batch update s3 objects metadata with proper mime-type

s3batch.rb

#!/usr/bin/env ruby

require 'aws-sdk-s3'
require 'rack'

class BucketListObjectsWrapper
  attr_reader :bucket

  def initialize(bucket)
    @bucket = bucket
  end

  # Lists object in a bucket and update it's mime type.
  #
  # @param start [Integer] The maximum number of objects to list.
  # @param max_objects [Integer] The maximum number of objects to list.
  # @return [Integer] The number of objects listed.
  def list_objects(start, max_objects)
    count = 0
    puts "The objects in #{@bucket.name} are:"
    @bucket.objects.each do |obj_summary|
      if count >= start
        source = Aws::S3::Object.new(obj_summary.bucket_name, obj_summary.key)
        target = source
        mime = Rack::Mime.mime_type('.' + obj_summary.key.split('.').last)
        puts "processing key: #{obj_summary.key} to set content type as: #{mime}"
        target.copy_from(source, content_type: mime, metadata_directive: 'REPLACE')
      end
      count += 1
      break if count == max_objects
    end
    count
  rescue Aws::Errors::ServiceError => e
    puts "Couldn't list objects in bucket #{bucket.name}. Here's why: #{e.message}"
    0
  end
end

def run_script
  bucket_name = "oscar-dev-document-service"
  
  if ARGV.length != 2
    puts "We need exactly two arguments. a start and limit. Example: ruby s3batch.rb 0 2"
    exit
  end
  bkt = Aws::S3::Bucket.new(bucket_name)
  wrapper = BucketListObjectsWrapper.new(bkt)
  count = wrapper.list_objects(ARGV[0].to_i, ARGV[1].to_i)
  puts "Updated #{count} objects out of #{bkt.objects.count}"
end

run_script if $PROGRAM_NAME == __FILE__