normalize_country 0.3.0 → 0.3.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
-SHA1:
-  metadata.gz: 88f20aa5f2af02a58b33f56348ada0dad0dc4c90
-  data.tar.gz: 45c41affc1a9373de1eb73a605d54b0acd55a809
+SHA256:
+  metadata.gz: 6124424b5e181869babd13d860f376b0f50f556897a1d9eb2d81137d12a85971
+  data.tar.gz: d0459074064cdcf9e8c9aa6a20f47a15b750be93b11406faccf281990dc2572a
 SHA512:
-  metadata.gz: 55ff643b54bf0d4fc14d4c1d18d70488774240cf3bebceee101115ccf900a91cde04b4c5a9e444532fa7dc536749805104c2b6618aeb3722eca464f8e37722cb
-  data.tar.gz: e1fad3759975aa1e7baee83daf3113043930240bd252dd17ebec3a9b3694211bcd4fad636a21e33c1c106e6438b3c3b37d931ff1ca5752113027a758d7384bef
+  metadata.gz: eb739c3d7b976793f8b479d4e271a704733d85183c3c55ea3b9e768b03c8992a8c5ed5fdc9d08a03e14757ae4ae5df821f0d5a671bda1a04d7acdf2f419583d9
+  data.tar.gz: bfdedc36669f192bd2c2efc85fd51e7b99340b14f614e2f6ffe8f726a0db29989dcc699c0fd4d41bb0f288487b4325e1474e41212604d6268824597754620743

data/README.rdoc

@@ -123,3 +123,7 @@ Upon further investigation I've found the following:
 * {country_codes}[https://github.com/SunDawg/country_codes] ISO country names and currency data
 * {i18n_data}[https://github.com/grosser/i18n_data]: ISO country names in different languages, includes alpha codes
 * {ModelUN}[https://github.com/uhhuhyeah/model_un]: Similar to this gem but with less support for conversion, it does include US states
+
+=== See Also
+
+* {National Colors}[https://github.com/sshaw/national_colors]

data/lib/normalize_country.rb

@@ -3,11 +3,11 @@
 require "yaml"
 
 module NormalizeCountry
-  VERSION = "0.3.0"
+  VERSION = "0.3.1"
   Countries = {}
 
   class << self
-    attr_accessor :to
+    attr_writer :to
 
     def to
       @to ||= :iso_name

data/lib/normalize_country/countries/en.yml

@@ -1897,6 +1897,8 @@ NI:
   emoji: "\U0001F1F3\U0001F1EE"
   shortcode: ":flag-ni:"
 NL:
+  aliases:
+  - Holland
   alpha2: NL
   alpha3: NLD
   fifa: NED
@@ -2627,6 +2629,8 @@ TZ:
   emoji: "\U0001F1F9\U0001F1FF"
   shortcode: ":flag-tz:"
 UA:
+  aliases:
+  - The Ukraine
   alpha2: UA
   alpha3: UKR
   fifa: UKR

data/lib/normalize_country/scanner.rb

@@ -0,0 +1,121 @@
+module NormalizeCountry
+  class Tokenizer
+    # Need numbers to match ISO codes
+    # Need "-" to match emoji names
+    # Need regex to match emojis
+    PATTERN = /[[:word:]]+/
+
+    def initialize(s)
+      @scanner = StringScanner.new(s)
+    end
+
+    def scan
+      @scanner.scan_until(PATTERN)
+      @scanner.matched
+    end
+
+    def peek
+      match = scan
+      @scanner.unscan if match
+      match
+    end
+
+    def end?
+      peek.nil?
+    end
+  end
+
+  class Scanner
+    def initialize(options = nil)
+      options ||= {}
+
+      @to = options[:to] || NormalizeCountry.to
+      @table = lookup_table(options[:from] || NormalizeCountry.formats)   # need aliases!!!
+    end
+
+    def convert(text)
+      s = Tokenizer.new(text)
+
+      matches = []
+      stack = []
+      match_position = @table
+
+      while !s.end?
+        word = s.peek.downcase
+        if !match_position[word]
+          s.scan
+          next
+        end
+
+        stack << s.scan
+        alternatives = match_position[stack[-1].downcase]
+
+        peek = s.peek
+        if alternatives && peek && alternatives[peek.downcase]
+          match_position = alternatives
+          next
+        end
+
+        if match_position[stack[-1].downcase][:match]
+          text = stack.join(" ")
+          matches << { :matched => text, :converted => NormalizeCountry(text, :to => @to) }
+
+        end
+
+        stack.clear
+        match_position = @table
+      end
+
+      matches
+    end
+
+    private
+
+    def lookup_table(from_formats)
+      table = {}
+
+      NormalizeCountry::Countries.values.uniq.each_with_object(table) do |country, o|
+        # country.name # eq all
+        # no way to get aliases
+        from_formats.each do |format|
+          name = country[format]
+          next unless name
+
+          head = o
+          parts = name.split(/[[:space:]]+/)
+          parts.each_with_index do |word, i|
+            # options[:case_sensitive_codes] = [x,y] # or true
+            # options[:case_sensitive_formats] = [x,y]
+            # options[:case_sensitive_formats] = true   # alpha2, alpha3, fifa, ioc
+            word.downcase! #unless parts.size == 1 && ABBRV_FORMATS.include?(format)
+
+            # if head[word]
+            #   if i == parts.size - 1
+            #     head[word][:match] = true
+            #   else
+            #     head = head[word]
+            #   end
+            # else
+            #   head[word] = {}
+            #   if i == parts.size - 1
+            #     head[word][:match] = true
+            #   else
+            #     head = head[word]
+            #   end
+            # end
+
+            head[word] = {} unless head[word]
+
+            if i == parts.size - 1
+              head[word][:match] = true
+            else
+              head = head[word]
+            end
+          end
+        end
+
+        o
+      end
+    end
+  end
+end

metadata

@@ -1,29 +1,29 @@
 --- !ruby/object:Gem::Specification
 name: normalize_country
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.3.1
 platform: ruby
 authors:
 - Skye Shaw
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2020-04-05 00:00:00.000000000 Z
+date: 2021-01-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '0.9'
+        version: 12.3.3
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '0.9'
+        version: 12.3.3
 - !ruby/object:Gem::Dependency
   name: minitest
   requirement: !ruby/object:Gem::Requirement
@@ -54,6 +54,7 @@ files:
 - bin/normalize_country
 - lib/normalize_country.rb
 - lib/normalize_country/countries/en.yml
+- lib/normalize_country/scanner.rb
 - spec/normalize_country_spec.rb
 homepage: http://github.com/sshaw/normalize_country
 licenses:
@@ -75,7 +76,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.6.14
+rubygems_version: 2.7.6
 signing_key: 
 specification_version: 4
 summary: Convert country names and codes to a standard