Fuzzy Matching Library

fuzzy_match.rb

class FuzzyMatch
  class Matcher
    require 'singleton'
    require 'fuzzystringmatch'
    include Singleton
    JW = FuzzyStringMatch::JaroWinkler.create

    def self.get_score(str1,str2)
      JW.getDistance(str1,str2)
    end
  end

  class List
    require 'set'
    def initialize(options={})
      @options = {
        :filter     => //,
        :score      => 0.8,
        :downcase   => true,
        :ascii_only => true,
        :fix_spaces => true
      }.merge!(options)
      @data = SortedSet.new
    end

    def to_a
      @data.to_a
    end

    def <<(item)
      item.downcase!              if @options[:downcase]
      item.gsub!(/[^a-z0-9 ]/,'') if @options[:ascii_only]
      item.gsub!(/\s{2,}/, '')    if @options[:fix_spaces]
      item.gsub!(@options[:filter], '')
      @data << item
    end

    def search(string)
      @data.to_a.select do |candidate|
        Matcher.get_score(string,candidate) > @options[:score]
      end
    end
  end
end