Rails runner friendly script to associate pids with a collection

README.md

The associated Ruby script is more of a proof of concept to demonstrate using pipes and parameters in conjunction with the Rails runner. I have not fully verified that it works.

echo '<work_pid>' | bundle exec rails runner associate_pids_with_collection.rb -c <collection_pid>

associate_pids_with_collection.rb

# !/usr/bin/env ruby

# In general this script does the following, assuming that it is run within a Rails context:
# ```ruby
# work = ActiveFedora::Base.find(patent_pid, cast: true)
# collection = LibraryCollection(collection_pid)
# return true if work.library_collection_ids.include?(collection.id)
# work.library_collection_ids += collection.id
# work.save!
# ```

require 'optparse'
config = {
  environment: 'development'
}

command_name = File.basename(__FILE__)

OptionParser.new do |options|
  # This banner is the first line of your help documentation.
  options.set_banner "Usage: bundle exec rails runner #{command_name} --collection_id [PID] [files]\n\n" \
    "Associate items identified by PIDs in the source files with the given --collection_id.\n\n" \
    "Each line of the given file is assumed to be a single pid\n\n " \
    "Examples:\n" \
    "\t$ bundle exec rails runner #{command_name} -c 'und:123456789' list_of_work_pids.txt\n" \
    "\t$ bundle exec rails runner #{command_name} -c 'und:123456789' list_of_pids.txt\n" \
    "\t$ bundle exec rails runner #{command_name} -c 'und:123456789' list_of_pids.txt | more\n" \
    "\t$ bundle exec rails runner #{command_name} -c 'und:123456789' list_of_pids.txt > output.txt\n" \
    "\t$ bundle exec rails runner #{command_name} -c 'und:123456789' list_of_pids.txt second_list_of_pids.txt\n" \
    "\t$ echo 'und:abcdefg' | bundle exec rails runner #{command_name} -c 'und:123456789'"

  # Separator just adds a new line with the specified text.
  options.separator ""
  options.separator "Specific options:"

  # Note this is provided to "parrot" the parameters of rails runner
  options.on("-e", "--environment=name", String, "Specifies the environment for the runner to operate under (test/development/production).", "Default: #{config[:environment]}") do |environment|
    config[:environment] = environment
  end

  options.on("-c", "--collection_pid [PID]", String, "The full PID of the target Collection (REQUIRED)") do |collection_pid|
    config[:collection_pid] = collection_pid
  end

  options.on_tail("-h", "--help", "You're looking at it!") do
    $stderr.puts options
    exit 1
  end
end.parse!

# Responsible for associating the PID to the config assigned collection
def associate_pid(pid, config)
  pid = pid.strip
  collection = config.fetch(:collection)
  work = nil
  begin
    work = ActiveFedora::Base.find(pid, cast: true)
  rescue ActiveFedora::ObjectNotFoundError
    $stderr.puts "Unable to find ActiveFedora::Base ID='#{pid}' and cannot add to the #{collection.class} ID='#{collection.id}'"
    return false
  end

  if work.library_collection_ids.include?(collection.id)
    $stdout.puts "#{work.class} ID='#{work.id}' already associated with #{collection.class} ID='#{collection.id}'"
    return true
  else
    work.library_collection_ids += [collection.id]
    work.save!
    $stdout.puts "Added #{work.class} ID='#{work.id}' to #{collection.class} ID='#{collection.id}'"
    return true
  end
end

unless config.key?(:collection_pid)
  $stderr.puts "Error: #{command_name} requires a --collection_pid option\nSee '#{command_name} -h' for more information"
  exit 2
end

begin
  $stdout.puts "Finding LibraryCollection ID='#{config.fetch(:collection_pid)}'"
  config[:collection] = LibraryCollection.find(config.fetch(:collection_pid))
rescue ActiveFedora::ObjectNotFoundError
  $stderr.puts "Unable to find LibraryCollection ID='#{config.fetch(:collection_pid)}'"
  exit 3
end

# Keep reading lines of input as long as they're coming.
while input = ARGF.gets
  input.each_line do |pid|
    begin
      associate_pid(pid, config)
    rescue Errno::EPIPE
      # sysexits(3) specifies that exit code 74 represent an IO error,
      # which is the likely situation
      exit(74)
    end
  end
end