twitter_cldr 1.0.0

data/lib/tokenizers/base.rb

@@ -0,0 +1,112 @@
+module TwitterCldr
+  module Tokenizers
+    class Base
+      attr_reader :resource, :locale
+      attr_reader :token_splitter_regex, :token_type_regexes, :paths
+      attr_accessor :type, :placeholders
+
+      def initialize(options = {})
+        @locale = (options[:locale] || TwitterCldr::DEFAULT_LOCALE).to_sym
+        self.init_resources
+        self.init_placeholders
+      end
+
+      protected
+
+      # Not to be confused with tokenize_pattern, which pulls out placeholders.  Tokenize_format actually splits a completely
+      # expanded format string into whatever parts are defined by the subclass's token type and token splitter regexes.
+      def tokenize_format(text)
+        final = []
+        text.split(self.token_splitter_regex).each_with_index do |token, index|
+          unless index == 0 && token == ""
+            self.token_type_regexes.each do |token_type|
+              if token =~ token_type[:regex]
+                final << Token.new(:value => token, :type => token_type[:type])
+                break
+              end
+            end
+          end
+        end
+        final
+      end
+
+      def tokens_for(key, type)
+        final = []
+        tokens = self.expand_pattern(self.pattern_for(self.traverse(key)), type)
+
+        tokens.each do |token|
+          if token.is_a?(Token)
+            final << token
+          else
+            final += tokenize_format(token[:value])
+          end
+        end
+
+        final
+      end
+
+      def init_placeholders
+        @placeholders = {}
+      end
+
+      def traverse(needle, haystack = @resource)
+        segments = needle.to_s.split(".")
+        final = haystack
+        segments.each { |segment| final = final[segment.to_sym] }
+        final
+      rescue NameError
+        nil
+      end
+
+      def expand_pattern(format_str, type)
+        if format_str.is_a?(Symbol)
+          # symbols mean another path was given
+          self.expand_pattern(self.pattern_for(self.traverse(format_str)), type)
+        else
+          parts = tokenize_pattern(format_str)
+          final = []
+
+          parts.each do |part|
+            case part[:type]
+              when :placeholder then
+                placeholder = self.choose_placeholder(part[:value], @placeholders)
+                final += placeholder ? placeholder.tokens(:type => type) : []
+              else
+                final << part
+            end
+          end
+
+          final
+        end
+      end
+
+      # Tokenize_pattern is supposed to take a pattern found in the YAML resource files and break it into placeholders and plaintext.
+      # Placeholders are delimited by single and double curly braces, plaintext is everything else.
+      def tokenize_pattern(pattern_str)
+        results = []
+        pattern_str.split(/(\{\{?\w*\}?\}|\'\w+\')/).each do |piece|
+          unless piece.empty?
+            case piece[0].chr
+              when "{"
+                results << { :value => piece, :type => :placeholder }
+              else
+                results << { :value => piece, :type => :plaintext }
+            end
+          end
+        end
+        results
+      end
+
+      def choose_placeholder(token, placeholders)
+        if token[0..1] == "{{"
+          token_value = token[2..-3]
+          found = placeholders.find { |placeholder| placeholder[:name].to_s == token_value }
+        else
+          found = placeholders[token[1..-2].to_i]
+        end
+
+        found ? found[:object] : nil
+      end
+    end
+  end
+end

data/lib/tokenizers/calendars/date_tokenizer.rb

@@ -0,0 +1,24 @@
+module TwitterCldr
+  module Tokenizers
+    class DateTokenizer < TwitterCldr::Tokenizers::DateTimeTokenizer
+      def initialize(options = {})
+        super(options)
+        @token_splitter_regex = /(\s*\'[\w\s-]+\'\s*|G{1,5}|y+|Y+|Q{1,4}|q{1,5}|M{1,5}|L{1,5}|d{1,2}|F{1}|E{1,5}|e{1,5}|c{1,5})/
+        @token_type_regexes = [{ :type => :pattern, :regex => /^(?:G{1,5}|y+|Y+|Q{1,4}|q{1,5}|M{1,5}|L{1,5}|d{1,2}|F{1}|E{1,5}|e{1,5}|c{1,5})/ },
+                               { :type => :plaintext, :regex => // }]
+        @paths = { :default => "calendars.gregorian.formats.date.default",
+                   :full => "calendars.gregorian.formats.date.full",
+                   :long => "calendars.gregorian.formats.date.long",
+                   :medium => "calendars.gregorian.formats.date.medium",
+                   :short => "calendars.gregorian.formats.date.short" }
+      end
+
+      protected
+
+      # must override this because DateTimeTokenizer will set them otherwise
+      def init_placeholders
+        @placeholders = {}
+      end
+    end
+  end
+end

data/lib/tokenizers/calendars/datetime_tokenizer.rb

@@ -0,0 +1,46 @@
+module TwitterCldr
+  module Tokenizers
+    class DateTimeTokenizer < Base
+      attr_reader :placeholders, :calendar_type
+
+      DEFAULT_CALENDAR_TYPE = :gregorian
+      VALID_TYPES = [:default, :full, :long, :medium, :short]
+
+      def initialize(options = {})
+        super(options)
+        @calendar_type = options[:calendar_type] || DEFAULT_CALENDAR_TYPE
+        @token_splitter_regex = //
+        @token_type_regexes = [{ :type => :plaintext, :regex => // }]
+        @paths = { :default => "calendars.gregorian.formats.datetime.default",
+                   :full => "calendars.gregorian.formats.datetime.full",
+                   :long => "calendars.gregorian.formats.datetime.long",
+                   :medium => "calendars.gregorian.formats.datetime.medium",
+                   :short => "calendars.gregorian.formats.datetime.short" }
+      end
+
+      def tokens(options = {})
+        type = options[:type] || :default
+        self.tokens_for(self.paths[type], type)
+      end
+
+      def calendar
+        @resource[:calendars][@calendar_type]
+      end
+
+      protected
+
+      def init_resources
+        @resource = TwitterCldr.resources.resource_for(@locale, "calendars")[TwitterCldr.convert_locale(@locale)]
+      end
+
+      def init_placeholders
+        @placeholders = [{ :name => :date, :object => TwitterCldr::Tokenizers::DateTokenizer.new(:locale => @locale) },
+                         { :name => :time, :object => TwitterCldr::Tokenizers::TimeTokenizer.new(:locale => @locale) }]
+      end
+
+      def pattern_for(resource)
+        resource.is_a?(Hash) ? resource[:pattern] : resource
+      end
+    end
+  end
+end

data/lib/tokenizers/calendars/time_tokenizer.rb

@@ -0,0 +1,24 @@
+module TwitterCldr
+  module Tokenizers
+    class TimeTokenizer < TwitterCldr::Tokenizers::DateTimeTokenizer
+      def initialize(options = {})
+        super(options)
+        @token_splitter_regex = /(\'[\w\s-]+\'|a{1}|h{1,2}|H{1,2}|K{1,2}|k{1,2}|m{1,2}|s{1,2}|S+|z{1,4}|Z{1,4})/
+        @token_type_regexes = [{ :type => :pattern, :regex => /^a{1}|h{1,2}|H{1,2}|K{1,2}|k{1,2}|m{1,2}|s{1,2}|S+|z{1,4}|Z{1,4}/ },
+                               { :type => :plaintext, :regex => // }]
+        @paths = { :default => "calendars.gregorian.formats.time.default",
+                   :full => "calendars.gregorian.formats.time.full",
+                   :long => "calendars.gregorian.formats.time.long",
+                   :medium => "calendars.gregorian.formats.time.medium",
+                   :short => "calendars.gregorian.formats.time.short" }
+      end
+
+      protected
+
+      # must override this because DateTimeTokenizer will set them otherwise
+      def init_placeholders
+        @placeholders = {}
+      end
+    end
+  end
+end

data/lib/tokenizers/key_path.rb

@@ -0,0 +1,30 @@
+module TwitterCldr
+  module Tokenizers
+    class KeyPath
+      class << self
+        def dirname(path)
+          self.join_path(self.split_path(path)[0..-2])
+        end
+
+        def join(*args)
+          final = ""
+          args.each do |arg|
+            fixed_arg = arg.chomp(".")
+            fixed_arg = arg[1..-1] if fixed_arg[0].chr == "."
+            final << "." if final.size > 0
+            final << fixed_arg
+          end
+          final
+        end
+
+        def split_path(path)
+          path.split(/\./)
+        end
+
+        def join_path(path_arr)
+          path_arr.join(".")
+        end
+      end
+    end
+  end
+end

data/lib/tokenizers/numbers/number_tokenizer.rb

@@ -0,0 +1,50 @@
+module TwitterCldr
+  module Tokenizers
+    class NumberTokenizer < Base
+      def initialize(options = {})
+        super(options)
+        @type = options[:type] || :decimal
+        @token_splitter_regex = /([^0*#,\.]*)([0#,\.]+)([^0*#,\.]*)$/
+        @token_type_regexes = [{ :type => :pattern, :regex => /[0?#,\.]*/ },
+                               { :type => :plaintext, :regex => // }]
+        @base_path = "numbers.formats"
+        @symbol_path = "numbers.symbols"
+        @paths = { :default => "patterns.default",
+                   :positive => "patterns.positive",
+                   :negative => "patterns.negative" }
+      end
+
+      def tokens(options = {})
+        unless self.traverse(self.full_path_for(:positive))
+          key_path = self.full_path_for(:default)
+          positive, negative = self.traverse(key_path).split(/;/)
+          insert_point = self.traverse(KeyPath.dirname(key_path))
+          insert_point[:positive] = positive
+          insert_point[:negative] = negative ? negative : "#{self.symbols[:minus] || '-'}#{positive}"
+        end
+
+        sign = options[:sign] || :positive
+        self.tokens_for(self.full_path_for(sign), nil)
+      end
+
+      def symbols
+        self.traverse(@symbol_path)
+      end
+
+      protected
+
+      def full_path_for(path)
+        KeyPath.join(@base_path, @type.to_s, self.paths[path])
+      end
+
+      def init_resources
+        @resource = TwitterCldr.resources.resource_for(@locale, "numbers")[TwitterCldr.convert_locale(@locale)]
+      end
+
+      def pattern_for(resource)
+        # can't go any deeper, so return original pattern (which should NOT be a hash, by the way)
+        resource
+      end
+    end
+  end
+end

data/lib/tokenizers/token.rb

@@ -0,0 +1,17 @@
+module TwitterCldr
+  module Tokenizers
+    class Token
+      attr_accessor :value, :type
+
+      def initialize(options = {})
+        options.each_pair do |key, val|
+          self.send("#{key.to_s}=", val)
+        end
+      end
+
+      def to_s
+        @value
+      end
+    end
+  end
+end

data/lib/twitter_cldr.rb

@@ -0,0 +1,104 @@
+$:.push(File.dirname(__FILE__))
+
+require 'yaml'
+require 'date'
+require 'time'
+
+require 'version'
+
+# patches for extending Ruby functionality
+require 'ext/localized_object'
+require 'ext/calendars/datetime'
+require 'ext/calendars/date'
+require 'ext/calendars/time'
+require 'ext/numbers/localized_number'
+require 'ext/numbers/bignum'
+require 'ext/numbers/fixnum'
+require 'ext/numbers/float'
+require 'ext/strings/symbol'
+
+# manages access to CLDR resources (yaml files in resources dir)
+require 'shared/resources'
+
+
+module TwitterCldr
+  DEFAULT_LOCALE = :en
+  RESOURCE_DIR = File.join(File.dirname(File.dirname(File.expand_path(__FILE__))), "resources")
+
+  # maps twitter locales to cldr locales
+  TWITTER_LOCALE_MAP = { :msa     => :ms,
+                         :'zh-cn' => :zh,
+                         :'zh-tw' => :'zh-Hant' }
+
+  @@resources = TwitterCldr::Shared::Resources.new
+
+  def self.get_resource_file(locale, resource)
+    File.join(RESOURCE_DIR, self.convert_locale(locale).to_s, "#{resource}.yml")
+  end
+
+  def self.resources
+    @@resources
+  end
+
+  def self.get_locale
+    if defined?(FastGettext)
+      locale = FastGettext.locale
+      locale = DEFAULT_LOCALE if locale.to_s.empty?
+    else
+      locale = DEFAULT_LOCALE
+    end
+
+    (self.supported_locale?(locale) ? locale : DEFAULT_LOCALE).to_sym
+  end
+
+  def self.convert_locale(locale)
+    locale = locale.to_sym
+    TWITTER_LOCALE_MAP.include?(locale) ? TWITTER_LOCALE_MAP[locale] : locale
+  end
+
+  def self.supported_locales
+    unless defined?(@@supported_locales)
+      rejectable = [:shared]
+      @@supported_locales = Dir.glob(File.join(File.dirname(File.dirname(__FILE__)), "resources/*")).map do |file|
+        File.basename(file).to_sym
+      end.reject { |file| rejectable.include?(file) }
+    end
+
+    @@supported_locales
+  end
+
+  def self.supported_locale?(locale)
+    locale = locale.to_sym
+    self.supported_locales.include?(locale) || self.supported_locales.include?(self.convert_locale(locale))
+  end
+end
+
+
+# other shared libraries (most access shared resource data in resources/shared)
+require 'shared/currencies'
+require 'shared/languages'
+
+# all tokenizers
+require 'tokenizers/base'
+require 'tokenizers/key_path'
+require 'tokenizers/token'
+require 'tokenizers/calendars/datetime_tokenizer'
+require 'tokenizers/calendars/date_tokenizer'
+require 'tokenizers/calendars/time_tokenizer'
+require 'tokenizers/numbers/number_tokenizer'
+
+# all formatters
+require 'formatters/base'
+require 'formatters/calendars/datetime_formatter'
+require 'formatters/calendars/date_formatter'
+require 'formatters/calendars/time_formatter'
+require 'formatters/numbers/number_formatter'
+require 'formatters/numbers/decimal_formatter'
+require 'formatters/numbers/currency_formatter'
+require 'formatters/numbers/percent_formatter'
+require 'formatters/plurals/rules'
+
+# formatter helpers
+require 'formatters/numbers/helpers/base'
+require 'formatters/numbers/helpers/fraction'
+require 'formatters/numbers/helpers/integer'

data/lib/version.rb

@@ -0,0 +1,3 @@
+module TwitterCldr
+  VERSION = "1.0.0"
+end

data/resources/ar/calendars.yml

@@ -0,0 +1,145 @@
+ar: 
+  calendars: 
+    gregorian: 
+      days: 
+        format: 
+          abbreviated: 
+            fri: جمعة
+            mon: اثنين
+            sat: سبت
+            sun: أحد
+            thu: خميس
+            tue: ثلاثاء
+            wed: أربعاء
+          narrow: :"calendars.gregorian.days.stand-alone.narrow"
+          wide: 
+            fri: الجمعة
+            mon: الاثنين
+            sat: السبت
+            sun: الأحد
+            thu: الخميس
+            tue: الثلاثاء
+            wed: الأربعاء
+        stand-alone: 
+          abbreviated: :"calendars.gregorian.days.format.abbreviated"
+          narrow: 
+            fri: ج
+            mon: ن
+            sat: س
+            sun: ح
+            thu: خ
+            tue: ث
+            wed: ر
+          wide: :"calendars.gregorian.days.format.wide"
+      fields: 
+        day: يوم
+        dayperiod: ص/م
+        era: العصر
+        hour: الساعات
+        minute: الدقائق
+        month: الشهر
+        second: الثواني
+        week: الأسبوع
+        weekday: اليوم
+        year: السنة
+        zone: التوقيت
+      formats: 
+        date: 
+          default: :"calendars.gregorian.formats.date.medium"
+          full: 
+            pattern: "EEEE، d MMMM، y"
+          long: 
+            pattern: "d MMMM، y"
+          medium: 
+            pattern: dd‏/MM‏/yyyy
+          short: 
+            pattern: d‏/M‏/yyyy
+        datetime: 
+          default: :"calendars.gregorian.formats.datetime.medium"
+          full: 
+            pattern: "{1} {0}"
+          long: 
+            pattern: "{1} {0}"
+          medium: 
+            pattern: "{1} {0}"
+          short: 
+            pattern: "{1} {0}"
+        time: 
+          default: :"calendars.gregorian.formats.time.medium"
+          full: 
+            pattern: "zzzz h:mm:ss a"
+          long: 
+            pattern: "z h:mm:ss a"
+          medium: 
+            pattern: "h:mm:ss a"
+          short: 
+            pattern: "h:mm a"
+      months: 
+        format: 
+          abbreviated: 
+            1: يناير
+            10: أكتوبر
+            11: نوفمبر
+            12: ديسمبر
+            2: فبراير
+            3: مارس
+            4: أبريل
+            5: مايو
+            6: يونيو
+            7: يوليو
+            8: أغسطس
+            9: سبتمبر
+          narrow: :"calendars.gregorian.months.stand-alone.narrow"
+          wide: 
+            1: يناير
+            10: أكتوبر
+            11: نوفمبر
+            12: ديسمبر
+            2: فبراير
+            3: مارس
+            4: أبريل
+            5: مايو
+            6: يونيو
+            7: يوليو
+            8: أغسطس
+            9: سبتمبر
+        stand-alone: 
+          abbreviated: :"calendars.gregorian.months.format.abbreviated"
+          narrow: 
+            1: ي
+            10: ك
+            11: ب
+            12: د
+            2: ف
+            3: م
+            4: أ
+            5: و
+            6: ن
+            7: ل
+            8: غ
+            9: س
+          wide: :"calendars.gregorian.months.format.wide"
+      periods: 
+        am: ص
+        pm: م
+      quarters: 
+        format: 
+          abbreviated: 
+            1: "الربع الأول"
+            2: "الربع الثاني"
+            3: "الربع الثالث"
+            4: "الربع الرابع"
+          narrow: :"calendars.gregorian.quarters.stand-alone.narrow"
+          wide: 
+            1: "الربع الأول"
+            2: "الربع الثاني"
+            3: "الربع الثالث"
+            4: "الربع الرابع"
+        stand-alone: 
+          abbreviated: :"calendars.gregorian.quarters.format.abbreviated"
+          narrow: 
+            1: ١
+            2: ٢
+            3: ٣
+            4: ٤
+          wide: :"calendars.gregorian.quarters.format.wide"