NPE fix

[bradNASA]

NPE fix, untested

[mattdailis]

Thank for the fix, I've run into this a number of times myself. This can occur if a daemon task spawns an activity, and that activity doesn't finish by the end of simulation

[mattdailis]

Discussed with Theresa, and we decided to refactor these two lines of code:

aerie/merlin-driver/src/main/java/gov/nasa/jpl/aerie/merlin/driver/engine/SimulationEngine.java

Line 831 in ac8720e

(activityParents.containsKey(span)) ? Optional.empty() : Optional.ofNullable(directiveId),

To this:

+(activityParents.containsKey(span)) ? Optional.empty() : Optional.ofNullable(directiveId)
-Optional.ofNullable(directiveId)

The ternary dates back to a time when we used the absence of a parent span to determine whether a span is "top level", and assumed that all top level spans came from the plan. Now, we have a more direct way to compute directive ids, which is robust to the situation where a parent-less span comes from a daemon task rather than a directive in the plan.

[dandelany]

Thanks all! Merging this fix.