GeezerVison is showing github.com — Problems caused by isinstance(list[int], type) returning True · Issue #89828 · python/cpython
Problems caused by isinstance(list[int], type) returning True #89828Description
Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.
Show more details
GitHub fields:
assignee = None
closed_at = None
created_at = <Date 2021-10-29.08:27:43.950>
labels = ['type-bug', 'library', '3.9', '3.10', '3.11']
title = 'Problems caused by isinstance(list[int], type) returning True'
updated_at = <Date 2022-01-06.16:58:47.655>
user = 'https://github.com/serhiy-storchaka' bugs.python.org fields:
activity = <Date 2022-01-06.16:58:47.655>
actor = 'AlexWaygood'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = ['Library (Lib)']
creation = <Date 2021-10-29.08:27:43.950>
creator = 'serhiy.storchaka'
dependencies = ['40296', '45438', '45662', '45663', '45664', '46032']
files = []
hgrepos = []
issue_num = 45665
keywords = []
message_count = 17.0
messages = ['405290', '405291', '405427', '405429', '408207', '409167', '409233', '409679', '409691', '409694', '409697', '409701', '409702', '409703', '409711', '409792', '409881']
nosy_count = 6.0
nosy_names = ['gvanrossum', 'serhiy.storchaka', 'joperez', 'kj', 'AlexWaygood', 'martinitus']
pr_nums = []
priority = 'normal'
resolution = None
stage = None
status = 'open'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue45665'
versions = ['Python 3.9', 'Python 3.10', 'Python 3.11'] Activityserhiy-storchaka commented on Oct 29, 2021 This is a meta-issue for problems caused by isinstance(list[int]) returning True.
See also discussion in bpo-45438. serhiy-storchaka commented on Oct 29, 2021 isinstance(x, type) returns True for instances of types.GenericAlias (like list[int]). While it may help in some cases related to typing, in many unrelated cases it causes problems if the value which is not a type passes checks for types.
Also, isinstance(x, type) been not equal to issubclass(type(x), type) can cause other problems. No matter what the result should be, it should be consistent.
There are many occurrences of isinstance(..., type) in the code.
$ find Lib -name '*.py' \! -path '*/test*' -exec egrep 'isinstance.*,
type\)' '{}' + | wc -l
55And all of them can potentially be broken if pass a types.GenericAlias instance. Many of them are safe, but every case should be analyzed. changed the title [-]Problems caused by isinstance(list[int]) returning True[/-] [+]Problems caused by isinstance(list[int], type) returning True[/+] on Oct 29, 2021changed the title [-]Problems caused by isinstance(list[int]) returning True[/-] [+]Problems caused by isinstance(list[int], type) returning True[/+] on Oct 29, 2021martinitus commented on Nov 1, 2021 One thing that probably should be considered in this context:
isinstance(arg, type) == issubclass(type(arg), type)
Holds True for arg in (Optional[X], Union[X, Y]). Both sides evaluate to False. (v3.10.0)
While I still think both sides evaluating to True would be more intuitive, this supports the proposed change.
Small test snippet:
from typing import Dict, List, Set, Tuple, Optional, Union
import pytest
@pytest.mark.parametrize('arg', [
list, List[int], list[int],
dict, Dict[str, int], dict[str, int],
set, Set[int], set[int],
tuple, Tuple[str, int], tuple[str, int],
Optional[int],
Union[int, str]
])
def test_invariant(arg):
same = isinstance(arg, type) == issubclass(type(arg), type)
result = "Check" if same else "Failed"
print(f"\n{result}: Testing: {arg=} with {type(arg)=}: {isinstance(arg, type)=} <> {issubclass(type(arg), type)=}")
assert same
Any other commonly used annotations that could be added to the checklist? pablogsal commented on Jan 5, 2022 Gentle ping, as the next alpha will be release soon AlexWaygood commented on Jan 6, 2022
We will still allow instantiating e.g. list[int], right?
I certainly hope so! That would be a much more breaking change if we were to change that, and I can't personally see any benefit to doing so. serhiy-storchaka commented on Jun 12, 2022 The cause what that list[int].__class__ returned type instead of types.GenericAlias.
I suppose also that there may be problems caused by relaying other dunder methods. There is a list of exceptions, but it is not complete. typing._GenericAlias in general does not relay dunder methods (except __name__ and __qualname__). johnslavik commented on Jan 17, 2023 At this point I'd like to report the following inconsistency between C PyType_Check(t) and Python's isinstance(t, type) that I've noticed. I don't know if it's fixed at the moment, just so that it doesn't get lost:
from abc import ABCMeta # the C implementation is required for this snippet
class MyABC(metaclass=ABCMeta):
pass
t = list[int]
isinstance(t, type) and issubclass(t, MyABC)causing the following error:
Traceback (most recent call last):
File "...", line 9, in <module>
isinstance(t, type) and issubclass(t, MyABC)
File "/usr/lib/python3.10/abc.py", line 123, in __subclasscheck__
return _abc_subclasscheck(cls, subclass)
TypeError: issubclass() arg 1 must be a class
But if we change abc import to _py_abc,
from _py_abc import ABCMeta # Python implementation
class MyABC(metaclass=ABCMeta):
pass
t = list[int]
isinstance(t, type) and issubclass(t, MyABC) the above absolutely does not produce any error.
This is of course clearly a GenericAlias-related problem. gvanrossum commented on Jan 18, 2023
At this point I'd like to report the following inconsistency between C PyType_Check(t) and Python's isinstance(t, type) that I've noticed. I don't know if it's fixed at the moment, just so that it doesn't get lost:
[...]
What Python version? The most recent version where I can repro that error is 3.10. The fix appears to be in 3.11 and on main (3.12 alpha). johnslavik commented on Jan 19, 2023 Oh, right. Yes, it was 3.10, like in the traceback.
Sorry for not specifying the version and not checking if the bug is not already fixed in versions >3.10.
Also sorry for causing unnecessary chaos.
Thanks and good luck! serhiy-storchaka commented on Jan 19, 2023 There is not much that we can do with this in 3.10. The workaround in the user code is:
isinstance(t, type) and not isinstance(t, types.GenericAlias) and issubclass(t, MyABC) The middle check is not needed in 3.11+.
I think this issue can now be closed. I have not found other sites in the stdlib with user visible differences.
But I have found an interesting difference in the behavior of issubclass(), which should rarely affect user code: #101162.
MetadataMetadataDevelopmentNo branches or pull requestsIssue actions
|