string_metric 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 32870ce170c62b4036f6b8eaf3f92e43280a0246
-  data.tar.gz: a7f38c88be3211f6ef1d82c71d701a0655081570
+  metadata.gz: 229ed953c5aadc5c929cb2b47ad7898cdc19d742
+  data.tar.gz: b41a66496dd7c0cb2627af22b9764d2bf8d06c37
 SHA512:
-  metadata.gz: 8668f405e8568c4254ef7856660a87bd4488d8e4ddbcea95f185575562787ad072696868a831431b821d95f868a8a1f67e1f2b057272375dbda3f0fbe73a6ccb
-  data.tar.gz: 9e9b25912b3f695161262fb331cc8f043fe8b269b29f1880ee9d8bce8dbf367c339de18b9c3756d15d9ee0e5ba677e1ff5b2837909869c0b1984797b75638f5b
+  metadata.gz: c0e932fbe3e03ec7b8268a2cc85f276b7b0041cdaf8fea5fb98d7e8a966ded52280fbe51b01a89da4e26430020a4995cc3b043110f2a95731e70c07ad3a30748
+  data.tar.gz: 19f270701a5e4383b84ab089ea534e7930fb980ff7dba494c51201436c69408cf54b0cca7d64098aefdd24214cc6e250168a1dc975abbf49e3672ba262b7eead

data/README.md

@@ -3,12 +3,14 @@
 [![Build Status](https://travis-ci.org/skroutz/string_metric.png?branch=master)](https://travis-ci.org/skroutz/string_metric)
 [![Code Climate](https://codeclimate.com/github/skroutz/string_metric.png)](https://codeclimate.com/github/skroutz/string_metric)
 [![Coverage Status](https://coveralls.io/repos/skroutz/string_metric/badge.png?branch=master)](https://coveralls.io/r/skroutz/string_metric?branch=master)
+[![Gem Version](https://badge.fury.io/rb/string_metric.png)](http://badge.fury.io/rb/string_metric)
+[![Dependency Status](https://gemnasium.com/skroutz/string_metric.png)](https://gemnasium.com/skroutz/string_metric)
 
 A simple library with String Metric algorithms. If you want to read more about
 String Metric algorithms please read [here](https://en.wikipedia.org/wiki/String_metric).
 
 This library wants to support __MRI__ (1.9.3, 2.0.0, 2.1.0), __JRuby__ and
-__Rubinious__.
+__Rubinius__.
 
 ## Installation
 
@@ -42,13 +44,14 @@ __Options__
 * `:deletion_cost`: It overrides the default (equals to 1) deletion penanty.
   Can be `Fixnum` or `Float`.
 
-* `:subsctitution_cost`: It overrides the default (equals to 1) substitution
+* `:substitution_cost`: It overrides the default (equals to 1) substitution
   penalty. Can be `Fixum` or `Float`.
 
 * `:strategy`: The desired strategy for Levenshtein distance. Supported
-  strategies are `:recursive`, `:two_matrix_rows`, `:full_matrix` and
-  `:experiment`. The default strategy is `:two_matrix_rows`. One should not
-  depend on `:experiment` strategy.
+  strategies are `:recursive`, `:two_matrix_rows`, `:two_matrix_rows_v2`,
+  `:full_matrix` and `:experiment`. The default strategy is
+  `:two_matrix_rows_v2` for MRI and `:two_matrix_rows` for other platforms
+  One should not depend on `:experiment` strategy.
 
 __Examples__
 
@@ -96,15 +99,15 @@ or you can choose to benchmark a specific algorithm like:
 
 __Levenshtein__
 
+Implementation                                   | User      | Real
+-------------------------------------------------|-----------|-----------
+Levenshtein::IterativeWithFullMatrix             | 2.260000  | 2.265873
+Levenshtein::IterativeWithTwoMatrixRows          | 1.970000  | 1.971205
+Levenshtein::Experiment                          | 1.680000  | 1.684419
+Levenshtein::IterativeWithTwoMatrixRowsOptimized | 1.270000  | 1.269643
+Text::Levenshtein (from gem text)                | 2.180000  | 2.186258
 
-Implementation                              | User      | Real
---------------------------------------------|-----------|-----------
-Levenshtein::IterativeWithFullMatrix        | 0.480000  | 0.475662
-Levenshtein::IterativeWithTwoMatrixRows     | 0.350000  | 0.352388
-Levenshtein::Experiment                     | 0.420000  | 0.420000
-Text::Levenshtein (from gem text)           | 0.400000  | 0.400346
-
-_Currently the set of fixtures is very small_
+_Currently the set of fixtures is very small - ruby 2.1.0 is used_
 
 ## Other implementations
 
@@ -131,4 +134,4 @@ __Various__
 
 ## Licence
 
-string_metric is licensed under MIT. See [License](LICENSE.txt)
+string_metric is licensed under MIT. See [License](LICENSE.txt)

data/lib/string_metric/levenshtein/iterative_with_two_matrix_rows_optimized.rb

@@ -9,13 +9,16 @@ module StringMetric
         return from.size if to.size.zero?
 
         max_distance      = options[:max_distance]
-        insertion_cost    = options.fetch(:insertion_cost, 1)
-        deletion_cost     = options.fetch(:deletion_cost, 1)
-        substitution_cost = options.fetch(:substitution_cost, 1)
+        insertion_cost    = options[:insertion_cost]    || 1
+        deletion_cost     = options[:deletion_cost]     || 1
+        substitution_cost = options[:substitution_cost] || 1
 
         m = from.length
         n = to.length
 
+        from = from.codepoints.to_a
+        to = to.codepoints.to_a
+
         v0 = (0..m).to_a
         v1 = []
         x = 0
@@ -28,14 +31,12 @@ module StringMetric
           m.times do |j|
             cost = (from[j] == to[i]) ? 0 : substitution_cost
 
-            ins_cell = v0[j+1]
+            ins_cell = v0[j + 1]
 
             x = [x + deletion_cost,         # deletion
                  ins_cell + insertion_cost, # insertion
                  sub_cell + cost            # substitution
-                ].sort!
-
-            x = x[0]
+                ].sort![0]
 
             v1[j + 1] = x
 

data/lib/string_metric/levenshtein.rb

@@ -35,7 +35,8 @@ module StringMetric
     #   new substitution cost is taken into account (be default is 1)
     # @option options [Symbol] :strategy The desired strategy for Levenshtein
     #   distance. Supported strategies are :recursive, :two_matrix_rows,
-    #   :full_matrix and :experiment. The default strategy is :two_matrix_rows.
+    #   :full_matrix, :two_matrix_rows_v2 and :experiment. The default strategy
+    #   is :two_matrix_rows_v2 for MRI and :two_matrix_rows for other platforms.
     #   One should not depend on :experiment strategy.
     # @return [Fixnum, Float] the Levenshtein Distance
     def distance(from, to, options = {})

data/lib/string_metric/version.rb

@@ -1,3 +1,3 @@
 module StringMetric
-  VERSION = "0.1.0"
+  VERSION = "0.1.1"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: string_metric
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Giorgos Tsiftsis
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-01-30 00:00:00.000000000 Z
+date: 2014-02-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler