Karthikeyan Singaravelan <[email protected]> added the comment:
This sounds like a good idea given the docs and the general agreement in the
thread. I would like to target this for 3.9 . Differentiation between the
normal patches and the ones from patch.dict is nice but given that docstring
indicates LIFO for all active patches storing patch.dict items also in
_patch.active_patches ensures we can do LIFO if there is a mixture of normal
and dictionary patches started with start method. Adding other maintainers for
thoughts.
diff --git Lib/unittest/mock.py Lib/unittest/mock.py
index cd5a2aeb60..96115e06ba 100644
--- Lib/unittest/mock.py
+++ Lib/unittest/mock.py
@@ -1853,8 +1853,21 @@ class _patch_dict(object):
self._unpatch_dict()
return False
- start = __enter__
- stop = __exit__
+ def start(self):
+ """Activate a patch, returning any created mock."""
+ result = self.__enter__()
+ _patch._active_patches.append(self)
+ return result
+
+ def stop(self):
+ """Stop an active patch."""
+ try:
+ _patch._active_patches.remove(self)
+ except ValueError:
+ # If the patch hasn't been started this will fail
+ pass
+
+ return self.__exit__()
----------
components: +Library (Lib) -Tests
nosy: +cjw296, lisroach, mariocj89, xtreak
versions: +Python 3.9 -Python 3.5
_______________________________________
Python tracker <[email protected]>
<https://bugs.python.org/issue21600>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
