[FIX] tests: fix remaining patcher in python 3.10.12

[Xavier-Do]

While deploying runbot 200-204 an error started to occur on TestEquipement

https://runbot.odoo.com/runbot/build/51753527
And also in the lunch single module
https://runbot.odoo.com/runbot/build/51708103

A patcher was remaining active at the end of TestEquipment, disabling it...

This is caused by a freeze_time on the previous executed class, TestAlarm (actually any class inheriting from odoo.addons.lunch.tests.common.TestsCommon)

It looks like the issue appeared between 3.10.6 and 3.10.12 (maybe because of python/cpython#99699): _class_cleanups is now on the TestCase. This is still unsure since we have a custom version of TestCase, could be a strange interaction.

Could be related to spulec/freezegun#485

Switching the class decorator inside the setupClass fixes the issue

Freezetime code managing decorator on TestCase as a reference:

    def decorate_class(self, klass):
        if issubclass(klass, unittest.TestCase):
            # If it's a TestCase, we assume you want to freeze the time for the
            # tests, from setUpClass to tearDownClass

            # Use getattr as in Python 2.6 they are optional
            orig_setUpClass = getattr(klass, 'setUpClass', None)
            orig_tearDownClass = getattr(klass, 'tearDownClass', None)

            # noinspection PyDecorator
            @classmethod
            def setUpClass(cls):
                self.start()
                if orig_setUpClass is not None:
                    orig_setUpClass()

            # noinspection PyDecorator
            @classmethod
            def tearDownClass(cls):
                if orig_tearDownClass is not None:
                    orig_tearDownClass()
                self.stop()

            klass.setUpClass = setUpClass
            klass.tearDownClass = tearDownClass

            return klass

[robodoo]

[xmo-odoo]

FWIW there's also freeze_time class decorators in mrp (to... today???), account_peppol, purchase_stock, sale, and a bunch of others, which I'd assume are also affected.

edit: semgrep config for catching all them pokemans:

rules:
- id: search
  message: found hits
  severity: INFO
  languages: [python]
  patterns:
  - pattern: |
      @freezegun.freeze_time(...)
      class $C:
        ...

[d-fence]

robodoo r+
The other cases can be handled in other Pr as they don't seem to break the tests.

[Xavier-Do]

FWIW there's also freeze_time class decorators in mrp (to... today???), account_peppol, purchase_stock, sale, and a bunch of others, which I'd assume are also affected.

Yes good catch, it may be a good idea to forbid that since it can be bug prone
Anyway, I think that the issue here is only when the freezetime is on an inherited class this is why we don't have to fix every places.

edit: semgrep config for catching all them pokemans:

rules:
- id: search
  message: found hits
  severity: INFO
  languages: [python]
  patterns:
  - pattern: |
      @freezegun.freeze_time(...)
      class $C:
        ...

Did it with a regexe @freeze_time.+\nclass, really need to dive into semgrep one day :)

[robodoo]

@Xavier-Do @d-fence because this PR has multiple commits, I need to know how to merge it:

• merge to merge directly, using the PR as merge commit message
• rebase-merge to rebase and merge, using the PR as merge commit message
• rebase-ff to rebase and fast-forward

[Xavier-Do]

@robodoo rebase-ff

[robodoo]

Merge method set to rebase and fast-forward.