[dataclass_transform] detect transform spec changes in incremental mo…

…de (#14695)

Adds support for triggering rechecking of downstream classes when
`@dataclass_transform` is added or removed from a function/class, as
well as when parameters to `dataclass_transform` are updated. These
changes aren't propagated normally since they don't change the type
signature of the `dataclass_transform` decorator.

Also adds new a new `find-grained-dataclass-transform.test` file to test
the new logic.

mypy/nodes.py

@@ -3907,7 +3907,7 @@ def serialize(self) -> JsonDict:
             "order_default": self.order_default,
             "kw_only_default": self.kw_only_default,
             "frozen_only_default": self.frozen_default,
-            "field_specifiers": self.field_specifiers,
+            "field_specifiers": list(self.field_specifiers),
         }
 
     @classmethod
@@ -3917,7 +3917,7 @@ def deserialize(cls, data: JsonDict) -> DataclassTransformSpec:
             order_default=data.get("order_default"),
             kw_only_default=data.get("kw_only_default"),
             frozen_default=data.get("frozen_default"),
-            field_specifiers=data.get("field_specifiers"),
+            field_specifiers=tuple(data.get("field_specifiers", [])),
         )
 
 

mypy/server/astdiff.py

@@ -73,6 +73,7 @@ class level -- these are handled at attribute level (say, 'mod.Cls.method'
     TypeVarTupleExpr,
     Var,
 )
+from mypy.semanal_shared import find_dataclass_transform_spec
 from mypy.types import (
     AnyType,
     CallableType,
@@ -230,6 +231,7 @@ def snapshot_definition(node: SymbolNode | None, common: SymbolSnapshot) -> Symb
         elif isinstance(node, OverloadedFuncDef) and node.impl:
             impl = node.impl.func if isinstance(node.impl, Decorator) else node.impl
         is_trivial_body = impl.is_trivial_body if impl else False
+        dataclass_transform_spec = find_dataclass_transform_spec(node)
         return (
             "Func",
             common,
@@ -239,6 +241,7 @@ def snapshot_definition(node: SymbolNode | None, common: SymbolSnapshot) -> Symb
             node.is_static,
             signature,
             is_trivial_body,
+            dataclass_transform_spec.serialize() if dataclass_transform_spec is not None else None,
         )
     elif isinstance(node, Var):
         return ("Var", common, snapshot_optional_type(node.type), node.is_final)
@@ -256,6 +259,10 @@ def snapshot_definition(node: SymbolNode | None, common: SymbolSnapshot) -> Symb
             snapshot_definition(node.func, common),
         )
     elif isinstance(node, TypeInfo):
+        dataclass_transform_spec = node.dataclass_transform_spec
+        if dataclass_transform_spec is None:
+            dataclass_transform_spec = find_dataclass_transform_spec(node)
+
         attrs = (
             node.is_abstract,
             node.is_enum,
@@ -280,6 +287,7 @@ def snapshot_definition(node: SymbolNode | None, common: SymbolSnapshot) -> Symb
             tuple(snapshot_type(tdef) for tdef in node.defn.type_vars),
             [snapshot_type(base) for base in node.bases],
             [snapshot_type(p) for p in node._promote],
+            dataclass_transform_spec.serialize() if dataclass_transform_spec is not None else None,
         )
         prefix = node.fullname
         symbol_table = snapshot_symbol_table(prefix, node.names)

test-data/unit/fine-grained-dataclass-transform.test

@@ -0,0 +1,92 @@
+[case updateDataclassTransformParameterViaDecorator]
+# flags: --python-version 3.11
+from m import my_dataclass
+
+@my_dataclass
+class Foo:
+    x: int
+
+foo = Foo(1)
+foo.x = 2
+
+[file m.py]
+from typing import dataclass_transform
+
+@dataclass_transform(frozen_default=False)
+def my_dataclass(cls): return cls
+
+[file m.py.2]
+from typing import dataclass_transform
+
+@dataclass_transform(frozen_default=True)
+def my_dataclass(cls): return cls
+
+[typing fixtures/typing-full.pyi]
+[builtins fixtures/dataclasses.pyi]
+
+[out]
+==
+main:9: error: Property "x" defined in "Foo" is read-only
+
+[case updateDataclassTransformParameterViaParentClass]
+# flags: --python-version 3.11
+from m import Dataclass
+
+class Foo(Dataclass):
+    x: int
+
+foo = Foo(1)
+foo.x = 2
+
+[file m.py]
+from typing import dataclass_transform
+
+@dataclass_transform(frozen_default=False)
+class Dataclass: ...
+
+[file m.py.2]
+from typing import dataclass_transform
+
+@dataclass_transform(frozen_default=True)
+class Dataclass: ...
+
+[typing fixtures/typing-full.pyi]
+[builtins fixtures/dataclasses.pyi]
+
+[out]
+==
+main:8: error: Property "x" defined in "Foo" is read-only
+
+[case updateBaseClassToUseDataclassTransform]
+# flags: --python-version 3.11
+from m import A
+
+class B(A):
+    y: int
+
+B(x=1, y=2)
+
+[file m.py]
+class Dataclass: ...
+
+class A(Dataclass):
+    x: int
+
+[file m.py.2]
+from typing import dataclass_transform
+
+@dataclass_transform()
+class Dataclass: ...
+
+class A(Dataclass):
+    x: int
+
+[typing fixtures/typing-full.pyi]
+[builtins fixtures/dataclasses.pyi]
+
+[out]
+main:7: error: Unexpected keyword argument "x" for "B"
+builtins.pyi:12: note: "B" defined here
+main:7: error: Unexpected keyword argument "y" for "B"
+builtins.pyi:12: note: "B" defined here
+==