japanese_names 0.2.0 → 0.2.1

data/lib/japanese_names/splitter.rb

@@ -15,6 +15,12 @@ module JapaneseNames
       kanji = kanji.strip
       kana  = kana.strip
 
+      # Short-circuit: Return last name if it can match the full string
+      if kanji.size <= 3 && kana.size <= 4
+        full_match = finder.find(kanji).detect { |d| d[0] == kanji && d[1] =~ /\A#{hk kana}\z/ }
+        return [[kanji, nil], [kana, nil]] if full_match
+      end
+
       # Partition kanji into candidate n-grams
       kanji_ngrams = Util::Ngram.ngram_partition(kanji)
 

data/lib/japanese_names/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module JapaneseNames
-  VERSION = '0.2.0'
+  VERSION = '0.2.1'
 end

data/spec/config.yml

@@ -313,6 +313,11 @@
   - 蓮 れん
 
 :skip:
+  - 杏那
+  - 三郎
+  - 四郎
+  - 陽子
+  - 龍之介
   - 青木 緑
   - 石井 真
   - 石井 延

data/spec/unit/splitter_spec.rb

@@ -9,14 +9,23 @@ RSpec.describe JapaneseNames::Splitter do
     config = YAML.load_file(File.join(File.dirname(__FILE__), '..', 'config.yml'))
     skip_list = config[:skip]
 
+    config[:last_names] + config[:first_names].each do |name|
+      kanji, kana = name.split(' ')
+      next if skip_list.index(kanji)
+
+      it "should match #{kanji} #{kana}" do
+        result = subject.split(kanji, kana)
+        expect(result).to eq [[kanji, nil], [kana, nil]]
+      end
+    end
+
     config[:last_names].each do |last_name|
       config[:first_names].each do |first_name|
         kanji_fam, kana_fam = last_name.split(' ')
         kanji_giv, kana_giv = first_name.split(' ')
+        next if skip_list.index([kanji_fam, kanji_giv].join(' '))
 
-        next if skip_list.index("#{kanji_fam} #{kanji_giv}")
-
-        it "should parse #{kanji_fam + kanji_giv} #{kana_fam + kana_giv}" do
+        it "should split #{kanji_fam + kanji_giv} #{kana_fam + kana_giv}" do
           result = subject.split(kanji_fam + kanji_giv, kana_fam + kana_giv)
           expect(result).to eq [[kanji_fam, kanji_giv], [kana_fam, kana_giv]]
         end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: japanese_names
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - Johnny Shields