[stubgen] Improve dataclass init signatures (#18430)

Remove generated incomplete `__init__` signatures for dataclasses. Keep
the field specifiers instead.

Author: cdce8p

mypy/plugins/dataclasses.py

@@ -79,6 +79,8 @@
 
 # The set of decorators that generate dataclasses.
 dataclass_makers: Final = {"dataclass", "dataclasses.dataclass"}
+# Default field specifiers for dataclasses
+DATACLASS_FIELD_SPECIFIERS: Final = ("dataclasses.Field", "dataclasses.field")
 
 
 SELF_TVAR_NAME: Final = "_DT"
@@ -87,7 +89,7 @@
     order_default=False,
     kw_only_default=False,
     frozen_default=False,
-    field_specifiers=("dataclasses.Field", "dataclasses.field"),
+    field_specifiers=DATACLASS_FIELD_SPECIFIERS,
 )
 _INTERNAL_REPLACE_SYM_NAME: Final = "__mypy-replace"
 _INTERNAL_POST_INIT_SYM_NAME: Final = "__mypy-post_init"

mypy/stubgen.py

@@ -95,6 +95,7 @@
     ImportFrom,
     IndexExpr,
     IntExpr,
+    LambdaExpr,
     ListExpr,
     MemberExpr,
     MypyFile,
@@ -113,6 +114,7 @@
     Var,
 )
 from mypy.options import Options as MypyOptions
+from mypy.plugins.dataclasses import DATACLASS_FIELD_SPECIFIERS
 from mypy.semanal_shared import find_dataclass_transform_spec
 from mypy.sharedparse import MAGIC_METHODS_POS_ARGS_ONLY
 from mypy.stubdoc import ArgSig, FunctionSig
@@ -342,11 +344,12 @@ def visit_index_expr(self, node: IndexExpr) -> str:
         base = node.base.accept(self)
         index = node.index.accept(self)
         if len(index) > 2 and index.startswith("(") and index.endswith(")"):
-            index = index[1:-1]
+            index = index[1:-1].rstrip(",")
         return f"{base}[{index}]"
 
     def visit_tuple_expr(self, node: TupleExpr) -> str:
-        return f"({', '.join(n.accept(self) for n in node.items)})"
+        suffix = "," if len(node.items) == 1 else ""
+        return f"({', '.join(n.accept(self) for n in node.items)}{suffix})"
 
     def visit_list_expr(self, node: ListExpr) -> str:
         return f"[{', '.join(n.accept(self) for n in node.items)}]"
@@ -368,6 +371,10 @@ def visit_op_expr(self, o: OpExpr) -> str:
     def visit_star_expr(self, o: StarExpr) -> str:
         return f"*{o.expr.accept(self)}"
 
+    def visit_lambda_expr(self, o: LambdaExpr) -> str:
+        # TODO: Required for among other things dataclass.field default_factory
+        return self.stubgen.add_name("_typeshed.Incomplete")
+
 
 def find_defined_names(file: MypyFile) -> set[str]:
     finder = DefinitionFinder()
@@ -482,6 +489,7 @@ def __init__(
         self.method_names: set[str] = set()
         self.processing_enum = False
         self.processing_dataclass = False
+        self.dataclass_field_specifier: tuple[str, ...] = ()
 
     @property
     def _current_class(self) -> ClassDef | None:
@@ -636,8 +644,8 @@ def visit_func_def(self, o: FuncDef) -> None:
         is_dataclass_generated = (
             self.analyzed and self.processing_dataclass and o.info.names[o.name].plugin_generated
         )
-        if is_dataclass_generated and o.name != "__init__":
-            # Skip methods generated by the @dataclass decorator (except for __init__)
+        if is_dataclass_generated:
+            # Skip methods generated by the @dataclass decorator
             return
         if (
             self.is_private_name(o.name, o.fullname)
@@ -793,8 +801,9 @@ def visit_class_def(self, o: ClassDef) -> None:
             self.add(f"{self._indent}{docstring}\n")
         n = len(self._output)
         self._vars.append([])
-        if self.analyzed and find_dataclass_transform_spec(o):
+        if self.analyzed and (spec := find_dataclass_transform_spec(o)):
             self.processing_dataclass = True
+            self.dataclass_field_specifier = spec.field_specifiers
         super().visit_class_def(o)
         self.dedent()
         self._vars.pop()
@@ -809,6 +818,7 @@ def visit_class_def(self, o: ClassDef) -> None:
             self._state = CLASS
         self.method_names = set()
         self.processing_dataclass = False
+        self.dataclass_field_specifier = ()
         self._class_stack.pop(-1)
         self.processing_enum = False
 
@@ -879,8 +889,9 @@ def is_dataclass_transform(self, expr: Expression) -> bool:
             expr = expr.callee
         if self.get_fullname(expr) in DATACLASS_TRANSFORM_NAMES:
             return True
-        if find_dataclass_transform_spec(expr) is not None:
+        if (spec := find_dataclass_transform_spec(expr)) is not None:
             self.processing_dataclass = True
+            self.dataclass_field_specifier = spec.field_specifiers
             return True
         return False
 
@@ -1259,8 +1270,14 @@ def get_assign_initializer(self, rvalue: Expression) -> str:
             and not isinstance(rvalue, TempNode)
         ):
             return " = ..."
-        if self.processing_dataclass and not (isinstance(rvalue, TempNode) and rvalue.no_rhs):
-            return " = ..."
+        if self.processing_dataclass:
+            if isinstance(rvalue, CallExpr):
+                fullname = self.get_fullname(rvalue.callee)
+                if fullname in (self.dataclass_field_specifier or DATACLASS_FIELD_SPECIFIERS):
+                    p = AliasPrinter(self)
+                    return f" = {rvalue.accept(p)}"
+            if not (isinstance(rvalue, TempNode) and rvalue.no_rhs):
+                return " = ..."
         # TODO: support other possible cases, where initializer is important
 
         # By default, no initializer is required:

test-data/unit/stubgen.test

@@ -3101,15 +3101,14 @@ import attrs
 
 @attrs.define
 class C:
-    x = attrs.field()
+    x: int = attrs.field()
 
 [out]
 import attrs
 
 @attrs.define
 class C:
-    x = ...
-    def __init__(self, x) -> None: ...
+    x: int = attrs.field()
 
 [case testNamedTupleInClass]
 from collections import namedtuple
@@ -4050,8 +4049,9 @@ def i(x=..., y=..., z=...) -> None: ...
 [case testDataclass]
 import dataclasses
 import dataclasses as dcs
-from dataclasses import dataclass, InitVar, KW_ONLY
+from dataclasses import dataclass, field, Field, InitVar, KW_ONLY
 from dataclasses import dataclass as dc
+from datetime import datetime
 from typing import ClassVar
 
 @dataclasses.dataclass
@@ -4066,6 +4066,10 @@ class X:
     h: int = 1
     i: InitVar[str]
     j: InitVar = 100
+    # Lambda not supported yet -> marked as Incomplete instead
+    k: str = Field(
+        default_factory=lambda: datetime.utcnow().isoformat(" ", timespec="seconds")
+    )
     non_field = None
 
 @dcs.dataclass
@@ -4083,7 +4087,8 @@ class V: ...
 [out]
 import dataclasses
 import dataclasses as dcs
-from dataclasses import InitVar, KW_ONLY, dataclass, dataclass as dc
+from _typeshed import Incomplete
+from dataclasses import Field, InitVar, KW_ONLY, dataclass, dataclass as dc, field
 from typing import ClassVar
 
 @dataclasses.dataclass
@@ -4092,12 +4097,13 @@ class X:
     b: str = ...
     c: ClassVar
     d: ClassVar = ...
-    f: list[int] = ...
-    g: int = ...
+    f: list[int] = field(init=False, default_factory=list)
+    g: int = field(default=2, kw_only=True)
     _: KW_ONLY
     h: int = ...
     i: InitVar[str]
     j: InitVar = ...
+    k: str = Field(default_factory=Incomplete)
     non_field = ...
 
 @dcs.dataclass
@@ -4110,8 +4116,9 @@ class W: ...
 class V: ...
 
 [case testDataclass_semanal]
-from dataclasses import InitVar, dataclass, field
+from dataclasses import Field, InitVar, dataclass, field
 from typing import ClassVar
+from datetime import datetime
 
 @dataclass
 class X:
@@ -4125,13 +4132,18 @@ class X:
     h: int = 1
     i: InitVar = 100
     j: list[int] = field(default_factory=list)
+    # Lambda not supported yet -> marked as Incomplete instead
+    k: str = Field(
+        default_factory=lambda: datetime.utcnow().isoformat(" ", timespec="seconds")
+    )
     non_field = None
 
 @dataclass(init=False, repr=False, frozen=True)
 class Y: ...
 
 [out]
-from dataclasses import InitVar, dataclass
+from _typeshed import Incomplete
+from dataclasses import Field, InitVar, dataclass, field
 from typing import ClassVar
 
 @dataclass
@@ -4141,13 +4153,13 @@ class X:
     c: str = ...
     d: ClassVar
     e: ClassVar = ...
-    f: list[int] = ...
-    g: int = ...
+    f: list[int] = field(init=False, default_factory=list)
+    g: int = field(default=2, kw_only=True)
     h: int = ...
     i: InitVar = ...
-    j: list[int] = ...
+    j: list[int] = field(default_factory=list)
+    k: str = Field(default_factory=Incomplete)
     non_field = ...
-    def __init__(self, a, b, c=..., *, g=..., h=..., i=..., j=...) -> None: ...
 
 @dataclass(init=False, repr=False, frozen=True)
 class Y: ...
@@ -4175,7 +4187,7 @@ class X:
 class Y: ...
 
 [out]
-from dataclasses import InitVar, KW_ONLY, dataclass
+from dataclasses import InitVar, KW_ONLY, dataclass, field
 from typing import ClassVar
 
 @dataclass
@@ -4184,14 +4196,13 @@ class X:
     b: str = ...
     c: ClassVar
     d: ClassVar = ...
-    f: list[int] = ...
-    g: int = ...
+    f: list[int] = field(init=False, default_factory=list)
+    g: int = field(default=2, kw_only=True)
     _: KW_ONLY
     h: int = ...
     i: InitVar[str]
     j: InitVar = ...
     non_field = ...
-    def __init__(self, a, b=..., *, g=..., h=..., i, j=...) -> None: ...
 
 @dataclass(init=False, repr=False, frozen=True)
 class Y: ...
@@ -4236,15 +4247,13 @@ from dataclasses import dataclass
 @dataclass
 class X(missing.Base):
     a: int
-    def __init__(self, *generated_args, a, **generated_kwargs) -> None: ...
 
 @dataclass
 class Y(missing.Base):
     generated_args: str
     generated_args_: str
     generated_kwargs: float
     generated_kwargs_: float
-    def __init__(self, *generated_args__, generated_args, generated_args_, generated_kwargs, generated_kwargs_, **generated_kwargs__) -> None: ...
 
 [case testDataclassTransform]
 # dataclass_transform detection only works with sementic analysis.
@@ -4298,6 +4307,7 @@ class Z(metaclass=DCMeta):
 
 [case testDataclassTransformDecorator_semanal]
 import typing_extensions
+from dataclasses import field
 
 @typing_extensions.dataclass_transform(kw_only_default=True)
 def create_model(cls):
@@ -4307,9 +4317,11 @@ def create_model(cls):
 class X:
     a: int
     b: str = "hello"
+    c: bool = field(default=True)
 
 [out]
 import typing_extensions
+from dataclasses import field
 
 @typing_extensions.dataclass_transform(kw_only_default=True)
 def create_model(cls): ...
@@ -4318,9 +4330,10 @@ def create_model(cls): ...
 class X:
     a: int
     b: str = ...
-    def __init__(self, *, a, b=...) -> None: ...
+    c: bool = field(default=True)
 
 [case testDataclassTransformClass_semanal]
+from dataclasses import field
 from typing_extensions import dataclass_transform
 
 @dataclass_transform(kw_only_default=True)
@@ -4329,8 +4342,10 @@ class ModelBase: ...
 class X(ModelBase):
     a: int
     b: str = "hello"
+    c: bool = field(default=True)
 
 [out]
+from dataclasses import field
 from typing_extensions import dataclass_transform
 
 @dataclass_transform(kw_only_default=True)
@@ -4339,28 +4354,42 @@ class ModelBase: ...
 class X(ModelBase):
     a: int
     b: str = ...
-    def __init__(self, *, a, b=...) -> None: ...
+    c: bool = field(default=True)
 
 [case testDataclassTransformMetaclass_semanal]
+from dataclasses import field
+from typing import Any
 from typing_extensions import dataclass_transform
 
-@dataclass_transform(kw_only_default=True)
+def custom_field(*, default: bool, kw_only: bool) -> Any: ...
+
+@dataclass_transform(kw_only_default=True, field_specifiers=(custom_field,))
 class DCMeta(type): ...
 
 class X(metaclass=DCMeta):
     a: int
     b: str = "hello"
+    c: bool = field(default=True)  # should be ignored, not field_specifier here
+
+class Y(X):
+    d: str = custom_field(default="Hello")
 
 [out]
+from typing import Any
 from typing_extensions import dataclass_transform
 
-@dataclass_transform(kw_only_default=True)
+def custom_field(*, default: bool, kw_only: bool) -> Any: ...
+
+@dataclass_transform(kw_only_default=True, field_specifiers=(custom_field,))
 class DCMeta(type): ...
 
 class X(metaclass=DCMeta):
     a: int
     b: str = ...
-    def __init__(self, *, a, b=...) -> None: ...
+    c: bool = ...
+
+class Y(X):
+    d: str = custom_field(default='Hello')
 
 [case testAlwaysUsePEP604Union]
 import typing
@@ -4662,4 +4691,3 @@ class DCMeta(type): ...
 
 class DC(metaclass=DCMeta):
     x: str
-    def __init__(self, x) -> None: ...