jaro_winkler 1.2.0 → 1.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 9a1a6510f807b126a00145ca02323746c1885409
-  data.tar.gz: 3cc743dde90a2dc0f682af51b506da4d4dc8e6ed
+  metadata.gz: 4a1b9c70518b0e53cf56495a67e2ffe90962b25f
+  data.tar.gz: 407eee20cb14e8e2b3fde69ad200547b34b27ebe
 SHA512:
-  metadata.gz: 02fec80fb44db8e6efed4b6cf083d55e4356cf706827c85850db8ef1c409a483e90c20e6de8437f0282d9621d46513a80061d1dd333d6b4b45b85d81f5c73bd5
-  data.tar.gz: e6e43be13fb520ddfa88206488689f4329c68003b0c4c1c40b69192c6df3d7bcc323f18316f21392ebf0da0dc2fce6949d43b8d8833ad72fe59ca288ebce3132
+  metadata.gz: 0ec1165eac3e38cbac6462d3920467b3bdf2e08957367f150b51553710dd147ffd3ec78774bd74408867a433bf47ca7e5a810750aaddc2bc92aa3cb7d8dc1d31
+  data.tar.gz: 918f8cc2603b09f6a23e8df6bbdd76d3ea295988f20a57000b9382d84f8e97f7d7205d3992d2b48fa1820e3b01bcb4de8473a0d9f1a39f4735efeac441c26d71

data/README.md

@@ -1,6 +1,6 @@
 # About
 
-It's a pure Ruby implementation of [Jaro-Winkler distance](http://en.wikipedia.org/wiki/Jaro%E2%80%93Winkler_distance) algorithm.
+It's a implementation of [Jaro-Winkler distance](http://en.wikipedia.org/wiki/Jaro%E2%80%93Winkler_distance) algorithm, it uses C extension and will fallback to pure Ruby version in JRuby.
 
 # Installation
 
@@ -19,9 +19,9 @@ JaroWinkler.distance "MARTHA", "marhta", case_match: true
 JaroWinkler.distance "MARTHA", "MARHTA", weight: 0.2
 # => 0.9778
 
-# Native
-JaroWinkler.c_distance "MARTHA", "MARHTA" # Recommended one, it's 7 times faster than the latter.
-JaroWinkler.distance "MARTHA", "MARHTA", native: true
+# Force the strategy
+JaroWinkler.c_distance "MARTHA", "MARHTA"
+JaroWinkler.r_distance "MARTHA", "MARHTA"
 ```
 
 **Both implementations support UTF-8 string.**
@@ -33,7 +33,6 @@ Name        | Type    | Default | Note
 case_match  | boolean | false   | All upper case characters are converted to lower case prior to the comparison.
 weight      | number  | 0.1     | A constant scaling factor for how much the score is adjusted upwards for having common prefixes.
 threshold   | number  | 0.7     | The prefix bonus is only added when the compared strings have a Jaro distance above a this.
-native      | boolean | false   | Use native version.
 
 # Why This?
 
@@ -85,5 +84,4 @@ end
 
 # Todo
 
-- Adjusting word table (Reference to original C implementation.)
-- Remove `#c_distance`, use C extension as default, and fallback to Ruby in Java platform.
+- Adjusting word table (Reference to original C implementation.)

data/benchmark/native.rb

@@ -5,21 +5,16 @@ ary = [['al', 'al'], ['martha', 'marhta'], ['jones', 'johnson'], ['abcvwxyz', 'c
 
 n = 100000
 Benchmark.bmbm do |x|
-  x.report '#c_distance(s1, s2)' do
+  x.report 'jaro_winkler' do
     n.times{ ary.each{ |str1, str2| JaroWinkler.c_distance(str1, str2) } }
   end
 
-  x.report '#distance(s1, s2, native: true)' do
-    n.times{ ary.each{ |str1, str2| JaroWinkler.distance(str1, str2, native: true) } }
-  end
-
   x.report 'fuzzystringmatch' do
     jarow = FuzzyStringMatch::JaroWinkler.create(:native)
     n.times{ ary.each{ |str1, str2| jarow.getDistance(str1, str2) } }
   end
 end
 
-#                                       user     system      total        real
-# #c_distance(s1, s2)               0.350000   0.000000   0.350000 (  0.349109)
-# #distance(s1, s2, native: true)   2.480000   0.050000   2.530000 (  2.526027)
-# fuzzystringmatch                  0.160000   0.000000   0.160000 (  0.155539)
+#                        user     system      total        real
+# jaro_winkler       0.380000   0.000000   0.380000 (  0.386071)
+# fuzzystringmatch   0.140000   0.000000   0.140000 (  0.138053)

data/benchmark/pure.rb

@@ -5,8 +5,8 @@ ary = [['al', 'al'], ['martha', 'marhta'], ['jones', 'johnson'], ['abcvwxyz', 'c
 
 n = 100000
 Benchmark.bmbm do |x|
-  x.report 'jaro_winkler    ' do
-    n.times{ ary.each{ |str1, str2| JaroWinkler.distance(str1, str2) } }
+  x.report 'jaro_winkler' do
+    n.times{ ary.each{ |str1, str2| JaroWinkler.r_distance(str1, str2) } }
   end
 
   x.report 'fuzzystringmatch' do

data/ext/jaro_winkler/jaro_winkler.c

@@ -17,6 +17,7 @@ VALUE rb_distance(int argc, VALUE *argv, VALUE self){
     VALUE threshold = rb_hash_aref(opt, ID2SYM(rb_intern("threshold")));
     VALUE case_match = rb_hash_aref(opt, ID2SYM(rb_intern("case_match")));
     if(!NIL_P(weight)) c_opt->weight = NUM2DBL(weight);
+    if(c_opt->weight > 0.25) rb_raise(rb_eRuntimeError, "Scaling factor should not exceed 0.25, otherwise the distance can become larger than 1.");
     if(!NIL_P(threshold)) c_opt->threshold = NUM2DBL(threshold);
     if(!NIL_P(case_match)) c_opt->case_match = (TYPE(case_match) == T_FALSE || NIL_P(case_match)) ? 0 : 1;
   }

data/lib/jaro_winkler.rb

@@ -42,9 +42,8 @@ module JaroWinkler
     matches == 0 ? 0 : (matches / length1 + matches / length2 + (matches - transpositions) / matches) / 3.0
   end
 
-  def distance s1, s2, options = {}
-    options = {weight: 0.1, threshold: 0.7, case_match: false, native: false}.merge options
-    return c_distance(s1, s2, options) if RUBY_PLATFORM != 'java' && options[:native]
+  def r_distance s1, s2, options = {}
+    options = {weight: 0.1, threshold: 0.7, case_match: false}.merge options
     weight, threshold, case_match = options[:weight], options[:threshold], options[:case_match]
     raise 'Scaling factor should not exceed 0.25, otherwise the distance can become larger than 1' if weight > 0.25
     s1, s2 = s1.downcase, s2.downcase if case_match
@@ -56,4 +55,13 @@ module JaroWinkler
     end
     distance < threshold ? distance : distance + ((prefix * weight) * (1 - distance))
   end
+
+  if RUBY_PLATFORM == 'java'
+    alias :distance :r_distance
+    alias :c_distance :r_distance
+    module_function :distance, :c_distance
+  else
+    alias :distance :c_distance
+    module_function :distance
+  end
 end

data/lib/jaro_winkler/version.rb

@@ -1,3 +1,3 @@
 module JaroWinkler
-  VERSION = "1.2.0"
+  VERSION = "1.2.1"
 end

data/spec/jaro_winkler_spec.rb

@@ -1,3 +1,4 @@
+# encoding: utf-8
 require 'jaro_winkler'
 include JaroWinkler
 describe JaroWinkler do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: jaro_winkler
 version: !ruby/object:Gem::Version
-  version: 1.2.0
+  version: 1.2.1
 platform: ruby
 authors:
 - Jian Weihang