phonetics 1.0.0 → 1.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: c1ad0894f34019de351475721cfba624c2873944993e82db0e72ee7dd6aef374
-  data.tar.gz: 4590f9791a8409e9ef380949cc675c75507864050e72f616d4b3ac99f4dc4439
+  metadata.gz: b9f2be3c1987040125914d8a028218b495e6cd8c4c004820b4f9ffbcc9995eb8
+  data.tar.gz: 9b787ce78e582bfff9f515ab42b1f2d43de2b4224cb5d1a31831ddc8d1b5a672
 SHA512:
-  metadata.gz: 60af744c78500372afb02dd22fdc1066cc7d2e83ab218eed798ff7d96f101cbc112eb0b19c6a694ec83d79e891c5ea41e0d8cda26ad5aab209aa87aed3ba9f2e
-  data.tar.gz: 971566027a0b60cfad064f7c54c871a94830dae6c3210c77c306b31dbe0410c2687415c87686caecf9dd065812b3027b7bc6c48b69ec7fee3b01c56bfeff2644
+  metadata.gz: 4cd0298205c7e28785cdee86b95233ff9547317f4aa1bdee33f9fc3ca93fae636a6f1caf2b7a25061c02cf553aca9142d059f2eae19cb20ffb22b34ac3c2679e
+  data.tar.gz: e25d349bb3b6430f0b5001902ec9613974d60d880ccf85af4abb3ba191c73e7fb1edb3fd8c1b8af9cec39173d707a51c71d00a375de88490d7fd01092e402795

data/Gemfile.lock

@@ -6,7 +6,16 @@ PATH
 GEM
   remote: https://rubygems.org/
   specs:
+    byebug (11.0.1)
+    coderay (1.1.2)
     diff-lcs (1.3)
+    method_source (0.9.2)
+    pry (0.12.2)
+      coderay (~> 1.1.0)
+      method_source (~> 0.9.0)
+    pry-byebug (3.7.0)
+      byebug (~> 11.0)
+      pry (~> 0.10)
     rake (12.3.2)
     rspec (3.8.0)
       rspec-core (~> 3.8.0)
@@ -28,6 +37,7 @@ PLATFORMS
 DEPENDENCIES
   bundler (~> 1.16)
   phonetics!
+  pry-byebug
   rake
   rspec (~> 3.0)
 

data/VERSION

@@ -1 +1 @@
-1.0.0
+1.1.0

data/lib/phonetics/levenshtein.rb

@@ -0,0 +1,119 @@
+require_relative '../phonetics'
+
+# Using the Damerau version of the Levenshtein algorithm, with phonetic feature
+# count used instead of a binary edit distance calculation
+#
+# This implementation is almost entirely taken from the damerau-levenshtein gem
+# (https://github.com/GlobalNamesArchitecture/damerau-levenshtein/tree/master/ext/damerau_levenshtein).
+# The implementation is modified based on "Using Phonologically Weighted
+# Levenshtein Distances for the Prediction of Microscopic Intelligibility" by
+# Lionel Fontan, Isabelle Ferrané, Jérôme Farinas, Julien Pinquier, Xavier
+# Aumont, 2016
+# https://hal.archives-ouvertes.fr/hal-01474904/document
+module Phonetics
+  class Levenshtein
+    def initialize(ipa_str1, ipa_str2)
+      @str1 = ipa_str1
+      @str2 = ipa_str2
+      @len1 = ipa_str1.size
+      @len2 = ipa_str2.size
+      prepare_matrix
+      set_edit_distances(ipa_str1, ipa_str2)
+    end
+
+    def distance
+      walk.last[:distance]
+    end
+
+    def self.distance(str1, str2)
+      new(str1, str2).distance
+    end
+
+    private
+
+    def walk
+      res = []
+      cell = [@len2, @len1]
+      while cell != [0, 0]
+        cell, char = char_data(cell)
+        res.unshift char
+      end
+      res
+    end
+
+    def set_edit_distances(str1, str2)
+      (1..@len2).each do |i|
+        (1..@len1).each do |j|
+          no_change(i, j) && next if str2[i - 1] == str1[j - 1]
+          @matrix[i][j] = [del(i, j) + 1.0, ins(i, j) + 1.0, subst(i, j)].min
+        end
+      end
+    end
+
+    def char_data(cell)
+      char = { distance: @matrix[cell[0]][cell[1]] }
+      val = find_previous(cell)
+      previous_value = val[0][0]
+      char[:type] = previous_value == char[:distance] ? :same : val[1]
+      cell = val.pop
+      [cell, char]
+    end
+
+    def find_previous(cell)
+      candidates = [
+        [
+          [ins(*cell), 1],
+          :ins,
+          [cell[0], cell[1] - 1],
+        ],
+        [
+          [del(*cell), 2],
+          :del,
+          [cell[0] - 1, cell[1]],
+        ],
+        [
+          [subst(*cell), 0],
+          :subst,
+          [cell[0] - 1, cell[1] - 1],
+        ],
+      ]
+      select_cell(candidates)
+    end
+
+    def select_cell(candidates)
+      candidates.select { |e| e[-1][0] >= 0 && e[-1][1] >= 0 }.
+        sort_by(&:first).first
+    end
+
+    # TODO: Score the edit distance lower if sonorant sounds are found in sequence.
+    def del(i, j)
+      @matrix[i - 1][j]
+    end
+
+    def ins(i, j)
+      @matrix[i][j - 1]
+    end
+
+    # This is where we implement the modifications to Damerau-Levenshtein according to
+    # https://hal.archives-ouvertes.fr/hal-01474904/document
+    def subst(i, j)
+      map = Phonetics.distance_map[@str1[j]]
+      score = map[@str2[i]] if map
+      score ||= 1.0
+      @matrix[i - 1][j - 1] + score
+    end
+
+    def no_change(i, j)
+      @matrix[i][j] = @matrix[i - 1][j - 1]
+    end
+
+    def prepare_matrix
+      @matrix = []
+      @matrix << (0..@len1).to_a
+      @len2.times do |i|
+        ary = [i + 1] + (1..@len1).map { nil }
+        @matrix << ary
+      end
+    end
+  end
+end

data/phonetics.gemspec

@@ -16,7 +16,8 @@ Gem::Specification.new do |spec|
   end
   spec.require_paths = ["lib"]
 
-  spec.add_development_dependency "bundler", "~> 1.16"
-  spec.add_development_dependency "rake"
-  spec.add_development_dependency "rspec", "~> 3.0"
+  spec.add_development_dependency 'bundler', '~> 1.16'
+  spec.add_development_dependency 'rake'
+  spec.add_development_dependency "pry-byebug"
+  spec.add_development_dependency 'rspec', '~> 3.0'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: phonetics
 version: !ruby/object:Gem::Version
-  version: 1.0.0
+  version: 1.1.0
 platform: ruby
 authors:
 - Jack Danger
@@ -38,6 +38,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: pry-byebug
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -69,8 +83,8 @@ files:
 - README.md
 - Rakefile
 - VERSION
-- lib/phonetic_levenshtein.rb
 - lib/phonetics.rb
+- lib/phonetics/levenshtein.rb
 - lib/phonetics/version.rb
 - phonetics.gemspec
 homepage: https://github.com/JackDanger/phonetics

data/lib/phonetic_levenshtein.rb

@@ -1,103 +0,0 @@
-require_relative './phonetics'
-
-# Using the Damerau version of the Levenshtein algorithm, with phonetic feature
-# count used instead of a binary edit distance calculation
-#
-# This implementation is almost entirely taken from the damerau-levenshtein gem
-# (https://github.com/GlobalNamesArchitecture/damerau-levenshtein/tree/master/ext/damerau_levenshtein).
-# The implementation is modified based on "Using Phonologically Weighted
-# Levenshtein Distances for the Prediction of Microscopic Intelligibility" by
-# Lionel Fontan, Isabelle Ferrané, Jérôme Farinas, Julien Pinquier, Xavier
-# Aumont, 2016
-# https://hal.archives-ouvertes.fr/hal-01474904/document
-class PhoneticLevenshtein
-  def initialize(ipa_str1, ipa_str2)
-    @str1 = ipa_str1
-    @str2 = ipa_str2
-    @len1 = ipa_str1.size
-    @len2 = ipa_str2.size
-    prepare_matrix
-    set_edit_distances(ipa_str1, ipa_str2)
-  end
-
-  def distance
-    walk.last[:distance]
-  end
-
-  def self.distance(str1, str2)
-    new(str1, str2).distance
-  end
-
-  private
-
-  def walk
-    res = []
-    cell = [@len2, @len1]
-    while cell != [0, 0]
-      cell, char = char_data(cell)
-      res.unshift char
-    end
-    res
-  end
-
-  def set_edit_distances(str1, str2)
-    (1..@len2).each do |i|
-      (1..@len1).each do |j|
-        no_change(i, j) && next if str2[i - 1] == str1[j - 1]
-        @matrix[i][j] = [del(i, j) + 1.0, ins(i, j) + 1.0, subst(i, j)].min
-      end
-    end
-  end
-
-  def char_data(cell)
-    char = { distance: @matrix[cell[0]][cell[1]] }
-    val = find_previous(cell)
-    previous_value = val[0][0]
-    char[:type] = previous_value == char[:distance] ? :same : val[1]
-    cell = val.pop
-    [cell, char]
-  end
-
-  def find_previous(cell)
-    candidates = [[[ins(*cell), 1], :ins, [cell[0], cell[1] - 1]],
-                  [[del(*cell), 2], :del, [cell[0] - 1, cell[1]]],
-                  [[subst(*cell), 0], :subst, [cell[0] - 1, cell[1] - 1]]]
-    select_cell(candidates)
-  end
-
-  def select_cell(candidates)
-    candidates.select { |e| e[-1][0] >= 0 && e[-1][1] >= 0 }.
-      sort_by(&:first).first
-  end
-
-  # TODO: Score the edit distance lower if sonorant sounds are found in sequence.
-  def del(i, j)
-    @matrix[i - 1][j]
-  end
-
-  def ins(i, j)
-    @matrix[i][j - 1]
-  end
-
-  # This is where we implement the modifications to Damerau-Levenshtein according to
-  # https://hal.archives-ouvertes.fr/hal-01474904/document
-  def subst(i, j)
-    map = Phonetics.distance_map[@str1[j]]
-    score = map[@str2[i]] if map
-    score ||= 1.0
-    @matrix[i - 1][j - 1] + score
-  end
-
-  def no_change(i, j)
-    @matrix[i][j] = @matrix[i - 1][j - 1]
-  end
-
-  def prepare_matrix
-    @matrix = []
-    @matrix << (0..@len1).to_a
-    @len2.times do |i|
-      ary = [i + 1] + (1..@len1).map { nil }
-      @matrix << ary
-    end
-  end
-end