Write comprehensive set of tests for testing COA nodes

[hneema08]

@yogeshVU : Please make sure to test COAs not just for the HelloWorld sample in cpswt-samples, but also for your own COA models. This will help you in getting the same functionality after the structure and handling of COA JSONs and COASelections is updated per the issue in cpswt-meta, viz. https://github.com/SimIntToolkit/cpswt-meta/issues/63

At present, I expect at least these COA elements to be working fine: SyncPt, Dur, RandomDur, Fork, ProbabilisticChoice, AwaitN, Action, and Outcome. However, you can definitely test other COA elements as well.

[hneema08]

It was determined during testing, that although the core COA executor is in place and the JSON files are getting generated (in more or less the way expected), the Federation Manager and COAExecutor are not prepared for running only the selected COA. A new will be filed for this.

[hneema08]

@yogeshVU : Please take it with priority to test the COAs thoroughly and report any issues:

1. Test and close if working:
   Handle execution of only the selected COA-Selection #85

2. Test and close if working:
   https://github.com/SimIntToolkit/cpswt-meta/issues/66

3. Test HelloWorld_with_simpleCOAs example end-to-end

4. Test other COA elements that are in the example "HelloWorld example with ALL COAs" to see which work fine and which needs to be implemented. If SyncPt, Duration, RandomDuration, Fork, ProbabilisticChoice, AwaitN, Action, and Outcome node types work, close this issue. Other COA elements if they don't work, please file an issue with an example.

5. Check for boundary conditions and wrong input data, e.g. no COAs in an experiment, wrongly specified COASelection, etc.

[hneema08]

@yogeshVU Numbers 1, 2, and 3 are done and seem to work fine. Numbers 4 and 5 above are still to be done. Below is the detailed report on testing so far:

The following COA elements work fine as per the latest test results:

• SyncPt (waits for a certain time to be reached from the beginning of the simulation)
• Action (injects an interaction into the running simulation)
• Dur (waits for the specified duration from the time this COA element is reached in the COA graph)
• Outcome (waits for a specified interaction from any federate in the federation)
• Fork (branches the execution in parallel paths each of which gets executed simultaneously)

COA elements that aren't tested yet:

• RandomDur (waits for a randomly generated duration b/n uniform distribution lower and upper bounds)
• OutcomeFilter (after an interaction for an Outcome node has been received, the filter checks the values against specified expression and proceeds only if expression returns true)
• ProbabilisticChoice (like 'Fork' but only one outgoing branch is executed that is determined by the probabilities on the outgoing branches)
• TerminateCOA (when reached ends the execution of the entire COA where the 'TerminateCOA' element is encountered during execution)
  -AwaitN (waits for a given number of incoming branches to finish prior to continuing execution)

The COA elements tested so far might be enough for what NIST might need for now, but it will be good to get others working in general (depending on availability of time).