abusing blocks

rake-task.rb

def update_subm(name, &block)
  puts "Updating submodules for '#{name}'...."
  if block.nil?
    sh %{cd #{name} ; #{GIT_SM_UP}}
  else
    sh %{cd #{name} ; #{GIT_SM_UP}} do |o, r|
      block.call(o, r)
    end
  end
end

desc 'This does a recursive git submodule update.'
task :update do
  update_subm '.' do |ok, res|
    puts "ERROR: #{ok}. #{res}. Continuing."
  end
  sh %{cd #{GEOEXPLORER_DIR}/app/static/externals/gxp/externals/TimeManager ; git reset --hard HEAD}
end

Basically, I'm trying to eat the error that the first shell state in the task throws. To do that, I want to define a block on the calling location, but just pass it untouched to the call to sh.

Originally, I'd been calling update_subm multiple times, so having it refactored into its own method made more sense.

I assume that what I'm doing here can't be right.