gh-89529: disallow default_factory for fields in dataclasses without __init__

Lib/dataclasses.py

@@ -1009,6 +1009,10 @@ def _process_class(cls, init, repr, eq, order, unsafe_hash, frozen,
             # Otherwise it's a field of some type.
             cls_fields.append(_get_field(cls, name, type, kw_only))
 
+    # Test whether '__init__' is to be auto-generated or if
+    # it is provided explicitly by the user.
+    has_init_method = init or '__init__' in cls.__dict__
+
     for f in cls_fields:
         fields[f.name] = f
 
@@ -1018,6 +1022,15 @@ def _process_class(cls, init, repr, eq, order, unsafe_hash, frozen,
         # sees a real default value, not a Field.
         if isinstance(getattr(cls, f.name, None), Field):
             if f.default is MISSING:
+                # https://github.com/python/cpython/issues/89529
+                if f.default_factory is not MISSING and not has_init_method:
+                    raise ValueError(
+                        f'specifying default_factory for {f.name!r}'
+                        f' requires the @dataclass decorator to be'
+                        f' called with init=True or to implement'
+                        f' an __init__ method'
+                    )
+
                 # If there's no default, delete the class attribute.
                 # This happens if we specify field(repr=False), for
                 # example (that is, we specified a field object, but

Lib/test/test_dataclasses/__init__.py

@@ -9,6 +9,7 @@
 import pickle
 import inspect
 import builtins
+import re
 import types
 import weakref
 import traceback
@@ -18,6 +19,7 @@
 from typing import get_type_hints
 from collections import deque, OrderedDict, namedtuple, defaultdict
 from functools import total_ordering
+from itertools import product
 
 import typing       # Needed for the string "typing.ClassVar[int]" to work as an annotation.
 import dataclasses  # Needed for the string "dataclasses.InitVar[int]" to work as an annotation.
@@ -1411,6 +1413,61 @@ class C:
         C().x
         self.assertEqual(factory.call_count, 2)
 
+    def test_default_factory_with_no_init_method(self):
+        # See https://github.com/python/cpython/issues/89529.
+
+        @dataclass
+        class BaseWithInit:
+            x: list
+
+        @dataclass(slots=True)
+        class BaseWithSlots:
+            x: list
+
+        @dataclass(init=False)
+        class BaseWithOutInit:
+            x: list
+
+        @dataclass(init=False, slots=True)
+        class BaseWithOutInitWithSlots:
+            x: list
+
+        err = re.escape(
+            "specifying default_factory for 'x' requires the "
+            "@dataclass decorator to be called with init=True "
+            "or to implement an __init__ method"
+        )
+
+        for base_class, slots, field_init in product(
+            (object, BaseWithInit, BaseWithSlots,
+             BaseWithOutInit, BaseWithOutInitWithSlots),
+            (True, False),
+            (True, False),
+        ):
+            with self.subTest('generated __init__', base_class=base_class,
+                              init=True, slots=slots, field_init=field_init):
+                @dataclass(init=True, slots=slots)
+                class C(base_class):
+                    x: list = field(init=field_init, default_factory=list)
+                self.assertListEqual(C().x, [])
+
+            with self.subTest('user-defined __init__', base_class=base_class,
+                              init=False, slots=slots, field_init=field_init):
+                @dataclass(init=False, slots=slots)
+                class C(base_class):
+                    x: list = field(init=field_init, default_factory=list)
+                    def __init__(self, *a, **kw):
+                        # deliberately use something else
+                        self.x = 'hello'
+                self.assertEqual(C().x, 'hello')
+
+            with self.subTest('no generated __init__', base_class=base_class,
+                              init=False, slots=slots, field_init=field_init):
+                with self.assertRaisesRegex(ValueError, err):
+                    @dataclass(init=False, slots=slots)
+                    class C(base_class):
+                        x: list = field(init=field_init, default_factory=list)
+
     def test_default_factory_not_called_if_value_given(self):
         # We need a factory that we can test if it's been called.
         factory = Mock()

Misc/NEWS.d/next/Library/2024-08-16-15-35-57.gh-issue-89529.ayfZ1n.rst

@@ -0,0 +1,2 @@
+Disallow ``default_factory`` for dataclass fields if the dataclass does not
+have an ``__init__`` method. Patch by Bénédikt Tran.