Fix deepcopy for BooleanExpression subclasses

Add __deepcopy__ to BooleanExpression that reconstructs the expression
via model_fields, and returns self for Singleton subclasses. Fixes
TypeError from Pydantic v2's BaseModel.__deepcopy__ calling
cls.__new__(cls) with no args on And, Or, and Not which require
positional arguments in __new__.

Author: robreeves

pyiceberg/expressions/__init__.py

@@ -17,6 +17,7 @@
 
 from __future__ import annotations
 
+import copy
 from abc import ABC, abstractmethod
 from collections.abc import Callable, Iterable, Sequence
 from functools import cached_property
@@ -51,6 +52,12 @@ def _to_literal(value: L | Literal[L]) -> Literal[L]:
 class BooleanExpression(IcebergBaseModel, ABC):
     """An expression that evaluates to a boolean."""
 
+    def __deepcopy__(self, memo: dict[int, Any]) -> BooleanExpression:
+        if isinstance(self, Singleton):
+            return self
+        fields = {name: copy.deepcopy(getattr(self, name), memo) for name in type(self).model_fields}
+        return type(self)(**fields)
+
     @abstractmethod
     def __invert__(self) -> BooleanExpression:
         """Transform the Expression into its negated version."""

tests/expressions/test_expressions.py

@@ -1336,6 +1336,18 @@ def test_deepcopy_always_true_then_pickle() -> None:
     assert restored is AlwaysTrue()
 
 
+def test_deepcopy_balanced_and() -> None:
+    expr = And(EqualTo("a", 1), EqualTo("b", 2), EqualTo("c", 3), EqualTo("d", 4))
+    copied = copy.deepcopy(expr)
+    assert copied == expr
+
+
+def test_deepcopy_balanced_or() -> None:
+    expr = Or(EqualTo("a", 1), EqualTo("b", 2), EqualTo("c", 3), EqualTo("d", 4))
+    copied = copy.deepcopy(expr)
+    assert copied == expr
+
+
 def test_deepcopy_nested_expression() -> None:
     expr = And(Or(EqualTo("a", 1), EqualTo("b", 2)), Not(EqualTo("c", 3)))
     copied = copy.deepcopy(expr)