How to ignore warnings coming from a particular module?

[vect-tmazzucotelli]

I want to treat warnings as errors in my tests suite, and ignore some of them.

Here are the warnings:

$ python -m pytest tests
========================================== test session starts ==========================================
platform linux -- Python 3.9.15, pytest-7.2.0, pluggy-1.0.0
rootdir: /home/tmazzucotelli/data/dev/sdk-python, configfile: pytest.ini
plugins: cov-4.0.0, anyio-3.6.2
collected 199 items / 3 errors                                                                          

================================================ ERRORS =================================================
____________________ ERROR collecting tests/ipynb/test_experiment_quick_reference.py ____________________
tests/ipynb/test_experiment_quick_reference.py:6: in <module>
    from testbook import testbook
.venv/lib/python3.9/site-packages/testbook/__init__.py:2: in <module>
    from .testbook import testbook
.venv/lib/python3.9/site-packages/testbook/testbook.py:7: in <module>
    from .client import TestbookNotebookClient
.venv/lib/python3.9/site-packages/testbook/client.py:6: in <module>
    from nbclient import NotebookClient
.venv/lib/python3.9/site-packages/nbclient/__init__.py:6: in <module>
    from .client import NotebookClient, execute  # noqa: F401
.venv/lib/python3.9/site-packages/nbclient/client.py:13: in <module>
    from jupyter_client import KernelManager
.venv/lib/python3.9/site-packages/jupyter_client/__init__.py:8: in <module>
    from .asynchronous import AsyncKernelClient  # noqa
.venv/lib/python3.9/site-packages/jupyter_client/asynchronous/__init__.py:1: in <module>
    from .client import AsyncKernelClient  # noqa
.venv/lib/python3.9/site-packages/jupyter_client/asynchronous/client.py:8: in <module>
    from jupyter_client.client import KernelClient
.venv/lib/python3.9/site-packages/jupyter_client/client.py:22: in <module>
    from .connect import ConnectionFileMixin
.venv/lib/python3.9/site-packages/jupyter_client/connect.py:27: in <module>
    from jupyter_core.paths import jupyter_data_dir
.venv/lib/python3.9/site-packages/jupyter_core/paths.py:208: in <module>
    deprecation(
.venv/lib/python3.9/site-packages/jupyter_core/utils/__init__.py:87: in deprecation
    warnings.warn(message, DeprecationWarning, stacklevel=stacklevel + 1)
E   DeprecationWarning: Jupyter is migrating its paths to use standard platformdirs
E   given by the platformdirs library.  To remove this warning and
E   see the appropriate new directories, set the environment variable
E   `JUPYTER_PLATFORM_DIRS=1` and then run `jupyter --paths`.
E   The use of platformdirs will be the default in `jupyter_core` v6
______________________ ERROR collecting tests/ipynb/test_tutorial_modelling_all.py ______________________
tests/ipynb/test_tutorial_modelling_all.py:6: in <module>
    from testbook import testbook
.venv/lib/python3.9/site-packages/testbook/__init__.py:2: in <module>
    from .testbook import testbook
.venv/lib/python3.9/site-packages/testbook/testbook.py:7: in <module>
    from .client import TestbookNotebookClient
.venv/lib/python3.9/site-packages/testbook/client.py:6: in <module>
    from nbclient import NotebookClient
.venv/lib/python3.9/site-packages/nbclient/__init__.py:6: in <module>
    from .client import NotebookClient, execute  # noqa: F401
.venv/lib/python3.9/site-packages/nbclient/client.py:13: in <module>
    from jupyter_client import KernelManager
.venv/lib/python3.9/site-packages/jupyter_client/__init__.py:8: in <module>
    from .asynchronous import AsyncKernelClient  # noqa
.venv/lib/python3.9/site-packages/jupyter_client/asynchronous/__init__.py:1: in <module>
    from .client import AsyncKernelClient  # noqa
.venv/lib/python3.9/site-packages/jupyter_client/asynchronous/client.py:8: in <module>
    from jupyter_client.client import KernelClient
.venv/lib/python3.9/site-packages/jupyter_client/client.py:22: in <module>
    from .connect import ConnectionFileMixin
.venv/lib/python3.9/site-packages/jupyter_client/connect.py:27: in <module>
    from jupyter_core.paths import jupyter_data_dir
.venv/lib/python3.9/site-packages/jupyter_core/paths.py:208: in <module>
    deprecation(
.venv/lib/python3.9/site-packages/jupyter_core/utils/__init__.py:87: in deprecation
    warnings.warn(message, DeprecationWarning, stacklevel=stacklevel + 1)
E   DeprecationWarning: Jupyter is migrating its paths to use standard platformdirs
E   given by the platformdirs library.  To remove this warning and
E   see the appropriate new directories, set the environment variable
E   `JUPYTER_PLATFORM_DIRS=1` and then run `jupyter --paths`.
E   The use of platformdirs will be the default in `jupyter_core` v6
_____________________ ERROR collecting tests/vectice/models/test_register_model.py ______________________
tests/vectice/models/test_register_model.py:9: in <module>
    from sklearn import datasets, linear_model
.venv/lib/python3.9/site-packages/sklearn/datasets/__init__.py:22: in <module>
    from ._twenty_newsgroups import fetch_20newsgroups
.venv/lib/python3.9/site-packages/sklearn/datasets/_twenty_newsgroups.py:46: in <module>
    from ..feature_extraction.text import CountVectorizer
.venv/lib/python3.9/site-packages/sklearn/feature_extraction/__init__.py:10: in <module>
    from . import text
.venv/lib/python3.9/site-packages/sklearn/feature_extraction/text.py:29: in <module>
    from ..preprocessing import normalize
.venv/lib/python3.9/site-packages/sklearn/preprocessing/__init__.py:31: in <module>
    from ._label import label_binarize
.venv/lib/python3.9/site-packages/sklearn/preprocessing/_label.py:25: in <module>
    from ..utils.multiclass import unique_labels
.venv/lib/python3.9/site-packages/sklearn/utils/multiclass.py:14: in <module>
    from scipy.sparse.base import spmatrix
.venv/lib/python3.9/site-packages/scipy/sparse/base.py:38: in __getattr__
    warnings.warn(f"Please use `{name}` from the `scipy.sparse` namespace, "
E   DeprecationWarning: Please use `spmatrix` from the `scipy.sparse` namespace, the `scipy.sparse.base` namespace is deprecated.
======================================== short test summary info ========================================
ERROR tests/ipynb/test_experiment_quick_reference.py - DeprecationWarning: Jupyter is migrating its paths to use standard platformdirs
ERROR tests/ipynb/test_tutorial_modelling_all.py - DeprecationWarning: Jupyter is migrating its paths to use standard platformdirs
ERROR tests/vectice/models/test_register_model.py - DeprecationWarning: Please use `spmatrix` from the `scipy.sparse` namespace, the `scipy.sparse.base`...
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Interrupted: 3 errors during collection !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
=========================================== 3 errors in 1.23s ===========================================

I tried to ignore them based on their category and module of origin, but it does not work:

filterwarnings =
    error
    ignore::DeprecationWarning:scipy.sparse.base
    ignore::DeprecationWarning:jupyter_core.utils

If I specify a message instead, warnings are correctly ignored:

filterwarnings =
    error
    ignore:Please use `spmatrix`:DeprecationWarning
    ignore:Jupyter is migrating its paths to use standard platformdirs:DeprecationWarning

I understand that module is a regex that the module's qualified name must match, but I can't seem to get it right.

Even this does not work:

filterwarnings =
    error
    ignore::DeprecationWarning:.*(scipy|sparse|base|jupyter_core|utils).*

Even weirder, if I use .*utils.* as regex, then the scipy warnings are ignored! But utils is nowhere to be found in .venv/lib/python3.9/site-packages/scipy/sparse/base.py!! What is happening 😂 ??

[vect-tmazzucotelli]

Ah, I think I understand: utils does appear in sklearn/utils/multiclass, the module just above in the stack.
Does it mean I always have to use the qualified name of the previous module in the stack? What if it's the same module in the whole stack 🤔? Unless the qualified name depends on the specified warning stack level 🤯? If Python itself behaves like that as well, its documentation is clearly lacking 😕

Would be nice to get it documented in Pytest in the meantime 🙂