framework/output: handle missing/corrupt result files.

Do not explode if a result file for a job is missing when loading a
RunOutput. Specify job status as "UNKNOWN" and add the exception from
attempting to load the file to the events.

wa/framework/output.py

@@ -64,8 +64,14 @@ class Output(object):
         self.events = []
 
     def reload(self):
+        try:
             pod = read_pod(self.resultfile)
             self.result = Result.from_pod(pod)
+        except Exception as e:
+            self.result = Result()
+            self.result.status = Status.UNKNOWN
+            self.add_event(str(e))
+
 
     def write_result(self):
         write_pod(self.result.to_pod(), self.resultfile)