Test::Unit::TestCase

This method will define a test method using the description as the test name (“pending function” => “test_pending_function”)

Instead of doing this:

def test_function_is_pending

pending "this test is pending"

end

You can just do this:

pending “function is pending”

This method can be called with a block passed, the same as the instance method

# File lib/pending.rb, line 76
def self.pending(description, &block)
  test_name = "test_#{description.gsub(/\s+/,'_')}".to_sym
  defined = instance_method(test_name) rescue false
  raise "#{test_name} is already defined in #{self}" if defined
  define_method(test_name) do
    pending(description) {self.instance_eval(&block)}
  end
end

The pending method lets you define a block of test code that is currently “pending” functionality.

You can use it two ways. One is simply put a string as the parameter:

def test_web_service_integration
  pending "This is not done yet..."
end

This will output a “P” in the test output alerting there is pending functionality.

You can also supply a block of code:

def test_new_helpers
  pending "New helpers for database display" do
    output = render_record(User.first)
    assert_equal "Jerry User (jerry@users.com)", output
  end
end

If the block doesn’t fail, then the test will flunk with output like:

<New helpers for database display> did not fail.

If the test fails (i.e., the functionality isn’t implemented), then it will not fail the surrounding test.

# File lib/pending.rb, line 34
def pending(description = "", &block)
  if block_given?
    failed = false

    begin
      block.call
    rescue
      failed = true
    end

    flunk("<#{description}> did not fail.") unless failed 
  end

  caller[0] =~ (/(.*):(.*):in `(.*)'/)
  @@pending_cases << "#{$3} at #{$1}, line #{$2}"
  print "P"
  
  @@at_exit ||= begin
    at_exit do
      puts "\nPending Cases:"
      @@pending_cases.each do |test_case|
        puts test_case
      end
    end
  end
end