mini-levenshtein 0.1.1 → 0.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 72484523445a8c05ff3be19bf9043bec3e7e5a1ddcd98cafa6f7eea96517e5c3
-  data.tar.gz: b52cdebc6caa82a3dba2152a7fef78e5509cb4b6a288bd1db7830d9f9192ced3
+  metadata.gz: 87d8dfd831b540e20d080df316abaf2195f371822d68e819e5c717fd2cbb637d
+  data.tar.gz: acd26d48b08feeaecea2ff7b9e4afe7f751e6590dafd54c87971fec244d79e28
 SHA512:
-  metadata.gz: 27479145fb061c23d178b42215121fdbe1b05a701fdd8c832b7df3ac28af01a6973201bed2b8d5350549e1e768642c0ee6c45d5cf4301c6aa89db3c4c94667b1
-  data.tar.gz: e2c95da36b1ec84593fe297360c9b8a029c1744d3c0a64b98eb19962e4b27b67113b04e525cbe60ad7fdfc3597f823ca81490d60af636baabc9716849789a1ac
+  metadata.gz: 4a380bcd7f95a4bc0ae68a871a40c0ec18beab58ef5af6bcc3313de38bd836f1e0e40538e491c9c0be46ba7882f25607b747cce8b579654d58cbac55303e67f1
+  data.tar.gz: f46268d96b7dfb651bb42db3d2f509bb5e2bdb908375832ee0a12a21ce388461cf52434156e042c8ba201c24041451dc03dd74d17b43cf3d2efec913089472e1

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    mini-levenshtein (0.1.0)
+    mini-levenshtein (0.1.2)
 
 GEM
   remote: https://rubygems.org/

data/README.md

@@ -5,9 +5,9 @@ Simple, fast, levenshtein distance
 # Synopsis
 
 ```ruby
-MiniLevenshtein.distance("levenshtein", "levenstien") # => 3
+MiniLevenshtein.edit_distance("levenshtein", "levenstien") # => 3
 
-MiniLevenshtein.ratio("levenshtein", "levenstien") # 0.857142...
+MiniLevenshtein.similarity("levenshtein", "levenstien") # 0.857142...
 ```
 
 # Credits

data/ext/mini_levenshtein/extconf.rb

@@ -7,4 +7,5 @@ require 'mkmf'
 dir_config('mini-levenshtein/mini_levenshtein')
 
 # Do the work
+create_header
 create_makefile('mini-levenshtein/mini_levenshtein')

data/ext/mini_levenshtein/mini_levenshtein.c

@@ -1,23 +1,20 @@
-#include <string.h>
-#include <math.h>
-#include <wchar.h>
-#include <stdlib.h>
+#include <ruby.h>
+#include "extconf.h"
 
-#include "ruby.h"
 #include "levenshtein.h"
 
-VALUE MiniLevenshteinInternal = Qnil;
+VALUE MiniLevenshtein = Qnil;
 
-VALUE method_internal_distance(VALUE self, VALUE s1, VALUE s2, VALUE xcost);
+VALUE rb_lev_edit_distance(VALUE self, VALUE s1, VALUE s2, VALUE xcost);
+VALUE rb_edit_distance(VALUE self, VALUE s1, VALUE s2);
+VALUE rb_similarity(VALUE self, VALUE s1, VALUE s2);
 
-void Init_mini_levenshtein()
-{
-  MiniLevenshteinInternal = rb_define_module("MiniLevenshteinInternal");
-  rb_define_method(MiniLevenshteinInternal, "internal_distance", method_internal_distance, 3);
-}
-
-VALUE method_internal_distance(VALUE self, VALUE s1, VALUE s2, VALUE xcost)
+VALUE rb_lev_edit_distance(VALUE self, VALUE s1, VALUE s2, VALUE xcost)
 {
+  Check_Type(s1, T_STRING);
+  Check_Type(s2, T_STRING);
+  Check_Type(xcost, T_FIXNUM);
+  
   size_t len1 = RSTRING_LEN(s1);
   size_t len2 = RSTRING_LEN(s2);
 
@@ -30,3 +27,83 @@ VALUE method_internal_distance(VALUE self, VALUE s1, VALUE s2, VALUE xcost)
 
   return INT2NUM(distance);
 }
+
+/*
+ * Compute absolute Levenshtein distance of two strings.
+ *
+ * @example it's hard to spell Levenshtein correctly
+ *   edit_distance('Levenshtein', 'Lenvinsten')  # => 4
+ *   edit_distance('Levenshtein', 'Levensthein') # => 2
+ *   edit_distance('Levenshtein', 'Levenshten')  # => 1
+ *   edit_distance('Levenshtein', 'Levenshtein') # => 0
+ *   "Yeah, we've managed it at last."
+ *
+ * @param string1 [String]
+ * @param string2 [String]
+ *
+ * @return [Integer]
+ */
+VALUE rb_edit_distance(VALUE self, VALUE string1, VALUE string2)
+{
+  Check_Type(string1, T_STRING);
+  Check_Type(string2, T_STRING);
+  
+  size_t len1 = RSTRING_LEN(string1);
+  size_t len2 = RSTRING_LEN(string2);
+
+  const lev_byte *str1 = StringValuePtr(string1);
+  const lev_byte *str2 = StringValuePtr(string2);
+
+  long distance = lev_edit_distance(len1, str1, len2, str2, 0);
+
+  return INT2NUM(distance);
+}
+
+/*
+ * Compute similarity of two strings.
+ *
+ * The similarity is a number between 0 and 1.
+ *
+ * @example
+ *   similarity('Hello World!', 'Holly grail!') # => 0.5833...
+ *
+ * @example
+ *   similarity('Holmes', 'Jack') # => 0.0
+ *
+ * @param string1 [String]
+ * @param string2 [String]
+ *
+ * @return [Float]
+ */
+VALUE rb_similarity(VALUE self, VALUE string1, VALUE string2)
+{
+  Check_Type(string1, T_STRING);
+  Check_Type(string2, T_STRING);
+  
+  size_t len1 = RSTRING_LEN(string1);
+  size_t len2 = RSTRING_LEN(string2);
+
+  const lev_byte *str1 = StringValuePtr(string1);
+  const lev_byte *str2 = StringValuePtr(string2);
+
+  size_t lensum = len1 + len2;
+  if (lensum == 0) {
+    return DBL2NUM(1.0);
+  }
+  
+  long distance = lev_edit_distance(len1, str1, len2, str2, 0);
+  if (distance == 0) {
+    return DBL2NUM(1.0);
+  }
+
+  return DBL2NUM((lensum - distance) / (double)lensum);
+}
+
+void Init_mini_levenshtein()
+{
+  MiniLevenshtein = rb_define_module("MiniLevenshtein");
+  rb_define_singleton_method(MiniLevenshtein, "lev_edit_distance", rb_lev_edit_distance, 3);
+  rb_define_singleton_method(MiniLevenshtein, "edit_distance", rb_edit_distance, 2);
+  rb_define_singleton_method(MiniLevenshtein, "similarity", rb_similarity, 2);
+}
+

data/lib/mini-levenshtein/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module MiniLevenshtein
-  VERSION = '0.1.1'
+  VERSION = '0.1.2'
 end

data/lib/mini-levenshtein.rb

@@ -1,62 +1,8 @@
 require 'mini-levenshtein/mini_levenshtein'
 
 module MiniLevenshtein
-  extend MiniLevenshteinInternal
-
   class << self
-    # Compute absolute Levenshtein distance of two strings.
-    #
-    # @example (it's hard to spell Levenshtein correctly):
-    #   distance('Levenshtein', 'Lenvinsten')  # => 4
-    #   distance('Levenshtein', 'Levensthein') # => 2
-    #   distance('Levenshtein', 'Levenshten')  # => 1
-    #   distance('Levenshtein', 'Levenshtein') # => 0
-    #   "Yeah, we've managed it at last."
-    #
-    # @param string1 [String]
-    # @param string2 [String]
-    #
-    # @return [Integer]
-    def edit_distance(string1, string2)
-      validate_string!(string1)
-      validate_string!(string2)
-      
-      internal_distance(string1, string2, 0)
-    end
     alias distance edit_distance
-
-    # Compute similarity of two strings.
-    #
-    # The similarity is a number between 0 and 1.
-    #
-    # @example
-    #   similarity('Hello World!', 'Holly grail!') # => 0.5833...
-    #
-    # @example
-    #   similarity('Holmes', 'Jack') # => 0.0
-    #
-    # @param string1 [String]
-    # @param string2 [String]
-    #
-    # @return [Float]
-    def similarity(string1, string2)
-      validate_string!(string1)
-      validate_string!(string2)
-
-      lensum = string1.length + string2.length
-      return 1.0 if lensum.zero?
-  
-      distance = internal_distance(string1, string2, 1)
-      return 1.0 if distance.zero?
-  
-      (lensum - distance) / lensum.to_f
-    end
     alias ratio similarity
-
-    private
-
-    def validate_string!(string)
-      raise TypeError, "no implicit conversion of #{string.class} to String" unless string.is_a?(String)
-    end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: mini-levenshtein
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors:
 - Delon Newman
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2022-05-04 00:00:00.000000000 Z
+date: 2022-05-05 00:00:00.000000000 Z
 dependencies: []
 description: Simple, fast, levenshtein distance
 email: contact@delonnewman.name