes = parser.parse(text)
+      nodes.length.times.map do |index|
+        Scanner.new(nodes[index..-1], rule: @rule).scan
+      end.compact
+    end
+
+    private
+
+    def parser
+      @parser ||= Parser.new
+    end
+  end
+end

data/lib/ikku/scanner.rb

@@ -0,0 +1,86 @@
+module Ikku
+  # Find one haiku that starts from the 1st node of given nodes.
+  class Scanner
+    DEFAULT_RULE = [5, 7, 5]
+
+    attr_writer :count
+
+    def initialize(nodes, exactly: false, rule: nil)
+      @exactly = exactly
+      @nodes = nodes
+      @rule = rule
+    end
+
+    # @note Pronounciation count
+    def count
+      @count ||= 0
+    end
+
+    def scan
+      if has_valid_first_node? && has_valid_last_node?
+        @nodes.each_with_index do |node, index|
+          if consume(node)
+            if has_full_count?
+              return phrases unless @exactly
+            end
+          else
+            return
+          end
+        end
+        phrases if has_full_count?
+      end
+    end
+
+    private
+
+    def consume(node)
+      case
+      when node.pronounciation_length > max_consumable_length
+        false
+      when first_of_phrase? && !node.first_of_phrase?
+        false
+      else
+        phrases[phrase_index] ||= []
+        phrases[phrase_index] << node
+        self.count += node.pronounciation_length
+        true
+      end
+    end
+
+    def first_of_phrase?
+      rule.inject([]) do |array, length|
+        array << array.last.to_i + length
+      end.include?(count)
+    end
+
+    def has_full_count?
+      count == rule.inject(0, :+)
+    end
+
+    def has_valid_first_node?
+      @nodes.first.first_of_ikku?
+    end
+
+    def has_valid_last_node?
+      @nodes.last.last_of_ikku?
+    end
+
+    def max_consumable_length
+      rule[0..phrase_index].inject(0, :+) - count
+    end
+
+    def phrase_index
+      rule.length.times.find do |index|
+        count < rule[0..index].inject(0, :+)
+      end || rule.length - 1
+    end
+
+    def phrases
+      @phrases ||= []
+    end
+
+    def rule
+      @rule || DEFAULT_RULE
+    end
+  end
+end

data/lib/ikku/version.rb

@@ -1,3 +1,3 @@
 module Ikku
-  VERSION = "0.0.4"
+  VERSION = "0.0.5"
 end

data/spec/ikku/reviewer_spec.rb

@@ -0,0 +1,113 @@
+require "spec_helper"
+
+RSpec.describe Ikku::Reviewer do
+  let(:instance) do
+    described_class.new(rule: rule)
+  end
+
+  let(:rule) do
+    nil
+  end
+
+  let(:text) do
+    "古池や蛙飛び込む水の音"
+  end
+
+  describe "#find" do
+    subject do
+      instance.find(text)
+    end
+
+    context "without ikku" do
+      let(:text) do
+        "test"
+      end
+
+      it { is_expected.to be_nil }
+    end
+
+    context "with valid ikku" do
+      it { is_expected.to be_a Array }
+    end
+
+    context "with text including ikku" do
+      let(:text) do
+        "ああ#{super()}ああ"
+      end
+
+      it { is_expected.to be_a Array }
+    end
+  end
+
+  describe "#judge" do
+    subject do
+      instance.judge(text)
+    end
+
+    context "with valid ikku" do
+      it { is_expected.to be true }
+    end
+
+    context "with invalid ikku" do
+      let(:text) do
+        "#{super()}ああ"
+      end
+
+      it { is_expected.to be false }
+    end
+
+    context "with rule option and valid ikku" do
+      let(:rule) do
+        [4, 3, 5]
+      end
+
+      let(:text) do
+        "すもももももももものうち"
+      end
+
+      it { is_expected.to be true }
+    end
+
+    context "with rule option and invalid ikku" do
+      let(:rule) do
+        [4, 3, 5]
+      end
+
+      it { is_expected.to be false }
+    end
+
+    context "with phrase starting with independent verb (歩く)" do
+      let(:text) do
+        "なぜ鳩は頭を振って歩くのか"
+      end
+
+      it { is_expected.to be true }
+    end
+  end
+
+  describe "#search" do
+    subject do
+      instance.search(text)
+    end
+
+    context "without ikku" do
+      let(:text) do
+        "test"
+      end
+
+      it { is_expected.to be_a Array }
+    end
+
+    context "with valid ikku" do
+      it { is_expected.to be_a Array }
+    end
+
+    context "with text including ikku" do
+      let(:text) do
+        "ああ#{super()}ああ"
+      end
+
+      it { is_expected.to be_a Array }
+    end
+  end
+end

data/spec/spec_helper.rb

@@ -0,0 +1,18 @@
+$LOAD_PATH.unshift File.expand_path("../../lib", __FILE__)
+require "ikku"
+
+RSpec.configure do |config|
+  config.expect_with :rspec do |expectations|
+    expectations.include_chain_clauses_in_custom_matcher_descriptions = true
+  end
+
+  config.mock_with :rspec do |mocks|
+    mocks.verify_partial_doubles = true
+  end
+
+  config.filter_run :focus
+  config.run_all_when_everything_filtered = true
+  config.disable_monkey_patching!
+  config.warnings = true
+  config.default_formatter = "doc" if config.files_to_run.one?
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ikku
 version: !ruby/object:Gem::Version
-  version: 0.0.4
+  version: 0.0.5
 platform: ruby
 authors:
 - Ryo Nakamura
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-02-14 00:00:00.000000000 Z
+date: 2015-02-15 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: natto
@@ -66,6 +66,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '10.0'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.2.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.2.0
 description: 
 email:
 - r7kamura@gmail.com
@@ -74,6 +88,7 @@ extensions: []
 extra_rdoc_files: []
 files:
 - ".gitignore"
+- ".rspec"
 - CHANGELOG.md
 - Gemfile
 - LICENSE.txt
@@ -81,7 +96,13 @@ files:
 - Rakefile
 - ikku.gemspec
 - lib/ikku.rb
+- lib/ikku/node.rb
+- lib/ikku/parser.rb
+- lib/ikku/reviewer.rb
+- lib/ikku/scanner.rb
 - lib/ikku/version.rb
+- spec/ikku/reviewer_spec.rb
+- spec/spec_helper.rb
 homepage: https://github.com/r7kamura/ikku
 licenses:
 - MIT
@@ -106,5 +127,7 @@ rubygems_version: 2.4.5
 signing_key: 
 specification_version: 4
 summary: Discover haiku from text.
-test_files: []
+test_files:
+- spec/ikku/reviewer_spec.rb
+- spec/spec_helper.rb
 has_rdoc: