feat(wip): port reference JS implementation to Python

[0-th]

The cooperative scheduler can run cooperatively run tasks

[0-th]

Running the tests in test/test_task.py would give the following output:

py-actress on  dev(wip) [?] is 📦 v0.0.1 via 🐍 v3.13.2 (.venv) 
❯ python test/test_task.py

======================================================================
TEST 1: Stack Class
======================================================================
Empty stack size: 0
After adding to active: 1
After adding to idle: 2
✓ Stack tests passed!

======================================================================
TEST 2: Primitives
======================================================================
Got reference: <generator object test_primitives.<locals>.task at 0x7863bf7faa80>
Active tasks: 0
Idle tasks: 1
✓ Primitives tests passed!

======================================================================
TEST 3: Multiple Tasks
======================================================================
Execution order:
  task1: start
  task2: start
  task1: middle
  task2: end
  task1: end
✓ Multiple tasks test passed!

======================================================================
TEST 4: Suspend and Resume
======================================================================
{<generator object test_suspend_resume.<locals>.task1 at 0x7863bf3eb850>}
Execution order:
  task1: before suspend
  task2: running
  task2: resuming task1
  task1: after resume
✓ Suspend/resume test passed!

======================================================================
TEST 5: Messages
======================================================================
Collected messages: ['message-1', 'message-2', 'message-3']
✓ Messages test passed!

======================================================================
ALL TESTS PASSED! Great work on Milestone 3!
======================================================================

py-actress on  dev(wip) [?] is 📦 v0.0.1 via 🐍 v3.13.2 (.venv) 
❯ 

[0-th]

59 tests from the reference implementation are passing, all 64 tests have been implemented except 2 type tests that were typescript specific (which I'd assume the type checker - mypy would handle in this Python implementation).

3 of the tests are currently deadlocking, potentially caused by differences in event loop behaviour in Python and JS. These 3 tests are currently commented out, and I'm currently debugging them for a fix.