amatch 0.3.1 → 0.4.0

data/lib/amatch/version.rb

@@ -1,6 +1,6 @@
 module Amatch
   # Amatch version
-  VERSION         = '0.3.1'
+  VERSION         = '0.4.0'
   VERSION_ARRAY   = VERSION.split('.').map(&:to_i) # :nodoc:
   VERSION_MAJOR   = VERSION_ARRAY[0] # :nodoc:
   VERSION_MINOR   = VERSION_ARRAY[1] # :nodoc:

data/tests/test_damerau_levenshtein.rb

@@ -0,0 +1,93 @@
+require 'test/unit'
+require 'amatch'
+
+class TestDamerauLevenshtein < Test::Unit::TestCase
+  include Amatch
+
+  def setup
+    @d        = 0.000001
+    @empty    = DamerauLevenshtein.new('')
+    @simple   = DamerauLevenshtein.new('test')
+    @long     = DamerauLevenshtein.new('A' * 160)
+  end
+
+  def test_version
+    assert_kind_of String, Amatch::VERSION
+  end
+
+  def test_match
+    assert_equal 4,     @simple.match('')
+    assert_equal 0,     @simple.match('test')
+    assert_equal 1,     @simple.match('testa')
+    assert_equal 1,     @simple.match('atest')
+    assert_equal 1,     @simple.match('teast')
+    assert_equal 1,     @simple.match('est')
+    assert_equal 1,     @simple.match('tes')
+    assert_equal 1,     @simple.match('tst')
+    assert_equal 1,     @simple.match('best')
+    assert_equal 1,     @simple.match('tost')
+    assert_equal 1,     @simple.match('tesa')
+    assert_equal 3,     @simple.match('taex')
+    assert_equal 6,     @simple.match('aaatestbbb')
+    assert_equal 1,     @simple.match('tset')
+  end
+
+  def test_search
+    assert_equal 4,     @simple.search('')
+    assert_equal 0,     @empty.search('')
+    assert_equal 0,     @empty.search('test')
+    assert_equal 0,     @simple.search('aaatestbbb')
+    assert_equal 3,     @simple.search('aaataexbbb')
+    assert_equal 4,     @simple.search('aaaaaaaaa')
+  end
+
+  def test_array_result
+    assert_equal [1, 0],    @simple.match(["tets", "test"])
+    assert_equal [1, 0],    @simple.search(["tetsaaa", "testaaa"])
+    assert_raises(TypeError) { @simple.match([:foo, "bar"]) }
+  end
+
+  def test_pattern_setting
+    assert_raises(TypeError) { @simple.pattern = :something }
+    assert_equal 0, @simple.match('test')
+    @simple.pattern = ''
+    assert_equal 4, @simple.match('test')
+    @simple.pattern = 'test'
+    assert_equal 0, @simple.match('test')
+  end
+
+  def test_similar
+    assert_in_delta 1, @empty.similar(''), @d
+    assert_in_delta 0, @empty.similar('not empty'), @d
+    assert_in_delta 0.0, @simple.similar(''), @d
+    assert_in_delta 1.0, @simple.similar('test'), @d
+    assert_in_delta 0.8, @simple.similar('testa'), @d
+    assert_in_delta 0.8, @simple.similar('atest'), @d
+    assert_in_delta 0.8, @simple.similar('teast'), @d
+    assert_in_delta 0.75, @simple.similar('est'), @d
+    assert_in_delta 0.75, @simple.similar('tes'), @d
+    assert_in_delta 0.75, @simple.similar('tst'), @d
+    assert_in_delta 0.75, @simple.similar('best'), @d
+    assert_in_delta 0.75, @simple.similar('tost'), @d
+    assert_in_delta 0.75, @simple.similar('tesa'), @d
+    assert_in_delta 0.25, @simple.similar('taex'), @d
+    assert_in_delta 0.4, @simple.similar('aaatestbbb'), @d
+    assert_in_delta 0.75, @simple.pattern.damerau_levenshtein_similar('est'), @d
+  end
+
+  def test_transpositions
+    assert_in_delta 1.0, 'atestatest'.damerau_levenshtein_similar('atestatest'), @d
+    assert_in_delta 0.9, 'atestatest'.damerau_levenshtein_similar('atetsatest'), @d
+    assert_in_delta 0.8, 'atestatest'.damerau_levenshtein_similar('atetsatset'), @d
+  end
+
+  def test_long
+    assert_in_delta 1.0, @long.similar(@long.pattern), @d
+  end
+
+  def test_long2
+    a = "lost this fantasy,  this fantasy,  this  fantasy,  this fantasy,  this fantasy,  this fantasy\r\n\r\nGood love  Neat work\r\n\r\nSuper job Fancy work\r\n\r\nPants job Cool work"
+    b = "lost\r\n\r\nGood love Neat work\r\n\r\nSuper  job Fancy work\r\n\r\nPants job Cool work"
+    assert a.damerau_levenshtein_similar(b)
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: amatch
 version: !ruby/object:Gem::Version
-  version: 0.3.1
+  version: 0.4.0
 platform: ruby
 authors:
 - Florian Frank
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-05-23 00:00:00.000000000 Z
+date: 2017-07-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: gem_hadar
@@ -52,17 +52,32 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.0'
+- !ruby/object:Gem::Dependency
+  name: mize
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: |
   Amatch is a library for approximate string matching and searching in strings.
   Several algorithms can be used to do this, and it's also possible to compute a
   similarity metric number between 0.0 and 1.0 for two given strings.
 email: flori@ping.de
 executables:
-- agrep.rb
+- agrep
+- dupfind
 extensions:
 - ext/extconf.rb
 extra_rdoc_files:
-- README.rdoc
+- README.md
 - lib/amatch.rb
 - lib/amatch/polite.rb
 - lib/amatch/rude.rb
@@ -75,22 +90,25 @@ files:
 - CHANGES
 - COPYING
 - Gemfile
-- README.rdoc
+- README.md
 - Rakefile
 - VERSION
 - amatch.gemspec
-- bin/agrep.rb
+- bin/agrep
+- bin/dupfind
 - ext/amatch_ext.c
 - ext/common.h
 - ext/extconf.rb
 - ext/pair.c
 - ext/pair.h
+- images/amatch_ext.png
 - install.rb
 - lib/amatch.rb
 - lib/amatch/.keep
 - lib/amatch/polite.rb
 - lib/amatch/rude.rb
 - lib/amatch/version.rb
+- tests/test_damerau_levenshtein.rb
 - tests/test_hamming.rb
 - tests/test_jaro.rb
 - tests/test_jaro_winkler.rb
@@ -101,14 +119,14 @@ files:
 - tests/test_sellers.rb
 homepage: http://github.com/flori/amatch
 licenses:
-- GPL
+- Apache-2.0
 metadata: {}
 post_install_message: 
 rdoc_options:
 - "--title"
 - Amatch - Approximate Matching
 - "--main"
-- README.rdoc
+- README.md
 require_paths:
 - lib
 - ext
@@ -129,6 +147,7 @@ signing_key:
 specification_version: 4
 summary: Approximate String Matching library
 test_files:
+- tests/test_damerau_levenshtein.rb
 - tests/test_hamming.rb
 - tests/test_jaro.rb
 - tests/test_jaro_winkler.rb

data/README.rdoc

@@ -1,128 +0,0 @@
-= amatch - Approximate Matching Extension for Ruby
-
-== Description
-
-This is a collection of classes that can be used for Approximate
-matching, searching, and comparing of Strings. They implement algorithms
-that compute the Levenshtein edit distance, Sellers edit distance, the
-Hamming distance, the longest common subsequence length, the longest common
-substring length, the pair distance metric, the Jaro-Winkler metric.
-
-== Download
-
-The latest version of <b>amatch</b> can be found at
-
-* http://rubyforge.org/frs/?group_id=390
-
-Online Documentation should be located at
-
-* http://amatch.rubyforge.org
-
-== Installation
-
-Just type into the command line as root:
-
- # ruby install.rb
-
-If you have installed rake (rake.rubyforge.org), you can also type:
-
- # rake install
-
-To install this extension as a gem type
-
- # gem install amatch
-
-== Examples
- require 'amatch'
- # => true
- include Amatch
- # => Object
- 
- m = Sellers.new("pattern")
- # => #<Amatch::Sellers:0x40366324>
- m.match("pattren")
- # => 2.0
- m.substitution = m.insertion = 3
- # => 3
- m.match("pattren")
- # => 4.0
- m.reset_weights
- # => #<Amatch::Sellers:0x40366324>
- m.match(["pattren","parent"])
- # => [2.0, 4.0]
- m.search("abcpattrendef")
- # => 2.0
- 
- m = Levenshtein.new("pattern")
- # => #<Amatch::Levenshtein:0x4035919c>
- m.match("pattren")
- # => 2
- m.search("abcpattrendef")
- # => 2
- "pattern language".levenshtein_similar("language of patterns")
- # => 0.2
- 
- m = Hamming.new("pattern")
- # => #<Amatch::Hamming:0x40350858>
- m.match("pattren")
- # => 2
- "pattern language".hamming_similar("language of patterns")
- # => 0.1
- 
- m = PairDistance.new("pattern")
- # => #<Amatch::PairDistance:0x40349be8>
- m.match("pattr en")
- # => 0.545454545454545
- m.match("pattr en", nil)
- # => 0.461538461538462
- m.match("pattr en", /t+/)
- # => 0.285714285714286
- "pattern language".pair_distance_similar("language of patterns")
- # => 0.928571428571429
- 
- m = LongestSubsequence.new("pattern")
- # => #<Amatch::LongestSubsequence:0x4033e900>
- m.match("pattren")
- # => 6
- "pattern language".longest_subsequence_similar("language of patterns")
- # => 0.4
- 
- m = LongestSubstring.new("pattern")
- # => #<Amatch::LongestSubstring:0x403378d0>
- m.match("pattren")
- # => 4
- "pattern language".longest_substring_similar("language of patterns")
- # => 0.4
-
- m = Jaro.new("pattern")
- # => #<Amatch::Jaro:0x363b70>
- m.match("paTTren")
- # => 0.952380952380952
- m.ignore_case = false
- m.match("paTTren")
- # => 0.742857142857143
- "pattern language".jaro_similar("language of patterns")
- # => 0.672222222222222
-
- m = JaroWinkler.new("pattern")
- # #<Amatch::JaroWinkler:0x3530b8>
- m.match("paTTren")
- # => 0.971428571712403
- m.ignore_case = false
- m.match("paTTren")
- # => 0.79428571505206
- m.scaling_factor = 0.05
- m.match("pattren")
- # => 0.961904762046678
- "pattern language".jarowinkler_similar("language of patterns")
- # => 0.672222222222222
-
-== Author
-
-Florian Frank mailto:flori@ping.de
-
-== License
-
-This is free software; you can redistribute it and/or modify it under
-the terms of the GNU General Public License Version 2 as published by
-the Free Software Foundation: http://www.gnu.org/copyleft/gpl.html