fix returning tuples from async fns #4407
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Fixes PyO3#4400 As the return value is ultimately communicated back via a StopIteration exception instance, a peculiar behavior of `PyErr::new` is encountered here: when the argument is a tuple `arg`, it is used to construct the exception as if calling from Python `Exception(*arg)` and not `Exception(arg)` like for every other type of argument. This comes from from CPython's `PyErr_SetObject` which ultimately calls `_PyErr_CreateException` where the "culprit" is found here: https://github.com/python/cpython/blob/main/Python/errors.c#L33 We can fix this particular bug in the invocation of `PyErr::new` but it is a more general question if we want to keep reflecting this somewhat surprising CPython behavior, or create a better API, introducing a breaking change.
LilyFoote
approved these changes
Aug 1, 2024
|
I think the CPython behaviour is a footgun, so I'd lean towards implementing a better API in PyO3. |
davidhewitt
pushed a commit
that referenced
this pull request
Sep 3, 2024
Fixes #4400 As the return value is ultimately communicated back via a StopIteration exception instance, a peculiar behavior of `PyErr::new` is encountered here: when the argument is a tuple `arg`, it is used to construct the exception as if calling from Python `Exception(*arg)` and not `Exception(arg)` like for every other type of argument. This comes from from CPython's `PyErr_SetObject` which ultimately calls `_PyErr_CreateException` where the "culprit" is found here: https://github.com/python/cpython/blob/main/Python/errors.c#L33 We can fix this particular bug in the invocation of `PyErr::new` but it is a more general question if we want to keep reflecting this somewhat surprising CPython behavior, or create a better API, introducing a breaking change.
davidhewitt
pushed a commit
that referenced
this pull request
Sep 3, 2024
Fixes #4400 As the return value is ultimately communicated back via a StopIteration exception instance, a peculiar behavior of `PyErr::new` is encountered here: when the argument is a tuple `arg`, it is used to construct the exception as if calling from Python `Exception(*arg)` and not `Exception(arg)` like for every other type of argument. This comes from from CPython's `PyErr_SetObject` which ultimately calls `_PyErr_CreateException` where the "culprit" is found here: https://github.com/python/cpython/blob/main/Python/errors.c#L33 We can fix this particular bug in the invocation of `PyErr::new` but it is a more general question if we want to keep reflecting this somewhat surprising CPython behavior, or create a better API, introducing a breaking change.
davidhewitt
pushed a commit
that referenced
this pull request
Sep 15, 2024
Fixes #4400 As the return value is ultimately communicated back via a StopIteration exception instance, a peculiar behavior of `PyErr::new` is encountered here: when the argument is a tuple `arg`, it is used to construct the exception as if calling from Python `Exception(*arg)` and not `Exception(arg)` like for every other type of argument. This comes from from CPython's `PyErr_SetObject` which ultimately calls `_PyErr_CreateException` where the "culprit" is found here: https://github.com/python/cpython/blob/main/Python/errors.c#L33 We can fix this particular bug in the invocation of `PyErr::new` but it is a more general question if we want to keep reflecting this somewhat surprising CPython behavior, or create a better API, introducing a breaking change.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes #4400
As the return value is ultimately communicated back via a StopIteration exception instance, a peculiar behavior of
PyErr::newis encountered here: when the argument is a tuplearg, it is used to construct the exception as if calling from PythonException(*arg)and notException(arg)like for every other type of argument.This comes from from CPython's
PyErr_SetObjectwhich ultimately calls_PyErr_CreateExceptionwhere the "culprit" is found here: https://github.com/python/cpython/blob/main/Python/errors.c#L33We can fix this particular bug in the invocation of
PyErr::newbut it is a more general question if we want to keep reflecting this somewhat surprising CPython behavior, or create a better API, introducing a breaking change.