scorer 1.0.0

data/lib/scorer.rb

@@ -0,0 +1,82 @@
+module Scorer
+  def score(abbreviation, fuzziness = nil)
+    # If the string is equal to the abbreviation, perfect match.
+    return 1 if self == abbreviation
+
+    string = self
+    string_length = string.length.to_f
+    total_character_score = 0
+    should_award_common_prefix_bonus = false
+    fuzzies = 1
+
+    # Walk through abbreviation and add up scores
+    abbreviation.chars.each_with_index do |c, i|
+      # Find the index of current character (case-insensitive) in remaining part
+      # of string.
+      indexes = [string.index(c.downcase), string.index(c.upcase)]
+      indexes = indexes - [nil]
+      index_in_string = indexes.min
+
+      if index_in_string
+        # Set base score for current character.
+        character_score = 0.1
+      else
+        return 0 unless fuzziness
+        fuzzies += 1 - fuzziness;
+        next
+      end
+
+      # Same case bonus.
+      if string[index_in_string].chr == c
+        character_score += 0.1
+      end
+
+      # Consecutive letter & start-of-string bonus
+      if index_in_string == 0
+        # Increase the score when matching first character of the
+        # remainder of the string
+        character_score += 0.6;
+
+        # If match is the first character of the string
+        # & the first character of abbreviation, add a
+        # start-of-string match bonus.
+        should_award_common_prefix_bonus = true if i == 0
+      end
+
+      # Acronym Bonus
+      # Weighing Logic: Typing the first character of an acronym is as if you
+      # preceded it with two perfect character matches.
+      if string[index_in_string - 1].chr == ' '
+        character_score += 0.8
+      end
+
+      # Left trim the matched part of the string
+      # (forces sequential matching).
+      string = string.slice(index_in_string + 1, string_length)
+
+      # Add to total character score.
+      total_character_score += character_score
+    end
+
+    abbreviation_score = total_character_score / abbreviation.length
+
+    # Reduce penalty for longer strings
+    percentage_of_matched_string = abbreviation.length / string_length
+    word_score = abbreviation_score * percentage_of_matched_string
+    final_score = (word_score + abbreviation_score) / 2.0
+
+    # Penalize any fuzzies
+    final_score = final_score / fuzzies;
+
+    # Award common prefix bonus if possible
+    if should_award_common_prefix_bonus && (final_score + 0.15 < 1)
+      final_score += 0.15
+    end
+
+    return final_score
+  end
+end
+
+class String
+  include Scorer
+end

data/lib/scorer/version.rb

@@ -0,0 +1,3 @@
+module Scorer
+  Version = VERSION = '1.0.0'
+end

data/spec/scorer_spec.rb

@@ -0,0 +1,81 @@
+require 'spec_helper'
+
+describe Scorer do
+  describe "score" do
+    it "returns a perfect score for identical strings" do
+      "Hello World".score("Hello World").should == 1
+    end
+
+    it "returns no score for abbreviations with inexistant characters" do
+      "hello world".score("hellx").should == 0
+      "hello world".score("hello_world").should == 0
+    end
+
+    it "returns no score for a non-sequential match" do
+      "Hello World".score("WH").should == 0
+    end
+
+    it "returns higher score for same case than for wrong case" do
+      "Hello World".score("hello").should be < "Hello World".score("Hello")
+    end
+
+    it "returns higher score for closer matches" do
+      "Hello World".score("H").should be < "Hello World".score("He")
+    end
+
+    it "returns a score even with wrong casing" do
+      "Hillsdale Michigan".score("himi").should be > 0
+    end
+
+    it "returns a higher score for closer matches" do
+      "hello world".score("e").should be          < "hello world".score("h")
+      "hello world".score("h").should be          < "hello world".score("he")
+      "hello world".score("hel").should be        < "hello world".score("hell")
+      "hello world".score("hell").should be       < "hello world".score("hello")
+      "hello world".score("hello").should be      < "hello world".score("helloworld")
+      "hello world".score("helloworl").should be  < "hello world".score("hello worl")
+      "hello world".score("hello worl").should be < "hello world".score("hello world")
+    end
+
+    it "gives a bonus for consecutive matches" do
+      "Hello World".score("Hel").should be  > "Hello World".score("Hld")
+    end
+
+    it "gives a bonus for acronyms" do
+      "Hello World".score("HW").should be > "Hello World".score("Ho")
+      "yet another Hello World".score("yaHW").should be > "Hello World".score("yet another")
+      "Hillsdale Michigan".score("HiMi").should be > "Hillsdale Michigan".score("Hil")
+      "Hillsdale Michigan".score("HiMi").should be > "Hillsdale Michigan".score("illsda")
+      "Hillsdale Michigan".score("HiMi").should be > "Hillsdale Michigan".score("hills")
+      "Hillsdale Michigan".score("HiMi").should be < "Hillsdale Michigan".score("hillsd")
+    end
+
+    it "gives a bonus for matching beginning of string" do
+      "Hillsdale".score("hi").should be > "Chippewa".score("hi")
+      "hello world".score("h").should be > "hello world".score("w")
+    end
+
+    it "weighs strings properly" do
+      "Research Resources North".score("res").should be > "Mary Conces".score('res')
+      "Research Resources North".score("res").should be > "Bonnie Strathern - Southwest Michigan Title Search".score('res')
+    end
+
+    describe "fuzzy scoring" do
+      it "scores fuzzy lower than non-fuzzy" do
+        "Hello World".score("Hz", 0.5).should be < "Hello World".score("H", 0.5)
+      end
+
+      it "scores mismatch lower" do
+        "hello world".score("hello worl", 0.5).should be > "hello world".score("hello wor1", 0.5)
+      end
+
+      it "returns fuzzy scores" do
+        "Hello World".score("jello", 0.5).should be > 0
+      end
+
+      it "returns higher scores for higher fuzziness" do
+        "Hello World".score("Hz", 0.9).should be > "Hello World".score("Hz", 0.5)
+      end
+    end
+  end
+end

data/spec/spec_helper.rb

@@ -0,0 +1 @@
+require 'scorer'

data/test/benchmark.rb

@@ -0,0 +1,26 @@
+require 'benchmark'
+require 'scorer'
+
+iterations = 4000
+puts "Benchmarking #{iterations} iterations..."
+Benchmark.bmbm do |x|
+  x.report("one letter") do
+    iterations.times { "hello world".score("h") }
+  end
+  x.report("two letters") do
+    iterations.times { "hello world".score("hw") }
+  end
+  x.report("full match") do
+    iterations.times { "hello world".score("hello world") }
+  end
+  x.report("13 char / 32 char") do
+    iterations.times do
+      "hello any world that will listen".score("hlo wrdthlstn")
+    end
+  end
+  x.report("70 char / 446 char") do
+    iterations.times do
+      "Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.".score("Lorem i dor coecadipg et, Duis aute irure dole nulla. qui ofa mot am l")
+    end
+  end
+end

data/test/benchmark.rbc

@@ -0,0 +1,816 @@
+!RBIX
+10937318184790222022
+x
+M
+1
+n
+n
+x
+10
+__script__
+i
+51
+5
+7
+0
+64
+47
+49
+1
+1
+15
+5
+7
+2
+64
+47
+49
+1
+1
+15
+7
+3
+19
+0
+15
+5
+7
+4
+20
+0
+47
+101
+5
+7
+6
+63
+3
+47
+49
+7
+1
+15
+45
+8
+9
+56
+10
+50
+11
+0
+15
+2
+11
+I
+5
+I
+1
+I
+0
+I
+0
+n
+p
+12
+s
+9
+benchmark
+x
+7
+require
+s
+5
+score
+I
+fa0
+s
+13
+Benchmarking 
+x
+4
+to_s
+s
+14
+ iterations...
+x
+4
+puts
+x
+9
+Benchmark
+n
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+59
+57
+19
+0
+15
+20
+0
+7
+0
+64
+56
+1
+50
+2
+1
+15
+20
+0
+7
+3
+64
+56
+4
+50
+2
+1
+15
+20
+0
+7
+5
+64
+56
+6
+50
+2
+1
+15
+20
+0
+7
+7
+64
+56
+8
+50
+2
+1
+15
+20
+0
+7
+9
+64
+56
+10
+50
+2
+1
+11
+I
+5
+I
+1
+I
+1
+I
+1
+n
+p
+11
+s
+10
+one letter
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+9
+21
+2
+0
+56
+0
+50
+1
+0
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+2
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+10
+7
+0
+64
+7
+1
+64
+49
+2
+1
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+3
+s
+11
+hello world
+s
+1
+h
+x
+5
+score
+p
+3
+I
+0
+I
+8
+I
+a
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+x
+5
+times
+p
+3
+I
+0
+I
+8
+I
+9
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+x
+6
+report
+s
+11
+two letters
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+9
+21
+2
+0
+56
+0
+50
+1
+0
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+2
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+10
+7
+0
+64
+7
+1
+64
+49
+2
+1
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+3
+s
+11
+hello world
+s
+2
+hw
+x
+5
+score
+p
+3
+I
+0
+I
+b
+I
+a
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+x
+5
+times
+p
+3
+I
+0
+I
+b
+I
+9
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+s
+10
+full match
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+9
+21
+2
+0
+56
+0
+50
+1
+0
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+2
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+10
+7
+0
+64
+7
+0
+64
+49
+1
+1
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+2
+s
+11
+hello world
+x
+5
+score
+p
+3
+I
+0
+I
+e
+I
+a
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+x
+5
+times
+p
+3
+I
+0
+I
+e
+I
+9
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+s
+17
+13 char / 32 char
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+9
+21
+2
+0
+56
+0
+50
+1
+0
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+2
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+10
+7
+0
+64
+7
+1
+64
+49
+2
+1
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+3
+s
+32
+hello any world that will listen
+s
+13
+hlo wrdthlstn
+x
+5
+score
+p
+3
+I
+0
+I
+12
+I
+a
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+x
+5
+times
+p
+3
+I
+0
+I
+11
+I
+9
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+s
+18
+70 char / 446 char
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+9
+21
+2
+0
+56
+0
+50
+1
+0
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+2
+M
+1
+p
+2
+x
+9
+for_block
+t
+n
+x
+9
+__block__
+i
+10
+7
+0
+64
+7
+1
+64
+49
+2
+1
+11
+I
+3
+I
+0
+I
+0
+I
+0
+I
+-2
+p
+3
+s
+446
+Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.
+s
+70
+Lorem i dor coecadipg et, Duis aute irure dole nulla. qui ofa mot am l
+x
+5
+score
+p
+3
+I
+0
+I
+17
+I
+a
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+x
+5
+times
+p
+3
+I
+0
+I
+16
+I
+9
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+0
+p
+13
+I
+0
+I
+6
+I
+4
+I
+7
+I
+f
+I
+a
+I
+1a
+I
+d
+I
+25
+I
+10
+I
+30
+I
+15
+I
+3b
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+1
+x
+1
+x
+x
+4
+bmbm
+p
+11
+I
+0
+I
+1
+I
+9
+I
+2
+I
+12
+I
+4
+I
+17
+I
+5
+I
+28
+I
+6
+I
+33
+x
+39
+/Users/matt/src/score/test/benchmark.rb
+p
+1
+x
+10
+iterations