require 'nokogiri'

class BigFiveResultsTextSerializer

  # big5 = BigFiveResultsTextSerializer.new('personality_test.html').to_h
  # post = BigFiveResultsPoster.new(big5, 'test+1114@gmail.com').post
  # post.token = UPnemRUJEaXxhdKuwMzxMSJe
  # post.results_hash = UPnemRUJEaXxhdKuwMzxMSJe

  def initialize(text)

    @file = Nokogiri::HTML(File.read(text))
    @hash = {}

  end

  def to_h

    add_name_to_hash

    # contains all domain, facets and their score in a rough format
    extracted_domains_and_facet = @file.css('.graph-txt')

    extracted_domains_and_facet.each do |domain_and_facets|

      # clean up the data by removing the information not required
      clean_domain_and_facts = clean_up_domain_and_facts(domain_and_facets)

      # prepare the hash in a format that is required
      prepare_domain_and_facts(clean_domain_and_facts)
    end

    @hash
  end


  private

  def add_name_to_hash
    @hash.merge!({'NAME' => 'Shairyar Baig'})
  end

  def clean_up_domain_and_facts(result)
    # get the personality node that we are interested in
    domain_and_facets = result.css('code').children.children

    # drop the first index as we dont need it
    domain_and_facets.shift
    domain_and_facets
  end

  def prepare_domain_and_facts(results)

    domain = prepare_domain(results)
    prepare_facets(domain, results)

  end

  def prepare_domain(results)
    domain = extract_domain_and_score(results[0])
    @hash.merge!(domain.first => {'Overall Score' => domain.last.to_i, 'Facets' => {}})

    domain
  end

  def prepare_facets(domain, results)

    results.each_with_index do |result, index|

      # first index is the domain and we only want to prepare the facets
      if index != 0
        extract_domain_and_score = extract_domain_and_score(result)
        @hash[domain.first]['Facets'].merge!({"#{extract_domain_and_score.first}" => extract_domain_and_score.last.to_i})
      end
    end

  end

  def extract_domain_and_score(result)
    # removes the special characters
    # and then converts the result into an array of domain/facet and score ["EXTRAVERSION", "43"]
    result.text.gsub('.', '|').split('|')
  end

end