feat!: Make columns non-nullable by default

dataframely/_deprecation.py

@@ -2,7 +2,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 
 import os
-import warnings
 from collections.abc import Callable
 from functools import wraps
 
@@ -26,14 +25,3 @@ def wrapper() -> None:
         return wrapper
 
     return decorator
-
-
-@skip_if(env="DATAFRAMELY_NO_FUTURE_WARNINGS")
-def warn_nullable_default_change() -> None:
-    warnings.warn(
-        "The 'nullable' argument was not explicitly set. In a future release, "
-        "'nullable=False' will be the default if 'nullable' is not specified. "
-        "Explicitly set 'nullable=True' if you want your column to be nullable.",
-        FutureWarning,
-        stacklevel=4,
-    )

dataframely/columns/_base.py

@@ -13,9 +13,6 @@
 import polars as pl
 
 from dataframely._compat import pa, sa, sa_TypeEngine
-from dataframely._deprecation import (
-    warn_nullable_default_change,
-)
 from dataframely._polars import PolarsDataType
 from dataframely.random import Generator
 
@@ -45,7 +42,7 @@ class Column(ABC):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         check: Check | None = None,
         alias: str | None = None,
@@ -55,8 +52,6 @@ def __init__(
         Args:
             nullable: Whether this column may contain null values.
                 Explicitly set `nullable=True` if you want your column to be nullable.
-                In a future release, `nullable=False` will be the default if `nullable`
-                is not specified.
             primary_key: Whether this column is part of the primary key of the schema.
                 If ``True``, ``nullable`` is automatically set to ``False``.
             check: A custom rule or multiple rules to run for this column. This can be:
@@ -80,13 +75,6 @@ def __init__(
         if nullable and primary_key:
             raise ValueError("Nullable primary key columns are not supported.")
 
-        if nullable is None:
-            if primary_key:
-                nullable = False
-            else:
-                warn_nullable_default_change()
-                nullable = True
-
         self.nullable = nullable
         self.primary_key = primary_key
         self.check = check

dataframely/columns/categorical.py

@@ -21,7 +21,7 @@ class Categorical(Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         check: Check | None = None,
         alias: str | None = None,

dataframely/columns/datetime.py

@@ -34,7 +34,7 @@ class Date(OrdinalMixin[dt.date], Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         min: dt.date | None = None,
         min_exclusive: dt.date | None = None,
@@ -157,7 +157,7 @@ class Time(OrdinalMixin[dt.time], Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         min: dt.time | None = None,
         min_exclusive: dt.time | None = None,
@@ -286,7 +286,7 @@ class Datetime(OrdinalMixin[dt.datetime], Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         min: dt.datetime | None = None,
         min_exclusive: dt.datetime | None = None,
@@ -433,7 +433,7 @@ class Duration(OrdinalMixin[dt.timedelta], Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         min: dt.timedelta | None = None,
         min_exclusive: dt.timedelta | None = None,

dataframely/columns/decimal.py

@@ -27,7 +27,7 @@ def __init__(
         precision: int | None = None,
         scale: int = 0,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         min: decimal.Decimal | None = None,
         min_exclusive: decimal.Decimal | None = None,

dataframely/columns/enum.py

@@ -26,7 +26,7 @@ def __init__(
         self,
         categories: pl.Series | Iterable[str] | type[enum.Enum],
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         check: Check | None = None,
         alias: str | None = None,

dataframely/columns/float.py

@@ -26,7 +26,7 @@ class _BaseFloat(OrdinalMixin[float], Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         allow_inf_nan: bool = False,
         min: float | None = None,

dataframely/columns/integer.py

@@ -24,7 +24,7 @@ class _BaseInteger(IsInMixin[int], OrdinalMixin[int], Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         min: int | None = None,
         min_exclusive: int | None = None,

dataframely/columns/list.py

@@ -31,7 +31,7 @@ def __init__(
         self,
         inner: Column,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         check: Check | None = None,
         alias: str | None = None,

dataframely/columns/string.py

@@ -22,7 +22,7 @@ class String(Column):
     def __init__(
         self,
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         min_length: int | None = None,
         max_length: int | None = None,

dataframely/columns/struct.py

@@ -29,7 +29,7 @@ def __init__(
         self,
         inner: dict[str, Column],
         *,
-        nullable: bool | None = None,
+        nullable: bool = False,
         primary_key: bool = False,
         check: Check | None = None,
         alias: str | None = None,

tests/column_types/test_array.py

@@ -12,8 +12,8 @@
 @pytest.mark.parametrize(
     "inner",
     [
-        (dy.Int64()),
-        (dy.Integer()),
+        (dy.Int64(nullable=True)),
+        (dy.Integer(nullable=True)),
     ],
 )
 def test_integer_array(inner: Column) -> None:
@@ -29,12 +29,12 @@ def test_integer_array(inner: Column) -> None:
 
 
 def test_invalid_inner_type() -> None:
-    schema = create_schema("test", {"a": dy.Array(dy.Int64(), 1)})
+    schema = create_schema("test", {"a": dy.Array(dy.Int64(nullable=True), 1)})
     assert not schema.is_valid(pl.DataFrame({"a": [["1"], ["2"], ["3"]]}))
 
 
 def test_invalid_shape() -> None:
-    schema = create_schema("test", {"a": dy.Array(dy.Int64(), 2)})
+    schema = create_schema("test", {"a": dy.Array(dy.Int64(nullable=True), 2)})
     assert not schema.is_valid(
         pl.DataFrame(
             {"a": [[1], [2], [3]]},
@@ -49,52 +49,52 @@ def test_invalid_shape() -> None:
     ("column", "dtype", "is_valid"),
     [
         (
-            dy.Array(dy.Int64(), 1),
+            dy.Array(dy.Int64(nullable=True), 1),
             pl.Array(pl.Int64(), 1),
             True,
         ),
         (
-            dy.Array(dy.String(), 1),
+            dy.Array(dy.String(nullable=True), 1),
             pl.Array(pl.Int64(), 1),
             False,
         ),
         (
-            dy.Array(dy.String(), 1),
+            dy.Array(dy.String(nullable=True), 1),
             pl.Array(pl.Int64(), 2),
             False,
         ),
         (
-            dy.Array(dy.Int64(), (1,)),
+            dy.Array(dy.Int64(nullable=True), (1,)),
             pl.Array(pl.Int64(), (1,)),
             True,
         ),
         (
-            dy.Array(dy.Int64(), (1,)),
+            dy.Array(dy.Int64(nullable=True), (1,)),
             pl.Array(pl.Int64(), (2,)),
             False,
         ),
         (
-            dy.Array(dy.String(), 1),
-            dy.Array(dy.String(), 1),
+            dy.Array(dy.String(nullable=True), 1),
+            dy.Array(dy.String(nullable=True), 1),
             False,
         ),
         (
-            dy.Array(dy.String(), 1),
+            dy.Array(dy.String(nullable=True), 1),
             dy.String(),
             False,
         ),
         (
-            dy.Array(dy.String(), 1),
+            dy.Array(dy.String(nullable=True), 1),
             pl.String(),
             False,
         ),
         (
-            dy.Array(dy.Array(dy.String(), 1), 1),
+            dy.Array(dy.Array(dy.String(nullable=True), 1), 1),
             pl.Array(pl.String(), (1, 1)),
             True,
         ),
         (
-            dy.Array(dy.String(), (1, 1)),
+            dy.Array(dy.String(nullable=True), (1, 1)),
             pl.Array(pl.Array(pl.String(), 1), 1),
             True,
         ),
@@ -105,7 +105,9 @@ def test_validate_dtype(column: Column, dtype: pl.DataType, is_valid: bool) -> N
 
 
 def test_nested_arrays() -> None:
-    schema = create_schema("test", {"a": dy.Array(dy.Array(dy.Int64(), 1), 1)})
+    schema = create_schema(
+        "test", {"a": dy.Array(dy.Array(dy.Int64(nullable=True), 1), 1)}
+    )
     assert schema.is_valid(
         pl.DataFrame(
             {"a": [[[1]], [[2]], [[3]]]},
@@ -117,7 +119,9 @@ def test_nested_arrays() -> None:
 
 
 def test_nested_array() -> None:
-    schema = create_schema("test", {"a": dy.Array(dy.Array(dy.Int64(), 1), 1)})
+    schema = create_schema(
+        "test", {"a": dy.Array(dy.Array(dy.Int64(nullable=True), 1), 1)}
+    )
     assert schema.is_valid(
         pl.DataFrame(
             {"a": [[[1]], [[2]], [[3]]]},
@@ -147,7 +151,7 @@ def test_array_with_rules() -> None:
 def test_outer_nullability() -> None:
     schema = create_schema(
         "test",
-        {"nullable": dy.Array(inner=dy.Integer(), shape=1, nullable=True)},
+        {"nullable": dy.Array(inner=dy.Integer(nullable=True), shape=1, nullable=True)},
     )
     df = pl.DataFrame({"nullable": [None, None]})
     schema.validate(df, cast=True)