Justify pending behaviour with mocks expectations.

Author: xaviershay

lib/rspec/core/pending.rb

@@ -132,7 +132,7 @@ def self.mark_pending!(example, message_or_bool)
 
       def self.mark_fixed!(example)
         example.metadata[:pending] = false
-        example.execution_result[:pending_fixed] = true
+        example.metadata[:execution_result][:pending_fixed] = true
       end
 
       def self.guarded?(*args)

spec/rspec/core/pending_example_spec.rb

@@ -207,9 +207,20 @@ def run_example(*pending_args)
         super(*pending_args) { expect("foo").to receive(:bar) }
       end
 
-      it 'passes' do
-        pending 'need to figure out a way to do this'
-        expect(run_example("just because")).to_not be_pending
+      it 'is pending' do
+        # This behaviour appears inconsistent. You would expect this to
+        # pass, since the pending block is "failing". It is problematic though
+        # because the failure is not actually present until the end of the
+        # example.
+        #
+        # There may be a solution here where the example is a run a second time
+        # treating pending blocks as skipped, but I haven't thought it through
+        # and seems probably more complex than it's worth. It's not clear to me
+        # what "expected behaviour" should even be.
+        #
+        # The fallback behaviour of always flagging the example pending doesn't
+        # seem so bad.
+        expect(run_example("just because")).to be_pending
       end
     end