do not execute boundary events in catch

[essweine]

This is a fairly minor change, but required lots of test updates.

Boundary events were formerly being run from catch; ideally tasks should only run through an explicit call. To fix boundary events, I had to also update subworkflow tasks not to immediately move to ready when the subworkflow completes (also a net positive since it removes a special case transition).

However, this means that they need to be updated before they become ready, and then they need to be run, which required lots of updates to our tests.

[sonarqubecloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

0.0% Coverage
0.0% Duplication

[burnettk]

if i replace, in backend's pyproject.toml:

SpiffWorkflow = {git = "https://github.com/sartography/SpiffWorkflow", rev = "main"}

with:

SpiffWorkflow = {git = "https://github.com/sartography/SpiffWorkflow", rev = "bugfix/run-boundary-events-from-engine-steps"}

and run poetry update SpiffWorkflow, then i get this when i run tests in backend:

FAILED tests/spiffworkflow_backend/unit/test_process_instance_processor.py::TestProcessInstanceProcessor::test_properly_saves_tasks_when_running - IndexError: list index out of range
FAILED tests/spiffworkflow_backend/unit/test_process_instance_processor.py::TestProcessInstanceProcessor::test_it_can_loopback_to_previous_bpmn_subprocess_with_gateway - assert 0 == 1
FAILED tests/spiffworkflow_backend/unit/test_process_model.py::TestProcessModel::test_can_run_process_model_with_call_activities_when_in_same_process_model_directory - AssertionError: assert 'waiting' == 'complete'
FAILED tests/spiffworkflow_backend/unit/test_process_model.py::TestProcessModel::test_can_run_process_model_with_call_activities_when_not_in_same_directory - AssertionError: assert 'waiting' == 'complete'
FAILED tests/spiffworkflow_backend/unit/test_process_model.py::TestProcessModel::test_can_run_process_model_with_call_activities_when_process_identifier_is_not_in_database - AssertionError: assert 'waiting' == 'complete'

the assertion that fails in test_can_run_process_model_with_call_activities_when_in_same_process_model_directory is:

assert process_instance.status == "complete" # and the instance is waiting instead of complete.

it seems like when a call activity is run with this branch in backend, activities inside of the called activity can all complete successfully, but the call activity itself (the caller) gets stuck in waiting instead of completing.

[essweine]

If you refresh waiting tasks, they should change to ready. I changed this in many of the spiffworkflow unit tests, and you probably have to do that in the spiffworkflow-backend tests too. Instead of the subworkflow completion event automatically changing the task's state, it now happens via the _update method (which is really where the update should have been done all along).