get source, etc. from DigiTool objects. Check against CSV and update matches

gistfile1.rb

#!/usr/bin/env ruby

require 'nokogiri'
require 'csv'

def get_stuff_from_dc(file)
  ret_values = {}
  my_xml = Nokogiri::XML(File.read(file))
  # I know these rescue nils are bad but I don't know a more efficient way
  ret_values[:pid] = my_xml.at_xpath('//xb:digital_entity/pid/text()', "xb" => "http://com/exlibris/digitool/repository/api/xmlbeans").content rescue nil
  dc_source_raw = Nokogiri::XML(my_xml.at_xpath("//md[type='dc']/value").content).xpath("//dc:source/text()","dc" => "http://purl.org/dc/elements/1.1/") rescue nil
  ret_values[:source] = dc_source_raw.map(&:content).select{|z| z.match(/^Photo\sNumber:\s/) }.map{|z| z.gsub(/^Photo\sNumber:\s/,"") } rescue nil
  return ret_values
end


f = nil; File.open('/home/jhbrown/shared_with_fedora/photodata.csv') do |file|
  f = CSV.parse(file.read.encode('UTF-8', :invalid => :replace)) # , :headers => true, :header_converters => :symbol)
end

my_files = Dir.chdir('/home/jhbrown/shared_with_fedora/digital_entities') { Dir.glob("./*").map{|x| File.expand_path(x) } }
# my_files = ['/home/jhbrown/shared_with_fedora/digital_entities/87688.xml']

my_files_parsed = my_files.map{|x| get_stuff_from_dc(x) }

csv_arr_1 = []
f.each do |x|
  my_pn = x[6] # x[:photo_number]
  pid_array = []
  my_files_parsed.each do |y|
    temp = y[:source].include?(my_pn) ? y[:pid] : nil rescue nil
    pid_array << temp unless temp.nil?  
  end
  pid_array.empty? ? csv_arr_1 << (x << nil) : csv_arr_1 << (x << pid_array.join("|"))   
end

CSV.open("temp.csv", "w") do |csv|
  csv_arr_1.each do |i|
    csv << i
  end
end