Remove elapsed time from pending results

spec/pending_result_spec.cr

@@ -36,10 +36,4 @@ describe Spectator::PendingResult do
       result.example.should eq(example)
     end
   end
-
-  describe "#elapsed" do
-    it "is zero" do
-      new_pending_result.elapsed.should eq(Time::Span.zero)
-    end
-  end
 end

spec/runner_spec.cr

@@ -148,7 +148,7 @@ describe Spectator::Runner do
         spy = SpyFormatter.new
         runner = Spectator::Runner.new(suite, spectator_test_config(spy))
         max_time = Time.measure { runner.run }
-        min_time = spy.end_example_calls.sum(&.elapsed)
+        min_time = spy.end_example_calls.each.map(&.as(Spectator::FinishedResult)).sum(&.elapsed)
         report = spy.end_suite_calls.first
         report.runtime.should be <= max_time
         report.runtime.should be >= min_time

src/spectator/pending_result.cr

@@ -5,11 +5,6 @@ module Spectator
   # A pending result means the example is not ready to run yet.
   # This can happen when the functionality to be tested is not implemented yet.
   class PendingResult < Result
-    # Length of time it took to run the example.
-    def elapsed : Time::Span
-      Time::Span.zero
-    end
-
     # Indicates that an example was run and it was successful.
     # NOTE: Examples with warnings count as successful.
     # This will always be false for this type of result.

src/spectator/result.cr

@@ -5,9 +5,6 @@ module Spectator
     # Example that was run that this result is for.
     getter example : Example
 
-    # Length of time it took to run the example.
-    abstract def elapsed : Time::Span
-
     # Indicates that an example was run and it was successful.
     # NOTE: Examples with warnings count as successful.
     abstract def passed? : Bool