sort_kana_jisx4061 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: d6f0c5c42b2e0b2dadfc08ed0ae0d2112283776d
-  data.tar.gz: 5bb7731e8526efc259bec4c9a95f4aac00db0365
+  metadata.gz: 1f398922e26d2aedfc4b2ee9f0241e641706c649
+  data.tar.gz: 9ee9eae6c2e20257cc226196667f975d66420e45
 SHA512:
-  metadata.gz: babaa39b79f465ffc4a1d76bb3c83bee419fcdb9bbb401d414c795294411e5168d39e9af75ae6bbf74e34c86fd5b72aad2e38f795258b1ccbe3f790deb0bcd25
-  data.tar.gz: 8ffac04a94ac36f11ba1d6332dedaea1ed1bf290ce851af7f14826714591cac37bee1be5d0c21f5001ff9f7971a916a9a3b32fe2edc28c45ff37f19a76a437f7
+  metadata.gz: 28e83f6e99ea1aaa151d3a177c77db9037f4c6933be059387b029679958887413025ba974899c94f2645ae1fe2e3ec4913171c72f6581dcb5da89a424ff9bf26
+  data.tar.gz: 49492ed88da38045a3e9f265652cf2cae6942aee48b11c1a4d55f4e32e2e4bf5540b0e3d8b20fe0b9afe16e05a9ef91ecf822084c7ab50129a620e15bad5a974

data/README.md

@@ -20,6 +20,7 @@ words_sorted = sort_kana_jisx4061_by(words) {|x| x[:yomi] }
 
 ## Note
 
+* Strings are exptected to be UTF-8
 * Sorting Kanji is not supported
 * Hiragana is converted to Katakana internally
 

data/lib/sort_kana_jisx4061/version.rb

@@ -1,3 +1,3 @@
 module SortKanaJisx4061
-  VERSION = "0.1.0"
+  VERSION = "0.2.0"
 end

data/lib/sort_kana_jisx4061.rb

@@ -18,23 +18,34 @@ module SortKanaJisx4061         # :nodoc:
   CC_GETA = 12                  # unused
 
   class << self
-    def process_yomi(str)
-      r = []
-      str.each_char do |c|
-        c = c.downcase
-        c = c.tr('ぁ-ん', 'ァ-ン') # Hiragana -> Katakana
-        c = base_char(c)
-        if c == 'ー'
-          c = process_onbiki(c, r.last[1])
+
+    # == Example
+    # 'ガッコウ1'
+    # →
+    # [[[CC_KANA, 'カ'], [CC_KANA, 'ツ'], [CC_KANA, 'コ'],
+    #   [CC_KANA, 'ウ'], [CC_NUMBER, '1'] ],
+    #  [[CC_KANA, 'ガ'], [CC_KANA, 'ッ'], [CC_KANA, 'コ'],
+    #   [CC_KANA, 'ウ'], [CC_NUMBER, '1'] ]]
+    def parse_yomi(str)
+      r1 = []                   # without collation properties
+      r2 = []                   # with collation properties
+      str.each_char do |c1|
+        c1 = c1.tr('ぁ-ん', 'ァ-ン') # Hiragana -> Katakana
+        if c1 == 'ー'
+          c1 = process_onbiki(c1, r1.last[1])
         end
-        r << [jisx4061_charclass(c), c]
+        c2 = c1
+        c1 = base_char(c1)
+        c2 = c2.swapcase  # WORKAROUND: downcase should be smaller than updase
+        ccls = jisx4061_charclass(c1)
+        r1 << [ccls, c1]; r2 << [ccls, c2]
       end
-      r
+      [r1, r2]
     end
 
     # for test
     def compare(str1, str2)
-      process_yomi(str1) <=> process_yomi(str2)
+      parse_yomi(str1) <=> parse_yomi(str2)
     end
 
     private
@@ -62,6 +73,7 @@ module SortKanaJisx4061         # :nodoc:
     }
 
     def base_char(c)
+      c = c.downcase
       BASE_CHAR_TBL[c] || c
     end
 
@@ -93,5 +105,5 @@ end
 #
 #  words_sorted = sort_kana_jisx4061_by(words) {|x| x[:yomi] }
 def sort_kana_jisx4061_by(enum)
-  enum.sort_by {|x| SortKanaJisx4061::process_yomi(yield(x)) }
+  enum.sort_by {|x| SortKanaJisx4061::parse_yomi(yield(x)) }
 end

data/sort_kana_jisx4061.gemspec

@@ -5,6 +5,7 @@ require "sort_kana_jisx4061/version"
 Gem::Specification.new do |spec|
   spec.name          = "sort_kana_jisx4061"
   spec.version       = SortKanaJisx4061::VERSION
+  spec.license       = 'MIT'
   spec.authors       = ["emasaka (Masakazu Takahashi)"]
   spec.email         = ["emasaka@gmail.com"]
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: sort_kana_jisx4061
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors:
 - emasaka (Masakazu Takahashi)
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-05-12 00:00:00.000000000 Z
+date: 2018-05-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -68,7 +68,8 @@ files:
 - lib/sort_kana_jisx4061/version.rb
 - sort_kana_jisx4061.gemspec
 homepage: https://github.com/emasaka/sort_kana_jisx4061
-licenses: []
+licenses:
+- MIT
 metadata: {}
 post_install_message: 
 rdoc_options: []
@@ -86,7 +87,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.4.8
+rubygems_version: 2.6.14.1
 signing_key: 
 specification_version: 4
 summary: sort Japanese Kana strings by JIS X 4061 order