jaro_winkler 1.2.2 → 1.2.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 7a1229af20c91c12d0f4aeb8fb13dd57cc617344
-  data.tar.gz: 919b78c919c60d15bdeacdfc84bd6e13952e6181
+  metadata.gz: 036b901c1de03b193a10e71951a6c241c4fa31ef
+  data.tar.gz: 3cbf7ec98a3d56e7f522345061f4395e18f1a7ca
 SHA512:
-  metadata.gz: 56011763a68a13b9b04973a33b3575c768f81c6257d42284e99bcdf94c66896493cf68fad4e33338081b251fb5d1d93f0f6815e82eee7a8e33b0e97d5707826a
-  data.tar.gz: 79e38669e7d5d4a7582af83a76bee06b49e3b34304708ac5a0d5c8be661dbfc4196a81157b636d8ed4266645af4fcc671b6afcde506588fa3af56d8178961b92
+  metadata.gz: 1fa3be807f2cac992dc419df83267ab73d8924d32b1f341cab9d6adfcf31211fd5a89b5b2badb6fe36c78617ccf1524d4f6d9fa785ba50c3ab46481d0033a14f
+  data.tar.gz: 54682556479a596049b1f927dd546cfccefb61c0db6ce0f9ae4fec1790780839910ffd7985ec8dccb5f8ed5ce8c1a36498202841bf5d1d7485b1bdc985dd8cdb

data/README.md

@@ -1,6 +1,6 @@
 # About
 
-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.
+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. Both implementation supports UTF-8 string.
 
 **Windows Issue**
 
@@ -34,7 +34,7 @@ JaroWinkler.r_distance "MARTHA", "MARHTA" # Pure Ruby
 
 Name        | Type    | Default | Note
 ----------- | ------  | ------- | ------------------------------------------------------------------------------------------------------------
-case_match  | boolean | false   | All upper case characters are converted to lower case prior to the comparison.
+case_match  | boolean | false   | All lower case characters are converted to upper 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 the threshold.
 

data/Rakefile

@@ -5,3 +5,15 @@ require "rake/extensiontask"
 Rake::ExtensionTask.new("jaro_winkler") do |ext|
   ext.lib_dir = "lib/jaro_winkler"
 end
+
+task :benchmark do
+  ROOT_PATH = File.expand_path('..', __FILE__)
+  LIB_PATH = File.join(ROOT_PATH, 'lib')
+  BENCHMARK_PATH = File.join(ROOT_PATH, 'benchmark')
+  Dir[File.join(BENCHMARK_PATH, '*.rb')].each do |path|
+    output_path = File.join(BENCHMARK_PATH, File.basename(path, '*.rb').sub('.rb', '.txt'))
+    cmd = "RUBYLIB=#{LIB_PATH} ruby #{path} > #{output_path}"
+    puts cmd
+    system(cmd)
+  end
+end

data/benchmark/native.rb

@@ -24,9 +24,3 @@ Benchmark.bmbm do |x|
     n.times{ ary.each{ |str1, str2| Amatch::Jaro.new(str1).match(str2) } }
   end
 end
-
-#                        user     system      total        real
-# jaro_winkler       0.420000   0.000000   0.420000 (  0.426742)
-# fuzzystringmatch   0.160000   0.000000   0.160000 (  0.160146)
-# hotwater           0.300000   0.000000   0.300000 (  0.297350)
-# amatch             0.980000   0.010000   0.990000 (  0.982874)

data/benchmark/native.txt

@@ -0,0 +1,5 @@
+Rehearsal ----------------------------------------------------
+jaro_winkler       0.380000   0.000000   0.380000 (  0.376303)
+fuzzystringmatch   0.350000   0.020000   0.370000 (  0.369344)
+hotwater           0.280000   0.000000   0.280000 (  0.281579)
+amatch           

data/benchmark/pure.rb

@@ -14,7 +14,3 @@ Benchmark.bmbm do |x|
     n.times{ ary.each{ |str1, str2| jarow.getDistance(str1, str2) } }
   end
 end
-
-#                        user     system      total        real
-# jaro_winkler      12.480000   0.010000  12.490000 ( 12.497828)
-# fuzzystringmatch  14.990000   0.010000  15.000000 ( 15.014898)

data/benchmark/pure.txt

@@ -0,0 +1,8 @@
+Rehearsal ----------------------------------------------------
+jaro_winkler      12.520000   0.020000  12.540000 ( 12.548948)
+fuzzystringmatch  15.370000   0.020000  15.390000 ( 15.408540)
+------------------------------------------ total: 27.930000sec
+
+                       user     system      total        real
+jaro_winkler      12.750000   0.030000  12.780000 ( 12.782842)
+fuzzystringmatch  16.240000   0.030000  16.270000 ( 16.287380)

data/ext/jaro_winkler/distance.c

@@ -21,11 +21,10 @@ static int char_bytes_num(char first_char){
   else return 1;
 }
 
-static unsigned long* codepoints(const char *str, int *ret_len){
-  int str_len = strlen(str);
-  unsigned long *ret = calloc(str_len, sizeof(long));
+static unsigned long* codepoints(const char *str, int byte_len, int *ret_len){
+  unsigned long *ret = calloc(byte_len, sizeof(long));
   int count = 0;
-  for(int i = 0; i < str_len;){
+  for(int i = 0; i < byte_len;){
     int bytes_num = char_bytes_num(str[i]);
     memcpy(&ret[count], &str[i], bytes_num);
     count++;
@@ -35,13 +34,13 @@ static unsigned long* codepoints(const char *str, int *ret_len){
   return ret;
 }
 
-double c_distance(char *s1, char *s2, Option *opt){
+double c_distance(char *s1, int byte_len1, char *s2, int byte_len2, Option *opt){
   // set default option if NULL passed
   int free_opt_flag = 0;
   if(!opt){ free_opt_flag = 1; opt = option_new(); }
 
   int ary_1_len, ary_2_len;
-  unsigned long *ary_1 = codepoints(s1, &ary_1_len), *ary_2 = codepoints(s2, &ary_2_len);
+  unsigned long *ary_1 = codepoints(s1, byte_len1, &ary_1_len), *ary_2 = codepoints(s2, byte_len2, &ary_2_len);
 
   if(opt->case_match){
     for(int i = 0; i < ary_1_len; ++i) if(ary_1[i] < 256 && islower(ary_1[i])) ary_1[i] -= 32;

data/ext/jaro_winkler/distance.h

@@ -6,7 +6,7 @@ typedef struct{
   double weight, threshold;
 } Option;
 
-double c_distance(char *s1, char *s2, Option *opt);
+double c_distance(char *s1, int byte_len1, char *s2, int byte_len2, Option *opt);
 Option* option_new();
 
 #endif /* DISTANCE_H */

data/ext/jaro_winkler/jaro_winkler.c

@@ -21,8 +21,7 @@ VALUE rb_distance(int argc, VALUE *argv, VALUE self){
     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;
   }
-  // TODO: replace StringValueCStr with StringValuePtr and RSTRING_LEN
-  VALUE ret = rb_float_new(c_distance(StringValueCStr(s1), StringValueCStr(s2), c_opt));
+  VALUE ret = rb_float_new(c_distance(StringValuePtr(s1), RSTRING_LEN(s1), StringValuePtr(s2), RSTRING_LEN(s2), c_opt));
   free(c_opt);
   return ret;
 }

data/jaro_winkler.gemspec

@@ -10,8 +10,8 @@ Gem::Specification.new do |spec|
   spec.authors       = ["Jian Weihang"]
   spec.email         = ["tonytonyjan@gmail.com"]
   spec.extensions    = ["ext/jaro_winkler/extconf.rb"] unless JaroWinkler.fallback?
-  spec.summary       = %q{Pure Ruby implementation of Jaro-Winkler distance algorithm.}
-  spec.description   = %q{Pure Ruby implementation of Jaro-Winkler distance algorithm.}
+  spec.summary       = %q{Ruby & C implementation of Jaro-Winkler distance algorithm which both support UTF-8 string.}
+  spec.description   = %q{It's a implementation of Jaro-Winkler distance algorithm, it uses C extension and will fallback to pure Ruby version in JRuby. Both implementation supports UTF-8 string.}
   spec.homepage      = "https://github.com/tonytonyjan/jaro_winkler"
   spec.license       = "MIT"
 

data/lib/jaro_winkler.rb

@@ -47,7 +47,7 @@ module JaroWinkler
     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
+    s1, s2     = s1.upcase, s2.upcase if case_match
     distance   = jaro_distance(s1, s2)
     prefix     = 0
     max_length = [4, s1.length, s2.length].min

data/lib/jaro_winkler/version.rb

@@ -1,3 +1,3 @@
 module JaroWinkler
-  VERSION = "1.2.2"
+  VERSION = "1.2.3"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: jaro_winkler
 version: !ruby/object:Gem::Version
-  version: 1.2.2
+  version: 1.2.3
 platform: ruby
 authors:
 - Jian Weihang
@@ -52,7 +52,9 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-description: Pure Ruby implementation of Jaro-Winkler distance algorithm.
+description: It's a implementation of Jaro-Winkler distance algorithm, it uses C extension
+  and will fallback to pure Ruby version in JRuby. Both implementation supports UTF-8
+  string.
 email:
 - tonytonyjan@gmail.com
 executables: []
@@ -67,7 +69,9 @@ files:
 - README.md
 - Rakefile
 - benchmark/native.rb
+- benchmark/native.txt
 - benchmark/pure.rb
+- benchmark/pure.txt
 - ext/jaro_winkler/distance.c
 - ext/jaro_winkler/distance.h
 - ext/jaro_winkler/extconf.rb
@@ -103,7 +107,8 @@ rubyforge_project:
 rubygems_version: 2.4.1
 signing_key: 
 specification_version: 4
-summary: Pure Ruby implementation of Jaro-Winkler distance algorithm.
+summary: Ruby & C implementation of Jaro-Winkler distance algorithm which both support
+  UTF-8 string.
 test_files:
 - spec/jaro_winkler_spec.rb
 - spec/spec_helper.rb