twitter_cldr 2.0.2 → 2.1.0

data/History.txt

@@ -1,3 +1,12 @@
+== 2.1.0
+
+* Significant performance improvements (memoization, resource preloading).
+* Number parsing.
+* Custom Hebrew units (thanks @yarons!)
+* Icelandic and Croatian support.
+* Global locale setter and fallbacks.
+* Support for territories from CLDR.
+
 == 2.0.2
 
 * Added support for Vietnamese.

data/NOTICE

@@ -151,4 +151,32 @@ OR PERFORMANCE OF THIS SOFTWARE.
 
 Except as contained in this notice, the name of a copyright holder shall not be
 used in advertising or otherwise to promote the sale, use or other dealings in
-this Software without prior written authorization of the copyright holder.
+this Software without prior written authorization of the copyright holder.
+
+
+
+
+
+The currencies data in twitter-cldr-rb has been adapted from the currencies gem.
+Here's the license that accompanied that code:
+
+Copyright (c) 2009 hexorx
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

data/README.md

@@ -42,7 +42,6 @@ TwitterCldr patches core Ruby objects like `Fixnum` and `Date` to make localizat
 # currencies, default USD
 1337.localize(:es).to_currency.to_s                        # 1.337,00 $
 1337.localize(:es).to_currency.to_s(:currency => "EUR")    # 1.337,00 €
-1337.localize(:es).to_currency.to_s(:currency => "Peru")   # 1.337,00 S/.
 
 # percentages
 1337.localize(:es).to_percent.to_s                         # 1.337%
@@ -63,20 +62,14 @@ num.to_currency.to_s  # ...etc
 
 #### More on Currencies
 
-If you're looking for a list of supported countries and currencies, use the `TwitterCldr::Shared::Currencies` class:
+If you're looking for a list of supported currencies, use the `TwitterCldr::Shared::Currencies` class:
 
 ```ruby
-# all supported countries
-TwitterCldr::Shared::Currencies.countries                  # ["Lithuania", "Philippines", ... ]
-
 # all supported currency codes
 TwitterCldr::Shared::Currencies.currency_codes             # ["LTL", "PHP" ... ]
 
-# data for a specific country
-TwitterCldr::Shared::Currencies.for_country("Canada")      # { :currency => "Dollar", :symbol => "$", :code => "CAD" }
-
 # data for a specific currency code
-TwitterCldr::Shared::Currencies.for_code("CAD")            # { :currency => "Dollar", :symbol => "$", :country => "Canada"}
+TwitterCldr::Shared::Currencies.for_code("CAD")            # { :currency => "Canadian dollar", :symbol => "$", :code_points => [36], :cldr_symbol => "CA$", :country => "Canada" }
 ```
 
 #### Short / Long Decimals
@@ -126,7 +119,7 @@ dt.to_short_s  # ...etc
 Besides the default date formats, CLDR supports a number of additional ones.  The list of available formats varys for each locale.  To get a full list, use the `additional_formats_for` method:
 
 ```ruby
-# ["EEEEd", "Ed", "GGGGyMd", "H", "Hm", "Hms", "M", "MEd", "MMM", "MMMEEEEd", "MMMEd", ... ] 
+# ["EEEEd", "Ed", "GGGGyMd", "H", "Hm", "Hms", "M", "MEd", "MMM", "MMMEEEEd", "MMMEd", ... ]
 TwitterCldr::Formatters::DateTimeFormatter.additional_formats_for(:ja)
 ```
 
@@ -620,7 +613,7 @@ TwitterCldr.convert_locale(:msa)              # :ms
 TwitterCldr.convert_locale(:ms)               # :ms
 ```
 
-There are a few functions in TwitterCLDR that don't require a locale code, and instead use the default locale by calling `TwitterCldr.get_locale`.  The `get_locale` function defers to `FastGettext.locale` when the FastGettext library is available, and falls back on :en (English) when it's not.  (Twitter uses the FastGettext gem to retrieve translations efficiently in Ruby).
+There are a few functions in TwitterCLDR that don't require a locale code, and instead use the default locale by calling `TwitterCldr.locale`.  The `locale` function defers to `FastGettext.locale` when the FastGettext library is available, and falls back on :en (English) when it's not.  (Twitter uses the FastGettext gem to retrieve translations efficiently in Ruby).
 
 ```ruby
 TwitterCldr.get_locale    # will return :en

data/lib/twitter_cldr/formatters/calendars/date_formatter.rb

@@ -7,7 +7,12 @@ module TwitterCldr
   module Formatters
     class DateFormatter < DateTimeFormatter
       def initialize(options = {})
-        @tokenizer = TwitterCldr::Tokenizers::DateTokenizer.new(:locale => extract_locale(options), :calendar_type => options[:calendar_type])
+        locale = extract_locale(options)
+        cache_key = TwitterCldr::Utils.compute_cache_key("date", locale, options[:calendar_type])
+        @tokenizer = tokenizer_cache[cache_key] ||= TwitterCldr::Tokenizers::DateTokenizer.new(
+          :locale => locale,
+          :calendar_type => options[:calendar_type]
+        )
       end
     end
   end

data/lib/twitter_cldr/formatters/calendars/datetime_formatter.rb

@@ -41,7 +41,12 @@ module TwitterCldr
       }
 
       def initialize(options = {})
-        @tokenizer = TwitterCldr::Tokenizers::DateTimeTokenizer.new(:locale => extract_locale(options), :calendar_type => options[:calendar_type])
+        locale = extract_locale(options)
+        cache_key = TwitterCldr::Utils.compute_cache_key("datetime", locale, options[:calendar_type])
+        @tokenizer = tokenizer_cache[cache_key] ||= TwitterCldr::Tokenizers::DateTimeTokenizer.new(
+          :locale => locale,
+          :calendar_type => options[:calendar_type]
+        )
       end
 
       def result_for_token(token, index, date)
@@ -58,6 +63,10 @@ module TwitterCldr
 
       protected
 
+      def tokenizer_cache
+        @@tokenizer_cache ||= {}
+      end
+
       # there is incomplete era data in CLDR for certain locales like Hindi
       # fall back if that happens
       def era(date, pattern, length)

data/lib/twitter_cldr/formatters/calendars/time_formatter.rb

@@ -7,7 +7,12 @@ module TwitterCldr
   module Formatters
     class TimeFormatter < DateTimeFormatter
       def initialize(options = {})
-        @tokenizer = TwitterCldr::Tokenizers::TimeTokenizer.new(:locale => extract_locale(options), :calendar_type => options[:calendar_type])
+        locale = extract_locale(options)
+        cache_key = TwitterCldr::Utils.compute_cache_key("time", locale, options[:calendar_type])
+        @tokenizer = tokenizer_cache[cache_key] ||= TwitterCldr::Tokenizers::TimeTokenizer.new(
+          :locale => locale,
+          :calendar_type => options[:calendar_type]
+        )
       end
     end
   end

data/lib/twitter_cldr/formatters/calendars/timespan_formatter.rb

@@ -21,14 +21,18 @@ module TwitterCldr
       }
 
       def initialize(options = {})
+        locale = extract_locale(options)
         @direction = options[:direction]
-        @tokenizer = TwitterCldr::Tokenizers::TimespanTokenizer.new(:locale => extract_locale(options))
+        cache_key = TwitterCldr::Utils.compute_cache_key(locale)
+        @tokenizer = tokenizer_cache[cache_key] ||= TwitterCldr::Tokenizers::TimespanTokenizer.new(
+          :locale => locale
+        )
       end
 
       def format(seconds, fmt_options = {})
         options = fmt_options.dup
         options[:type]      ||= DEFAULT_TYPE
-        options[:direction] ||= @direction || (seconds < 0 ? :ago : :until)
+        options[:direction] ||= (seconds < 0 ? :ago : :until)
         options[:unit]      ||= calculate_unit(seconds.abs, options)
 
         options[:number] = calculate_time(seconds.abs, options[:unit])
@@ -38,6 +42,12 @@ module TwitterCldr
         strings.join.gsub(/\{[0-9]\}/, options[:number].to_s)
       end
 
+      protected
+
+      def tokenizer_cache
+        @@tokenizer_cache ||= {}
+      end
+
       def calculate_unit(seconds, unit_options = {})
         options = unit_options.dup
         options[:approximate] = false if options[:approximate].nil?

data/lib/twitter_cldr/formatters/numbers/currency_formatter.rb

@@ -12,17 +12,26 @@ module TwitterCldr
       def format(number, options = {})
         options[:currency] ||= "USD"
         currency = TwitterCldr::Shared::Currencies.for_code(options[:currency])
-        currency ||= TwitterCldr::Shared::Currencies.for_country(options[:currency])
-        currency ||= { :symbol => options[:currency], :currency => options[:currency] }
+        currency ||= {
+          :currency    => options[:currency],
+          :symbol      => options[:currency],
+          :cldr_symbol => options[:currency]
+        }
+
+        # overwrite with explicit symbol if given
+        currency[:symbol] = options[:symbol] if options[:symbol]
 
         digits_and_rounding = resource(options[:currency])
         options[:precision] ||= digits_and_rounding[:digits]
         options[:rounding] ||= digits_and_rounding[:rounding]
 
-        super(number, options).gsub('¤', (currency[:symbol] || currency[:currency].to_s))
+        symbol = options[:use_cldr_symbol] ? currency[:cldr_symbol] : currency[:symbol]
+        symbol ||= currency[:currency].to_s
+        super(number, options).gsub('¤', symbol)
       end
 
       private
+
       def resource(code)
         @resource ||= TwitterCldr.get_resource(:shared, :currency_digits_and_rounding)
         @resource[code.to_sym] || @resource[:DEFAULT]

data/lib/twitter_cldr/formatters/numbers/number_formatter.rb

@@ -11,7 +11,11 @@ module TwitterCldr
       DEFAULT_SYMBOLS = { :group => ',', :decimal => '.', :plus_sign => '+', :minus_sign => '-' }
 
       def initialize(options = {})
-        @tokenizer = TwitterCldr::Tokenizers::NumberTokenizer.new(:locale => extract_locale(options))
+        locale = extract_locale(options)
+        cache_key = TwitterCldr::Utils.compute_cache_key(locale)
+        @tokenizer = tokenizer_cache[cache_key] ||= TwitterCldr::Tokenizers::NumberTokenizer.new(
+          :locale => locale
+        )
         @symbols = DEFAULT_SYMBOLS.merge(tokenizer.symbols)
       end
 
@@ -28,6 +32,10 @@ module TwitterCldr
 
       protected
 
+      def tokenizer_cache
+        @@tokenizer_cache ||= {}
+      end
+
       def transform_number(number)
         number  # noop for base class
       end

data/lib/twitter_cldr/formatters/plurals/rules.rb

@@ -11,7 +11,7 @@ module TwitterCldr
         class << self
 
           def all
-            all_for(TwitterCldr.get_locale)
+            all_for(TwitterCldr.locale)
           end
 
           def all_for(locale)
@@ -20,7 +20,7 @@ module TwitterCldr
             nil
           end
 
-          def rule_for(number, locale = TwitterCldr.get_locale)
+          def rule_for(number, locale = TwitterCldr.locale)
             get_resource(locale)[:rule].call(number)
           rescue
             :other
@@ -30,7 +30,12 @@ module TwitterCldr
 
           def get_resource(locale)
             locale = TwitterCldr.convert_locale(locale)
-            eval(TwitterCldr.get_locale_resource(locale, :plurals)[locale])[locale][:i18n][:plural]
+            cache_key = TwitterCldr::Utils.compute_cache_key(locale)
+            locale_cache[cache_key] ||= eval(TwitterCldr.get_locale_resource(locale, :plurals)[locale])[locale][:i18n][:plural]
+          end
+
+          def locale_cache
+            @locale_cache ||= {}
           end
 
         end

data/lib/twitter_cldr/localized/localized_date.rb

@@ -12,6 +12,19 @@ module TwitterCldr
         LocalizedDateTime.new(DateTime.parse("#{@base_obj.strftime("%Y-%m-%d")}T#{time_obj.strftime("%H:%M:%S%z")}"), @locale, :calendar_type => @calendar_type)
       end
 
+      def to_time(base = Time.now)
+        time = Time.gm(
+          @base_obj.year,
+          @base_obj.month,
+          @base_obj.day,
+          base.hour,
+          base.min,
+          base.sec
+        )
+
+        LocalizedTime.new(time, @locale, :calendar_type => @calendar_type)
+      end
+
       protected
 
       def formatter_const

data/lib/twitter_cldr/localized/localized_datetime.rb

@@ -27,15 +27,15 @@ module TwitterCldr
       end
 
       def ago(options = {})
-        base_time = options[:base_time] || Time.now
-        seconds = self.to_time.base_obj.to_i - base_time.to_i
+        base_time = (options[:base_time] || Time.now).gmtime
+        seconds = self.to_time(base_time).base_obj.gmtime.to_i - base_time.to_i
         raise ArgumentError.new('Start date is after end date. Consider using "until" function.') if seconds > 0
         TwitterCldr::Localized::LocalizedTimespan.new(seconds, options.merge(:locale => @locale))
       end
 
       def until(options = {})
-        base_time = options[:base_time] || Time.now
-        seconds = self.to_time.base_obj.to_i - base_time.to_i
+        base_time = (options[:base_time] || Time.now).gmtime
+        seconds = self.to_time(base_time).base_obj.gmtime.to_i - base_time.to_i
         raise ArgumentError.new('End date is before start date. Consider using "ago" function.') if seconds < 0
         TwitterCldr::Localized::LocalizedTimespan.new(seconds, options.merge(:locale => @locale))
       end
@@ -49,11 +49,24 @@ module TwitterCldr
       end
 
       def to_date
-        LocalizedDate.new(Date.parse(@base_obj.strftime("%Y-%m-%dT%H:%M:%S%z")), @locale, :calendar_type => @calendar_type)
+        date = Date.new(@base_obj.year, @base_obj.month, @base_obj.day)
+        LocalizedDate.new(date, @locale, :calendar_type => @calendar_type)
       end
 
-      def to_time
-        LocalizedTime.new(Time.parse(@base_obj.strftime("%Y-%m-%dT%H:%M:%S%z")), @locale, :calendar_type => @calendar_type)
+      def to_time(base = Time.now)
+        utc_dt = @base_obj.new_offset(0)
+
+        time = Time.gm(
+          utc_dt.year,
+          utc_dt.month,
+          utc_dt.day,
+          utc_dt.hour,
+          utc_dt.min,
+          utc_dt.sec,
+          utc_dt.sec_fraction * (RUBY_VERSION < '1.9' ? 86400000000 : 1000000)
+        )
+
+        LocalizedTime.new(time, @locale, :calendar_type => @calendar_type)
       end
 
       protected

data/lib/twitter_cldr/localized/localized_object.rb

@@ -27,7 +27,7 @@ module TwitterCldr
 
       def self.localize(klass)
         klass.class_eval <<-LOCALIZE, __FILE__, __LINE__ + 1
-          def localize(locale = TwitterCldr.get_locale, options = {})
+          def localize(locale = TwitterCldr.locale, options = {})
             #{self}.new(self, locale, options)
           end
         LOCALIZE

data/lib/twitter_cldr/localized/localized_string.rb

@@ -36,6 +36,20 @@ module TwitterCldr
         @base_obj.dup
       end
 
+      def to_i(options = {})
+        to_f(options).to_i
+      end
+
+      def to_f(options = {})
+        if TwitterCldr::Parsers::NumberParser.is_numeric?(@base_obj)
+          TwitterCldr::Parsers::NumberParser.new(@locale).try_parse(@base_obj, options) do |result|
+            result || @base_obj.to_f
+          end
+        else
+          @base_obj.to_f
+        end
+      end
+
       def size
         code_points.size
       end

data/lib/twitter_cldr/localized/localized_time.rb

@@ -12,6 +12,18 @@ module TwitterCldr
         LocalizedDateTime.new(DateTime.parse("#{date_obj.strftime("%Y-%m-%d")}T#{@base_obj.strftime("%H:%M:%S%z")}"), @locale, :calendar_type => @calendar_type)
       end
 
+      def to_time(base_time = Time.now)
+        self
+      end
+
+      def gmtime
+        LocalizedTime.new(@base_obj.gmtime, @locale, :calendar_type => @calendar_type)
+      end
+
+      def localtime
+        LocalizedTime.new(@base_obj.localtime, @locale, :calendar_type => @calendar_type)
+      end
+
       protected
 
       def formatter_const

data/lib/twitter_cldr/localized/localized_timespan.rb

@@ -8,7 +8,7 @@ module TwitterCldr
 
     class LocalizedTimespan < LocalizedObject
       def initialize(seconds, options = {})
-        super(seconds, options[:locale] || TwitterCldr.get_locale, options)
+        super(seconds, options[:locale] || TwitterCldr.locale, options)
       end
 
       def to_s(options = {})

data/lib/twitter_cldr/parsers/number_parser.rb

@@ -0,0 +1,113 @@
+# encoding: UTF-8
+
+# Copyright 2012 Twitter, Inc
+# http://www.apache.org/licenses/LICENSE-2.0
+
+module TwitterCldr
+  module Parsers
+
+    class InvalidNumberError < StandardError; end
+
+    class NumberParser
+
+      SEPARATOR_CHARS = ['.', ',', ' '].map do |char|
+        char == ' ' ? '\s' : Regexp.escape(char)
+      end.join
+
+      def initialize(locale = TwitterCldr.locale)
+        @locale = locale
+      end
+
+      def parse(number_text, options = {})
+        options[:strict] = true unless options.include?(:strict)
+        group, decimal = separators(options[:strict])
+        tokens = tokenize(number_text, group, decimal)
+
+        num_list, punct_list = tokens.partition { |t| t[:type] == :numeric }
+        raise InvalidNumberError unless punct_valid?(punct_list)
+        raise InvalidNumberError unless tokens.last && tokens.last[:type] == :numeric
+
+        if punct_list.last && punct_list.last[:type] == :decimal
+          result = num_list[0..-2].map { |num| num[:value] }.join.to_i
+          result + num_list.last[:value].to_i / (10.0 ** num_list.last[:value].size)
+        else
+          num_list.map { |num| num[:value] }.join.to_i
+        end
+      end
+
+      def try_parse(number_text, default = nil, options = {})
+        begin
+          result = parse(number_text, options)
+        rescue InvalidNumberError
+          result = nil
+        end
+
+        if block_given?
+          yield(result)
+        else
+          result || default
+        end
+      end
+
+      def valid?(number_text, options = {})
+        parse(number_text, options)
+        true
+      rescue
+        false
+      end
+
+      def self.is_numeric?(text, separators = SEPARATOR_CHARS)
+        !!(text =~ /\A[0-9#{separators}]+\Z/)
+      end
+
+      protected
+
+      def punct_valid?(punct_list)
+        # all group, allowed one decimal at end
+        punct_list.each_with_index.all? do |punct, index|
+          punct[:type] == :group || (index == (punct_list.size - 1) && punct[:type] == :decimal)
+        end
+      end
+
+      def separators(strict = false)
+        group = strict ? group_separator : SEPARATOR_CHARS
+        decimal = strict ? decimal_separator : SEPARATOR_CHARS
+        [group, decimal]
+      end
+
+      def tokenize(number_text, group, decimal)
+        match_data = number_text.scan(/([\d]*)([#{group}]{0,1})([\d]*)([#{decimal}]{0,1})([\d]*)/)
+        (match_data.flatten || []).reject(&:empty?).map { |match| identify(match, group, decimal) }
+      end
+
+      def identify(text, group, decimal)
+        result = { :value => text }
+        result[:type] = if self.class.is_numeric?(result[:value], "")
+          :numeric
+        else
+          if result[:value] =~ /[#{group}]/
+            :group
+          elsif result[:value] =~ /[#{decimal}]/
+            :decimal
+          else
+            nil
+          end
+        end
+        result
+      end
+
+      def decimal_separator
+        @decimal_separator ||= Regexp.escape(resource[:symbols][:decimal])
+      end
+
+      def group_separator
+        @group_separator ||= Regexp.escape(resource[:symbols][:group])
+      end
+
+      def resource
+        @resource ||= TwitterCldr.get_locale_resource(@locale, "numbers")[@locale][:numbers]
+      end
+
+    end
+  end
+end

data/lib/twitter_cldr/parsers.rb

@@ -0,0 +1,10 @@
+# encoding: UTF-8
+
+# Copyright 2012 Twitter, Inc
+# http://www.apache.org/licenses/LICENSE-2.0
+
+module TwitterCldr
+  module Parsers
+    autoload :NumberParser, 'twitter_cldr/parsers/number_parser'
+  end
+end

data/lib/twitter_cldr/resources/loader.rb

@@ -13,11 +13,58 @@ module TwitterCldr
       end
 
       def get_locale_resource(locale, resource_name)
-        get_resource(:locales, TwitterCldr.convert_locale(locale), resource_name)
+        get_resource(*locale_resource_path(locale, resource_name))
+      end
+
+      def resource_loaded?(*path)
+        resources_cache.include?(resource_file_path(path))
+      end
+
+      def locale_resource_loaded?(locale, resource_name)
+        resource_loaded?(*locale_resource_path(locale, resource_name))
+      end
+
+      def resource_types
+        @resource_types ||= Dir.glob(File.join(RESOURCES_DIR, 'locales/en', '*')).map do |file|
+          File.basename(file).chomp(File.extname(file)).to_sym
+        end
+      end
+
+      def preload_resources_for_locale(locale, *resources)
+        if resources.size > 0
+          resources = resource_types if resources.first == :all
+          resources.each { |res| get_locale_resource(locale, res) }
+        end
+        nil
+      end
+
+      def preload_resource_for_locales(resource, *locales)
+        locales.each do |locale|
+          preload_resources_for_locale(locale, resource)
+        end
+        nil
+      end
+
+      def preload_resources_for_all_locales(*resources)
+        TwitterCldr.supported_locales.each do |locale|
+          preload_resources_for_locale(locale, *resources)
+        end
+        nil
+      end
+
+      def preload_all_resources
+        TwitterCldr.supported_locales.each do |locale|
+          preload_resources_for_locale(locale, :all)
+        end
+        nil
       end
 
       private
 
+      def locale_resource_path(locale, resource_name)
+        [:locales, TwitterCldr.convert_locale(locale), resource_name]
+      end
+
       def resources_cache
         @resources_cache ||= Hash.new { |hash, path| hash[path] = load_resource(path) }
       end
@@ -54,4 +101,4 @@ module TwitterCldr
     end
 
   end
-end
+end

data/lib/twitter_cldr/resources/locales_resources_importer.rb

@@ -12,7 +12,7 @@ module TwitterCldr
 
     class LocalesResourcesImporter
 
-      COMPONENTS = %w[calendars languages numbers units plurals lists layout currencies]
+      COMPONENTS = %w[calendars languages numbers units plurals lists layout currencies territories]
 
       # Arguments:
       #
@@ -49,6 +49,7 @@ module TwitterCldr
         Cldr::Export.export(:locales => TwitterCldr.supported_locales, :components => COMPONENTS, :target => File.join(@output_path, 'locales')) do |component, locale, path|
           add_buddhist_calendar(component, locale, path)
           process_plurals(component, locale, path)
+          downcase_territory_codes(component, locale, path)
           deep_symbolize(component, locale, path)
         end
 
@@ -68,6 +69,25 @@ module TwitterCldr
         end
       end
 
+      # CLDR stores territory codes uppercase. For consistency with how we
+      # handle territory codes in methods relating to phone and postal codes,
+      # we downcase them here.
+      #
+      # (There is also some trickery relating to three-digit UN "area codes"
+      # used by CLDR; see comment of Utils::Territories::deep_normalize_territory_code_keys.)
+      def downcase_territory_codes(component, locale, path)
+        return unless component == 'Territories'
+
+        data = YAML.load(File.read(path))
+        data.keys.each do |l|
+          data[l] = TwitterCldr::Utils::Territories.deep_normalize_territory_code_keys(data[l])
+        end
+
+        File.open(path, 'w:utf-8') do |output|
+          output.write(YAML.dump(data))
+        end
+      end
+
       def process_plurals(component, locale, path)
         return unless component == 'Plurals'
 
@@ -106,4 +126,4 @@ module TwitterCldr
     end
 
   end
-end
+end