E001 update tests

[syntron]

(E001) update tests (v4.x.x)

[test_*] reorder imports
[tests_ModelicaDoE*] fix pylint hint
* use .items()
[tests_*] use OMSessionABC.get_version()
[test_ModelicaSystemCmd] use get_model_name() instead of access to private variable _model_name
[test_ModelicaSystemOMC] read file using utf-8 encoding / linter fix
[test_ModelicaSystemRunner] update test case
* ModelicaSystemRunner & OMCPath
* ModelicaSystemRunner & OMPathRunnerLocal
* ModelicaSystemRunner & OMPathRunnerBash
* ModelicaSystemRunner & OMPathRunnerBash using docker
* ModelicaSystemRunner & OMPathRunnerBash using WSL (not tested!)
[test_OMCPath] update test case
* OMCPath & OMCSessionZMQ
* OMCPath & OMCSessionLocal
* OMCPath & OMCSessionDocker
* OMCPath & OMCSessionWSL (not tested!)
* OMPathLocal & OMCSessionRunner
* OMPathBash & OMCSessionRunner
* OMPathBash & OMCSessionRunner in docker
* OMPathBash & OMCSessionRunner in WSL (not tested!)
add workflow to run unittests in ./tests
[test_OMParser] use only the public interface => om_parser_basic()
[test_OMTypedParser] rename file / use om_parser_typed()
update tests - do NOT run test_FMIRegression.py
reason:
* it is only a test for OMC / not OMPython specific
* furthermore, it is run automatically via cron job (= FMITest)
[test_ModelExecutionCmd] rename from test_ModelicaSystemCmd

[adeas31]

The simulation on Windows takes more time. Increase the default timeout of 10s in OMCSession/ModelicaSystem.

[syntron]

it is now down to docker / windows failing

[adeas31]

I updated the timeout to 5 mins but it still fails. I think something is stuck when running tests/test_linearization.py::test_getters. I will try to run it locally on machine and debug it.

[adeas31]

Running the test alone works fine https://github.com/OpenModelica/OMPython/actions/runs/26285383965/job/77371637784?pr=439

I will debug more.

[adeas31]

If i run pytest with --log-cli-level=DEBUG then it works fine. This is very strange.

[adeas31]

I still haven't got enough time to look into this. Will try to debug this more asap.

[syntron]

@adeas31 Any point where I can help?

[adeas31]

@adeas31 Any point where I can help?

I will let you know.
I am running all the test locally now on my Windows machine. Lets see if I can find something.

[adeas31]

This seems to work fine with increased timeout of workflow timeout-minutes: 120.

The FMI import tests were failing for Windows. I am trying to get them fixed in OpenModelica OpenModelica/OpenModelica#15864. For now I have skipped the test to see if this PR passes.

[adeas31]

This works finally.

I added inputs for pytest in the workflow and increased the workflow timeout.
Also increased the command execution timeout from 10 secs to 5 minutes. I noticed that the timeout is defined in OMCSession, ModelExecutionCmd and ModelicaSystemCmd. AFAIU we only need it defined in OMCSession. ModelExecutionCmd uses the timeout from OMCSession and ModelicaSystemCmd then uses ModelExecutionCmd. Perhaps we can improve this in another PR. For now I will merge this one.

[syntron]

This works finally.

I added inputs for pytest in the workflow and increased the workflow timeout. Also increased the command execution timeout from 10 secs to 5 minutes. I noticed that the timeout is defined in OMCSession, ModelExecutionCmd and ModelicaSystemCmd. AFAIU we only need it defined in OMCSession. ModelExecutionCmd uses the timeout from OMCSession and ModelicaSystemCmd then uses ModelExecutionCmd. Perhaps we can improve this in another PR. For now I will merge this one.

timeout values are handled in different places but - as you wrote - defined only in some selected ones; I will check it again - it should be possible to get it down to ONE point with the definition (300s)

[syntron]

@adeas31 see PR #467

[adeas31]

@adeas31 see PR #467

Thanks!