[dataclass_transform] minimal implementation of dataclass_transform #14523
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
… dataclasses plugin This is a very simple first step to implementing [PEP 0681](https://peps.python.org/pep-0681/#decorator-function-example), which will allow MyPy to recognize user-defined types that behave similarly to dataclasses. This initial implementation is very limited: we only support decorator-style use of `typing.dataclass_transform` and do not support passing additional options to the transform (such as `freeze` or `init`). Within MyPy, we add a new `is_dataclass_transform` field to `FuncBase` which is populated during semantic analysis. When we check for plugin hooks later, we add new special cases to use the dataclasses plugin if a class decorator is marked with `is_dataclass_transform`. Ideally we would use a proper plugin API; the hacky special case here can be replaced in subsequent iterations.
wesleywright
commented
Jan 25, 2023
| @@ -19,7 +19,7 @@ | |||
| Var, | |||
| ) | |||
| from mypy.plugin import CheckerPluginInterface, ClassDefContext, SemanticAnalyzerPluginInterface | |||
| from mypy.semanal import ALLOW_INCOMPATIBLE_OVERRIDE, set_callable_name | |||
| from mypy.semanal_shared import ALLOW_INCOMPATIBLE_OVERRIDE, set_callable_name | |||
There was a problem hiding this comment.
this move was needed to avoid a cyclic dependency between semanal.py and plugins/common.py
|
I see only |
This comment has been minimized.
This comment has been minimized.
I think this is a gap currently, thanks for the catch (this is my first PR adding this kind of feature haha) |
wesleywright
commented
Jan 25, 2023
Comment on lines
+40
to
+44
| class list(Generic[_T], Sequence[_T]): | ||
| def __contains__(self, item: object) -> int: pass | ||
| def __getitem__(self, key: int) -> _T: pass | ||
| def __iter__(self) -> Iterator[_T]: pass | ||
|
|
There was a problem hiding this comment.
needed to use dataclasses.pyi and typing-full.pyi at the same time in a test case
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
JukkaL
approved these changes
Jan 26, 2023
|
According to mypy_primer, this change has no effect on the checked open source code. 🤖🎉 |
hauntsaninja
approved these changes
Jan 26, 2023
hauntsaninja
pushed a commit
to hauntsaninja/mypy
that referenced
this pull request
Jan 26, 2023
…ython#14523) This is a very simple first step to implementing [PEP 0681](https://peps.python.org/pep-0681/#decorator-function-example), which will allow MyPy to recognize user-defined types that behave similarly to dataclasses. This initial implementation is very limited: we only support decorator-style use of `typing.dataclass_transform` and do not support passing additional options to the transform (such as `freeze` or `init`). Within MyPy, we add a new `is_dataclass_transform` field to `FuncBase` which is populated during semantic analysis. When we check for plugin hooks later, we add new special cases to use the existing dataclasses plugin if a class decorator is marked with `is_dataclass_transform`. Ideally we would use a proper plugin API; the hacky special case here can be replaced in subsequent iterations. Co-authored-by: Wesley Wright <wesleyw@dropbox.com>
hauntsaninja
added a commit
that referenced
this pull request
Jan 29, 2023
…#14532) This is a very simple first step to implementing [PEP 0681](https://peps.python.org/pep-0681/#decorator-function-example), which will allow MyPy to recognize user-defined types that behave similarly to dataclasses. This initial implementation is very limited: we only support decorator-style use of `typing.dataclass_transform` and do not support passing additional options to the transform (such as `freeze` or `init`). Within MyPy, we add a new `is_dataclass_transform` field to `FuncBase` which is populated during semantic analysis. When we check for plugin hooks later, we add new special cases to use the existing dataclasses plugin if a class decorator is marked with `is_dataclass_transform`. Ideally we would use a proper plugin API; the hacky special case here can be replaced in subsequent iterations. Co-authored-by: Wesley Collin Wright <deinlebenandern@gmail.com> Co-authored-by: Wesley Wright <wesleyw@dropbox.com>
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.
This is a very simple first step to implementing PEP 0681, which will allow MyPy to recognize user-defined types that behave similarly to dataclasses.
This initial implementation is very limited: we only support decorator-style use of
typing.dataclass_transformand do not support passing additional options to the transform (such asfreezeorinit).Within MyPy, we add a new
is_dataclass_transformfield toFuncBasewhich is populated during semantic analysis. When we check for plugin hooks later, we add new special cases to use the existing dataclasses plugin if a class decorator is marked withis_dataclass_transform. Ideally we would use a proper plugin API; the hacky special case here can be replaced in subsequent iterations.