phonetic 1.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: cb22f0be1272e5b72586a964943808d93f99c5c2
+  data.tar.gz: cfab5ebba7adc3823f9c74a92b5877681e73d36d
+SHA512:
+  metadata.gz: da460d1d048d38d39af6970b3c456551217551d3aa1275d8ec83cc86ea6ae5fcf454842ccb0300d93726303ab9009639bd72c894d99e30d9773edd269e0f41c8
+  data.tar.gz: 3b94a722e973f37eccd62841d0cd64be3e0dbce0ab27766cd4384a01e375440334f74cc22b14c21804489a5e87f6052911f3decf5b1a2c1f4f24e7b6459306cc

data/.gitignore

@@ -0,0 +1,17 @@
+*.gem
+*.rbc
+.bundle
+.config
+.yardoc
+Gemfile.lock
+InstalledFiles
+_yardoc
+coverage
+doc/
+lib/bundler/man
+pkg
+rdoc
+spec/reports
+test/tmp
+test/version_tmp
+tmp

data/.rspec

@@ -0,0 +1,2 @@
+--color
+--format doc

data/Gemfile

@@ -0,0 +1,3 @@
+source 'https://rubygems.org'
+
+gemspec

data/LICENSE.txt

@@ -0,0 +1,22 @@
+Copyright (c) 2013 n7v
+
+MIT License
+
+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

@@ -0,0 +1,98 @@
+# Phonetic
+
+Ruby library for phonetic algorithms.
+It supports Soundex, Metaphone, Double Metaphone, Caverphone, NYSIIS and others.
+
+## Installation
+
+Add this line to your application's Gemfile:
+
+    gem 'phonetic'
+
+And then execute:
+
+```shell
+$ bundle
+```
+
+Or install it yourself as:
+
+```shell
+$ gem install phonetic
+```
+
+## Usage
+
+```ruby
+require 'phonetic'
+```
+
+### Soundex
+
+```ruby
+'Ackerman'.soundex # => 'A265'
+'ammonium'.soundex # => 'A500'
+'implementation'.soundex # => 'I514'
+```
+
+### Refined Soundex
+
+```ruby
+'Caren'.refined_soundex   # => 'C30908'
+'Hayers'.refined_soundex  # => 'H093'
+'Lambard'.refined_soundex # => 'L7081096'
+```
+
+### Metaphone
+
+```ruby
+'Accola'.metaphone # => 'AKKL'
+'Nikki'.metaphone # => 'NK'
+'Wright'.metaphone #=> 'RT'
+```
+
+### Double Metaphone
+
+```ruby
+'czerny'.double_metaphone # => ['SRN', 'XRN']
+'dumb'.double_metaphone   # => ['TM', 'TM']
+'edgar'.double_metaphone  # => ['ATKR', 'ATKR']
+```
+
+or use alias:
+
+```ruby
+'czerny'.metaphone2 # => ['SRN', 'XRN']
+'dumb'.metaphone2   # => ['TM', 'TM']
+'edgar'.metaphone2  # => ['ATKR', 'ATKR']
+```
+
+### Caverphone
+
+```ruby
+'Lashaunda'.caverphone # => 'LSNT11'
+'Vidaurri'.caverphone # => 'FTR111'
+````
+
+### Caverphone 2
+
+```ruby
+'Stevenson'.caverphone2 # => 'STFNSN1111'
+'Peter'.caverphone2 # => 'PTA1111111'
+```
+
+### NYSIIS
+
+```ruby
+'Alexandra'.nysiis # => 'ALAXANDR'
+'Aumont'.nysiis # => 'AANAD'
+'Bonnie'.nysiis # => 'BANY'
+```
+
+## Contributing
+
+1. Fork it
+2. Create your feature branch (`git checkout -b my-new-feature`)
+3. Commit your changes (`git commit -am 'Add some feature'`)
+4. Push to the branch (`git push origin my-new-feature`)
+5. Create new Pull Request

data/Rakefile

@@ -0,0 +1 @@
+require "bundler/gem_tasks"

data/lib/phonetic.rb

@@ -0,0 +1,10 @@
+require 'phonetic/version'
+require 'phonetic/nysiis'
+require 'phonetic/soundex'
+require 'phonetic/refined_soundex'
+require 'phonetic/metaphone'
+require 'phonetic/double_metaphone'
+require 'phonetic/metaphone2'
+require 'phonetic/caverphone'
+require 'phonetic/caverphone2'
+require 'phonetic/core_ext/string'

data/lib/phonetic/algorithm.rb

@@ -0,0 +1,24 @@
+module Phonetic
+  # Base class for phonetic algorithms.
+  class Algorithm
+    # Generic method for encoding single word. Override it in your algorithm class.
+    # @param [String] word the word to encode
+    # @param [Hash] options the options for the algorithm
+    # @return [String] the word
+    def self.encode_word(word, options = {})
+      word
+    end
+
+    # Generic method for encoding string.
+    # Splits string by words and encodes it with {Algorithm.encode_word}.
+    #
+    # @param [String] str the string to encode.
+    # @param [Hash] options the options for algorithm.
+    # @return [String] the space separated codes of words from input string.
+    def self.encode(str, options = {})
+      str.scan(/\p{Word}+/).map do |word|
+        encode_word(word, options)
+      end.compact.reject(&:empty?).join(' ')
+    end
+  end
+end

data/lib/phonetic/caverphone.rb

@@ -0,0 +1,68 @@
+require 'phonetic/algorithm'
+
+module Phonetic
+  # Caverphone created by the Caversham Project at the University of Otago.
+  # @see http://caversham.otago.ac.nz/files/working/ctp060902.pdf Caverphone: Phonetic Matching algorithm by David Hood (2002)
+  # This class implements this algorithm.
+  # @example
+  #    Phonetic::Caverphone.encode('Charmain') # => 'KMN111'
+  #    Phonetic::Caverphone.encode('Ellett')   # => 'ALT111'
+  #    Phonetic::Caverphone.encode('Siegmund') # => 'SKMNT1'
+  class Caverphone < Algorithm
+    MAP = {
+      /^(cou|rou|tou|enou)gh/ => '\12f',
+      /^gn/ => '2n',
+      /mb$/ => 'mb',
+      'cq' => '2q',
+      /c([iey])/ => 's\1',
+      'tch' => '2ch',
+      /[cqx]/ => 'k',
+      'v' => 'f',
+      'dg' => '2g',
+      /ti([oa])/ => 'si\1',
+      'd' => 't',
+      'ph' => 'fh',
+      'b' => 'p',
+      'sh' => 's2',
+      'z' => 's',
+      /^[aeiou]/ => 'A',
+      /[aeiou]/ => '3',
+      '3gh3' => '3kh3',
+      'gh' => '22',
+      'g' => 'k',
+      /s+/ => 'S',
+      /t+/ => 'T',
+      /p+/ => 'P',
+      /k+/ => 'K',
+      /f+/ => 'F',
+      /m+/ => 'M',
+      /n+/ => 'N',
+      'w3' => 'W3',
+      /wy/ => 'Wy',
+      'wh3' => 'Wh3',
+      'why' => 'Why',
+      'w' => '2',
+      /^h/ => 'A',
+      'h' => '2',
+      'r3' => 'R3',
+      'ry' => 'Ry',
+      'r' => '2',
+      'l3' => 'L3',
+      'ly' => 'Ly',
+      'l' => '2',
+      'j' => 'y',
+      'y3' => 'Y3',
+      'y' => '2',
+      '2' => '',
+      '3' => ''
+    }
+
+    # Encode word to its Caverphone code
+    def self.encode_word(word, options = {})
+      w = word.strip.downcase.gsub(/[^a-z]/, '')
+      MAP.each { |r, v| w.gsub!(r, v) }
+      w = w + '1' * 6
+      w[0..5]
+    end
+  end
+end

data/lib/phonetic/caverphone2.rb

@@ -0,0 +1,69 @@
+require 'phonetic/algorithm'
+
+module Phonetic
+  # Caverphone 2.0 created by the Caversham Project at the University of Otago.
+  # @see http://caversham.otago.ac.nz/files/working/ctp150804.pdf Caverphone Revisited by David Hood (2004)
+  # This class implements this algorithm.
+  # @example
+  #    Phonetic::Caverphone2.encode('Stevenson') # => 'STFNSN1111'
+  #    Phonetic::Caverphone2.encode('Peter') # => 'PTA1111111'
+  class Caverphone2 < Algorithm
+    MAP = {
+      /e$/ => '',
+      /^(cou|rou|tou|enou|trou)gh/ => '\12f',
+      /^gn/ => '2n',
+      /mb$/ => 'mb',
+      'cq' => '2q',
+      /c([iey])/ => 's\1',
+      'tch' => '2ch',
+      /[cqx]/ => 'k',
+      'v' => 'f',
+      'dg' => '2g',
+      /ti([oa])/ => 'si\1',
+      'd' => 't',
+      'ph' => 'fh',
+      'b' => 'p',
+      'sh' => 's2',
+      'z' => 's',
+      /^[aeiou]/ => 'A',
+      /[aeiou]/ => '3',
+      'j' => 'y',
+      /^y3/ => 'Y3',
+      /^y/ => 'A',
+      /y/ => '3',
+      '3gh3' => '3kh3',
+      'gh' => '22',
+      'g' => 'k',
+      /s+/ => 'S',
+      /t+/ => 'T',
+      /p+/ => 'P',
+      /k+/ => 'K',
+      /f+/ => 'F',
+      /m+/ => 'M',
+      /n+/ => 'N',
+      'w3' => 'W3',
+      'wh3' => 'Wh3',
+      /w$/ => '3',
+      'w' => '2',
+      /^h/ => 'A',
+      'h' => '2',
+      'r3' => 'R3',
+      /r$/ => '3',
+      'r' => '2',
+      'l3' => 'L3',
+      /l$/ => '3',
+      'l' => '2',
+      '2' => '',
+      /3$/ => 'A',
+      '3' => ''
+    }
+
+    # Encode word to its Caverphone 2 code
+    def self.encode_word(word, options = {})
+      w = word.strip.downcase.gsub(/[^a-z]/, '')
+      MAP.each { |r, v| w.gsub!(r, v) }
+      w = w + '1' * 10
+      w[0..9]
+    end
+  end
+end

data/lib/phonetic/core_ext/string.rb

@@ -0,0 +1,3 @@
+Gem.find_files('phonetic/core_ext/string/*.rb')
+   .reject{|path| path =~ /_spec/}
+   .each { |path| require path }

data/lib/phonetic/core_ext/string/caverphone.rb

@@ -0,0 +1,11 @@
+require 'phonetic/caverphone'
+
+class String
+  # Caverphone value of string
+  # @example
+  #    'Lashaunda'.caverphone # => 'LSNT11'
+  #    'Vidaurri'.caverphone # => 'FTR111'
+  def caverphone(options = {})
+    Phonetic::Caverphone.encode(self, options)
+  end
+end

data/lib/phonetic/core_ext/string/caverphone2.rb

@@ -0,0 +1,11 @@
+require 'phonetic/caverphone2'
+
+class String
+  # Caverphone 2 value of string
+  # @example
+  #    'Stevenson'.caverphone2 # => 'STFNSN1111'
+  #    'Peter'.caverphone2 # => 'PTA1111111'
+  def caverphone2(options = {})
+    Phonetic::Caverphone2.encode(self, options)
+  end
+end

data/lib/phonetic/core_ext/string/double_metaphone.rb

@@ -0,0 +1,18 @@
+require 'phonetic/double_metaphone'
+
+class String
+  # Double Metahpone code of string.
+  # @example
+  #    'czerny'.double_metaphone # => ['SRN', 'XRN']
+  #    'dumb'.double_metaphone   # => ['TM', 'TM']
+  #    'edgar'.double_metaphone  # => ['ATKR', 'ATKR']
+  #    # or use alias:
+  #    'czerny'.metaphone2 # => ['SRN', 'XRN']
+  #    'dumb'.metaphone2   # => ['TM', 'TM']
+  #    'edgar'.metaphone2  # => ['ATKR', 'ATKR']
+  def double_metaphone(options = { size: 4 })
+    Phonetic::DoubleMetaphone.encode(self, options)
+  end
+
+  alias_method :metaphone2, :double_metaphone
+end

data/lib/phonetic/core_ext/string/metaphone.rb

@@ -0,0 +1,12 @@
+require 'phonetic/metaphone'
+
+class String
+  # Metaphone value of string.
+  # @example
+  #    'Accola'.metaphone # => 'AKKL'
+  #    'Nikki'.metaphone # => 'NK'
+  #    'Wright'.metaphone #=> 'RT'
+  def metaphone(options = { size: 4 })
+    Phonetic::Metaphone.encode(self, options)
+  end
+end

data/lib/phonetic/core_ext/string/nysiis.rb

@@ -0,0 +1,12 @@
+require 'phonetic/nysiis'
+
+class String
+  # Caverphone value of string.
+  # @example
+  #    'Alexandra'.nysiis # => 'ALAXANDR'
+  #    'Aumont'.nysiis # => 'AANAD'
+  #    'Bonnie'.nysiis # => 'BANY'
+  def nysiis(options = { trim: true })
+    Phonetic::NYSIIS.encode(self, options)
+  end
+end

data/lib/phonetic/core_ext/string/refined_soundex.rb

@@ -0,0 +1,12 @@
+require 'phonetic/refined_soundex'
+
+class String
+  # Refined Soundex value of string.
+  # @example
+  #    'Caren'.refined_soundex   # => 'C30908'
+  #    'Hayers'.refined_soundex  # => 'H093'
+  #    'Lambard'.refined_soundex # => 'L7081096'
+  def refined_soundex(options = { trim: true })
+    Phonetic::RefinedSoundex.encode(self, options)
+  end
+end

data/lib/phonetic/core_ext/string/soundex.rb

@@ -0,0 +1,12 @@
+require 'phonetic/soundex'
+
+class String
+  # Soundex value of string
+  # @example
+  #    'Ackerman'.soundex # => 'A265'
+  #    'ammonium'.soundex # => 'A500'
+  #    'implementation'.soundex # => 'I514'
+  def soundex(options = { trim: true })
+    Phonetic::Soundex.encode(self, options)
+  end
+end

data/lib/phonetic/double_metaphone.rb

@@ -0,0 +1,640 @@
+# encoding: utf-8
+
+require 'phonetic/algorithm'
+
+module Phonetic
+  # The Double Metaphone phonetic encoding algorithm is the second generation
+  # of the Metaphone algorithm. Its original implementation was described
+  # by Lawrence Philips in the June 2000 issue of C/C++ Users Journal.
+  #
+  # This implementation based on the PHP implementation by Stephen Woodbridge
+  # and contains modifications of algorithm by Kevin Atkinson.
+  # @see http://swoodbridge.com/DoubleMetaPhone/ PHP implementation by Stephen Woodbridge
+  # @see http://aspell.net/metaphone/dmetaph.cpp C++ implementation with modifications by Kevin Atkinson
+  # @example
+  #    Phonetic::DoubleMetaphone.encode('czerny') # => ['SRN', 'XRN']
+  #    Phonetic::DoubleMetaphone.encode('dumb')   # => ['TM', 'TM']
+  #    Phonetic::DoubleMetaphone.encode('edgar')  # => ['ATKR', 'ATKR']
+  #    # or use alias:
+  #    Phonetic::Metaphone2.encode('czerny') # => ['SRN', 'XRN']
+  #    Phonetic::Metaphone2.encode('dumb')   # => ['TM', 'TM']
+  #    Phonetic::Metaphone2.encode('edgar')  # => ['ATKR', 'ATKR']
+  class DoubleMetaphone < Algorithm
+    VOWELS = 'AEIOUY'
+
+    # Encode word to its Double Metaphone code.
+    def self.encode_word(word, options = { size: 4 })
+      code_size = options[:size] || 4
+      w = word.strip.upcase
+      primary = ''
+      secondary = ''
+      i = 0
+      len = w.size
+      last = len - 1
+      # pad the original string so that we can index beyond the edge of the world
+      w += ' ' * 5
+      # skip these when at start of word
+      i += 1 if ['GN','KN','PN','WR','PS'].include? w[0, 2]
+      # initial 'X' is pronounced 'Z' e.g. 'Xavier'
+      if w[0] == 'X'
+        primary += 'S'
+        secondary += 'S'
+        i += 1
+      end
+      while i < len && (primary.size < code_size || primary.size < code_size)
+        case w[i]
+        when 'A', 'E', 'I', 'O', 'U', 'Y'
+          if i == 0
+            # all init vowels now map to 'A'
+            primary += 'A'
+            secondary += 'A'
+          end
+          i += 1
+        when 'B'
+          # "-mb", e.g", "dumb", already skipped over...
+          primary += 'P'
+          secondary += 'P'
+          i += (w[i + 1] == 'B') ? 2 : 1
+        when 'Ç', 'ç'
+          primary += 'S'
+          secondary += 'S'
+          i += 1
+        when 'C'
+          # various germanic
+          if i > 1 && !vowel?(w[i - 2]) && w[i - 1, 3] == 'ACH' &&
+             (w[i + 2] != 'I' && (w[i + 2] != 'E' || w[i - 2, 6] =~ /[BM]ACHER/))
+            primary += 'K'
+            secondary += 'K'
+            i += 2
+          # special case 'caesar'
+          elsif i == 0 && w[i, 6] == 'CAESAR'
+            primary += 'S'
+            secondary += 'S'
+            i += 2
+          # italian 'chianti'
+          elsif w[i, 4] == 'CHIA'
+            primary += 'K'
+            secondary += 'K'
+            i += 2
+          elsif w[i, 2] == 'CH'
+            # find 'michael'
+            if i > 0 && w[i, 4] == 'CHAE'
+              primary += 'K'
+              secondary += 'X'
+              i += 2
+            # greek roots e.g. 'chemistry', 'chorus'
+            elsif i == 0 && (w[i + 1, 5] =~ /HARAC|HARIS/ || w[i + 1, 3] =~ /HOR|HYM|HIA|HEM/) &&
+                  w[0, 5] != 'CHORE'
+              primary += 'K'
+              secondary += 'K'
+              i += 2
+            else
+              # germanic, greek, or otherwise 'ch' for 'kh' sound
+              if (w[0, 4] =~ /(VAN|VON)\s/ || w[0, 3] == 'SCH') ||
+                 # 'architect but not 'arch', 'orchestra', 'orchid'
+                 (i > 1 && w[i - 2, 6] =~ /ORCHES|ARCHIT|ORCHID/) ||
+                 (w[i + 2] =~ /[TS]/) ||
+                 ((i > 0 && w[i - 1] =~ /[AOUE]/) || i == 0) &&
+                 # e.g., 'wachtler', 'wechsler', but not 'tichner'
+                 (w[i + 2] =~ /[LRNMBHFVW ]/ || i + 2 >= len)
+                primary += 'K'
+                secondary += 'K'
+              else
+                if i > 0
+                  if w[0, 2] == 'MC'
+                    # e.g., "McHugh"
+                    primary += 'K'
+                    secondary += 'K'
+                  else
+                    primary += 'X'
+                    secondary += 'K'
+                  end
+                else
+                  primary += 'X'
+                  secondary += 'X'
+                end
+              end
+              i += 2
+            end
+          elsif w[i, 2] == 'CZ' && !(i > 1 && w[i - 2, 4] == 'WICZ')
+            # e.g, 'czerny'
+            primary += 'S'
+            secondary += 'X'
+            i += 2
+          elsif w[i + 1, 3] == 'CIA'
+            # e.g., 'focaccia'
+            primary += 'X'
+            secondary += 'X'
+            i += 3
+          # double 'C', but not if e.g. 'McClellan'
+          elsif w[i, 2] == 'CC' && !(i == 1 && w[0] == 'M')
+            # 'bellocchio' but not 'bacchus'
+            if w[i + 2, 1] =~ /[IEH]/ && w[i + 2, 2] != 'HU'
+              # 'accident', 'accede' 'succeed'
+              if i == 1 && w[i - 1] == 'A' || w[i - 1, 5] =~ /UCCEE|UCCES/
+                # 'bacci', 'bertucci', other italian
+                primary += 'KS'
+                secondary += 'KS'
+              else
+                primary += 'X'
+                secondary += 'X'
+              end
+              i += 3
+            else
+              # Pierce's rule
+              primary += 'K'
+              secondary += 'K'
+              i += 2
+            end
+          elsif w[i, 2] =~ /CK|CG|CQ/
+            primary += 'K'
+            secondary += 'K'
+            i += 2
+          elsif w[i, 2] =~ /CI|CE|CY/
+            # italian vs. english
+            if w[i, 3] =~ /CIO|CIE|CIA/
+              primary += 'S'
+              secondary += 'X'
+            else
+              primary += 'S'
+              secondary += 'S'
+            end
+            i += 2
+          else
+            primary += 'K'
+            secondary += 'K'
+            # name sent in 'mac caffrey', 'mac gregor'
+            if w[i + 1, 2] =~ /\s[CQG]/
+              i += 3
+            else
+              if w[i + 1] =~ /[CKQ]/ && !(w[i + 1, 2] =~ /CE|CI/)
+                i += 2
+              else
+                i += 1
+              end
+            end
+          end
+        when 'D'
+          if w[i, 2] == 'DG'
+            if w[i + 2] =~ /[IEY]/
+              # e.g. 'edge'
+              primary += 'J'
+              secondary += 'J'
+              i += 3
+            else
+              # e.g. 'edgar'
+              primary += 'TK'
+              secondary += 'TK'
+              i += 2
+            end
+          elsif w[i, 2] =~ /DT|DD/
+            primary += 'T'
+            secondary += 'T'
+            i += 2
+          else
+            primary += 'T'
+            secondary += 'T'
+            i += 1
+          end
+        when 'F'
+          if w[i + 1] == 'F'
+            i += 2
+          else
+            i += 1
+          end
+          primary += 'F'
+          secondary += 'F'
+        when 'G'
+          if w[i + 1] == 'H'
+            if i > 0 && !vowel?(w[i - 1])
+              primary += 'K'
+              secondary += 'K'
+              i += 2
+            elsif i == 0
+              # ghislane, ghiradelli
+              if w[i + 2] == 'I'
+                primary += 'J'
+                secondary += 'J'
+              else
+                primary += 'K'
+                secondary += 'K'
+              end
+              i += 2
+            # Parker's rule (with some further refinements) - e.g., 'hugh'
+            elsif (i > 1 && w[i - 2] =~ /[BHD]/) ||
+                  # e.g., 'bough'
+                  (i > 2 && w[i - 3] =~ /[BHD]/) ||
+                  # e.g., 'broughton'
+                  (i > 3 && w[i - 4] =~ /[BH]/)
+              i += 2
+            else
+              # e.g., 'laugh', 'McLaughlin', 'cough', 'gough', 'rough', 'tough'
+              if i > 2 && w[i - 1] == 'U' && w[i - 3] =~ /[CGLRT]/
+                primary += 'F'
+                secondary += 'F'
+              else
+                if i > 0 && w[i - 1] != 'I'
+                  primary += 'K'
+                  secondary += 'K'
+                end
+              end
+              i += 2
+            end
+          elsif w[i + 1] == 'N'
+            if i == 1 && vowel?(w[0]) && !slavo_germanic?(w)
+              primary += 'KN'
+              secondary += 'N'
+            else
+              # not e.g. 'cagney'
+              if w[i + 2, 2] != 'EY' && w[i + 1] != 'Y' && !slavo_germanic?(w)
+                primary += 'N'
+                secondary += 'KN'
+              else
+                primary += 'KN'
+                secondary += 'KN'
+              end
+            end
+            i += 2
+          # 'tagliaro'
+          elsif w[i + 1, 2] == 'LI' && !slavo_germanic?(w)
+            primary += 'KL'
+            secondary += 'L'
+            i += 2
+          # -ges-,-gep-,-gel-, -gie- at beginning
+          elsif i == 0 && (w[i + 1] == 'Y' || w[i + 1, 2] =~ /ES|EP|EB|EL|EY|IB|IL|IN|IE|EI|ER/)
+            primary += 'K'
+            secondary += 'J'
+            i += 2
+          # -ger-,  -gy-
+          elsif (w[i + 1, 2] == 'ER' || w[i + 1] == 'Y') &&
+                !(w[0, 6] =~ /[DRM]ANGER/) &&
+                !(i > 0 && w[i - 1] =~ /[EI]/) &&
+                !(i > 0 && w[i - 1, 3] =~ /RGY|OGY/)
+            primary += 'K'
+            secondary += 'J'
+            i += 2
+          # italian e.g, 'biaggi'
+          elsif w[i + 1] =~ /[EIY]/ || (i > 0 && w[i - 1, 4] =~ /[AO]GGI/)
+            if w[0, 4] =~ /(VAN|VON)\s/ || w[0, 3] == 'SCH' || w[i + 1, 2] == 'ET'
+              primary += 'K'
+              secondary += 'K'
+            else
+              if w[i + 1, 4] =~ /IER\s/
+                primary += 'J'
+                secondary += 'J'
+              else
+                primary += 'J'
+                secondary += 'K'
+              end
+            end
+            i += 2
+          else
+            if w[i + 1] == 'G'
+              i += 2
+            else
+              i += 1
+            end
+            primary += 'K'
+            secondary += 'K'
+          end
+        when 'H'
+          # only keep if first & before vowel or btw. 2 vowels
+          if (i == 0 || (i > 0 && vowel?(w[i - 1]))) && vowel?(w[i + 1])
+            primary += 'H'
+            secondary += 'H'
+            i += 2
+          else # also takes care of 'HH'
+            i += 1
+          end
+        when 'J'
+          # obvious spanish, 'jose', 'san jacinto'
+          if w[i, 4] == 'JOSE' || w[0, 4] =~ /SAN\s/
+            if i == 0 && w[i + 4] == ' ' || w[0, 4] =~ /SAN\s/
+              primary += 'H'
+              secondary += 'H'
+            else
+              primary += 'J'
+              secondary += 'H'
+            end
+            i += 1
+          else
+            if i == 0 && w[i, 4] != 'JOSE'
+              primary += 'J'
+              secondary += 'A'
+              # Yankelovich/Jankelowicz
+            else
+              # spanish pron. of e.g. 'bajador'
+              if i > 0 && vowel?(w[i - 1]) && !slavo_germanic?(w) && (w[i + 1] == 'A' || w[i + 1] == 'O')
+                primary += 'J'
+                secondary += 'H'
+              else
+                if i == last
+                  primary += 'J'
+                  #secondary += ' '
+                else
+                  if !(w[i + 1] =~ /[LTKSNMBZ]/) && !(i > 0 && w[i - 1] =~ /[SKL]/)
+                    primary += 'J'
+                    secondary += 'J'
+                  end
+                end
+              end
+            end
+            if w[i + 1] == 'J'
+              i += 2
+            else
+              i += 1
+            end
+          end
+        when 'K'
+          if w[i + 1] == 'K'
+            i += 2
+          else
+            i += 1
+          end
+          primary += 'K'
+          secondary += 'K'
+        when 'L'
+          if w[i + 1] == 'L'
+            # spanish e.g. 'cabrillo', 'gallegos'
+            if (i == len - 3 && i > 0 && w[i - 1, 4] =~ /ILLO|ILLA|ALLE/) ||
+               ((last > 0 && w[last - 1, 2] =~ /AS|OS/ || w[last] =~ /[AO]/) &&
+               (i > 0 && w[i - 1, 4] == 'ALLE'))
+              primary += 'L'
+              i += 2
+              next
+            end
+            i += 2
+          else
+            i += 1
+          end
+          primary += 'L'
+          secondary += 'L'
+        when 'M'
+          if (i > 0 && w[i - 1, 3] == 'UMB' && (i + 1 == last || w[i + 2, 2] == "ER")) ||
+             # 'dumb','thumb'
+             w[i + 1] == 'M'
+            i += 2
+          else
+            i += 1
+          end
+          primary += 'M'
+          secondary += 'M'
+        when 'N'
+          if w[i + 1] == 'N'
+            i += 2
+          else
+            i += 1
+          end
+          primary += 'N'
+          secondary += 'N'
+        when 'Ñ', 'ñ'
+          i += 1;
+          primary += 'N'
+          secondary += 'N'
+        when 'P'
+          if w[i + 1] == 'H'
+            primary += 'F'
+            secondary += 'F'
+            i += 2
+          else
+            # also account for "campbell", "raspberry"
+            if w[i + 1] =~ /[PB]/
+              i += 2
+            else
+              i += 1
+            end
+            primary += 'P'
+            secondary += 'P'
+          end
+        when 'Q'
+          if w[i + 1] == 'Q'
+            i += 2
+          else
+            i += 1
+          end
+          primary += 'K'
+          secondary += 'K'
+        when 'R'
+          # french e.g. 'rogier', but exclude 'hochmeier'
+          if i == last && !slavo_germanic?(w) &&
+             (i > 1 && w[i - 2, 2] == "IE") &&
+             !(i > 3 && w[i - 4, 2] =~ /M[EA]/)
+            secondary += 'R'
+          else
+            primary += 'R'
+            secondary += 'R'
+          end
+          if w[i + 1] == 'R'
+            i += 2
+          else
+            i += 1
+          end
+        when 'S'
+          # special cases 'island', 'isle', 'carlisle', 'carlysle'
+          if i > 0 && w[i - 1, 3] =~ /ISL|YSL/
+            i += 1
+          # special case 'sugar-'
+          elsif i == 0 && w[i, 5] == 'SUGAR'
+            primary += 'X'
+            secondary += 'S'
+            i += 1
+          elsif w[i, 2] == 'SH'
+            # germanic
+            if w[i + 1, 4] =~ /HEIM|HOEK|HOLM|HOLZ/
+              primary += 'S'
+              secondary += 'S'
+            else
+              primary += 'X'
+              secondary += 'X'
+            end
+            i += 2
+          # italian & armenian
+          elsif w[i, 3] =~ /SIO|SIA/ || w[i, 4] == 'SIAN'
+            if !slavo_germanic?(w)
+              primary += 'S'
+              secondary += 'X'
+            else
+              primary += 'S'
+              secondary += 'S'
+            end
+            i += 3
+          # german & anglicisations, e.g. 'smith' match 'schmidt', 'snider' match 'schneider'
+          # also, -sz- in slavic language altho in hungarian it is pronounced 's'
+          elsif (i == 0 && w[i + 1] =~ /[MNLW]/) || w[i + 1] == 'Z'
+            primary += 'S'
+            secondary += 'X'
+            if w[i + 1] == 'Z'
+              i += 2
+            else
+              i += 1
+            end
+          elsif w[i, 2] == 'SC'
+            # Schlesinger's rule
+            if w[i + 2] == 'H'
+              # dutch origin, e.g. 'school', 'schooner'
+              if w[i + 3, 2] =~ /OO|ER|EN|UY|ED|EM/
+                # 'schermerhorn', 'schenker'
+                if w[i + 3, 2] =~ /ER|EN/
+                  primary += 'X'
+                  secondary += 'SK'
+                else
+                  primary += 'SK'
+                  secondary += 'SK'
+                end
+                i += 3
+              else
+                if i == 0 && !vowel?(w[3]) && w[3] != 'W'
+                  primary += 'X'
+                  secondary += 'S'
+                else
+                  primary += 'X'
+                  secondary += 'X'
+                end
+                i += 3
+              end
+            elsif w[i + 2, 1] =~ /[IEY]/
+              primary += 'S'
+              secondary += 'S'
+              i += 3
+            else
+              primary += 'SK'
+              secondary += 'SK'
+              i += 3
+            end
+          else
+            # french e.g. 'resnais', 'artois'
+            if i == last && i > 1 && w[i - 2, 2] =~ /AI|OI/
+              secondary += 'S'
+            else
+              primary += 'S'
+              secondary += 'S'
+            end
+            if w[i + 1] =~ /[SZ]/
+              i += 2
+            else
+              i += 1
+            end
+          end
+        when 'T'
+          if w[i, 4] == 'TION'
+            primary += 'X'
+            secondary += 'X'
+            i += 3
+          elsif w[i, 3] =~ /TIA|TCH/
+            primary += 'X'
+            secondary += 'X'
+            i += 3
+          elsif w[i, 2] == 'TH' || w[i, 3] == 'TTH'
+            # special case 'thomas', 'thames' or germanic
+            if w[i + 2, 2] =~ /OM|AM/ || w[0, 4] =~ /VAN|VON\s/ || w[0, 3] == 'SCH'
+              primary += 'T'
+              secondary += 'T'
+            else
+              primary += '0'
+              secondary += 'T'
+            end
+            i += 2
+          else
+            if w[i + 1] =~ /[TD]/
+              i += 2
+            else
+              i += 1
+            end
+            primary += 'T'
+            secondary += 'T'
+          end
+        when 'V'
+          if w[i + 1] == 'V'
+            i += 2
+          else
+            i += 1
+          end
+          primary += 'F'
+          secondary += 'F'
+        when 'W'
+          # can also be in middle of word
+          if w[i, 2] == 'WR'
+            primary += 'R'
+            secondary += 'R'
+            i += 2
+          else
+            if i == 0 && (vowel?(w[i + 1]) || w[i, 2] == 'WH')
+              # Wasserman should match Vasserman
+              if vowel?(w[i + 1])
+                primary += 'A'
+                secondary += 'F'
+              else
+                # need Uomo to match Womo
+                primary += 'A'
+                secondary += 'A'
+              end
+            end
+            # Arnow should match Arnoff
+            if i == last && i > 0 && vowel?(w[i - 1]) ||
+               (i > 0 && w[i - 1, 5] =~ /EWSKI|EWSKY|OWSKI|OWSKY/) || w[0, 3] == 'SCH'
+              secondary += 'F'
+              i += 1
+            elsif w[i, 4] =~ /WICZ|WITZ/
+              # polish e.g. 'filipowicz'
+              primary += 'TS'
+              secondary += 'FX'
+              i += 4
+            else
+              i += 1
+            end
+          end
+        when 'X'
+          # french e.g. breaux
+          if !(i == last && ((i > 2 && w[i - 3, 3] =~ /IAU|EAU/) || (i > 1 && w[i - 2, 2] =~ /AU|OU/)))
+            primary += 'KS'
+            secondary += 'KS'
+          end
+          if w[i + 1] =~ /[CX]/
+            i += 2
+          else
+            i += 1
+          end
+        when 'Z'
+          # chinese pinyin e.g. 'zhao'
+          if w[i + 1] == 'H'
+            primary += 'J'
+            secondary += 'J'
+            i += 2
+          else
+            if w[i + 1, 2] =~ /ZO|ZI|ZA/ || slavo_germanic?(w) && (i > 0 && w[i - 1] != 'T')
+              primary += 'S'
+              secondary += 'TS';
+            else
+              primary += 'S'
+              secondary += 'S';
+            end
+            if w[i + 1] == 'Z'
+              i += 2
+            else
+              i += 1
+            end
+          end
+        else
+          i += 1
+        end
+      end
+      [primary[0, code_size], secondary[0, code_size]]
+    end
+
+    def self.encode(str, options = { size: 4 })
+      encode_word(str, options)
+    end
+
+    private
+
+    def self.slavo_germanic?(str)
+      !!(str[/W|K|CZ|WITZ/])
+    end
+
+    def self.vowel?(char)
+      c = VOWELS[char.to_s]
+      !c.nil? && !c.empty?
+    end
+
+  end
+end