Question: How to fail a leaky test

[tonybaloney]

Feature Request

I've got some scripts using memray that run some critical code 100,000 times, I use this in the following way:

mkdir -p .profiles
PYTHONMALLOC=malloc memray run -o .profiles/memray_logrecord.py.bin -f --native memray_logrecord.py
memray flamegraph --leaks -f .profiles/memray_logrecord.py.bin
python -m http.server --directory .profiles 8000

Then I manually look at the HTML report for leaky lines and raise bugs against them. Is there any way to automate this using this pytest package?

The test script is just something like:

def do():
    # setup... 

    for _ in range(100_000):
        ... # do code things

    # assert


if __name__ == "__main__":
    do()

[gaborbernat]

Any reason the https://pytest-memray.readthedocs.io/en/latest/usage.html#limit-memory does not help?

[seebi]

@gaborbernat thanks for the link. Is there a way to set a default limit for all tests (without the need for adding this to all tests manually) - something like a memory base fence ...?

[gaborbernat]

There is not at the moment, but PR welcome. And you can also set it to your project via a pytest hook.

[seebi]

Do you have a link to an example for this as well ? :-)

[sarahmonod]

You can use any of those hooks: https://docs.pytest.org/en/7.1.x/reference/reference.html

But I would personally recommend using an autouse fixture instead: https://docs.pytest.org/en/6.2.x/fixture.html#autouse-fixtures-fixtures-you-don-t-have-to-request

[pablogsal]

I'm adding a new marker in #52 for checking memory leaks. Please @tonybaloney check if this would work for your use case.