gistfile1.rb

#!/bin/env ruby
# encoding: utf-8

module Normalizers
  class SongDataNormalizer

#    def initialize(object)
#      @object = object
#      @tags = {}
#    end

#  end

#  class TitleNormalizer

    DASHES = '\p{Pd}'
    STOPPERS = '' + DASHES

    def initialize(text)
      @text = text
      @tags = {}
    end
    
    def scan!
      remove_quotes!
      token = /\s\p{Pd}\s/i

      if @text =~ token
        parts = @text.split(/\s[-–]\s/, 2)
        normalized = [parts[0], parts[1]]

        [0, 1].each do |index|
          extract_features(normalized[index])
          extract_remix(normalized[index])
          extract_producer(normalized[index])
        end

        @tags[:artist] = normalized[0]
        @tags[:title] = normalized[1]
      end

      @tags.inspect
    end

    def remove_quotes!
      @text.gsub! /["\A\Z]/m, ''
    end

    def extract_features(text)
      if match = text.match(/[\[\(]?\s*f[ea]{0,2}t\.?\s([^\[\]\)\(]+)[\[\]\)\(]?/i)
        @tags[:features] ||= match[1]
        text.gsub!(match[0], '')
      end
    end

    def extract_remix(text)
      if match = text.match(/\s[\(\[](.*?)\s?(Remix|Rmx|Mix)[\)\]]/i)
        @tags[:is_remix] = !match[2].nil? 
        @tags[:remixed_by] = match[1]
        text.gsub!(match[0], '')
      end
    end

    def extract_producer(text)
      if match = text.match(/\s[\(\[]\s?(Produced by|prod[^\s]+)\s?(.*?)[\)\]]/i)
        @tags[:producer] = match[2]
        text.gsub!(match[0], '')
      end
    end
  end
end