-
Notifications
You must be signed in to change notification settings - Fork 2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[BUG]: enum should not be silently converted to float in method call #5020
Labels
triage
New bug, unverified
Comments
Based on this test code it seems like implicit conversion to int is intended behavior. def test_enum_to_int():
m.test_enum_to_int(m.Flags.Read)
m.test_enum_to_int(m.ClassWithUnscopedEnum.EMode.EFirstMode)
m.test_enum_to_int(m.ScopedCharEnum.Positive)
m.test_enum_to_int(m.ScopedBoolEnum.TRUE)
# ... This is fine, but
|
codeinred
changed the title
[BUG]: enum should NOT be implicitly converted to int, float in method call
[BUG]: enum should not be silectly converted to float in method call
Feb 14, 2024
codeinred
changed the title
[BUG]: enum should not be silectly converted to float in method call
[BUG]: enum should not be silently converted to float in method call
Feb 14, 2024
codeinred
added a commit
to codeinred/pybind11
that referenced
this issue
Feb 14, 2024
Enums (including enum class) are silently converted to floats during method calls. This can cause overload resolution to fail. I believe this behavior is a bug. This commit adds tests for that behavior. These tests fail right now, but they should pass once the behavior is fixed. See: pybind#5020
codeinred
added a commit
to codeinred/pybind11
that referenced
this issue
Feb 14, 2024
Enums (including enum class) are silently converted to floats during method calls. This can cause overload resolution to fail. I believe this behavior is a bug. This commit adds tests for that behavior. These tests fail right now, but they should pass once the behavior is fixed. See: pybind#5020
Hey, I just wanted to check if theres an update on this in regards to disabling the implicit conversion from enum to int since I am running into the same problem where I want two separate constructors: template <typename T>
constructLayer(std::unordered_map<int, py::array_t<T>>& channel_mapping);
template <typename T>
constructLayer(std::unordered_map<Enum::ChannelID, py::array_t<T>>& channel_mapping); but it appears to always call the first ctor even when calling it with these args from python # Passing these to the ctor always calls the int overload
image_dict = {}
image_dict[0] = np.full((height, width), 255, np.uint8)
image_dict[1] = np.full((height, width), 0, np.uint8)
image_dict[2] = np.full((height, width), 0, np.uint8)
image_dict_id = {}
image_dict_id[psapi.enum.ChannelID.red] = np.full((height, width), 255, np.uint8)
image_dict_id[psapi.enum.ChannelID.green] = np.full((height, width), 0, np.uint8)
image_dict_id[psapi.enum.ChannelID.blue] = np.full((height, width), 0, np.uint8) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Required prerequisites
What version (or hash if on master) of pybind11 are you using?
v2.11.1
Problem description
Enums (including
enum class
) will be implicitly converted to ints or floats during method calls. This behavior is surprising and can lead to problems where the wrong function is selected during overload resolution.There should be a way to disable implicit conversion of enums to other types.
Consider the following case:
Calling
f(0, Color.RED)
from python incorrectly calls thef(double, double)
overload. If theColor
enum were not implicitly convertible, this would not occur.Backwards Compatibility Concerns
I understand that there may be concerns about backwards compatibility. If these concerns are significant enough, then there should at least be a way to opt out of implicit conversion for enums
Reproducible example code
Full example code (C++):
Full example code (Python):
Output of python:
Is this a regression? Put the last known working version here if it is.
Not a regression
The text was updated successfully, but these errors were encountered: