Make parameterized.expand play nice with underlying mock patches

[alexpizarroj]

Fixes issue #66.

This is a workaround for https://bugs.python.org/issue40126 when using parameterized.expand along old mock versions.

Avoid fully stripping out patches from source methods and replace them with a "dummy" no-op patch instead.

[alexpizarroj]

Greetings, @wolever

I've put together a PR that fixes the long-standing issue #66. I'd love to get some feedback on it / hear your thoughts. 😃

Also, is there a way for me to trigger CI tests for my PR? Nevermind, just managed to run them and they all passed 🎆 : https://app.circleci.com/pipelines/github/alexpizarroj/parameterized/2/workflows/eb15ed9c-5118-443e-a33d-ad70414aa6d9

Thanks in advance.

[alexpizarroj]

Just realized that you may have already implemented something very similar to this workaround, @wolever . I'm talking about 737c1e8. The "hack" in the commit is very similar to mine, though you also introduced changes to the reapplication of patches. Was there any reason to not pursue this further?

[wandering-tales]

👍

[alexpizarroj]

@wolever Please take a look when you can, or perhaps point me to someone else maintaining this project who can help push this forward... It's been a couple of months since I posted and I'd hate for this to get stale again 🙏🏻

[apizarro-paypal]

@wolever Would appreciate some eyes on this, please. It's already been several months 🙏🏻

[abacigalup-eb]

bump, +1

[wolever]

Ah! Thank you very much for the time you've put into this, and I'm extremely sorry it took so long to get merged.

It's in there now, and will be released as part of 0.9.0.

Thank you again!

[wolever]

Hey folks! Just a heads up: 0.9.0 is out: https://pypi.org/project/parameterized/0.9.0/