Add support for dataclass fields init (#1163)

Co-authored-by: sydney-runkle <[email protected]>
Co-authored-by: Sydney Runkle <[email protected]>
Co-authored-by: David Hewitt <[email protected]>

Author: 3 people

python/pydantic_core/core_schema.py

@@ -2985,6 +2985,7 @@ class DataclassField(TypedDict, total=False):
     name: Required[str]
     schema: Required[CoreSchema]
     kw_only: bool  # default: True
+    init: bool  # default: True
     init_only: bool  # default: False
     frozen: bool  # default: False
     validation_alias: Union[str, List[Union[str, int]], List[List[Union[str, int]]]]
@@ -2998,6 +2999,7 @@ def dataclass_field(
     schema: CoreSchema,
     *,
     kw_only: bool | None = None,
+    init: bool | None = None,
     init_only: bool | None = None,
     validation_alias: str | list[str | int] | list[list[str | int]] | None = None,
     serialization_alias: str | None = None,
@@ -3023,6 +3025,7 @@ def dataclass_field(
         name: The name to use for the argument parameter
         schema: The schema to use for the argument parameter
         kw_only: Whether the field can be set with a positional argument as well as a keyword argument
+        init: Whether the field should be validated during initialization
         init_only: Whether the field should be omitted  from `__dict__` and passed to `__post_init__`
         validation_alias: The alias(es) to use to find the field in the validation data
         serialization_alias: The alias to use as a key when serializing
@@ -3035,6 +3038,7 @@ def dataclass_field(
         name=name,
         schema=schema,
         kw_only=kw_only,
+        init=init,
         init_only=init_only,
         validation_alias=validation_alias,
         serialization_alias=serialization_alias,

src/validators/dataclass.rs

@@ -26,6 +26,7 @@ struct Field {
     kw_only: bool,
     name: String,
     py_name: Py<PyString>,
+    init: bool,
     init_only: bool,
     lookup_key: LookupKey,
     validator: CombinedValidator,
@@ -107,6 +108,7 @@ impl BuildValidator for DataclassArgsValidator {
                 py_name: py_name.into(),
                 lookup_key,
                 validator,
+                init: field.get_as(intern!(py, "init"))?.unwrap_or(true),
                 init_only: field.get_as(intern!(py, "init_only"))?.unwrap_or(false),
                 frozen: field.get_as::<bool>(intern!(py, "frozen"))?.unwrap_or(false),
             });
@@ -176,6 +178,23 @@ impl Validator for DataclassArgsValidator {
                     ($args:ident, $get_method:ident, $get_macro:ident, $slice_macro:ident) => {{
                         // go through fields getting the value from args or kwargs and validating it
                         for (index, field) in self.fields.iter().enumerate() {
+                            if (!field.init) {
+                                match field.validator.default_value(py, Some(field.name.as_str()), state) {
+                                    Ok(Some(value)) => {
+                                        // Default value exists, and passed validation if required
+                                        set_item!(field, value);
+                                    },
+                                    Ok(None) | Err(ValError::Omit) => continue,
+                                    // Note: this will always use the field name even if there is an alias
+                                    // However, we don't mind so much because this error can only happen if the
+                                    // default value fails validation, which is arguably a developer error.
+                                    // We could try to "fix" this in the future if desired.
+                                    Err(ValError::LineErrors(line_errors)) => errors.extend(line_errors),
+                                    Err(err) => return Err(err),
+                                };
+                                continue;
+                            };
+
                             let mut pos_value = None;
                             if let Some(args) = $args.args {
                                 if !field.kw_only {

tests/validators/test_dataclasses.py

@@ -1592,3 +1592,124 @@ def _wrap_validator(cls, v, validator, info):
     gc.collect()
 
     assert ref() is None
+
+
+init_test_cases = [
+    ({'a': 'hello', 'b': 'bye'}, 'ignore', {'a': 'hello', 'b': 'HELLO'}),
+    ({'a': 'hello'}, 'ignore', {'a': 'hello', 'b': 'HELLO'}),
+    # note, for the case below, we don't actually support this case in Pydantic
+    # it's disallowed in Pydantic to have a model with extra='allow' and a field
+    # with init=False, so this case isn't really possible at the momment
+    # however, no conflict arises here because we don't pass in the value for b
+    # to __init__
+    ({'a': 'hello'}, 'allow', {'a': 'hello', 'b': 'HELLO'}),
+    (
+        {'a': 'hello', 'b': 'bye'},
+        'forbid',
+        Err(
+            'Unexpected keyword argument',
+            errors=[
+                {
+                    'type': 'unexpected_keyword_argument',
+                    'loc': ('b',),
+                    'msg': 'Unexpected keyword argument',
+                    'input': 'bye',
+                }
+            ],
+        ),
+    ),
+    ({'a': 'hello'}, 'forbid', {'a': 'hello', 'b': 'HELLO'}),
+]
+
+
+@pytest.mark.parametrize(
+    'input_value,extra_behavior,expected',
+    [
+        *init_test_cases,
+        # special case - when init=False, extra='allow', and the value is provided
+        # currently, it's disallowed in Pydantic to have a model with extra='allow'
+        # and a field with init=False, so this case isn't really possible at the momment
+        # TODO: open to changing this behavior, and changes won't be significantly breaking
+        # because we currently don't support this case
+        ({'a': 'hello', 'b': 'bye'}, 'allow', {'a': 'hello', 'b': 'HELLO'}),
+    ],
+)
+def test_dataclass_args_init(input_value, extra_behavior, expected):
+    @dataclasses.dataclass
+    class Foo:
+        a: str
+        b: str
+
+        def __post_init__(self):
+            self.b = self.a.upper()
+
+    schema = core_schema.dataclass_schema(
+        Foo,
+        core_schema.dataclass_args_schema(
+            'Foo',
+            [
+                core_schema.dataclass_field(name='a', schema=core_schema.str_schema()),
+                core_schema.dataclass_field(name='b', schema=core_schema.str_schema(), init=False),
+            ],
+            extra_behavior=extra_behavior,
+        ),
+        ['a', 'b'],
+        post_init=True,
+    )
+
+    v = SchemaValidator(schema)
+
+    if isinstance(expected, Err):
+        with pytest.raises(ValidationError, match=re.escape(expected.message)) as exc_info:
+            v.validate_python(input_value)
+
+        if expected.errors is not None:
+            assert exc_info.value.errors(include_url=False) == expected.errors
+    else:
+        assert dataclasses.asdict(v.validate_python(input_value)) == expected
+
+
+@pytest.mark.parametrize(
+    'input_value,extra_behavior,expected',
+    [
+        *init_test_cases,
+        # special case - allow override of default, even when init=False, if extra='allow'
+        # TODO: we haven't really decided if this should be allowed or not
+        # currently, it's disallowed in Pydantic to have a model with extra='allow'
+        # and a field with init=False, so this case isn't really possible at the momment
+        ({'a': 'hello', 'b': 'bye'}, 'allow', {'a': 'hello', 'b': 'bye'}),
+    ],
+)
+def test_dataclass_args_init_with_default(input_value, extra_behavior, expected):
+    @dataclasses.dataclass
+    class Foo:
+        a: str
+        b: str
+
+    schema = core_schema.dataclass_schema(
+        Foo,
+        core_schema.dataclass_args_schema(
+            'Foo',
+            [
+                core_schema.dataclass_field(name='a', schema=core_schema.str_schema()),
+                core_schema.dataclass_field(
+                    name='b',
+                    schema=core_schema.with_default_schema(schema=core_schema.str_schema(), default='HELLO'),
+                    init=False,
+                ),
+            ],
+            extra_behavior=extra_behavior,
+        ),
+        ['a', 'b'],
+    )
+
+    v = SchemaValidator(schema)
+
+    if isinstance(expected, Err):
+        with pytest.raises(ValidationError, match=re.escape(expected.message)) as exc_info:
+            v.validate_python(input_value)
+
+        if expected.errors is not None:
+            assert exc_info.value.errors(include_url=False) == expected.errors
+    else:
+        assert dataclasses.asdict(v.validate_python(input_value)) == expected