edits 0.3.0 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 1c829d27a322f8c9b3ad50437a3b66d2548ccfc40d435ebaee4c362b22802fb8
-  data.tar.gz: dd351b715125154bd7f9b37b98fd1f75aea504163fe550795c0615af1fea9b65
+  metadata.gz: d51ead5c11a725d7c2efcdd8030098c67e20c765ae1c1d4a17843bb4f993f781
+  data.tar.gz: dcec6dfc9b48f29050989edf108e9d930b49c66d73dbd29800466adf307b68eb
 SHA512:
-  metadata.gz: 8d5b3cd301621b913768f63c50f4876a340f620c22c2fbe68d1659bd1566141f6a737264dfb3da15f0b7a7e1a567a1036dcd978cd9c2001210f843a1e5da8f93
-  data.tar.gz: 821c4cd4acf9e2b9e25c21ce2f61a4f489838c4252fd6c4e0fd323fe9b5f3baaa1948e3ab87b0c1448830e39c9ff2362aa457e4fe9ce902994545dd8c986137e
+  metadata.gz: b7b82962d4b3fbbf84fa73d79052f449a7016529f5dd3b82960c1a9898df066f566ea3b768eec218500c65cdd55b57b34eb729f142a6fc2a0eced43415e35ef4
+  data.tar.gz: 0a0c8e0f1593847a74e25c585c2c4265b4790a641b6ce5d9cebb5c9fda0e4d971313e03efb314e937c29eeef4b168459eb1e8c6f75fd22348635110f8467b07f

data/lib/edits/damerau_levenshtein.rb

@@ -21,70 +21,66 @@ module Edits
       seq1, seq2 = seq2, seq1 if seq1.length > seq2.length
 
       # array of codepoints outperforms String
-      seq1 = seq1.codepoints if seq1.is_a? String
-      seq2 = seq2.codepoints if seq2.is_a? String
+      if seq1.is_a?(String) && seq2.is_a?(String)
+        seq1 = seq1.codepoints
+        seq2 = seq2.codepoints
+      end
 
       rows = seq1.length
       cols = seq2.length
       return cols if rows == 0
       return rows if cols == 0
 
-      # 'infinite' edit distance for padding cost matrix.
-      # Can be any value > max[rows, cols]
-      inf = rows + cols
-
-      # Initialize first two rows of cost matrix.
-      # The full initial state where cols=3, rows=2 (inf=5) would be:
-      #   [[5, 5, 5, 5, 5],
-      #    [5, 0, 1, 2, 3],
-      #    [5, 1, 0, 0, 0],
-      #    [5, 2, 0, 0, 0]]
-      matrix = [Array.new(cols + 2, inf)]
-      matrix << 0.upto(cols).to_a.unshift(inf)
+      # 'infinite' edit distance to pad cost matrix.
+      # Any value > max[rows, cols]
+      inf = cols + 1
 
       # element => last row seen
-      item_history = Hash.new(0)
+      row_history = Hash.new(0)
 
-      1.upto(rows) do |row|
-        # generate next row of cost matrix
-        new_row = Array.new(cols + 2, 0)
-        new_row[0] = inf
-        new_row[1] = row
-        matrix << new_row
+      # initialize alphabet-keyed cost matrix
+      matrix = {}
+      curr_row = 0.upto(cols).to_a
 
-        last_match_col = 0
-        seq1_item = seq1[row - 1]
+      rows.times do |row|
+        seq1_item = seq1[row]
+        match_col = 0
 
-        1.upto(cols) do |col|
-          seq2_item = seq2[col - 1]
-          last_match_row = item_history[seq2_item]
+        # rotate row arrays & generate next
+        matrix[seq1_item] = last_row = curr_row
+        curr_row = Array.new(cols + 1, inf)
+        curr_row[0] = row + 1
 
+        cols.times do |col|
+          seq2_item = seq2[col]
           sub_cost = seq1_item == seq2_item ? 0 : 1
 
-          transposition = 1 + matrix[last_match_row][last_match_col]
-          transposition += row - last_match_row - 1
-          transposition += col - last_match_col - 1
-
-          # TODO: do insertion/deletion need to be considered when
-          # seq1_item == seq2_item ?
-          #
-          # substitution, deletion, insertion, transposition
+          # | Xs | Xd |
+          # | Xi | ?  |
+          # substitution, deletion, insertion
           cost = [
-            matrix[row][col] + sub_cost,
-            matrix[row][col + 1] + 1,
-            matrix[row + 1][col] + 1,
-            transposition
+            last_row[col] + sub_cost,
+            last_row[col + 1] + 1,
+            curr_row[col] + 1
           ].min
 
-          matrix[row + 1][col + 1] = cost
-
-          last_match_col = col if sub_cost == 0
+          # transposition cost
+          # skip missed matrix lookup (inf cost)
+          if sub_cost > 0 && row > 0 && (m = matrix[seq2_item])
+            transpose = 1 + m[match_col] \
+              + (row - row_history[seq2_item] - 1) \
+              + (col - match_col - 1)
+            cost = transpose if transpose < cost
+          end
+
+          match_col = col if sub_cost == 0
+          curr_row[col + 1] = cost
         end
 
-        item_history[seq1_item] = row
+        row_history[seq1_item] = row
       end
 
-      matrix[rows + 1][cols + 1]
+      curr_row[cols]
     end
   end
 end

data/lib/edits/version.rb

@@ -2,5 +2,5 @@
 
 module Edits
   # Current gem version
-  VERSION = "0.3.0"
+  VERSION = "0.4.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: edits
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.0
 platform: ruby
 authors:
 - Tom Crouch
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2020-05-21 00:00:00.000000000 Z
+date: 2020-10-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: benchmark-ips
@@ -174,8 +174,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.7.7
+rubygems_version: 3.0.8
 signing_key: 
 specification_version: 4
 summary: A collection of edit distance algorithms.