activesupport 4.0.12 → 7.0.2.4

data/lib/active_support/inflector/methods.rb

@@ -1,7 +1,7 @@
-# encoding: utf-8
+# frozen_string_literal: true
 
-require 'active_support/inflector/inflections'
-require 'active_support/inflections'
+require "active_support/inflections"
+require "active_support/core_ext/object/blank"
 
 module ActiveSupport
   # The Inflector transforms words from singular to plural, class names to table
@@ -23,153 +23,212 @@ module ActiveSupport
     # pluralized using rules defined for that language. By default,
     # this parameter is set to <tt>:en</tt>.
     #
-    #   'post'.pluralize             # => "posts"
-    #   'octopus'.pluralize          # => "octopi"
-    #   'sheep'.pluralize            # => "sheep"
-    #   'words'.pluralize            # => "words"
-    #   'CamelOctopus'.pluralize     # => "CamelOctopi"
-    #   'ley'.pluralize(:es)         # => "leyes"
+    #   pluralize('post')             # => "posts"
+    #   pluralize('octopus')          # => "octopi"
+    #   pluralize('sheep')            # => "sheep"
+    #   pluralize('words')            # => "words"
+    #   pluralize('CamelOctopus')     # => "CamelOctopi"
+    #   pluralize('ley', :es)         # => "leyes"
     def pluralize(word, locale = :en)
-      apply_inflections(word, inflections(locale).plurals)
+      apply_inflections(word, inflections(locale).plurals, locale)
     end
 
-    # The reverse of +pluralize+, returns the singular form of a word in a
+    # The reverse of #pluralize, returns the singular form of a word in a
     # string.
     #
     # If passed an optional +locale+ parameter, the word will be
-    # pluralized using rules defined for that language. By default,
+    # singularized using rules defined for that language. By default,
     # this parameter is set to <tt>:en</tt>.
     #
-    #   'posts'.singularize            # => "post"
-    #   'octopi'.singularize           # => "octopus"
-    #   'sheep'.singularize            # => "sheep"
-    #   'word'.singularize             # => "word"
-    #   'CamelOctopi'.singularize      # => "CamelOctopus"
-    #   'leyes'.singularize(:es)       # => "ley"
+    #   singularize('posts')            # => "post"
+    #   singularize('octopi')           # => "octopus"
+    #   singularize('sheep')            # => "sheep"
+    #   singularize('word')             # => "word"
+    #   singularize('CamelOctopi')      # => "CamelOctopus"
+    #   singularize('leyes', :es)       # => "ley"
     def singularize(word, locale = :en)
-      apply_inflections(word, inflections(locale).singulars)
+      apply_inflections(word, inflections(locale).singulars, locale)
     end
 
-    # By default, +camelize+ converts strings to UpperCamelCase. If the argument
-    # to +camelize+ is set to <tt>:lower</tt> then +camelize+ produces
+    # Converts strings to UpperCamelCase.
+    # If the +uppercase_first_letter+ parameter is set to false, then produces
     # lowerCamelCase.
     #
-    # +camelize+ will also convert '/' to '::' which is useful for converting
+    # Also converts '/' to '::' which is useful for converting
     # paths to namespaces.
     #
-    #   'active_model'.camelize                # => "ActiveModel"
-    #   'active_model'.camelize(:lower)        # => "activeModel"
-    #   'active_model/errors'.camelize         # => "ActiveModel::Errors"
-    #   'active_model/errors'.camelize(:lower) # => "activeModel::Errors"
+    #   camelize('active_model')                # => "ActiveModel"
+    #   camelize('active_model', false)         # => "activeModel"
+    #   camelize('active_model/errors')         # => "ActiveModel::Errors"
+    #   camelize('active_model/errors', false)  # => "activeModel::Errors"
     #
     # As a rule of thumb you can think of +camelize+ as the inverse of
-    # +underscore+, though there are cases where that does not hold:
+    # #underscore, though there are cases where that does not hold:
     #
-    #   'SSLError'.underscore.camelize # => "SslError"
+    #   camelize(underscore('SSLError'))        # => "SslError"
     def camelize(term, uppercase_first_letter = true)
       string = term.to_s
-      if uppercase_first_letter
-        string = string.sub(/^[a-z\d]*/) { inflections.acronyms[$&] || $&.capitalize }
+      # String#camelize takes a symbol (:upper or :lower), so here we also support :lower to keep the methods consistent.
+      if !uppercase_first_letter || uppercase_first_letter == :lower
+        string = string.sub(inflections.acronyms_camelize_regex) { |match| match.downcase! || match }
       else
-        string = string.sub(/^(?:#{inflections.acronym_regex}(?=\b|[A-Z_])|\w)/) { $&.downcase }
+        string = string.sub(/^[a-z\d]*/) { |match| inflections.acronyms[match] || match.capitalize! || match }
+      end
+      string.gsub!(/(?:_|(\/))([a-z\d]*)/i) do
+        word = $2
+        substituted = inflections.acronyms[word] || word.capitalize! || word
+        $1 ? "::#{substituted}" : substituted
       end
-      string.gsub(/(?:_|(\/))([a-z\d]*)/i) { "#{$1}#{inflections.acronyms[$2] || $2.capitalize}" }.gsub('/', '::')
+      string
     end
 
     # Makes an underscored, lowercase form from the expression in the string.
     #
     # Changes '::' to '/' to convert namespaces to paths.
     #
-    #   'ActiveModel'.underscore         # => "active_model"
-    #   'ActiveModel::Errors'.underscore # => "active_model/errors"
+    #   underscore('ActiveModel')         # => "active_model"
+    #   underscore('ActiveModel::Errors') # => "active_model/errors"
     #
     # As a rule of thumb you can think of +underscore+ as the inverse of
-    # +camelize+, though there are cases where that does not hold:
+    # #camelize, though there are cases where that does not hold:
     #
-    #   'SSLError'.underscore.camelize # => "SslError"
+    #   camelize(underscore('SSLError'))  # => "SslError"
     def underscore(camel_cased_word)
-      word = camel_cased_word.to_s.dup
-      word.gsub!('::', '/')
-      word.gsub!(/(?:([A-Za-z\d])|^)(#{inflections.acronym_regex})(?=\b|[^a-z])/) { "#{$1}#{$1 && '_'}#{$2.downcase}" }
-      word.gsub!(/([A-Z\d]+)([A-Z][a-z])/,'\1_\2')
-      word.gsub!(/([a-z\d])([A-Z])/,'\1_\2')
+      return camel_cased_word.to_s unless /[A-Z-]|::/.match?(camel_cased_word)
+      word = camel_cased_word.to_s.gsub("::", "/")
+      word.gsub!(inflections.acronyms_underscore_regex) { "#{$1 && '_' }#{$2.downcase}" }
+      word.gsub!(/([A-Z]+)(?=[A-Z][a-z])|([a-z\d])(?=[A-Z])/) { ($1 || $2) << "_" }
       word.tr!("-", "_")
       word.downcase!
       word
     end
 
-    # Capitalizes the first word and turns underscores into spaces and strips a
-    # trailing "_id", if any. Like +titleize+, this is meant for creating pretty
-    # output.
+    # Tweaks an attribute name for display to end users.
+    #
+    # Specifically, performs these transformations:
+    #
+    # * Applies human inflection rules to the argument.
+    # * Deletes leading underscores, if any.
+    # * Removes an "_id" suffix if present.
+    # * Replaces underscores with spaces, if any.
+    # * Downcases all words except acronyms.
+    # * Capitalizes the first word.
+    # The capitalization of the first word can be turned off by setting the
+    # +:capitalize+ option to false (default is true).
+    #
+    # The trailing '_id' can be kept and capitalized by setting the
+    # optional parameter +keep_id_suffix+ to true (default is false).
     #
-    #   'employee_salary'.humanize # => "Employee salary"
-    #   'author_id'.humanize       # => "Author"
-    def humanize(lower_case_and_underscored_word)
+    #   humanize('employee_salary')                  # => "Employee salary"
+    #   humanize('author_id')                        # => "Author"
+    #   humanize('author_id', capitalize: false)     # => "author"
+    #   humanize('_id')                              # => "Id"
+    #   humanize('author_id', keep_id_suffix: true)  # => "Author id"
+    #
+    # If "SSL" was defined to be an acronym:
+    #
+    #   humanize('ssl_error') # => "SSL error"
+    #
+    def humanize(lower_case_and_underscored_word, capitalize: true, keep_id_suffix: false)
       result = lower_case_and_underscored_word.to_s.dup
+
       inflections.humans.each { |(rule, replacement)| break if result.sub!(rule, replacement) }
-      result.gsub!(/_id$/, "")
-      result.tr!('_', ' ')
-      result.gsub(/([a-z\d]*)/i) { |match|
-        "#{inflections.acronyms[match] || match.downcase}"
-      }.gsub(/^\w/) { $&.upcase }
+
+      result.tr!("_", " ")
+      result.lstrip!
+      unless keep_id_suffix
+        result.delete_suffix!(" id")
+      end
+
+      result.gsub!(/([a-z\d]+)/i) do |match|
+        match.downcase!
+        inflections.acronyms[match] || match
+      end
+
+      if capitalize
+        result.sub!(/\A\w/) do |match|
+          match.upcase!
+          match
+        end
+      end
+
+      result
+    end
+
+    # Converts just the first character to uppercase.
+    #
+    #   upcase_first('what a Lovely Day') # => "What a Lovely Day"
+    #   upcase_first('w')                 # => "W"
+    #   upcase_first('')                  # => ""
+    def upcase_first(string)
+      string.length > 0 ? string[0].upcase.concat(string[1..-1]) : ""
     end
 
     # Capitalizes all the words and replaces some characters in the string to
     # create a nicer looking title. +titleize+ is meant for creating pretty
     # output. It is not used in the Rails internals.
     #
+    # The trailing '_id','Id'.. can be kept and capitalized by setting the
+    # optional parameter +keep_id_suffix+ to true.
+    # By default, this parameter is false.
+    #
     # +titleize+ is also aliased as +titlecase+.
     #
-    #   'man from the boondocks'.titleize   # => "Man From The Boondocks"
-    #   'x-men: the last stand'.titleize    # => "X Men: The Last Stand"
-    #   'TheManWithoutAPast'.titleize       # => "The Man Without A Past"
-    #   'raiders_of_the_lost_ark'.titleize  # => "Raiders Of The Lost Ark"
-    def titleize(word)
-      humanize(underscore(word)).gsub(/\b(?<!['’`])[a-z]/) { $&.capitalize }
+    #   titleize('man from the boondocks')                       # => "Man From The Boondocks"
+    #   titleize('x-men: the last stand')                        # => "X Men: The Last Stand"
+    #   titleize('TheManWithoutAPast')                           # => "The Man Without A Past"
+    #   titleize('raiders_of_the_lost_ark')                      # => "Raiders Of The Lost Ark"
+    #   titleize('string_ending_with_id', keep_id_suffix: true)  # => "String Ending With Id"
+    def titleize(word, keep_id_suffix: false)
+      humanize(underscore(word), keep_id_suffix: keep_id_suffix).gsub(/\b(?<!\w['’`()])[a-z]/) do |match|
+        match.capitalize
+      end
     end
 
-    # Create the name of a table like Rails does for models to table names. This
-    # method uses the +pluralize+ method on the last word in the string.
+    # Creates the name of a table like Rails does for models to table names.
+    # This method uses the #pluralize method on the last word in the string.
     #
-    #   'RawScaledScorer'.tableize # => "raw_scaled_scorers"
-    #   'egg_and_ham'.tableize     # => "egg_and_hams"
-    #   'fancyCategory'.tableize   # => "fancy_categories"
+    #   tableize('RawScaledScorer') # => "raw_scaled_scorers"
+    #   tableize('ham_and_egg')     # => "ham_and_eggs"
+    #   tableize('fancyCategory')   # => "fancy_categories"
     def tableize(class_name)
       pluralize(underscore(class_name))
     end
 
-    # Create a class name from a plural table name like Rails does for table
+    # Creates a class name from a plural table name like Rails does for table
     # names to models. Note that this returns a string and not a Class (To
-    # convert to an actual class follow +classify+ with +constantize+).
+    # convert to an actual class follow +classify+ with #constantize).
     #
-    #   'egg_and_hams'.classify # => "EggAndHam"
-    #   'posts'.classify        # => "Post"
+    #   classify('ham_and_eggs') # => "HamAndEgg"
+    #   classify('posts')        # => "Post"
     #
     # Singular names are not handled correctly:
     #
-    #   'business'.classify     # => "Busines"
+    #   classify('calculus')     # => "Calculu"
     def classify(table_name)
       # strip out any leading schema name
-      camelize(singularize(table_name.to_s.sub(/.*\./, '')))
+      camelize(singularize(table_name.to_s.sub(/.*\./, "")))
     end
 
     # Replaces underscores with dashes in the string.
     #
-    #   'puni_puni'.dasherize # => "puni-puni"
+    #   dasherize('puni_puni') # => "puni-puni"
     def dasherize(underscored_word)
-      underscored_word.tr('_', '-')
+      underscored_word.tr("_", "-")
     end
 
     # Removes the module part from the expression in the string.
     #
-    #   'ActiveRecord::CoreExtensions::String::Inflections'.demodulize # => "Inflections"
-    #   'Inflections'.demodulize                                       # => "Inflections"
+    #   demodulize('ActiveSupport::Inflector::Inflections') # => "Inflections"
+    #   demodulize('Inflections')                           # => "Inflections"
+    #   demodulize('::Inflections')                         # => "Inflections"
+    #   demodulize('')                                      # => ""
     #
-    # See also +deconstantize+.
+    # See also #deconstantize.
     def demodulize(path)
       path = path.to_s
-      if i = path.rindex('::')
-        path[(i+2)..-1]
+      if i = path.rindex("::")
+        path[(i + 2)..-1]
       else
         path
       end
@@ -177,32 +236,32 @@ module ActiveSupport
 
     # Removes the rightmost segment from the constant expression in the string.
     #
-    #   'Net::HTTP'.deconstantize   # => "Net"
-    #   '::Net::HTTP'.deconstantize # => "::Net"
-    #   'String'.deconstantize      # => ""
-    #   '::String'.deconstantize    # => ""
-    #   ''.deconstantize            # => ""
+    #   deconstantize('Net::HTTP')   # => "Net"
+    #   deconstantize('::Net::HTTP') # => "::Net"
+    #   deconstantize('String')      # => ""
+    #   deconstantize('::String')    # => ""
+    #   deconstantize('')            # => ""
     #
-    # See also +demodulize+.
+    # See also #demodulize.
     def deconstantize(path)
-      path.to_s[0...(path.rindex('::') || 0)] # implementation based on the one in facets' Module#spacename
+      path.to_s[0, path.rindex("::") || 0] # implementation based on the one in facets' Module#spacename
     end
 
     # Creates a foreign key name from a class name.
     # +separate_class_name_and_id_with_underscore+ sets whether
     # the method should put '_' between the name and 'id'.
     #
-    #   'Message'.foreign_key        # => "message_id"
-    #   'Message'.foreign_key(false) # => "messageid"
-    #   'Admin::Post'.foreign_key    # => "post_id"
+    #   foreign_key('Message')        # => "message_id"
+    #   foreign_key('Message', false) # => "messageid"
+    #   foreign_key('Admin::Post')    # => "post_id"
     def foreign_key(class_name, separate_class_name_and_id_with_underscore = true)
       underscore(demodulize(class_name)) + (separate_class_name_and_id_with_underscore ? "_id" : "id")
     end
 
     # Tries to find a constant with the name specified in the argument string.
     #
-    #   'Module'.constantize     # => Module
-    #   'Test::Unit'.constantize # => Test::Unit
+    #   constantize('Module')   # => Module
+    #   constantize('Foo::Bar') # => Foo::Bar
     #
     # The name is assumed to be the one of a top-level constant, no matter
     # whether it starts with "::" or not. No lexical context is taken into
@@ -211,42 +270,20 @@ module ActiveSupport
     #   C = 'outside'
     #   module M
     #     C = 'inside'
-    #     C               # => 'inside'
-    #     'C'.constantize # => 'outside', same as ::C
+    #     C                # => 'inside'
+    #     constantize('C') # => 'outside', same as ::C
     #   end
     #
     # NameError is raised when the name is not in CamelCase or the constant is
     # unknown.
     def constantize(camel_cased_word)
-      names = camel_cased_word.split('::')
-      names.shift if names.empty? || names.first.empty?
-
-      names.inject(Object) do |constant, name|
-        if constant == Object
-          constant.const_get(name)
-        else
-          candidate = constant.const_get(name)
-          next candidate if constant.const_defined?(name, false)
-          next candidate unless Object.const_defined?(name)
-
-          # Go down the ancestors to check it it's owned
-          # directly before we reach Object or the end of ancestors.
-          constant = constant.ancestors.inject do |const, ancestor|
-            break const    if ancestor == Object
-            break ancestor if ancestor.const_defined?(name, false)
-            const
-          end
-
-          # owner is in Object, so raise
-          constant.const_get(name, false)
-        end
-      end
+      Object.const_get(camel_cased_word)
     end
 
     # Tries to find a constant with the name specified in the argument string.
     #
-    #   'Module'.safe_constantize     # => Module
-    #   'Test::Unit'.safe_constantize # => Test::Unit
+    #   safe_constantize('Module')   # => Module
+    #   safe_constantize('Foo::Bar') # => Foo::Bar
     #
     # The name is assumed to be the one of a top-level constant, no matter
     # whether it starts with "::" or not. No lexical context is taken into
@@ -255,23 +292,24 @@ module ActiveSupport
     #   C = 'outside'
     #   module M
     #     C = 'inside'
-    #     C                    # => 'inside'
-    #     'C'.safe_constantize # => 'outside', same as ::C
+    #     C                     # => 'inside'
+    #     safe_constantize('C') # => 'outside', same as ::C
     #   end
     #
     # +nil+ is returned when the name is not in CamelCase or the constant (or
     # part of it) is unknown.
     #
-    #   'blargle'.safe_constantize  # => nil
-    #   'UnknownModule'.safe_constantize  # => nil
-    #   'UnknownModule::Foo::Bar'.safe_constantize  # => nil
+    #   safe_constantize('blargle')                  # => nil
+    #   safe_constantize('UnknownModule')            # => nil
+    #   safe_constantize('UnknownModule::Foo::Bar')  # => nil
     def safe_constantize(camel_cased_word)
       constantize(camel_cased_word)
     rescue NameError => e
-      raise unless e.message =~ /(uninitialized constant|wrong constant name) #{const_regexp(camel_cased_word)}$/ ||
-        e.name.to_s == camel_cased_word.to_s
-    rescue ArgumentError => e
-      raise unless e.message =~ /not missing constant #{const_regexp(camel_cased_word)}\!$/
+      raise if e.name && !(camel_cased_word.to_s.split("::").include?(e.name.to_s) ||
+        e.name.to_s == camel_cased_word.to_s)
+    rescue LoadError => e
+      message = e.respond_to?(:original_message) ? e.original_message : e.message
+      raise unless /Unable to autoload constant #{const_regexp(camel_cased_word)}/.match?(message)
     end
 
     # Returns the suffix that should be added to a number to denote the position
@@ -284,18 +322,7 @@ module ActiveSupport
     #   ordinal(-11)   # => "th"
     #   ordinal(-1021) # => "st"
     def ordinal(number)
-      abs_number = number.to_i.abs
-
-      if (11..13).include?(abs_number % 100)
-        "th"
-      else
-        case abs_number % 10
-          when 1; "st"
-          when 2; "nd"
-          when 3; "rd"
-          else    "th"
-        end
-      end
+      I18n.translate("number.nth.ordinals", number: number)
     end
 
     # Turns a number into an ordinal string used to denote the position in an
@@ -308,35 +335,43 @@ module ActiveSupport
     #   ordinalize(-11)   # => "-11th"
     #   ordinalize(-1021) # => "-1021st"
     def ordinalize(number)
-      "#{number}#{ordinal(number)}"
+      I18n.translate("number.nth.ordinalized", number: number)
     end
 
     private
+      # Mounts a regular expression, returned as a string to ease interpolation,
+      # that will match part by part the given constant.
+      #
+      #   const_regexp("Foo::Bar::Baz") # => "Foo(::Bar(::Baz)?)?"
+      #   const_regexp("::")            # => "::"
+      def const_regexp(camel_cased_word)
+        parts = camel_cased_word.split("::")
+
+        return Regexp.escape(camel_cased_word) if parts.blank?
 
-    # Mount a regular expression that will match part by part of the constant.
-    # For instance, Foo::Bar::Baz will generate Foo(::Bar(::Baz)?)?
-    def const_regexp(camel_cased_word) #:nodoc:
-      parts = camel_cased_word.split("::")
-      last  = parts.pop
+        last = parts.pop
 
-      parts.reverse.inject(last) do |acc, part|
-        part.empty? ? acc : "#{part}(::#{acc})?"
+        parts.reverse!.inject(last) do |acc, part|
+          part.empty? ? acc : "#{part}(::#{acc})?"
+        end
       end
-    end
 
-    # Applies inflection rules for +singularize+ and +pluralize+.
-    #
-    #  apply_inflections('post', inflections.plurals)    # => "posts"
-    #  apply_inflections('posts', inflections.singulars) # => "post"
-    def apply_inflections(word, rules)
-      result = word.to_s.dup
+      # Applies inflection rules for +singularize+ and +pluralize+.
+      #
+      # If passed an optional +locale+ parameter, the uncountables will be
+      # found for that locale.
+      #
+      #  apply_inflections('post', inflections.plurals, :en)    # => "posts"
+      #  apply_inflections('posts', inflections.singulars, :en) # => "post"
+      def apply_inflections(word, rules, locale = :en)
+        result = word.to_s.dup
 
-      if word.empty? || inflections.uncountables.include?(result.downcase[/\b\w+\Z/])
-        result
-      else
-        rules.each { |(rule, replacement)| break if result.sub!(rule, replacement) }
-        result
+        if word.empty? || inflections(locale).uncountables.uncountable?(result)
+          result
+        else
+          rules.each { |(rule, replacement)| break if result.sub!(rule, replacement) }
+          result
+        end
       end
-    end
   end
 end

data/lib/active_support/inflector/transliterate.rb

@@ -1,9 +1,11 @@
-# encoding: utf-8
-require 'active_support/core_ext/string/multibyte'
-require 'active_support/i18n'
+# frozen_string_literal: true
+
+require "active_support/core_ext/string/multibyte"
+require "active_support/i18n"
 
 module ActiveSupport
   module Inflector
+    ALLOWED_ENCODINGS_FOR_TRANSLITERATE = [Encoding::UTF_8, Encoding::US_ASCII, Encoding::GB18030].freeze
 
     # Replaces non-ASCII characters with an ASCII approximation, or if none
     # exists, a replacement character which defaults to "?".
@@ -51,47 +53,95 @@ module ActiveSupport
     #
     # Now you can have different transliterations for each locale:
     #
-    #   I18n.locale = :en
-    #   transliterate('Jürgen')
+    #   transliterate('Jürgen', locale: :en)
     #   # => "Jurgen"
     #
-    #   I18n.locale = :de
-    #   transliterate('Jürgen')
+    #   transliterate('Jürgen', locale: :de)
     #   # => "Juergen"
-    def transliterate(string, replacement = "?")
-      I18n.transliterate(ActiveSupport::Multibyte::Unicode.normalize(
-        ActiveSupport::Multibyte::Unicode.tidy_bytes(string), :c),
-          :replacement => replacement)
+    #
+    # Transliteration is restricted to UTF-8, US-ASCII and GB18030 strings
+    # Other encodings will raise an ArgumentError.
+    def transliterate(string, replacement = "?", locale: nil)
+      string = string.dup if string.frozen?
+      raise ArgumentError, "Can only transliterate strings. Received #{string.class.name}" unless string.is_a?(String)
+      raise ArgumentError, "Cannot transliterate strings with #{string.encoding} encoding" unless ALLOWED_ENCODINGS_FOR_TRANSLITERATE.include?(string.encoding)
+
+      input_encoding = string.encoding
+
+      # US-ASCII is a subset of UTF-8 so we'll force encoding as UTF-8 if
+      # US-ASCII is given. This way we can let tidy_bytes handle the string
+      # in the same way as we do for UTF-8
+      string.force_encoding(Encoding::UTF_8) if string.encoding == Encoding::US_ASCII
+
+      # GB18030 is Unicode compatible but is not a direct mapping so needs to be
+      # transcoded. Using invalid/undef :replace will result in loss of data in
+      # the event of invalid characters, but since tidy_bytes will replace
+      # invalid/undef with a "?" we're safe to do the same beforehand
+      string.encode!(Encoding::UTF_8, invalid: :replace, undef: :replace) if string.encoding == Encoding::GB18030
+
+      transliterated = I18n.transliterate(
+        ActiveSupport::Multibyte::Unicode.tidy_bytes(string).unicode_normalize(:nfc),
+        replacement: replacement,
+        locale: locale
+      )
+
+      # Restore the string encoding of the input if it was not UTF-8.
+      # Apply invalid/undef :replace as tidy_bytes does
+      transliterated.encode!(input_encoding, invalid: :replace, undef: :replace) if input_encoding != transliterated.encoding
+
+      transliterated
     end
 
     # Replaces special characters in a string so that it may be used as part of
     # a 'pretty' URL.
     #
-    #   class Person
-    #     def to_param
-    #       "#{id}-#{name.parameterize}"
-    #     end
-    #   end
-    #
-    #   @person = Person.find(1)
-    #   # => #<Person id: 1, name: "Donald E. Knuth">
-    #
-    #   <%= link_to(@person.name, person_path(@person)) %>
-    #   # => <a href="/person/1-donald-e-knuth">Donald E. Knuth</a>
-    def parameterize(string, sep = '-')
-      # replace accented chars with their ascii equivalents
-      parameterized_string = transliterate(string)
-      # Turn unwanted chars into the separator
-      parameterized_string.gsub!(/[^a-z0-9\-_]+/i, sep)
-      unless sep.nil? || sep.empty?
-        re_sep = Regexp.escape(sep)
+    #   parameterize("Donald E. Knuth") # => "donald-e-knuth"
+    #   parameterize("^très|Jolie-- ")  # => "tres-jolie"
+    #
+    # To use a custom separator, override the +separator+ argument.
+    #
+    #   parameterize("Donald E. Knuth", separator: '_') # => "donald_e_knuth"
+    #   parameterize("^très|Jolie__ ", separator: '_')  # => "tres_jolie"
+    #
+    # To preserve the case of the characters in a string, use the +preserve_case+ argument.
+    #
+    #   parameterize("Donald E. Knuth", preserve_case: true) # => "Donald-E-Knuth"
+    #   parameterize("^très|Jolie-- ", preserve_case: true) # => "tres-Jolie"
+    #
+    # It preserves dashes and underscores unless they are used as separators:
+    #
+    #   parameterize("^très|Jolie__ ")                 # => "tres-jolie__"
+    #   parameterize("^très|Jolie-- ", separator: "_") # => "tres_jolie--"
+    #   parameterize("^très_Jolie-- ", separator: ".") # => "tres_jolie--"
+    #
+    # If the optional parameter +locale+ is specified,
+    # the word will be parameterized as a word of that language.
+    # By default, this parameter is set to <tt>nil</tt> and it will use
+    # the configured <tt>I18n.locale</tt>.
+    def parameterize(string, separator: "-", preserve_case: false, locale: nil)
+      # Replace accented chars with their ASCII equivalents.
+      parameterized_string = transliterate(string, locale: locale)
+
+      # Turn unwanted chars into the separator.
+      parameterized_string.gsub!(/[^a-z0-9\-_]+/i, separator)
+
+      unless separator.nil? || separator.empty?
+        if separator == "-"
+          re_duplicate_separator        = /-{2,}/
+          re_leading_trailing_separator = /^-|-$/i
+        else
+          re_sep = Regexp.escape(separator)
+          re_duplicate_separator        = /#{re_sep}{2,}/
+          re_leading_trailing_separator = /^#{re_sep}|#{re_sep}$/i
+        end
         # No more than one of the separator in a row.
-        parameterized_string.gsub!(/#{re_sep}{2,}/, sep)
+        parameterized_string.gsub!(re_duplicate_separator, separator)
         # Remove leading/trailing separator.
-        parameterized_string.gsub!(/^#{re_sep}|#{re_sep}$/i, '')
+        parameterized_string.gsub!(re_leading_trailing_separator, "")
       end
-      parameterized_string.downcase
-    end
 
+      parameterized_string.downcase! unless preserve_case
+      parameterized_string
+    end
   end
 end