isodoc-i18n 1.0.0 → 1.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 0cb9429df0523ad853ce2c16221b6ca2da9df7f12be71ec1d9d31dfe8a4507c7
-  data.tar.gz: 0a454f447146ec95be0a421d3b38844d3c502aabf998cef8dfa9158e731e65de
+  metadata.gz: e51e4427e97bf2e0fde14e6f995a35f56771180c5480de0231484927442aa80b
+  data.tar.gz: 7516734ee389ab585094598bab5b8ecc696cf488f91cedf1454f192809dcb075
 SHA512:
-  metadata.gz: 98607331dc8eea71395667abccf85d4e97ce7c91223ccc1e4b7bb860f596de4262cd642fa94d68693a9e77c61d045c402e6919e71b23a4571e519bc0e6fdcc92
-  data.tar.gz: e69e12e61a39badb700850998267e473b5acbe51cf3f11de332ed6a08cc6e50b73de8c6811bcb998999b6b07739b78c390368cdeacd5ad4b6dcf97b55787d5f2
+  metadata.gz: 876730b932cb38b6fcaa5af50e54b1a7f9b9be6c016ac0f82ad38fd00c0f510a15de812e09be5d733e81a6a16976cf547fc54a9cb113313c82e6762407a2e150
+  data.tar.gz: cabf6cd3651fb0314a701d5bb08aafdd607eed4f90d2a2df74d7ea7895e4f90d6566abad20c56e816cdf9c247e865131a7d27cbd2ea258a4e9b171b6afbe9117

data/.github/workflows/rake.yml

@@ -0,0 +1,15 @@
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
+name: rake
+
+on:
+  push:
+    branches: [ master, main ]
+    tags: [ v* ]
+  pull_request:
+
+jobs:
+  rake:
+    uses: metanorma/metanorma-build-scripts/.github/workflows/generic-rake.yml@main
+    secrets:
+      pat_token: ${{ secrets.METANORMA_CI_PAT_TOKEN }}

data/.hound.yml

@@ -0,0 +1,5 @@
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
+ruby:
+  enabled: true
+  config_file: .rubocop.yml

data/isodoc-i18n.gemspec

@@ -24,6 +24,7 @@ Gem::Specification.new do |spec|
 
   spec.add_dependency "htmlentities", "~> 4.3.4"
   spec.add_dependency "metanorma-utils", "~> 1.2.8"
+  spec.add_dependency "twitter_cldr"
 
   spec.add_development_dependency "debug"
   spec.add_development_dependency "equivalent-xml", "~> 0.6"

data/lib/isodoc/i18n.rb

@@ -1,6 +1,7 @@
 require "yaml"
 require "htmlentities"
 require "metanorma-utils"
+require "twitter_cldr"
 
 module IsoDoc
   class I18n
@@ -100,13 +101,28 @@ module IsoDoc
       xml.traverse do |n|
         next unless n.text?
 
-        n.replace(cleanup_entities(n.text.gsub(/ /, "").gsub(/:/, "：")
-          .gsub(/,/, "、").gsub(/\(/, "（").gsub(/\)/, "）")
-          .gsub(/\[/, "【").gsub(/\]/, "】"), is_xml: false))
+        n.replace(cleanup_entities(l10_zh1(n.text), is_xml: false))
       end
       xml.to_xml.gsub(/<b>/, "").gsub("</b>", "").gsub(/<\?[^>]+>/, "")
     end
 
+    ZH_CHAR = "\\p{Han}|\\p{In CJK Symbols And Punctuation}|"\
+              "\\p{In Halfwidth And Fullwidth Forms}".freeze
+
+    # note: we can't differentiate comma from enumeration comma 、
+    def l10_zh1(text)
+      [":：", ",，", ".。", ")）", "]】", ":：", ";；", "?？", "!！"].each do |m|
+        text = text.gsub(/(?<=#{ZH_CHAR})#{Regexp.quote m[0]}/, m[1])
+      end
+      ["(（", "[【"].each do |m|
+        text = text.gsub(/#{Regexp.quote m[0]}(?=#{ZH_CHAR})/, m[1])
+      end
+      text.gsub(/(?<=#{ZH_CHAR}) (?=#{ZH_CHAR})/o, "")
+        .gsub(/(?<=\d) (?=#{ZH_CHAR})/o, "")
+        .gsub(/(?<=#{ZH_CHAR}) (?=\d)/o, "")
+        .gsub(/(?<=#{ZH_CHAR}) (?=[A-Za-z](#{ZH_CHAR}|$))/o, "")
+    end
+
     def boolean_conj(list, conn)
       case list.size
       when 0 then ""
@@ -131,5 +147,32 @@ module IsoDoc
         c.encode(c.decode(text), :hexadecimal)
       end
     end
+
+    # ord class is either SpelloutRules or OrdinalRules
+    def inflect_ordinal(num, term, ord_class)
+      if @labels["ordinal_keys"].nil? || @labels["ordinal_keys"].empty?
+        num.localize(tw_cldr_lang).to_rbnf_s(ord_class, @labels[ord_class])
+      else
+        num.localize(tw_cldr_lang)
+          .to_rbnf_s(ord_class, @labels[ord_class][ordinal_key(term)])
+      end
+    end
+
+    def ordinal_key(term)
+      @labels["ordinal_keys"].each_with_object([]) do |k, m|
+        m << case k
+             when "gender" then term["gender"]
+             when "number" then term["number"] || "sg"
+             when "case" then term["case"] || "nom"
+             end
+      end.join(".")
+    end
+
+    def tw_cldr_lang
+      if @lang == "zh" && @script == "Hans" then :"zh-cn"
+      elsif @lang == "zh" && @script == "Hant" then :"zh-tw"
+      else @lang.to_sym
+      end
+    end
   end
 end

data/lib/isodoc/version.rb

@@ -1,5 +1,5 @@
 module IsoDoc
   class I18n
-    VERSION = "1.0.0".freeze
+    VERSION = "1.0.3".freeze
   end
 end

data/spec/assets/de.yaml

@@ -0,0 +1,12 @@
+ordinal_keys: [gender,number]
+SpelloutRules:
+  m.sg: spellout-ordinal-r
+  f.sg: spellout-ordinal
+  n.sg: spellout-ordinal-s
+  m.pl: spellout-ordinal
+  f.pl: spellout-ordinal
+  n.pl: spellout-ordinal
+edition: Auflage
+inflection:
+  Auflage:
+    gender: f

data/spec/assets/zh-Hans.yaml

@@ -0,0 +1,7 @@
+ordinal_keys: []
+SpelloutRules: spellout-ordinal
+edition: Auflage
+inflection:
+  Auflage:
+    gender: f
+

data/spec/isodoc/base_spec.rb

@@ -48,9 +48,11 @@ RSpec.describe IsoDoc::I18n do
   it "does Chinese localisation" do
     c = IsoDoc::I18n.new("zh", "Hans")
     expect(c.l10n("Code (hello, world.)"))
-      .to be_equivalent_to "Code（hello、world.）"
-    expect(c.l10n("<a>Code (he<b>l</b>lo, world.)</a>"))
-      .to be_equivalent_to "<a>Code（hello、world.）</a>"
+      .to be_equivalent_to "Code (hello, world.)"
+    expect(c.l10n("计算机代码 (你好, 世界.)"))
+      .to be_equivalent_to " 计算机代码（你好，世界。）"
+    expect(c.l10n("<a>计算机代码</a> (<b>你好,</b> 世界.)"))
+      .to be_equivalent_to "<a>计算机代码</a> (你好， 世界。）"
   end
 
   it "does Hebrew RTL localisation" do
@@ -83,4 +85,18 @@ RSpec.describe IsoDoc::I18n do
     expect(c.boolean_conj(%w(a b c), "and")).to eq "a, b, and c"
     expect(c.boolean_conj(%w(a b c d), "and")).to eq "a, b, c, and d"
   end
+
+  it "does German ordinals" do
+    c = IsoDoc::I18n.new("de", "Latn", "spec/assets/de.yaml")
+    term = c.inflection[c.edition]
+    expect(c.inflect_ordinal(5, term, "SpelloutRules"))
+      .to eq "fünfte"
+  end
+
+  it "does Chinese ordinals" do
+    c = IsoDoc::I18n.new("zh", "Hans", "spec/assets/zh-Hans.yaml")
+    term = c.inflection[c.edition]
+    expect(c.inflect_ordinal(5, term, "SpelloutRules"))
+      .to eq "第五"
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: isodoc-i18n
 version: !ruby/object:Gem::Version
-  version: 1.0.0
+  version: 1.0.3
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2022-03-23 00:00:00.000000000 Z
+date: 2022-04-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: htmlentities
@@ -38,6 +38,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: 1.2.8
+- !ruby/object:Gem::Dependency
+  name: twitter_cldr
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: debug
   requirement: !ruby/object:Gem::Requirement
@@ -187,6 +201,8 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- ".github/workflows/rake.yml"
+- ".hound.yml"
 - ".rubocop.yml"
 - CODE_OF_CONDUCT.md
 - Gemfile
@@ -199,7 +215,9 @@ files:
 - lib/isodoc-yaml/i18n-en.yaml
 - lib/isodoc/i18n.rb
 - lib/isodoc/version.rb
+- spec/assets/de.yaml
 - spec/assets/new.yaml
+- spec/assets/zh-Hans.yaml
 - spec/isodoc/base_spec.rb
 - spec/spec_helper.rb
 homepage: https://github.com/metanorma/isodoc-i18n