Logic to handle premature ready_to_run_testitems event

[SBuercklin]

I ran into a corner case where the ready_to_run_testitems event could be encountered before the get_procs_for_testrun event could complete.

The first change is to make the run_testitems event conditional on populating the testitem_ids_by_proc mapping with the relevant process ID.

However, this fix led to an issue for me where the line to determine the test process Julia version would hang (read(Cmd(...), String) further up in this file, line 146). In order to make sure that command actually gets scheduled, we need to yield, giving the read(Cmd(...)) bit an opportunity to run, and then resuming the core loop of execute_test_run.

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 5 lines in your changes missing coverage. Please review.

Project coverage is 0.47%. Comparing base (4ce762b) to head (3a6cf05).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/testitemcontroller.jl	0.00%	5 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##            main      #2      +/-   ##
========================================
- Coverage   0.47%   0.47%   -0.01%     
========================================
  Files          5       5              
  Lines        419     422       +3     
========================================
  Hits           2       2              
- Misses       417     420       +3     

Flag	Coverage Δ
unittests	0.47% <0.00%> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.