Automate interaction of successful tests

[benjamin051000]

Option (or default) to run all students' scripts automatically, only go back to the ones that have issues (compile or sim load errors). For each students' project that loads and runs TB properly, just output a little table that shows each name and score. Way less interactivity, faster results.

[benjamin051000]

Have an option to automatically run through all sims and yield a report at the very end (no need to press Return for each student). Will be tricky with sims that break/never finish. Not entirely sure how to handle all cases here, may just end up being a "use with caution" flag

[benjamin051000]

Not sure if this could be done with a timeout parameter in the TCL, but it could be done with multiprocessing, and just have a parent process that kills procs that take too long.

[benjamin051000]

This relies on #10 to work well, since separating submissions into their own projects allows for simple parallelization.

[benjamin051000]

Now that #10 is working out nicely, this is becoming something I've been thinking about more and more. I could see the terminal interface showing the status of each simulation process and its final scores (or an X for failed), and you could select each one to see details (e.g., transcript log, etc). This would also tie in to #19

[benjamin051000]

Testbench results will be confusing to understand because they aren't unified in their grading structures. This needs to change

[benjamin051000]

We don't actually need #19 in order for this to work. Just have it collect the stdout, parse it, and print a little table. That's a lot easier.