Run MicroPython in box tests on CI

[krzema12]

See #51.

Proposed goal: run box tests for MP next to regular Python, with separate report files. A separate Gradle task can be created: microPythonTest.

• [X] run box tests for MicroPython
• [x] generate SVG git plot, including MicroPython
• [X] run e2e tests with MicroPython - https://github.com/krzema12/kotlin-python/blob/python-backend/python/e2e-tests/run.sh#L25

  • 84

  • 85

[krzema12]

Starting with exposing a flag that makes the box tests use micropython binary.

[krzema12]

• [x] update blog post with new address to the box tests plot - it's impossible to edit the post

[krzema12]

Using MicroPython in e2e tests fails with:

Traceback (most recent call last):
  File "python/e2e-tests/out/python-consumer.py", line 1, in <module>
  File "/home/piotr/repos/priv/kotlin-python/python/e2e-tests/out/compiled.py", line 12986, in <module>
MemoryError: memory allocation failed, allocating 26032 bytes

At this line, there's:

def invokeSuspendSuperTypeWithReceiverAndParam(self, receiver, param, completion):
    raise NotImplementedError('It is intrinsic method')

I think using MP in e2e tests has to be postponed until these MemoryErrors are solved.

[LouisCAD]

Just thought about this: the issue might be about the implications of instantiating a NotImplementedError. Maybe it's worth trying to raise a custom thing/error/exception to see what causes this?

Isn't that related to the cost of building the stacktrace? I don't know how that works in Python, but I wouldn't be surprised to learn that it works differently in MicroPython, or that it's just not supported.

[krzema12]

Thanks, we'll investigate it in https://github.com/krzema12/kotlin-python/issues/84.