gh-104231: make str(x) a str, bytes(x) a bytes
#104247
Conversation
| @@ -1050,7 +1050,7 @@ class A: | |||
| def __bytes__(self): | |||
| return OtherBytesSubclass(b'abc') | |||
| self.assertEqual(bytes(A()), b'abc') | |||
| self.assertIs(type(bytes(A())), OtherBytesSubclass) | |||
| self.assertIs(type(bytes(A())), bytes) | |||
There was a problem hiding this comment.
The test being changed here was added in Python 3.4, in 1509580 (by @serhiy-storchaka). The commit was a fix for issue #68919.
(I haven't thought deeply about whether the change here is correct or desirable; just providing context.)
There was a problem hiding this comment.
The test was added to cover all paths, so the behavior change could not be unnoticed. I think that the new behavior is more correct.
But, for consistency with numeric types there should be a warning here. It may be a separate PR.
There was a problem hiding this comment.
@serhiy-storchaka Do you think we should go on with it, or wait for more opinions?
|
Thanks for the ping! I don't have much of an opinion here, and I don't think I'm a good person for review. |
|
Closed in favour of #108814 |
Implicitly converting subclass of
__str__and__bytes__.__init__onstr(x)/bytes(x)when__new__returns an instance ofstr/bytes' subclass #104231