Maybe a good way to make resque job definition just a tad simpler.

sugared_resque_job.rb

module SugaredResqueJob
  def self.new(queue, resque=Resque, &perform)
    Module.new do
      extend self
      define_method :queue   do queue end
      define_method :enqueue do |*args| resque.enqueue(self, *args) end
      define_method :perform do |*args| perform.call(        *args) end
    end
  end
end

# for mocking
resque = Object.new
SimpleJob = SugaredResqueJob.new(:queue_name, resque) do |*args|
  "performing on *args: #{args.inspect}"
end

describe SimpleJob do

  it "should enqueue job with Resque" do
    resque.should_receive(:enqueue).with(SimpleJob, :arg1, :arg2, :arg3)
    SimpleJob.enqueue(:arg1, :arg2, :arg3)
  end

  it "should assign job to the appropriate queue" do
    SimpleJob.queue.should == :queue_name
  end

  it "should assign job with the appropriate receiver name" do
    # "cheating" by using ruby's semi-magical Module#to_s constant lookup
    SimpleJob.to_s.should == "SimpleJob"
  end

  it "should run the appropriate code when performed" do
    result = SimpleJob.perform(:arg1, :arg2, :arg3)
    result.should == "performing on *args: [:arg1, :arg2, :arg3]"
  end

end

Note, I haven't tested this in production... yet. But I really like reducing my Job definitions (which almost always follow this pattern) to three lines of code. If I wanted to include or extend other modules into this, I can: (e.g. for the resque-retry plugin SimpleJob.extend Resque::Plugins::Retry should work). Class-inheritance based plugins (e.g. resque-status) simply wouldn't use this (presumably they come with their own sugar).

DRY resque job creation from 6 lines down to 3: normally my resque job creation looks like a less metaprogrammy version of lines 3..8, but now they would like like lines 14..16.

I like it