amakanize 0.2.0 → 0.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 29a3fe6104923c2dce5236cdd426331fac81ef9f
-  data.tar.gz: 48c1308766b0e8530b3b4653cdbc1778cd04f96d
+  metadata.gz: cd42f278e55fdace156f91282b117c2f32075cd8
+  data.tar.gz: 697300a04b3bc7a54ef611dad793bd1dd6bed448
 SHA512:
-  metadata.gz: 3de8f1cc840ae35402f9d726bb6a81452ece1becf040fb78ec28f688208c51005f6adc35c4b253016d5292392a555551ccbe61947c69fa0ebbff3b81a261a49f
-  data.tar.gz: d38495ab3280abd4f4c6eaaa3e2731be07d8496467015273fa0b6117f0a93cdb81198e4a835ef5817712a23310166d379db8c5a122f8361d52d1d357b98c68ea
+  metadata.gz: 792a8ec7d0e7715a81409a7ae847c5f4313bae35de67c3bb50bcfa3f854572793d96bb6984f9efdf9feaa68c7a89b9de3d48c1169f3cb1eec5baf94af0a8329d
+  data.tar.gz: 0aca318813fa1f33c5e459424ce30b88c1234d8eb86e0502911cd5807a6cd81fff4c89ce10a6360d215885d4578b2a7a9130be103070215bb89a020f5721eaf0

data/CHANGELOG.md

@@ -1,3 +1,8 @@
+## 0.2.1
+- Normalize angle brackets
+- Remove trailing LvN and Lv.N as volume number
+- Fix volume number detection pattern
+
 ## 0.2.0
 - Fix author name separator
 

data/lib/amakanize/filters/angle_brackets_after_word_normalization_filter.rb

@@ -0,0 +1,12 @@
+module Amakanize
+  module Filters
+    class AngleBracketsAfterWordNormalizationFilter < BaseFilter
+      # @note Override
+      # @param string [String] e.g. `"IS〈インフィニット・ストラトス〉 1 (オーバーラップ文庫)"`
+      # @return [String] e.g. `"IS<インフィニット・ストラトス> 1 (オーバーラップ文庫)"`
+      def call(string)
+        string.gsub(/([[:alnum:]])[〈《](.+?)[〉》]/, '\1<\2>')
+      end
+    end
+  end
+end

data/lib/amakanize/filters/obvious_volume_number_deletion_filter.rb

@@ -5,8 +5,8 @@ module Amakanize
       # @param string [String] e.g. `"刀語 第十一話 毒刀・鍍"`, `"アニウッド大通り 1: アニメ監督一家物語"`
       # @return [String] e.g. `"刀語"`, `"アニウッド大通り"`
       def call(string)
-        string.gsub(/\s*第?#{Amakanize::PATTERN_OF_NUMERIC_CHARACTER}+(?:話|巻|版).*/, "")
-          .gsub(/\s+第?#{Amakanize::PATTERN_OF_NUMERIC_CHARACTER}+(?:話|巻|版)?:\s+.*/, "")
+        string.gsub(/\s*第?#{Amakanize::PATTERN_OF_NUMERIC_CHARACTERS}(?:話|巻|版).*/, "")
+          .gsub(/\s+第?#{Amakanize::PATTERN_OF_NUMERIC_CHARACTERS}(?:話|巻|版)?:\s+.*/, "")
       end
     end
   end

data/lib/amakanize/filters/trailing_volume_number_deletion_filter.rb

@@ -2,10 +2,10 @@ module Amakanize
   module Filters
     class TrailingVolumeNumberDeletionFilter < BaseFilter
       # @note Override
-      # @param string [String] e.g. `"やはり俺の青春ラブコメはまちがっている。4"`
-      # @return [String] e.g. `"やはり俺の青春ラブコメはまちがっている。"`
+      # @param string [String] e.g. `"やはり俺の青春ラブコメはまちがっている。4"`, `"ネトゲの嫁は女の子じゃないと思った? Lv.2"`
+      # @return [String] e.g. `"やはり俺の青春ラブコメはまちがっている。"`, `"ネトゲの嫁は女の子じゃないと思った?"`
       def call(string)
-        string.gsub(/\s*第?#{Amakanize::PATTERN_OF_NUMERIC_CHARACTER}+(?:話|巻|版)?\z/, "")
+        string.gsub(/\s*(?:第|Lv\.?)?#{Amakanize::PATTERN_OF_NUMERIC_CHARACTERS}(?:話|巻|版)?\z/, "")
       end
     end
   end

data/lib/amakanize/series_name.rb

@@ -8,6 +8,7 @@ module Amakanize
           ::Amakanize::Filters::NormalizationFilter.new,
           ::Amakanize::Filters::HyphenMinusNormalizationFilter.new,
           ::Amakanize::Filters::DashBetweenAlhabetsNormalizationFilter.new,
+          ::Amakanize::Filters::AngleBracketsAfterWordNormalizationFilter.new,
           ::Amakanize::Filters::SpacesBetweenExclamationsDeletionFilter.new,
           ::Amakanize::Filters::ObviousVolumeNumberDeletionFilter.new,
           ::Amakanize::Filters::TrailingParenthesesDeletionFilter.new,

data/lib/amakanize/version.rb

@@ -1,3 +1,3 @@
 module Amakanize
-  VERSION = "0.2.0"
+  VERSION = "0.2.1"
 end

data/lib/amakanize.rb

@@ -1,6 +1,7 @@
 require "amakanize/author_name"
 require "amakanize/author_names"
 require "amakanize/filters/base_filter"
+require "amakanize/filters/angle_brackets_after_word_normalization_filter"
 require "amakanize/filters/dash_between_alphabets_normalization_filter"
 require "amakanize/filters/html_unescape_filter"
 require "amakanize/filters/hyphen_minus_normalization_filter"
@@ -21,5 +22,21 @@ require "amakanize/series_name"
 require "amakanize/version"
 
 module Amakanize
-  PATTERN_OF_NUMERIC_CHARACTER = /[\diIvVxX１-９①②③④⑤⑥⑦⑧⑨⑩〇一二三四五六七八九十百千万零壱弍参肆伍陸漆捌玖壹貳參拾佰仟萬]/
+  PATTERN_OF_NUMERIC_CHARACTERS = /
+    \d+
+    |[ivx]+
+    |[IVX]+
+    |[０-９]+
+    |①
+    |②
+    |③
+    |④
+    |⑤
+    |⑥
+    |⑦
+    |⑧
+    |⑨
+    |⑩
+    |[〇一二三四五六七八九十百千万零壱弍参肆伍陸漆捌玖壹貳參拾佰仟萬]+
+  /x
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: amakanize
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - r7kamura
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2016-08-21 00:00:00.000000000 Z
+date: 2016-08-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -87,6 +87,7 @@ files:
 - lib/amakanize.rb
 - lib/amakanize/author_name.rb
 - lib/amakanize/author_names.rb
+- lib/amakanize/filters/angle_brackets_after_word_normalization_filter.rb
 - lib/amakanize/filters/base_filter.rb
 - lib/amakanize/filters/dash_between_alphabets_normalization_filter.rb
 - lib/amakanize/filters/html_unescape_filter.rb