Update cherrypy==18.9.0

lib/pydantic/_internal/_config.py

@@ -0,0 +1,322 @@
+from __future__ import annotations as _annotations
+
+import warnings
+from contextlib import contextmanager
+from typing import (
+    TYPE_CHECKING,
+    Any,
+    Callable,
+    cast,
+)
+
+from pydantic_core import core_schema
+from typing_extensions import (
+    Literal,
+    Self,
+)
+
+from ..aliases import AliasGenerator
+from ..config import ConfigDict, ExtraValues, JsonDict, JsonEncoder, JsonSchemaExtraCallable
+from ..errors import PydanticUserError
+from ..warnings import PydanticDeprecatedSince20
+
+if not TYPE_CHECKING:
+    # See PyCharm issues https://youtrack.jetbrains.com/issue/PY-21915
+    # and https://youtrack.jetbrains.com/issue/PY-51428
+    DeprecationWarning = PydanticDeprecatedSince20
+
+if TYPE_CHECKING:
+    from .._internal._schema_generation_shared import GenerateSchema
+
+DEPRECATION_MESSAGE = 'Support for class-based `config` is deprecated, use ConfigDict instead.'
+
+
+class ConfigWrapper:
+    """Internal wrapper for Config which exposes ConfigDict items as attributes."""
+
+    __slots__ = ('config_dict',)
+
+    config_dict: ConfigDict
+
+    # all annotations are copied directly from ConfigDict, and should be kept up to date, a test will fail if they
+    # stop matching
+    title: str | None
+    str_to_lower: bool
+    str_to_upper: bool
+    str_strip_whitespace: bool
+    str_min_length: int
+    str_max_length: int | None
+    extra: ExtraValues | None
+    frozen: bool
+    populate_by_name: bool
+    use_enum_values: bool
+    validate_assignment: bool
+    arbitrary_types_allowed: bool
+    from_attributes: bool
+    # whether to use the actual key provided in the data (e.g. alias or first alias for "field required" errors) instead of field_names
+    # to construct error `loc`s, default `True`
+    loc_by_alias: bool
+    alias_generator: Callable[[str], str] | AliasGenerator | None
+    ignored_types: tuple[type, ...]
+    allow_inf_nan: bool
+    json_schema_extra: JsonDict | JsonSchemaExtraCallable | None
+    json_encoders: dict[type[object], JsonEncoder] | None
+
+    # new in V2
+    strict: bool
+    # whether instances of models and dataclasses (including subclass instances) should re-validate, default 'never'
+    revalidate_instances: Literal['always', 'never', 'subclass-instances']
+    ser_json_timedelta: Literal['iso8601', 'float']
+    ser_json_bytes: Literal['utf8', 'base64']
+    ser_json_inf_nan: Literal['null', 'constants']
+    # whether to validate default values during validation, default False
+    validate_default: bool
+    validate_return: bool
+    protected_namespaces: tuple[str, ...]
+    hide_input_in_errors: bool
+    defer_build: bool
+    plugin_settings: dict[str, object] | None
+    schema_generator: type[GenerateSchema] | None
+    json_schema_serialization_defaults_required: bool
+    json_schema_mode_override: Literal['validation', 'serialization', None]
+    coerce_numbers_to_str: bool
+    regex_engine: Literal['rust-regex', 'python-re']
+    validation_error_cause: bool
+
+    def __init__(self, config: ConfigDict | dict[str, Any] | type[Any] | None, *, check: bool = True):
+        if check:
+            self.config_dict = prepare_config(config)
+        else:
+            self.config_dict = cast(ConfigDict, config)
+
+    @classmethod
+    def for_model(cls, bases: tuple[type[Any], ...], namespace: dict[str, Any], kwargs: dict[str, Any]) -> Self:
+        """Build a new `ConfigWrapper` instance for a `BaseModel`.
+
+        The config wrapper built based on (in descending order of priority):
+        - options from `kwargs`
+        - options from the `namespace`
+        - options from the base classes (`bases`)
+
+        Args:
+            bases: A tuple of base classes.
+            namespace: The namespace of the class being created.
+            kwargs: The kwargs passed to the class being created.
+
+        Returns:
+            A `ConfigWrapper` instance for `BaseModel`.
+        """
+        config_new = ConfigDict()
+        for base in bases:
+            config = getattr(base, 'model_config', None)
+            if config:
+                config_new.update(config.copy())
+
+        config_class_from_namespace = namespace.get('Config')
+        config_dict_from_namespace = namespace.get('model_config')
+
+        if config_class_from_namespace and config_dict_from_namespace:
+            raise PydanticUserError('"Config" and "model_config" cannot be used together', code='config-both')
+
+        config_from_namespace = config_dict_from_namespace or prepare_config(config_class_from_namespace)
+
+        config_new.update(config_from_namespace)
+
+        for k in list(kwargs.keys()):
+            if k in config_keys:
+                config_new[k] = kwargs.pop(k)
+
+        return cls(config_new)
+
+    # we don't show `__getattr__` to type checkers so missing attributes cause errors
+    if not TYPE_CHECKING:  # pragma: no branch
+
+        def __getattr__(self, name: str) -> Any:
+            try:
+                return self.config_dict[name]
+            except KeyError:
+                try:
+                    return config_defaults[name]
+                except KeyError:
+                    raise AttributeError(f'Config has no attribute {name!r}') from None
+
+    def core_config(self, obj: Any) -> core_schema.CoreConfig:
+        """Create a pydantic-core config, `obj` is just used to populate `title` if not set in config.
+
+        Pass `obj=None` if you do not want to attempt to infer the `title`.
+
+        We don't use getattr here since we don't want to populate with defaults.
+
+        Args:
+            obj: An object used to populate `title` if not set in config.
+
+        Returns:
+            A `CoreConfig` object created from config.
+        """
+
+        def dict_not_none(**kwargs: Any) -> Any:
+            return {k: v for k, v in kwargs.items() if v is not None}
+
+        core_config = core_schema.CoreConfig(
+            **dict_not_none(
+                title=self.config_dict.get('title') or (obj and obj.__name__),
+                extra_fields_behavior=self.config_dict.get('extra'),
+                allow_inf_nan=self.config_dict.get('allow_inf_nan'),
+                populate_by_name=self.config_dict.get('populate_by_name'),
+                str_strip_whitespace=self.config_dict.get('str_strip_whitespace'),
+                str_to_lower=self.config_dict.get('str_to_lower'),
+                str_to_upper=self.config_dict.get('str_to_upper'),
+                strict=self.config_dict.get('strict'),
+                ser_json_timedelta=self.config_dict.get('ser_json_timedelta'),
+                ser_json_bytes=self.config_dict.get('ser_json_bytes'),
+                ser_json_inf_nan=self.config_dict.get('ser_json_inf_nan'),
+                from_attributes=self.config_dict.get('from_attributes'),
+                loc_by_alias=self.config_dict.get('loc_by_alias'),
+                revalidate_instances=self.config_dict.get('revalidate_instances'),
+                validate_default=self.config_dict.get('validate_default'),
+                str_max_length=self.config_dict.get('str_max_length'),
+                str_min_length=self.config_dict.get('str_min_length'),
+                hide_input_in_errors=self.config_dict.get('hide_input_in_errors'),
+                coerce_numbers_to_str=self.config_dict.get('coerce_numbers_to_str'),
+                regex_engine=self.config_dict.get('regex_engine'),
+                validation_error_cause=self.config_dict.get('validation_error_cause'),
+            )
+        )
+        return core_config
+
+    def __repr__(self):
+        c = ', '.join(f'{k}={v!r}' for k, v in self.config_dict.items())
+        return f'ConfigWrapper({c})'
+
+
+class ConfigWrapperStack:
+    """A stack of `ConfigWrapper` instances."""
+
+    def __init__(self, config_wrapper: ConfigWrapper):
+        self._config_wrapper_stack: list[ConfigWrapper] = [config_wrapper]
+
+    @property
+    def tail(self) -> ConfigWrapper:
+        return self._config_wrapper_stack[-1]
+
+    @contextmanager
+    def push(self, config_wrapper: ConfigWrapper | ConfigDict | None):
+        if config_wrapper is None:
+            yield
+            return
+
+        if not isinstance(config_wrapper, ConfigWrapper):
+            config_wrapper = ConfigWrapper(config_wrapper, check=False)
+
+        self._config_wrapper_stack.append(config_wrapper)
+        try:
+            yield
+        finally:
+            self._config_wrapper_stack.pop()
+
+
+config_defaults = ConfigDict(
+    title=None,
+    str_to_lower=False,
+    str_to_upper=False,
+    str_strip_whitespace=False,
+    str_min_length=0,
+    str_max_length=None,
+    # let the model / dataclass decide how to handle it
+    extra=None,
+    frozen=False,
+    populate_by_name=False,
+    use_enum_values=False,
+    validate_assignment=False,
+    arbitrary_types_allowed=False,
+    from_attributes=False,
+    loc_by_alias=True,
+    alias_generator=None,
+    ignored_types=(),
+    allow_inf_nan=True,
+    json_schema_extra=None,
+    strict=False,
+    revalidate_instances='never',
+    ser_json_timedelta='iso8601',
+    ser_json_bytes='utf8',
+    ser_json_inf_nan='null',
+    validate_default=False,
+    validate_return=False,
+    protected_namespaces=('model_',),
+    hide_input_in_errors=False,
+    json_encoders=None,
+    defer_build=False,
+    plugin_settings=None,
+    schema_generator=None,
+    json_schema_serialization_defaults_required=False,
+    json_schema_mode_override=None,
+    coerce_numbers_to_str=False,
+    regex_engine='rust-regex',
+    validation_error_cause=False,
+)
+
+
+def prepare_config(config: ConfigDict | dict[str, Any] | type[Any] | None) -> ConfigDict:
+    """Create a `ConfigDict` instance from an existing dict, a class (e.g. old class-based config) or None.
+
+    Args:
+        config: The input config.
+
+    Returns:
+        A ConfigDict object created from config.
+    """
+    if config is None:
+        return ConfigDict()
+
+    if not isinstance(config, dict):
+        warnings.warn(DEPRECATION_MESSAGE, DeprecationWarning)
+        config = {k: getattr(config, k) for k in dir(config) if not k.startswith('__')}
+
+    config_dict = cast(ConfigDict, config)
+    check_deprecated(config_dict)
+    return config_dict
+
+
+config_keys = set(ConfigDict.__annotations__.keys())
+
+
+V2_REMOVED_KEYS = {
+    'allow_mutation',
+    'error_msg_templates',
+    'fields',
+    'getter_dict',
+    'smart_union',
+    'underscore_attrs_are_private',
+    'json_loads',
+    'json_dumps',
+    'copy_on_model_validation',
+    'post_init_call',
+}
+V2_RENAMED_KEYS = {
+    'allow_population_by_field_name': 'populate_by_name',
+    'anystr_lower': 'str_to_lower',
+    'anystr_strip_whitespace': 'str_strip_whitespace',
+    'anystr_upper': 'str_to_upper',
+    'keep_untouched': 'ignored_types',
+    'max_anystr_length': 'str_max_length',
+    'min_anystr_length': 'str_min_length',
+    'orm_mode': 'from_attributes',
+    'schema_extra': 'json_schema_extra',
+    'validate_all': 'validate_default',
+}
+
+
+def check_deprecated(config_dict: ConfigDict) -> None:
+    """Check for deprecated config keys and warn the user.
+
+    Args:
+        config_dict: The input config.
+    """
+    deprecated_removed_keys = V2_REMOVED_KEYS & config_dict.keys()
+    deprecated_renamed_keys = V2_RENAMED_KEYS.keys() & config_dict.keys()
+    if deprecated_removed_keys or deprecated_renamed_keys:
+        renamings = {k: V2_RENAMED_KEYS[k] for k in sorted(deprecated_renamed_keys)}
+        renamed_bullets = [f'* {k!r} has been renamed to {v!r}' for k, v in renamings.items()]
+        removed_bullets = [f'* {k!r} has been removed' for k in sorted(deprecated_removed_keys)]
+        message = '\n'.join(['Valid config keys have changed in V2:'] + renamed_bullets + removed_bullets)
+        warnings.warn(message, UserWarning)

lib/pydantic/_internal/_core_metadata.py

@@ -0,0 +1,92 @@
+from __future__ import annotations as _annotations
+
+import typing
+from typing import Any
+
+import typing_extensions
+
+if typing.TYPE_CHECKING:
+    from ._schema_generation_shared import (
+        CoreSchemaOrField as CoreSchemaOrField,
+    )
+    from ._schema_generation_shared import (
+        GetJsonSchemaFunction,
+    )
+
+
+class CoreMetadata(typing_extensions.TypedDict, total=False):
+    """A `TypedDict` for holding the metadata dict of the schema.
+
+    Attributes:
+        pydantic_js_functions: List of JSON schema functions.
+        pydantic_js_prefer_positional_arguments: Whether JSON schema generator will
+            prefer positional over keyword arguments for an 'arguments' schema.
+    """
+
+    pydantic_js_functions: list[GetJsonSchemaFunction]
+    pydantic_js_annotation_functions: list[GetJsonSchemaFunction]
+
+    # If `pydantic_js_prefer_positional_arguments` is True, the JSON schema generator will
+    # prefer positional over keyword arguments for an 'arguments' schema.
+    pydantic_js_prefer_positional_arguments: bool | None
+
+    pydantic_typed_dict_cls: type[Any] | None  # TODO: Consider moving this into the pydantic-core TypedDictSchema
+
+
+class CoreMetadataHandler:
+    """Because the metadata field in pydantic_core is of type `Any`, we can't assume much about its contents.
+
+    This class is used to interact with the metadata field on a CoreSchema object in a consistent
+    way throughout pydantic.
+    """
+
+    __slots__ = ('_schema',)
+
+    def __init__(self, schema: CoreSchemaOrField):
+        self._schema = schema
+
+        metadata = schema.get('metadata')
+        if metadata is None:
+            schema['metadata'] = CoreMetadata()
+        elif not isinstance(metadata, dict):
+            raise TypeError(f'CoreSchema metadata should be a dict; got {metadata!r}.')
+
+    @property
+    def metadata(self) -> CoreMetadata:
+        """Retrieves the metadata dict from the schema, initializing it to a dict if it is None
+        and raises an error if it is not a dict.
+        """
+        metadata = self._schema.get('metadata')
+        if metadata is None:
+            self._schema['metadata'] = metadata = CoreMetadata()
+        if not isinstance(metadata, dict):
+            raise TypeError(f'CoreSchema metadata should be a dict; got {metadata!r}.')
+        return metadata
+
+
+def build_metadata_dict(
+    *,  # force keyword arguments to make it easier to modify this signature in a backwards-compatible way
+    js_functions: list[GetJsonSchemaFunction] | None = None,
+    js_annotation_functions: list[GetJsonSchemaFunction] | None = None,
+    js_prefer_positional_arguments: bool | None = None,
+    typed_dict_cls: type[Any] | None = None,
+    initial_metadata: Any | None = None,
+) -> Any:
+    """Builds a dict to use as the metadata field of a CoreSchema object in a manner that is consistent
+    with the CoreMetadataHandler class.
+    """
+    if initial_metadata is not None and not isinstance(initial_metadata, dict):
+        raise TypeError(f'CoreSchema metadata should be a dict; got {initial_metadata!r}.')
+
+    metadata = CoreMetadata(
+        pydantic_js_functions=js_functions or [],
+        pydantic_js_annotation_functions=js_annotation_functions or [],
+        pydantic_js_prefer_positional_arguments=js_prefer_positional_arguments,
+        pydantic_typed_dict_cls=typed_dict_cls,
+    )
+    metadata = {k: v for k, v in metadata.items() if v is not None}
+
+    if initial_metadata is not None:
+        metadata = {**initial_metadata, **metadata}
+
+    return metadata

lib/pydantic/_internal/_core_utils.py

@@ -0,0 +1,570 @@
+from __future__ import annotations
+
+import os
+from collections import defaultdict
+from typing import (
+    Any,
+    Callable,
+    Hashable,
+    TypeVar,
+    Union,
+)
+
+from pydantic_core import CoreSchema, core_schema
+from pydantic_core import validate_core_schema as _validate_core_schema
+from typing_extensions import TypeAliasType, TypeGuard, get_args, get_origin
+
+from . import _repr
+from ._typing_extra import is_generic_alias
+
+AnyFunctionSchema = Union[
+    core_schema.AfterValidatorFunctionSchema,
+    core_schema.BeforeValidatorFunctionSchema,
+    core_schema.WrapValidatorFunctionSchema,
+    core_schema.PlainValidatorFunctionSchema,
+]
+
+
+FunctionSchemaWithInnerSchema = Union[
+    core_schema.AfterValidatorFunctionSchema,
+    core_schema.BeforeValidatorFunctionSchema,
+    core_schema.WrapValidatorFunctionSchema,
+]
+
+CoreSchemaField = Union[
+    core_schema.ModelField, core_schema.DataclassField, core_schema.TypedDictField, core_schema.ComputedField
+]
+CoreSchemaOrField = Union[core_schema.CoreSchema, CoreSchemaField]
+
+_CORE_SCHEMA_FIELD_TYPES = {'typed-dict-field', 'dataclass-field', 'model-field', 'computed-field'}
+_FUNCTION_WITH_INNER_SCHEMA_TYPES = {'function-before', 'function-after', 'function-wrap'}
+_LIST_LIKE_SCHEMA_WITH_ITEMS_TYPES = {'list', 'set', 'frozenset'}
+
+_DEFINITIONS_CACHE_METADATA_KEY = 'pydantic.definitions_cache'
+
+TAGGED_UNION_TAG_KEY = 'pydantic.internal.tagged_union_tag'
+"""
+Used in a `Tag` schema to specify the tag used for a discriminated union.
+"""
+HAS_INVALID_SCHEMAS_METADATA_KEY = 'pydantic.internal.invalid'
+"""Used to mark a schema that is invalid because it refers to a definition that was not yet defined when the
+schema was first encountered.
+"""
+
+
+def is_core_schema(
+    schema: CoreSchemaOrField,
+) -> TypeGuard[CoreSchema]:
+    return schema['type'] not in _CORE_SCHEMA_FIELD_TYPES
+
+
+def is_core_schema_field(
+    schema: CoreSchemaOrField,
+) -> TypeGuard[CoreSchemaField]:
+    return schema['type'] in _CORE_SCHEMA_FIELD_TYPES
+
+
+def is_function_with_inner_schema(
+    schema: CoreSchemaOrField,
+) -> TypeGuard[FunctionSchemaWithInnerSchema]:
+    return schema['type'] in _FUNCTION_WITH_INNER_SCHEMA_TYPES
+
+
+def is_list_like_schema_with_items_schema(
+    schema: CoreSchema,
+) -> TypeGuard[core_schema.ListSchema | core_schema.SetSchema | core_schema.FrozenSetSchema]:
+    return schema['type'] in _LIST_LIKE_SCHEMA_WITH_ITEMS_TYPES
+
+
+def get_type_ref(type_: type[Any], args_override: tuple[type[Any], ...] | None = None) -> str:
+    """Produces the ref to be used for this type by pydantic_core's core schemas.
+
+    This `args_override` argument was added for the purpose of creating valid recursive references
+    when creating generic models without needing to create a concrete class.
+    """
+    origin = get_origin(type_) or type_
+
+    args = get_args(type_) if is_generic_alias(type_) else (args_override or ())
+    generic_metadata = getattr(type_, '__pydantic_generic_metadata__', None)
+    if generic_metadata:
+        origin = generic_metadata['origin'] or origin
+        args = generic_metadata['args'] or args
+
+    module_name = getattr(origin, '__module__', '<No __module__>')
+    if isinstance(origin, TypeAliasType):
+        type_ref = f'{module_name}.{origin.__name__}:{id(origin)}'
+    else:
+        try:
+            qualname = getattr(origin, '__qualname__', f'<No __qualname__: {origin}>')
+        except Exception:
+            qualname = getattr(origin, '__qualname__', '<No __qualname__>')
+        type_ref = f'{module_name}.{qualname}:{id(origin)}'
+
+    arg_refs: list[str] = []
+    for arg in args:
+        if isinstance(arg, str):
+            # Handle string literals as a special case; we may be able to remove this special handling if we
+            # wrap them in a ForwardRef at some point.
+            arg_ref = f'{arg}:str-{id(arg)}'
+        else:
+            arg_ref = f'{_repr.display_as_type(arg)}:{id(arg)}'
+        arg_refs.append(arg_ref)
+    if arg_refs:
+        type_ref = f'{type_ref}[{",".join(arg_refs)}]'
+    return type_ref
+
+
+def get_ref(s: core_schema.CoreSchema) -> None | str:
+    """Get the ref from the schema if it has one.
+    This exists just for type checking to work correctly.
+    """
+    return s.get('ref', None)
+
+
+def collect_definitions(schema: core_schema.CoreSchema) -> dict[str, core_schema.CoreSchema]:
+    defs: dict[str, CoreSchema] = {}
+
+    def _record_valid_refs(s: core_schema.CoreSchema, recurse: Recurse) -> core_schema.CoreSchema:
+        ref = get_ref(s)
+        if ref:
+            defs[ref] = s
+        return recurse(s, _record_valid_refs)
+
+    walk_core_schema(schema, _record_valid_refs)
+
+    return defs
+
+
+def define_expected_missing_refs(
+    schema: core_schema.CoreSchema, allowed_missing_refs: set[str]
+) -> core_schema.CoreSchema | None:
+    if not allowed_missing_refs:
+        # in this case, there are no missing refs to potentially substitute, so there's no need to walk the schema
+        # this is a common case (will be hit for all non-generic models), so it's worth optimizing for
+        return None
+
+    refs = collect_definitions(schema).keys()
+
+    expected_missing_refs = allowed_missing_refs.difference(refs)
+    if expected_missing_refs:
+        definitions: list[core_schema.CoreSchema] = [
+            # TODO: Replace this with a (new) CoreSchema that, if present at any level, makes validation fail
+            #   Issue: https://github.com/pydantic/pydantic-core/issues/619
+            core_schema.none_schema(ref=ref, metadata={HAS_INVALID_SCHEMAS_METADATA_KEY: True})
+            for ref in expected_missing_refs
+        ]
+        return core_schema.definitions_schema(schema, definitions)
+    return None
+
+
+def collect_invalid_schemas(schema: core_schema.CoreSchema) -> bool:
+    invalid = False
+
+    def _is_schema_valid(s: core_schema.CoreSchema, recurse: Recurse) -> core_schema.CoreSchema:
+        nonlocal invalid
+        if 'metadata' in s:
+            metadata = s['metadata']
+            if HAS_INVALID_SCHEMAS_METADATA_KEY in metadata:
+                invalid = metadata[HAS_INVALID_SCHEMAS_METADATA_KEY]
+                return s
+        return recurse(s, _is_schema_valid)
+
+    walk_core_schema(schema, _is_schema_valid)
+    return invalid
+
+
+T = TypeVar('T')
+
+
+Recurse = Callable[[core_schema.CoreSchema, 'Walk'], core_schema.CoreSchema]
+Walk = Callable[[core_schema.CoreSchema, Recurse], core_schema.CoreSchema]
+
+# TODO: Should we move _WalkCoreSchema into pydantic_core proper?
+#   Issue: https://github.com/pydantic/pydantic-core/issues/615
+
+
+class _WalkCoreSchema:
+    def __init__(self):
+        self._schema_type_to_method = self._build_schema_type_to_method()
+
+    def _build_schema_type_to_method(self) -> dict[core_schema.CoreSchemaType, Recurse]:
+        mapping: dict[core_schema.CoreSchemaType, Recurse] = {}
+        key: core_schema.CoreSchemaType
+        for key in get_args(core_schema.CoreSchemaType):
+            method_name = f"handle_{key.replace('-', '_')}_schema"
+            mapping[key] = getattr(self, method_name, self._handle_other_schemas)
+        return mapping
+
+    def walk(self, schema: core_schema.CoreSchema, f: Walk) -> core_schema.CoreSchema:
+        return f(schema, self._walk)
+
+    def _walk(self, schema: core_schema.CoreSchema, f: Walk) -> core_schema.CoreSchema:
+        schema = self._schema_type_to_method[schema['type']](schema.copy(), f)
+        ser_schema: core_schema.SerSchema | None = schema.get('serialization')  # type: ignore
+        if ser_schema:
+            schema['serialization'] = self._handle_ser_schemas(ser_schema, f)
+        return schema
+
+    def _handle_other_schemas(self, schema: core_schema.CoreSchema, f: Walk) -> core_schema.CoreSchema:
+        sub_schema = schema.get('schema', None)
+        if sub_schema is not None:
+            schema['schema'] = self.walk(sub_schema, f)  # type: ignore
+        return schema
+
+    def _handle_ser_schemas(self, ser_schema: core_schema.SerSchema, f: Walk) -> core_schema.SerSchema:
+        schema: core_schema.CoreSchema | None = ser_schema.get('schema', None)
+        if schema is not None:
+            ser_schema['schema'] = self.walk(schema, f)  # type: ignore
+        return_schema: core_schema.CoreSchema | None = ser_schema.get('return_schema', None)
+        if return_schema is not None:
+            ser_schema['return_schema'] = self.walk(return_schema, f)  # type: ignore
+        return ser_schema
+
+    def handle_definitions_schema(self, schema: core_schema.DefinitionsSchema, f: Walk) -> core_schema.CoreSchema:
+        new_definitions: list[core_schema.CoreSchema] = []
+        for definition in schema['definitions']:
+            if 'schema_ref' in definition and 'ref' in definition:
+                # This indicates a purposely indirect reference
+                # We want to keep such references around for implications related to JSON schema, etc.:
+                new_definitions.append(definition)
+                # However, we still need to walk the referenced definition:
+                self.walk(definition, f)
+                continue
+
+            updated_definition = self.walk(definition, f)
+            if 'ref' in updated_definition:
+                # If the updated definition schema doesn't have a 'ref', it shouldn't go in the definitions
+                # This is most likely to happen due to replacing something with a definition reference, in
+                # which case it should certainly not go in the definitions list
+                new_definitions.append(updated_definition)
+        new_inner_schema = self.walk(schema['schema'], f)
+
+        if not new_definitions and len(schema) == 3:
+            # This means we'd be returning a "trivial" definitions schema that just wrapped the inner schema
+            return new_inner_schema
+
+        new_schema = schema.copy()
+        new_schema['schema'] = new_inner_schema
+        new_schema['definitions'] = new_definitions
+        return new_schema
+
+    def handle_list_schema(self, schema: core_schema.ListSchema, f: Walk) -> core_schema.CoreSchema:
+        items_schema = schema.get('items_schema')
+        if items_schema is not None:
+            schema['items_schema'] = self.walk(items_schema, f)
+        return schema
+
+    def handle_set_schema(self, schema: core_schema.SetSchema, f: Walk) -> core_schema.CoreSchema:
+        items_schema = schema.get('items_schema')
+        if items_schema is not None:
+            schema['items_schema'] = self.walk(items_schema, f)
+        return schema
+
+    def handle_frozenset_schema(self, schema: core_schema.FrozenSetSchema, f: Walk) -> core_schema.CoreSchema:
+        items_schema = schema.get('items_schema')
+        if items_schema is not None:
+            schema['items_schema'] = self.walk(items_schema, f)
+        return schema
+
+    def handle_generator_schema(self, schema: core_schema.GeneratorSchema, f: Walk) -> core_schema.CoreSchema:
+        items_schema = schema.get('items_schema')
+        if items_schema is not None:
+            schema['items_schema'] = self.walk(items_schema, f)
+        return schema
+
+    def handle_tuple_schema(self, schema: core_schema.TupleSchema, f: Walk) -> core_schema.CoreSchema:
+        schema['items_schema'] = [self.walk(v, f) for v in schema['items_schema']]
+        return schema
+
+    def handle_dict_schema(self, schema: core_schema.DictSchema, f: Walk) -> core_schema.CoreSchema:
+        keys_schema = schema.get('keys_schema')
+        if keys_schema is not None:
+            schema['keys_schema'] = self.walk(keys_schema, f)
+        values_schema = schema.get('values_schema')
+        if values_schema:
+            schema['values_schema'] = self.walk(values_schema, f)
+        return schema
+
+    def handle_function_schema(self, schema: AnyFunctionSchema, f: Walk) -> core_schema.CoreSchema:
+        if not is_function_with_inner_schema(schema):
+            return schema
+        schema['schema'] = self.walk(schema['schema'], f)
+        return schema
+
+    def handle_union_schema(self, schema: core_schema.UnionSchema, f: Walk) -> core_schema.CoreSchema:
+        new_choices: list[CoreSchema | tuple[CoreSchema, str]] = []
+        for v in schema['choices']:
+            if isinstance(v, tuple):
+                new_choices.append((self.walk(v[0], f), v[1]))
+            else:
+                new_choices.append(self.walk(v, f))
+        schema['choices'] = new_choices
+        return schema
+
+    def handle_tagged_union_schema(self, schema: core_schema.TaggedUnionSchema, f: Walk) -> core_schema.CoreSchema:
+        new_choices: dict[Hashable, core_schema.CoreSchema] = {}
+        for k, v in schema['choices'].items():
+            new_choices[k] = v if isinstance(v, (str, int)) else self.walk(v, f)
+        schema['choices'] = new_choices
+        return schema
+
+    def handle_chain_schema(self, schema: core_schema.ChainSchema, f: Walk) -> core_schema.CoreSchema:
+        schema['steps'] = [self.walk(v, f) for v in schema['steps']]
+        return schema
+
+    def handle_lax_or_strict_schema(self, schema: core_schema.LaxOrStrictSchema, f: Walk) -> core_schema.CoreSchema:
+        schema['lax_schema'] = self.walk(schema['lax_schema'], f)
+        schema['strict_schema'] = self.walk(schema['strict_schema'], f)
+        return schema
+
+    def handle_json_or_python_schema(self, schema: core_schema.JsonOrPythonSchema, f: Walk) -> core_schema.CoreSchema:
+        schema['json_schema'] = self.walk(schema['json_schema'], f)
+        schema['python_schema'] = self.walk(schema['python_schema'], f)
+        return schema
+
+    def handle_model_fields_schema(self, schema: core_schema.ModelFieldsSchema, f: Walk) -> core_schema.CoreSchema:
+        extras_schema = schema.get('extras_schema')
+        if extras_schema is not None:
+            schema['extras_schema'] = self.walk(extras_schema, f)
+        replaced_fields: dict[str, core_schema.ModelField] = {}
+        replaced_computed_fields: list[core_schema.ComputedField] = []
+        for computed_field in schema.get('computed_fields', ()):
+            replaced_field = computed_field.copy()
+            replaced_field['return_schema'] = self.walk(computed_field['return_schema'], f)
+            replaced_computed_fields.append(replaced_field)
+        if replaced_computed_fields:
+            schema['computed_fields'] = replaced_computed_fields
+        for k, v in schema['fields'].items():
+            replaced_field = v.copy()
+            replaced_field['schema'] = self.walk(v['schema'], f)
+            replaced_fields[k] = replaced_field
+        schema['fields'] = replaced_fields
+        return schema
+
+    def handle_typed_dict_schema(self, schema: core_schema.TypedDictSchema, f: Walk) -> core_schema.CoreSchema:
+        extras_schema = schema.get('extras_schema')
+        if extras_schema is not None:
+            schema['extras_schema'] = self.walk(extras_schema, f)
+        replaced_computed_fields: list[core_schema.ComputedField] = []
+        for computed_field in schema.get('computed_fields', ()):
+            replaced_field = computed_field.copy()
+            replaced_field['return_schema'] = self.walk(computed_field['return_schema'], f)
+            replaced_computed_fields.append(replaced_field)
+        if replaced_computed_fields:
+            schema['computed_fields'] = replaced_computed_fields
+        replaced_fields: dict[str, core_schema.TypedDictField] = {}
+        for k, v in schema['fields'].items():
+            replaced_field = v.copy()
+            replaced_field['schema'] = self.walk(v['schema'], f)
+            replaced_fields[k] = replaced_field
+        schema['fields'] = replaced_fields
+        return schema
+
+    def handle_dataclass_args_schema(self, schema: core_schema.DataclassArgsSchema, f: Walk) -> core_schema.CoreSchema:
+        replaced_fields: list[core_schema.DataclassField] = []
+        replaced_computed_fields: list[core_schema.ComputedField] = []
+        for computed_field in schema.get('computed_fields', ()):
+            replaced_field = computed_field.copy()
+            replaced_field['return_schema'] = self.walk(computed_field['return_schema'], f)
+            replaced_computed_fields.append(replaced_field)
+        if replaced_computed_fields:
+            schema['computed_fields'] = replaced_computed_fields
+        for field in schema['fields']:
+            replaced_field = field.copy()
+            replaced_field['schema'] = self.walk(field['schema'], f)
+            replaced_fields.append(replaced_field)
+        schema['fields'] = replaced_fields
+        return schema
+
+    def handle_arguments_schema(self, schema: core_schema.ArgumentsSchema, f: Walk) -> core_schema.CoreSchema:
+        replaced_arguments_schema: list[core_schema.ArgumentsParameter] = []
+        for param in schema['arguments_schema']:
+            replaced_param = param.copy()
+            replaced_param['schema'] = self.walk(param['schema'], f)
+            replaced_arguments_schema.append(replaced_param)
+        schema['arguments_schema'] = replaced_arguments_schema
+        if 'var_args_schema' in schema:
+            schema['var_args_schema'] = self.walk(schema['var_args_schema'], f)
+        if 'var_kwargs_schema' in schema:
+            schema['var_kwargs_schema'] = self.walk(schema['var_kwargs_schema'], f)
+        return schema
+
+    def handle_call_schema(self, schema: core_schema.CallSchema, f: Walk) -> core_schema.CoreSchema:
+        schema['arguments_schema'] = self.walk(schema['arguments_schema'], f)
+        if 'return_schema' in schema:
+            schema['return_schema'] = self.walk(schema['return_schema'], f)
+        return schema
+
+
+_dispatch = _WalkCoreSchema().walk
+
+
+def walk_core_schema(schema: core_schema.CoreSchema, f: Walk) -> core_schema.CoreSchema:
+    """Recursively traverse a CoreSchema.
+
+    Args:
+        schema (core_schema.CoreSchema): The CoreSchema to process, it will not be modified.
+        f (Walk): A function to apply. This function takes two arguments:
+          1. The current CoreSchema that is being processed
+             (not the same one you passed into this function, one level down).
+          2. The "next" `f` to call. This lets you for example use `f=functools.partial(some_method, some_context)`
+             to pass data down the recursive calls without using globals or other mutable state.
+
+    Returns:
+        core_schema.CoreSchema: A processed CoreSchema.
+    """
+    return f(schema.copy(), _dispatch)
+
+
+def simplify_schema_references(schema: core_schema.CoreSchema) -> core_schema.CoreSchema:  # noqa: C901
+    definitions: dict[str, core_schema.CoreSchema] = {}
+    ref_counts: dict[str, int] = defaultdict(int)
+    involved_in_recursion: dict[str, bool] = {}
+    current_recursion_ref_count: dict[str, int] = defaultdict(int)
+
+    def collect_refs(s: core_schema.CoreSchema, recurse: Recurse) -> core_schema.CoreSchema:
+        if s['type'] == 'definitions':
+            for definition in s['definitions']:
+                ref = get_ref(definition)
+                assert ref is not None
+                if ref not in definitions:
+                    definitions[ref] = definition
+                recurse(definition, collect_refs)
+            return recurse(s['schema'], collect_refs)
+        else:
+            ref = get_ref(s)
+            if ref is not None:
+                new = recurse(s, collect_refs)
+                new_ref = get_ref(new)
+                if new_ref:
+                    definitions[new_ref] = new
+                return core_schema.definition_reference_schema(schema_ref=ref)
+            else:
+                return recurse(s, collect_refs)
+
+    schema = walk_core_schema(schema, collect_refs)
+
+    def count_refs(s: core_schema.CoreSchema, recurse: Recurse) -> core_schema.CoreSchema:
+        if s['type'] != 'definition-ref':
+            return recurse(s, count_refs)
+        ref = s['schema_ref']
+        ref_counts[ref] += 1
+
+        if ref_counts[ref] >= 2:
+            # If this model is involved in a recursion this should be detected
+            # on its second encounter, we can safely stop the walk here.
+            if current_recursion_ref_count[ref] != 0:
+                involved_in_recursion[ref] = True
+            return s
+
+        current_recursion_ref_count[ref] += 1
+        recurse(definitions[ref], count_refs)
+        current_recursion_ref_count[ref] -= 1
+        return s
+
+    schema = walk_core_schema(schema, count_refs)
+
+    assert all(c == 0 for c in current_recursion_ref_count.values()), 'this is a bug! please report it'
+
+    def can_be_inlined(s: core_schema.DefinitionReferenceSchema, ref: str) -> bool:
+        if ref_counts[ref] > 1:
+            return False
+        if involved_in_recursion.get(ref, False):
+            return False
+        if 'serialization' in s:
+            return False
+        if 'metadata' in s:
+            metadata = s['metadata']
+            for k in (
+                'pydantic_js_functions',
+                'pydantic_js_annotation_functions',
+                'pydantic.internal.union_discriminator',
+            ):
+                if k in metadata:
+                    # we need to keep this as a ref
+                    return False
+        return True
+
+    def inline_refs(s: core_schema.CoreSchema, recurse: Recurse) -> core_schema.CoreSchema:
+        if s['type'] == 'definition-ref':
+            ref = s['schema_ref']
+            # Check if the reference is only used once, not involved in recursion and does not have
+            # any extra keys (like 'serialization')
+            if can_be_inlined(s, ref):
+                # Inline the reference by replacing the reference with the actual schema
+                new = definitions.pop(ref)
+                ref_counts[ref] -= 1  # because we just replaced it!
+                # put all other keys that were on the def-ref schema into the inlined version
+                # in particular this is needed for `serialization`
+                if 'serialization' in s:
+                    new['serialization'] = s['serialization']
+                s = recurse(new, inline_refs)
+                return s
+            else:
+                return recurse(s, inline_refs)
+        else:
+            return recurse(s, inline_refs)
+
+    schema = walk_core_schema(schema, inline_refs)
+
+    def_values = [v for v in definitions.values() if ref_counts[v['ref']] > 0]  # type: ignore
+
+    if def_values:
+        schema = core_schema.definitions_schema(schema=schema, definitions=def_values)
+    return schema
+
+
+def _strip_metadata(schema: CoreSchema) -> CoreSchema:
+    def strip_metadata(s: CoreSchema, recurse: Recurse) -> CoreSchema:
+        s = s.copy()
+        s.pop('metadata', None)
+        if s['type'] == 'model-fields':
+            s = s.copy()
+            s['fields'] = {k: v.copy() for k, v in s['fields'].items()}
+            for field_name, field_schema in s['fields'].items():
+                field_schema.pop('metadata', None)
+                s['fields'][field_name] = field_schema
+            computed_fields = s.get('computed_fields', None)
+            if computed_fields:
+                s['computed_fields'] = [cf.copy() for cf in computed_fields]
+                for cf in computed_fields:
+                    cf.pop('metadata', None)
+            else:
+                s.pop('computed_fields', None)
+        elif s['type'] == 'model':
+            # remove some defaults
+            if s.get('custom_init', True) is False:
+                s.pop('custom_init')
+            if s.get('root_model', True) is False:
+                s.pop('root_model')
+            if {'title'}.issuperset(s.get('config', {}).keys()):
+                s.pop('config', None)
+
+        return recurse(s, strip_metadata)
+
+    return walk_core_schema(schema, strip_metadata)
+
+
+def pretty_print_core_schema(
+    schema: CoreSchema,
+    include_metadata: bool = False,
+) -> None:
+    """Pretty print a CoreSchema using rich.
+    This is intended for debugging purposes.
+
+    Args:
+        schema: The CoreSchema to print.
+        include_metadata: Whether to include metadata in the output. Defaults to `False`.
+    """
+    from rich import print  # type: ignore  # install it manually in your dev env
+
+    if not include_metadata:
+        schema = _strip_metadata(schema)
+
+    return print(schema)
+
+
+def validate_core_schema(schema: CoreSchema) -> CoreSchema:
+    if 'PYDANTIC_SKIP_VALIDATING_CORE_SCHEMAS' in os.environ:
+        return schema
+    return _validate_core_schema(schema)

lib/pydantic/_internal/_dataclasses.py

@@ -0,0 +1,225 @@
+"""Private logic for creating pydantic dataclasses."""
+from __future__ import annotations as _annotations
+
+import dataclasses
+import typing
+import warnings
+from functools import partial, wraps
+from typing import Any, Callable, ClassVar
+
+from pydantic_core import (
+    ArgsKwargs,
+    SchemaSerializer,
+    SchemaValidator,
+    core_schema,
+)
+from typing_extensions import TypeGuard
+
+from ..errors import PydanticUndefinedAnnotation
+from ..fields import FieldInfo
+from ..plugin._schema_validator import create_schema_validator
+from ..warnings import PydanticDeprecatedSince20
+from . import _config, _decorators, _typing_extra
+from ._fields import collect_dataclass_fields
+from ._generate_schema import GenerateSchema
+from ._generics import get_standard_typevars_map
+from ._mock_val_ser import set_dataclass_mocks
+from ._schema_generation_shared import CallbackGetCoreSchemaHandler
+from ._signature import generate_pydantic_signature
+
+if typing.TYPE_CHECKING:
+    from ..config import ConfigDict
+
+    class StandardDataclass(typing.Protocol):
+        __dataclass_fields__: ClassVar[dict[str, Any]]
+        __dataclass_params__: ClassVar[Any]  # in reality `dataclasses._DataclassParams`
+        __post_init__: ClassVar[Callable[..., None]]
+
+        def __init__(self, *args: object, **kwargs: object) -> None:
+            pass
+
+    class PydanticDataclass(StandardDataclass, typing.Protocol):
+        """A protocol containing attributes only available once a class has been decorated as a Pydantic dataclass.
+
+        Attributes:
+            __pydantic_config__: Pydantic-specific configuration settings for the dataclass.
+            __pydantic_complete__: Whether dataclass building is completed, or if there are still undefined fields.
+            __pydantic_core_schema__: The pydantic-core schema used to build the SchemaValidator and SchemaSerializer.
+            __pydantic_decorators__: Metadata containing the decorators defined on the dataclass.
+            __pydantic_fields__: Metadata about the fields defined on the dataclass.
+            __pydantic_serializer__: The pydantic-core SchemaSerializer used to dump instances of the dataclass.
+            __pydantic_validator__: The pydantic-core SchemaValidator used to validate instances of the dataclass.
+        """
+
+        __pydantic_config__: ClassVar[ConfigDict]
+        __pydantic_complete__: ClassVar[bool]
+        __pydantic_core_schema__: ClassVar[core_schema.CoreSchema]
+        __pydantic_decorators__: ClassVar[_decorators.DecoratorInfos]
+        __pydantic_fields__: ClassVar[dict[str, FieldInfo]]
+        __pydantic_serializer__: ClassVar[SchemaSerializer]
+        __pydantic_validator__: ClassVar[SchemaValidator]
+
+else:
+    # See PyCharm issues https://youtrack.jetbrains.com/issue/PY-21915
+    # and https://youtrack.jetbrains.com/issue/PY-51428
+    DeprecationWarning = PydanticDeprecatedSince20
+
+
+def set_dataclass_fields(cls: type[StandardDataclass], types_namespace: dict[str, Any] | None = None) -> None:
+    """Collect and set `cls.__pydantic_fields__`.
+
+    Args:
+        cls: The class.
+        types_namespace: The types namespace, defaults to `None`.
+    """
+    typevars_map = get_standard_typevars_map(cls)
+    fields = collect_dataclass_fields(cls, types_namespace, typevars_map=typevars_map)
+
+    cls.__pydantic_fields__ = fields  # type: ignore
+
+
+def complete_dataclass(
+    cls: type[Any],
+    config_wrapper: _config.ConfigWrapper,
+    *,
+    raise_errors: bool = True,
+    types_namespace: dict[str, Any] | None,
+) -> bool:
+    """Finish building a pydantic dataclass.
+
+    This logic is called on a class which has already been wrapped in `dataclasses.dataclass()`.
+
+    This is somewhat analogous to `pydantic._internal._model_construction.complete_model_class`.
+
+    Args:
+        cls: The class.
+        config_wrapper: The config wrapper instance.
+        raise_errors: Whether to raise errors, defaults to `True`.
+        types_namespace: The types namespace.
+
+    Returns:
+        `True` if building a pydantic dataclass is successfully completed, `False` otherwise.
+
+    Raises:
+        PydanticUndefinedAnnotation: If `raise_error` is `True` and there is an undefined annotations.
+    """
+    if hasattr(cls, '__post_init_post_parse__'):
+        warnings.warn(
+            'Support for `__post_init_post_parse__` has been dropped, the method will not be called', DeprecationWarning
+        )
+
+    if types_namespace is None:
+        types_namespace = _typing_extra.get_cls_types_namespace(cls)
+
+    set_dataclass_fields(cls, types_namespace)
+
+    typevars_map = get_standard_typevars_map(cls)
+    gen_schema = GenerateSchema(
+        config_wrapper,
+        types_namespace,
+        typevars_map,
+    )
+
+    # This needs to be called before we change the __init__
+    sig = generate_pydantic_signature(
+        init=cls.__init__,
+        fields=cls.__pydantic_fields__,  # type: ignore
+        config_wrapper=config_wrapper,
+        is_dataclass=True,
+    )
+
+    # dataclass.__init__ must be defined here so its `__qualname__` can be changed since functions can't be copied.
+    def __init__(__dataclass_self__: PydanticDataclass, *args: Any, **kwargs: Any) -> None:
+        __tracebackhide__ = True
+        s = __dataclass_self__
+        s.__pydantic_validator__.validate_python(ArgsKwargs(args, kwargs), self_instance=s)
+
+    __init__.__qualname__ = f'{cls.__qualname__}.__init__'
+
+    cls.__init__ = __init__  # type: ignore
+    cls.__pydantic_config__ = config_wrapper.config_dict  # type: ignore
+    cls.__signature__ = sig  # type: ignore
+    get_core_schema = getattr(cls, '__get_pydantic_core_schema__', None)
+    try:
+        if get_core_schema:
+            schema = get_core_schema(
+                cls,
+                CallbackGetCoreSchemaHandler(
+                    partial(gen_schema.generate_schema, from_dunder_get_core_schema=False),
+                    gen_schema,
+                    ref_mode='unpack',
+                ),
+            )
+        else:
+            schema = gen_schema.generate_schema(cls, from_dunder_get_core_schema=False)
+    except PydanticUndefinedAnnotation as e:
+        if raise_errors:
+            raise
+        set_dataclass_mocks(cls, cls.__name__, f'`{e.name}`')
+        return False
+
+    core_config = config_wrapper.core_config(cls)
+
+    try:
+        schema = gen_schema.clean_schema(schema)
+    except gen_schema.CollectedInvalid:
+        set_dataclass_mocks(cls, cls.__name__, 'all referenced types')
+        return False
+
+    # We are about to set all the remaining required properties expected for this cast;
+    # __pydantic_decorators__ and __pydantic_fields__ should already be set
+    cls = typing.cast('type[PydanticDataclass]', cls)
+    # debug(schema)
+
+    cls.__pydantic_core_schema__ = schema
+    cls.__pydantic_validator__ = validator = create_schema_validator(
+        schema, cls, cls.__module__, cls.__qualname__, 'dataclass', core_config, config_wrapper.plugin_settings
+    )
+    cls.__pydantic_serializer__ = SchemaSerializer(schema, core_config)
+
+    if config_wrapper.validate_assignment:
+
+        @wraps(cls.__setattr__)
+        def validated_setattr(instance: Any, __field: str, __value: str) -> None:
+            validator.validate_assignment(instance, __field, __value)
+
+        cls.__setattr__ = validated_setattr.__get__(None, cls)  # type: ignore
+
+    return True
+
+
+def is_builtin_dataclass(_cls: type[Any]) -> TypeGuard[type[StandardDataclass]]:
+    """Returns True if a class is a stdlib dataclass and *not* a pydantic dataclass.
+
+    We check that
+    - `_cls` is a dataclass
+    - `_cls` does not inherit from a processed pydantic dataclass (and thus have a `__pydantic_validator__`)
+    - `_cls` does not have any annotations that are not dataclass fields
+    e.g.
+    ```py
+    import dataclasses
+
+    import pydantic.dataclasses
+
+    @dataclasses.dataclass
+    class A:
+        x: int
+
+    @pydantic.dataclasses.dataclass
+    class B(A):
+        y: int
+    ```
+    In this case, when we first check `B`, we make an extra check and look at the annotations ('y'),
+    which won't be a superset of all the dataclass fields (only the stdlib fields i.e. 'x')
+
+    Args:
+        cls: The class.
+
+    Returns:
+        `True` if the class is a stdlib dataclass, `False` otherwise.
+    """
+    return (
+        dataclasses.is_dataclass(_cls)
+        and not hasattr(_cls, '__pydantic_validator__')
+        and set(_cls.__dataclass_fields__).issuperset(set(getattr(_cls, '__annotations__', {})))
+    )

lib/pydantic/_internal/_decorators.py

@@ -0,0 +1,791 @@
+"""Logic related to validators applied to models etc. via the `@field_validator` and `@model_validator` decorators."""
+from __future__ import annotations as _annotations
+
+from collections import deque
+from dataclasses import dataclass, field
+from functools import cached_property, partial, partialmethod
+from inspect import Parameter, Signature, isdatadescriptor, ismethoddescriptor, signature
+from itertools import islice
+from typing import TYPE_CHECKING, Any, Callable, ClassVar, Generic, Iterable, TypeVar, Union
+
+from pydantic_core import PydanticUndefined, core_schema
+from typing_extensions import Literal, TypeAlias, is_typeddict
+
+from ..errors import PydanticUserError
+from ._core_utils import get_type_ref
+from ._internal_dataclass import slots_true
+from ._typing_extra import get_function_type_hints
+
+if TYPE_CHECKING:
+    from ..fields import ComputedFieldInfo
+    from ..functional_validators import FieldValidatorModes
+
+
+@dataclass(**slots_true)
+class ValidatorDecoratorInfo:
+    """A container for data from `@validator` so that we can access it
+    while building the pydantic-core schema.
+
+    Attributes:
+        decorator_repr: A class variable representing the decorator string, '@validator'.
+        fields: A tuple of field names the validator should be called on.
+        mode: The proposed validator mode.
+        each_item: For complex objects (sets, lists etc.) whether to validate individual
+            elements rather than the whole object.
+        always: Whether this method and other validators should be called even if the value is missing.
+        check_fields: Whether to check that the fields actually exist on the model.
+    """
+
+    decorator_repr: ClassVar[str] = '@validator'
+
+    fields: tuple[str, ...]
+    mode: Literal['before', 'after']
+    each_item: bool
+    always: bool
+    check_fields: bool | None
+
+
+@dataclass(**slots_true)
+class FieldValidatorDecoratorInfo:
+    """A container for data from `@field_validator` so that we can access it
+    while building the pydantic-core schema.
+
+    Attributes:
+        decorator_repr: A class variable representing the decorator string, '@field_validator'.
+        fields: A tuple of field names the validator should be called on.
+        mode: The proposed validator mode.
+        check_fields: Whether to check that the fields actually exist on the model.
+    """
+
+    decorator_repr: ClassVar[str] = '@field_validator'
+
+    fields: tuple[str, ...]
+    mode: FieldValidatorModes
+    check_fields: bool | None
+
+
+@dataclass(**slots_true)
+class RootValidatorDecoratorInfo:
+    """A container for data from `@root_validator` so that we can access it
+    while building the pydantic-core schema.
+
+    Attributes:
+        decorator_repr: A class variable representing the decorator string, '@root_validator'.
+        mode: The proposed validator mode.
+    """
+
+    decorator_repr: ClassVar[str] = '@root_validator'
+    mode: Literal['before', 'after']
+
+
+@dataclass(**slots_true)
+class FieldSerializerDecoratorInfo:
+    """A container for data from `@field_serializer` so that we can access it
+    while building the pydantic-core schema.
+
+    Attributes:
+        decorator_repr: A class variable representing the decorator string, '@field_serializer'.
+        fields: A tuple of field names the serializer should be called on.
+        mode: The proposed serializer mode.
+        return_type: The type of the serializer's return value.
+        when_used: The serialization condition. Accepts a string with values `'always'`, `'unless-none'`, `'json'`,
+            and `'json-unless-none'`.
+        check_fields: Whether to check that the fields actually exist on the model.
+    """
+
+    decorator_repr: ClassVar[str] = '@field_serializer'
+    fields: tuple[str, ...]
+    mode: Literal['plain', 'wrap']
+    return_type: Any
+    when_used: core_schema.WhenUsed
+    check_fields: bool | None
+
+
+@dataclass(**slots_true)
+class ModelSerializerDecoratorInfo:
+    """A container for data from `@model_serializer` so that we can access it
+    while building the pydantic-core schema.
+
+    Attributes:
+        decorator_repr: A class variable representing the decorator string, '@model_serializer'.
+        mode: The proposed serializer mode.
+        return_type: The type of the serializer's return value.
+        when_used: The serialization condition. Accepts a string with values `'always'`, `'unless-none'`, `'json'`,
+            and `'json-unless-none'`.
+    """
+
+    decorator_repr: ClassVar[str] = '@model_serializer'
+    mode: Literal['plain', 'wrap']
+    return_type: Any
+    when_used: core_schema.WhenUsed
+
+
+@dataclass(**slots_true)
+class ModelValidatorDecoratorInfo:
+    """A container for data from `@model_validator` so that we can access it
+    while building the pydantic-core schema.
+
+    Attributes:
+        decorator_repr: A class variable representing the decorator string, '@model_serializer'.
+        mode: The proposed serializer mode.
+    """
+
+    decorator_repr: ClassVar[str] = '@model_validator'
+    mode: Literal['wrap', 'before', 'after']
+
+
+DecoratorInfo: TypeAlias = """Union[
+    ValidatorDecoratorInfo,
+    FieldValidatorDecoratorInfo,
+    RootValidatorDecoratorInfo,
+    FieldSerializerDecoratorInfo,
+    ModelSerializerDecoratorInfo,
+    ModelValidatorDecoratorInfo,
+    ComputedFieldInfo,
+]"""
+
+ReturnType = TypeVar('ReturnType')
+DecoratedType: TypeAlias = (
+    'Union[classmethod[Any, Any, ReturnType], staticmethod[Any, ReturnType], Callable[..., ReturnType], property]'
+)
+
+
+@dataclass  # can't use slots here since we set attributes on `__post_init__`
+class PydanticDescriptorProxy(Generic[ReturnType]):
+    """Wrap a classmethod, staticmethod, property or unbound function
+    and act as a descriptor that allows us to detect decorated items
+    from the class' attributes.
+
+    This class' __get__ returns the wrapped item's __get__ result,
+    which makes it transparent for classmethods and staticmethods.
+
+    Attributes:
+        wrapped: The decorator that has to be wrapped.
+        decorator_info: The decorator info.
+        shim: A wrapper function to wrap V1 style function.
+    """
+
+    wrapped: DecoratedType[ReturnType]
+    decorator_info: DecoratorInfo
+    shim: Callable[[Callable[..., Any]], Callable[..., Any]] | None = None
+
+    def __post_init__(self):
+        for attr in 'setter', 'deleter':
+            if hasattr(self.wrapped, attr):
+                f = partial(self._call_wrapped_attr, name=attr)
+                setattr(self, attr, f)
+
+    def _call_wrapped_attr(self, func: Callable[[Any], None], *, name: str) -> PydanticDescriptorProxy[ReturnType]:
+        self.wrapped = getattr(self.wrapped, name)(func)
+        return self
+
+    def __get__(self, obj: object | None, obj_type: type[object] | None = None) -> PydanticDescriptorProxy[ReturnType]:
+        try:
+            return self.wrapped.__get__(obj, obj_type)
+        except AttributeError:
+            # not a descriptor, e.g. a partial object
+            return self.wrapped  # type: ignore[return-value]
+
+    def __set_name__(self, instance: Any, name: str) -> None:
+        if hasattr(self.wrapped, '__set_name__'):
+            self.wrapped.__set_name__(instance, name)  # pyright: ignore[reportFunctionMemberAccess]
+
+    def __getattr__(self, __name: str) -> Any:
+        """Forward checks for __isabstractmethod__ and such."""
+        return getattr(self.wrapped, __name)
+
+
+DecoratorInfoType = TypeVar('DecoratorInfoType', bound=DecoratorInfo)
+
+
+@dataclass(**slots_true)
+class Decorator(Generic[DecoratorInfoType]):
+    """A generic container class to join together the decorator metadata
+    (metadata from decorator itself, which we have when the
+    decorator is called but not when we are building the core-schema)
+    and the bound function (which we have after the class itself is created).
+
+    Attributes:
+        cls_ref: The class ref.
+        cls_var_name: The decorated function name.
+        func: The decorated function.
+        shim: A wrapper function to wrap V1 style function.
+        info: The decorator info.
+    """
+
+    cls_ref: str
+    cls_var_name: str
+    func: Callable[..., Any]
+    shim: Callable[[Any], Any] | None
+    info: DecoratorInfoType
+
+    @staticmethod
+    def build(
+        cls_: Any,
+        *,
+        cls_var_name: str,
+        shim: Callable[[Any], Any] | None,
+        info: DecoratorInfoType,
+    ) -> Decorator[DecoratorInfoType]:
+        """Build a new decorator.
+
+        Args:
+            cls_: The class.
+            cls_var_name: The decorated function name.
+            shim: A wrapper function to wrap V1 style function.
+            info: The decorator info.
+
+        Returns:
+            The new decorator instance.
+        """
+        func = get_attribute_from_bases(cls_, cls_var_name)
+        if shim is not None:
+            func = shim(func)
+        func = unwrap_wrapped_function(func, unwrap_partial=False)
+        if not callable(func):
+            # This branch will get hit for classmethod properties
+            attribute = get_attribute_from_base_dicts(cls_, cls_var_name)  # prevents the binding call to `__get__`
+            if isinstance(attribute, PydanticDescriptorProxy):
+                func = unwrap_wrapped_function(attribute.wrapped)
+        return Decorator(
+            cls_ref=get_type_ref(cls_),
+            cls_var_name=cls_var_name,
+            func=func,
+            shim=shim,
+            info=info,
+        )
+
+    def bind_to_cls(self, cls: Any) -> Decorator[DecoratorInfoType]:
+        """Bind the decorator to a class.
+
+        Args:
+            cls: the class.
+
+        Returns:
+            The new decorator instance.
+        """
+        return self.build(
+            cls,
+            cls_var_name=self.cls_var_name,
+            shim=self.shim,
+            info=self.info,
+        )
+
+
+def get_bases(tp: type[Any]) -> tuple[type[Any], ...]:
+    """Get the base classes of a class or typeddict.
+
+    Args:
+        tp: The type or class to get the bases.
+
+    Returns:
+        The base classes.
+    """
+    if is_typeddict(tp):
+        return tp.__orig_bases__  # type: ignore
+    try:
+        return tp.__bases__
+    except AttributeError:
+        return ()
+
+
+def mro(tp: type[Any]) -> tuple[type[Any], ...]:
+    """Calculate the Method Resolution Order of bases using the C3 algorithm.
+
+    See https://www.python.org/download/releases/2.3/mro/
+    """
+    # try to use the existing mro, for performance mainly
+    # but also because it helps verify the implementation below
+    if not is_typeddict(tp):
+        try:
+            return tp.__mro__
+        except AttributeError:
+            # GenericAlias and some other cases
+            pass
+
+    bases = get_bases(tp)
+    return (tp,) + mro_for_bases(bases)
+
+
+def mro_for_bases(bases: tuple[type[Any], ...]) -> tuple[type[Any], ...]:
+    def merge_seqs(seqs: list[deque[type[Any]]]) -> Iterable[type[Any]]:
+        while True:
+            non_empty = [seq for seq in seqs if seq]
+            if not non_empty:
+                # Nothing left to process, we're done.
+                return
+            candidate: type[Any] | None = None
+            for seq in non_empty:  # Find merge candidates among seq heads.
+                candidate = seq[0]
+                not_head = [s for s in non_empty if candidate in islice(s, 1, None)]
+                if not_head:
+                    # Reject the candidate.
+                    candidate = None
+                else:
+                    break
+            if not candidate:
+                raise TypeError('Inconsistent hierarchy, no C3 MRO is possible')
+            yield candidate
+            for seq in non_empty:
+                # Remove candidate.
+                if seq[0] == candidate:
+                    seq.popleft()
+
+    seqs = [deque(mro(base)) for base in bases] + [deque(bases)]
+    return tuple(merge_seqs(seqs))
+
+
+_sentinel = object()
+
+
+def get_attribute_from_bases(tp: type[Any] | tuple[type[Any], ...], name: str) -> Any:
+    """Get the attribute from the next class in the MRO that has it,
+    aiming to simulate calling the method on the actual class.
+
+    The reason for iterating over the mro instead of just getting
+    the attribute (which would do that for us) is to support TypedDict,
+    which lacks a real __mro__, but can have a virtual one constructed
+    from its bases (as done here).
+
+    Args:
+        tp: The type or class to search for the attribute. If a tuple, this is treated as a set of base classes.
+        name: The name of the attribute to retrieve.
+
+    Returns:
+        Any: The attribute value, if found.
+
+    Raises:
+        AttributeError: If the attribute is not found in any class in the MRO.
+    """
+    if isinstance(tp, tuple):
+        for base in mro_for_bases(tp):
+            attribute = base.__dict__.get(name, _sentinel)
+            if attribute is not _sentinel:
+                attribute_get = getattr(attribute, '__get__', None)
+                if attribute_get is not None:
+                    return attribute_get(None, tp)
+                return attribute
+        raise AttributeError(f'{name} not found in {tp}')
+    else:
+        try:
+            return getattr(tp, name)
+        except AttributeError:
+            return get_attribute_from_bases(mro(tp), name)
+
+
+def get_attribute_from_base_dicts(tp: type[Any], name: str) -> Any:
+    """Get an attribute out of the `__dict__` following the MRO.
+    This prevents the call to `__get__` on the descriptor, and allows
+    us to get the original function for classmethod properties.
+
+    Args:
+        tp: The type or class to search for the attribute.
+        name: The name of the attribute to retrieve.
+
+    Returns:
+        Any: The attribute value, if found.
+
+    Raises:
+        KeyError: If the attribute is not found in any class's `__dict__` in the MRO.
+    """
+    for base in reversed(mro(tp)):
+        if name in base.__dict__:
+            return base.__dict__[name]
+    return tp.__dict__[name]  # raise the error
+
+
+@dataclass(**slots_true)
+class DecoratorInfos:
+    """Mapping of name in the class namespace to decorator info.
+
+    note that the name in the class namespace is the function or attribute name
+    not the field name!
+    """
+
+    validators: dict[str, Decorator[ValidatorDecoratorInfo]] = field(default_factory=dict)
+    field_validators: dict[str, Decorator[FieldValidatorDecoratorInfo]] = field(default_factory=dict)
+    root_validators: dict[str, Decorator[RootValidatorDecoratorInfo]] = field(default_factory=dict)
+    field_serializers: dict[str, Decorator[FieldSerializerDecoratorInfo]] = field(default_factory=dict)
+    model_serializers: dict[str, Decorator[ModelSerializerDecoratorInfo]] = field(default_factory=dict)
+    model_validators: dict[str, Decorator[ModelValidatorDecoratorInfo]] = field(default_factory=dict)
+    computed_fields: dict[str, Decorator[ComputedFieldInfo]] = field(default_factory=dict)
+
+    @staticmethod
+    def build(model_dc: type[Any]) -> DecoratorInfos:  # noqa: C901 (ignore complexity)
+        """We want to collect all DecFunc instances that exist as
+        attributes in the namespace of the class (a BaseModel or dataclass)
+        that called us
+        But we want to collect these in the order of the bases
+        So instead of getting them all from the leaf class (the class that called us),
+        we traverse the bases from root (the oldest ancestor class) to leaf
+        and collect all of the instances as we go, taking care to replace
+        any duplicate ones with the last one we see to mimic how function overriding
+        works with inheritance.
+        If we do replace any functions we put the replacement into the position
+        the replaced function was in; that is, we maintain the order.
+        """
+        # reminder: dicts are ordered and replacement does not alter the order
+        res = DecoratorInfos()
+        for base in reversed(mro(model_dc)[1:]):
+            existing: DecoratorInfos | None = base.__dict__.get('__pydantic_decorators__')
+            if existing is None:
+                existing = DecoratorInfos.build(base)
+            res.validators.update({k: v.bind_to_cls(model_dc) for k, v in existing.validators.items()})
+            res.field_validators.update({k: v.bind_to_cls(model_dc) for k, v in existing.field_validators.items()})
+            res.root_validators.update({k: v.bind_to_cls(model_dc) for k, v in existing.root_validators.items()})
+            res.field_serializers.update({k: v.bind_to_cls(model_dc) for k, v in existing.field_serializers.items()})
+            res.model_serializers.update({k: v.bind_to_cls(model_dc) for k, v in existing.model_serializers.items()})
+            res.model_validators.update({k: v.bind_to_cls(model_dc) for k, v in existing.model_validators.items()})
+            res.computed_fields.update({k: v.bind_to_cls(model_dc) for k, v in existing.computed_fields.items()})
+
+        to_replace: list[tuple[str, Any]] = []
+
+        for var_name, var_value in vars(model_dc).items():
+            if isinstance(var_value, PydanticDescriptorProxy):
+                info = var_value.decorator_info
+                if isinstance(info, ValidatorDecoratorInfo):
+                    res.validators[var_name] = Decorator.build(
+                        model_dc, cls_var_name=var_name, shim=var_value.shim, info=info
+                    )
+                elif isinstance(info, FieldValidatorDecoratorInfo):
+                    res.field_validators[var_name] = Decorator.build(
+                        model_dc, cls_var_name=var_name, shim=var_value.shim, info=info
+                    )
+                elif isinstance(info, RootValidatorDecoratorInfo):
+                    res.root_validators[var_name] = Decorator.build(
+                        model_dc, cls_var_name=var_name, shim=var_value.shim, info=info
+                    )
+                elif isinstance(info, FieldSerializerDecoratorInfo):
+                    # check whether a serializer function is already registered for fields
+                    for field_serializer_decorator in res.field_serializers.values():
+                        # check that each field has at most one serializer function.
+                        # serializer functions for the same field in subclasses are allowed,
+                        # and are treated as overrides
+                        if field_serializer_decorator.cls_var_name == var_name:
+                            continue
+                        for f in info.fields:
+                            if f in field_serializer_decorator.info.fields:
+                                raise PydanticUserError(
+                                    'Multiple field serializer functions were defined '
+                                    f'for field {f!r}, this is not allowed.',
+                                    code='multiple-field-serializers',
+                                )
+                    res.field_serializers[var_name] = Decorator.build(
+                        model_dc, cls_var_name=var_name, shim=var_value.shim, info=info
+                    )
+                elif isinstance(info, ModelValidatorDecoratorInfo):
+                    res.model_validators[var_name] = Decorator.build(
+                        model_dc, cls_var_name=var_name, shim=var_value.shim, info=info
+                    )
+                elif isinstance(info, ModelSerializerDecoratorInfo):
+                    res.model_serializers[var_name] = Decorator.build(
+                        model_dc, cls_var_name=var_name, shim=var_value.shim, info=info
+                    )
+                else:
+                    from ..fields import ComputedFieldInfo
+
+                    isinstance(var_value, ComputedFieldInfo)
+                    res.computed_fields[var_name] = Decorator.build(
+                        model_dc, cls_var_name=var_name, shim=None, info=info
+                    )
+                to_replace.append((var_name, var_value.wrapped))
+        if to_replace:
+            # If we can save `__pydantic_decorators__` on the class we'll be able to check for it above
+            # so then we don't need to re-process the type, which means we can discard our descriptor wrappers
+            # and replace them with the thing they are wrapping (see the other setattr call below)
+            # which allows validator class methods to also function as regular class methods
+            setattr(model_dc, '__pydantic_decorators__', res)
+            for name, value in to_replace:
+                setattr(model_dc, name, value)
+        return res
+
+
+def inspect_validator(validator: Callable[..., Any], mode: FieldValidatorModes) -> bool:
+    """Look at a field or model validator function and determine whether it takes an info argument.
+
+    An error is raised if the function has an invalid signature.
+
+    Args:
+        validator: The validator function to inspect.
+        mode: The proposed validator mode.
+
+    Returns:
+        Whether the validator takes an info argument.
+    """
+    try:
+        sig = signature(validator)
+    except ValueError:
+        # builtins and some C extensions don't have signatures
+        # assume that they don't take an info argument and only take a single argument
+        # e.g. `str.strip` or `datetime.datetime`
+        return False
+    n_positional = count_positional_params(sig)
+    if mode == 'wrap':
+        if n_positional == 3:
+            return True
+        elif n_positional == 2:
+            return False
+    else:
+        assert mode in {'before', 'after', 'plain'}, f"invalid mode: {mode!r}, expected 'before', 'after' or 'plain"
+        if n_positional == 2:
+            return True
+        elif n_positional == 1:
+            return False
+
+    raise PydanticUserError(
+        f'Unrecognized field_validator function signature for {validator} with `mode={mode}`:{sig}',
+        code='validator-signature',
+    )
+
+
+def inspect_field_serializer(
+    serializer: Callable[..., Any], mode: Literal['plain', 'wrap'], computed_field: bool = False
+) -> tuple[bool, bool]:
+    """Look at a field serializer function and determine if it is a field serializer,
+    and whether it takes an info argument.
+
+    An error is raised if the function has an invalid signature.
+
+    Args:
+        serializer: The serializer function to inspect.
+        mode: The serializer mode, either 'plain' or 'wrap'.
+        computed_field: When serializer is applied on computed_field. It doesn't require
+            info signature.
+
+    Returns:
+        Tuple of (is_field_serializer, info_arg).
+    """
+    sig = signature(serializer)
+
+    first = next(iter(sig.parameters.values()), None)
+    is_field_serializer = first is not None and first.name == 'self'
+
+    n_positional = count_positional_params(sig)
+    if is_field_serializer:
+        # -1 to correct for self parameter
+        info_arg = _serializer_info_arg(mode, n_positional - 1)
+    else:
+        info_arg = _serializer_info_arg(mode, n_positional)
+
+    if info_arg is None:
+        raise PydanticUserError(
+            f'Unrecognized field_serializer function signature for {serializer} with `mode={mode}`:{sig}',
+            code='field-serializer-signature',
+        )
+    if info_arg and computed_field:
+        raise PydanticUserError(
+            'field_serializer on computed_field does not use info signature', code='field-serializer-signature'
+        )
+
+    else:
+        return is_field_serializer, info_arg
+
+
+def inspect_annotated_serializer(serializer: Callable[..., Any], mode: Literal['plain', 'wrap']) -> bool:
+    """Look at a serializer function used via `Annotated` and determine whether it takes an info argument.
+
+    An error is raised if the function has an invalid signature.
+
+    Args:
+        serializer: The serializer function to check.
+        mode: The serializer mode, either 'plain' or 'wrap'.
+
+    Returns:
+        info_arg
+    """
+    sig = signature(serializer)
+    info_arg = _serializer_info_arg(mode, count_positional_params(sig))
+    if info_arg is None:
+        raise PydanticUserError(
+            f'Unrecognized field_serializer function signature for {serializer} with `mode={mode}`:{sig}',
+            code='field-serializer-signature',
+        )
+    else:
+        return info_arg
+
+
+def inspect_model_serializer(serializer: Callable[..., Any], mode: Literal['plain', 'wrap']) -> bool:
+    """Look at a model serializer function and determine whether it takes an info argument.
+
+    An error is raised if the function has an invalid signature.
+
+    Args:
+        serializer: The serializer function to check.
+        mode: The serializer mode, either 'plain' or 'wrap'.
+
+    Returns:
+        `info_arg` - whether the function expects an info argument.
+    """
+    if isinstance(serializer, (staticmethod, classmethod)) or not is_instance_method_from_sig(serializer):
+        raise PydanticUserError(
+            '`@model_serializer` must be applied to instance methods', code='model-serializer-instance-method'
+        )
+
+    sig = signature(serializer)
+    info_arg = _serializer_info_arg(mode, count_positional_params(sig))
+    if info_arg is None:
+        raise PydanticUserError(
+            f'Unrecognized model_serializer function signature for {serializer} with `mode={mode}`:{sig}',
+            code='model-serializer-signature',
+        )
+    else:
+        return info_arg
+
+
+def _serializer_info_arg(mode: Literal['plain', 'wrap'], n_positional: int) -> bool | None:
+    if mode == 'plain':
+        if n_positional == 1:
+            # (__input_value: Any) -> Any
+            return False
+        elif n_positional == 2:
+            # (__model: Any, __input_value: Any) -> Any
+            return True
+    else:
+        assert mode == 'wrap', f"invalid mode: {mode!r}, expected 'plain' or 'wrap'"
+        if n_positional == 2:
+            # (__input_value: Any, __serializer: SerializerFunctionWrapHandler) -> Any
+            return False
+        elif n_positional == 3:
+            # (__input_value: Any, __serializer: SerializerFunctionWrapHandler, __info: SerializationInfo) -> Any
+            return True
+
+    return None
+
+
+AnyDecoratorCallable: TypeAlias = (
+    'Union[classmethod[Any, Any, Any], staticmethod[Any, Any], partialmethod[Any], Callable[..., Any]]'
+)
+
+
+def is_instance_method_from_sig(function: AnyDecoratorCallable) -> bool:
+    """Whether the function is an instance method.
+
+    It will consider a function as instance method if the first parameter of
+    function is `self`.
+
+    Args:
+        function: The function to check.
+
+    Returns:
+        `True` if the function is an instance method, `False` otherwise.
+    """
+    sig = signature(unwrap_wrapped_function(function))
+    first = next(iter(sig.parameters.values()), None)
+    if first and first.name == 'self':
+        return True
+    return False
+
+
+def ensure_classmethod_based_on_signature(function: AnyDecoratorCallable) -> Any:
+    """Apply the `@classmethod` decorator on the function.
+
+    Args:
+        function: The function to apply the decorator on.
+
+    Return:
+        The `@classmethod` decorator applied function.
+    """
+    if not isinstance(
+        unwrap_wrapped_function(function, unwrap_class_static_method=False), classmethod
+    ) and _is_classmethod_from_sig(function):
+        return classmethod(function)  # type: ignore[arg-type]
+    return function
+
+
+def _is_classmethod_from_sig(function: AnyDecoratorCallable) -> bool:
+    sig = signature(unwrap_wrapped_function(function))
+    first = next(iter(sig.parameters.values()), None)
+    if first and first.name == 'cls':
+        return True
+    return False
+
+
+def unwrap_wrapped_function(
+    func: Any,
+    *,
+    unwrap_partial: bool = True,
+    unwrap_class_static_method: bool = True,
+) -> Any:
+    """Recursively unwraps a wrapped function until the underlying function is reached.
+    This handles property, functools.partial, functools.partialmethod, staticmethod and classmethod.
+
+    Args:
+        func: The function to unwrap.
+        unwrap_partial: If True (default), unwrap partial and partialmethod decorators, otherwise don't.
+            decorators.
+        unwrap_class_static_method: If True (default), also unwrap classmethod and staticmethod
+            decorators. If False, only unwrap partial and partialmethod decorators.
+
+    Returns:
+        The underlying function of the wrapped function.
+    """
+    all: set[Any] = {property, cached_property}
+
+    if unwrap_partial:
+        all.update({partial, partialmethod})
+
+    if unwrap_class_static_method:
+        all.update({staticmethod, classmethod})
+
+    while isinstance(func, tuple(all)):
+        if unwrap_class_static_method and isinstance(func, (classmethod, staticmethod)):
+            func = func.__func__
+        elif isinstance(func, (partial, partialmethod)):
+            func = func.func
+        elif isinstance(func, property):
+            func = func.fget  # arbitrary choice, convenient for computed fields
+        else:
+            # Make coverage happy as it can only get here in the last possible case
+            assert isinstance(func, cached_property)
+            func = func.func  # type: ignore
+
+    return func
+
+
+def get_function_return_type(
+    func: Any, explicit_return_type: Any, types_namespace: dict[str, Any] | None = None
+) -> Any:
+    """Get the function return type.
+
+    It gets the return type from the type annotation if `explicit_return_type` is `None`.
+    Otherwise, it returns `explicit_return_type`.
+
+    Args:
+        func: The function to get its return type.
+        explicit_return_type: The explicit return type.
+        types_namespace: The types namespace, defaults to `None`.
+
+    Returns:
+        The function return type.
+    """
+    if explicit_return_type is PydanticUndefined:
+        # try to get it from the type annotation
+        hints = get_function_type_hints(
+            unwrap_wrapped_function(func), include_keys={'return'}, types_namespace=types_namespace
+        )
+        return hints.get('return', PydanticUndefined)
+    else:
+        return explicit_return_type
+
+
+def count_positional_params(sig: Signature) -> int:
+    return sum(1 for param in sig.parameters.values() if can_be_positional(param))
+
+
+def can_be_positional(param: Parameter) -> bool:
+    return param.kind in (Parameter.POSITIONAL_ONLY, Parameter.POSITIONAL_OR_KEYWORD)
+
+
+def ensure_property(f: Any) -> Any:
+    """Ensure that a function is a `property` or `cached_property`, or is a valid descriptor.
+
+    Args:
+        f: The function to check.
+
+    Returns:
+        The function, or a `property` or `cached_property` instance wrapping the function.
+    """
+    if ismethoddescriptor(f) or isdatadescriptor(f):
+        return f
+    else:
+        return property(f)

lib/pydantic/_internal/_decorators_v1.py

@@ -0,0 +1,181 @@
+"""Logic for V1 validators, e.g. `@validator` and `@root_validator`."""
+from __future__ import annotations as _annotations
+
+from inspect import Parameter, signature
+from typing import Any, Dict, Tuple, Union, cast
+
+from pydantic_core import core_schema
+from typing_extensions import Protocol
+
+from ..errors import PydanticUserError
+from ._decorators import can_be_positional
+
+
+class V1OnlyValueValidator(Protocol):
+    """A simple validator, supported for V1 validators and V2 validators."""
+
+    def __call__(self, __value: Any) -> Any:
+        ...
+
+
+class V1ValidatorWithValues(Protocol):
+    """A validator with `values` argument, supported for V1 validators and V2 validators."""
+
+    def __call__(self, __value: Any, values: dict[str, Any]) -> Any:
+        ...
+
+
+class V1ValidatorWithValuesKwOnly(Protocol):
+    """A validator with keyword only `values` argument, supported for V1 validators and V2 validators."""
+
+    def __call__(self, __value: Any, *, values: dict[str, Any]) -> Any:
+        ...
+
+
+class V1ValidatorWithKwargs(Protocol):
+    """A validator with `kwargs` argument, supported for V1 validators and V2 validators."""
+
+    def __call__(self, __value: Any, **kwargs: Any) -> Any:
+        ...
+
+
+class V1ValidatorWithValuesAndKwargs(Protocol):
+    """A validator with `values` and `kwargs` arguments, supported for V1 validators and V2 validators."""
+
+    def __call__(self, __value: Any, values: dict[str, Any], **kwargs: Any) -> Any:
+        ...
+
+
+V1Validator = Union[
+    V1ValidatorWithValues, V1ValidatorWithValuesKwOnly, V1ValidatorWithKwargs, V1ValidatorWithValuesAndKwargs
+]
+
+
+def can_be_keyword(param: Parameter) -> bool:
+    return param.kind in (Parameter.POSITIONAL_OR_KEYWORD, Parameter.KEYWORD_ONLY)
+
+
+def make_generic_v1_field_validator(validator: V1Validator) -> core_schema.WithInfoValidatorFunction:
+    """Wrap a V1 style field validator for V2 compatibility.
+
+    Args:
+        validator: The V1 style field validator.
+
+    Returns:
+        A wrapped V2 style field validator.
+
+    Raises:
+        PydanticUserError: If the signature is not supported or the parameters are
+            not available in Pydantic V2.
+    """
+    sig = signature(validator)
+
+    needs_values_kw = False
+
+    for param_num, (param_name, parameter) in enumerate(sig.parameters.items()):
+        if can_be_keyword(parameter) and param_name in ('field', 'config'):
+            raise PydanticUserError(
+                'The `field` and `config` parameters are not available in Pydantic V2, '
+                'please use the `info` parameter instead.',
+                code='validator-field-config-info',
+            )
+        if parameter.kind is Parameter.VAR_KEYWORD:
+            needs_values_kw = True
+        elif can_be_keyword(parameter) and param_name == 'values':
+            needs_values_kw = True
+        elif can_be_positional(parameter) and param_num == 0:
+            # value
+            continue
+        elif parameter.default is Parameter.empty:  # ignore params with defaults e.g. bound by functools.partial
+            raise PydanticUserError(
+                f'Unsupported signature for V1 style validator {validator}: {sig} is not supported.',
+                code='validator-v1-signature',
+            )
+
+    if needs_values_kw:
+        # (v, **kwargs), (v, values, **kwargs), (v, *, values, **kwargs) or (v, *, values)
+        val1 = cast(V1ValidatorWithValues, validator)
+
+        def wrapper1(value: Any, info: core_schema.ValidationInfo) -> Any:
+            return val1(value, values=info.data)
+
+        return wrapper1
+    else:
+        val2 = cast(V1OnlyValueValidator, validator)
+
+        def wrapper2(value: Any, _: core_schema.ValidationInfo) -> Any:
+            return val2(value)
+
+        return wrapper2
+
+
+RootValidatorValues = Dict[str, Any]
+# technically tuple[model_dict, model_extra, fields_set] | tuple[dataclass_dict, init_vars]
+RootValidatorFieldsTuple = Tuple[Any, ...]
+
+
+class V1RootValidatorFunction(Protocol):
+    """A simple root validator, supported for V1 validators and V2 validators."""
+
+    def __call__(self, __values: RootValidatorValues) -> RootValidatorValues:
+        ...
+
+
+class V2CoreBeforeRootValidator(Protocol):
+    """V2 validator with mode='before'."""
+
+    def __call__(self, __values: RootValidatorValues, __info: core_schema.ValidationInfo) -> RootValidatorValues:
+        ...
+
+
+class V2CoreAfterRootValidator(Protocol):
+    """V2 validator with mode='after'."""
+
+    def __call__(
+        self, __fields_tuple: RootValidatorFieldsTuple, __info: core_schema.ValidationInfo
+    ) -> RootValidatorFieldsTuple:
+        ...
+
+
+def make_v1_generic_root_validator(
+    validator: V1RootValidatorFunction, pre: bool
+) -> V2CoreBeforeRootValidator | V2CoreAfterRootValidator:
+    """Wrap a V1 style root validator for V2 compatibility.
+
+    Args:
+        validator: The V1 style field validator.
+        pre: Whether the validator is a pre validator.
+
+    Returns:
+        A wrapped V2 style validator.
+    """
+    if pre is True:
+        # mode='before' for pydantic-core
+        def _wrapper1(values: RootValidatorValues, _: core_schema.ValidationInfo) -> RootValidatorValues:
+            return validator(values)
+
+        return _wrapper1
+
+    # mode='after' for pydantic-core
+    def _wrapper2(fields_tuple: RootValidatorFieldsTuple, _: core_schema.ValidationInfo) -> RootValidatorFieldsTuple:
+        if len(fields_tuple) == 2:
+            # dataclass, this is easy
+            values, init_vars = fields_tuple
+            values = validator(values)
+            return values, init_vars
+        else:
+            # ugly hack: to match v1 behaviour, we merge values and model_extra, then split them up based on fields
+            # afterwards
+            model_dict, model_extra, fields_set = fields_tuple
+            if model_extra:
+                fields = set(model_dict.keys())
+                model_dict.update(model_extra)
+                model_dict_new = validator(model_dict)
+                for k in list(model_dict_new.keys()):
+                    if k not in fields:
+                        model_extra[k] = model_dict_new.pop(k)
+            else:
+                model_dict_new = validator(model_dict)
+            return model_dict_new, model_extra, fields_set
+
+    return _wrapper2

lib/pydantic/_internal/_discriminated_union.py

@@ -0,0 +1,506 @@
+from __future__ import annotations as _annotations
+
+from typing import TYPE_CHECKING, Any, Hashable, Sequence
+
+from pydantic_core import CoreSchema, core_schema
+
+from ..errors import PydanticUserError
+from . import _core_utils
+from ._core_utils import (
+    CoreSchemaField,
+    collect_definitions,
+    simplify_schema_references,
+)
+
+if TYPE_CHECKING:
+    from ..types import Discriminator
+
+CORE_SCHEMA_METADATA_DISCRIMINATOR_PLACEHOLDER_KEY = 'pydantic.internal.union_discriminator'
+
+
+class MissingDefinitionForUnionRef(Exception):
+    """Raised when applying a discriminated union discriminator to a schema
+    requires a definition that is not yet defined
+    """
+
+    def __init__(self, ref: str) -> None:
+        self.ref = ref
+        super().__init__(f'Missing definition for ref {self.ref!r}')
+
+
+def set_discriminator_in_metadata(schema: CoreSchema, discriminator: Any) -> None:
+    schema.setdefault('metadata', {})
+    metadata = schema.get('metadata')
+    assert metadata is not None
+    metadata[CORE_SCHEMA_METADATA_DISCRIMINATOR_PLACEHOLDER_KEY] = discriminator
+
+
+def apply_discriminators(schema: core_schema.CoreSchema) -> core_schema.CoreSchema:
+    definitions: dict[str, CoreSchema] | None = None
+
+    def inner(s: core_schema.CoreSchema, recurse: _core_utils.Recurse) -> core_schema.CoreSchema:
+        nonlocal definitions
+
+        s = recurse(s, inner)
+        if s['type'] == 'tagged-union':
+            return s
+
+        metadata = s.get('metadata', {})
+        discriminator = metadata.pop(CORE_SCHEMA_METADATA_DISCRIMINATOR_PLACEHOLDER_KEY, None)
+        if discriminator is not None:
+            if definitions is None:
+                definitions = collect_definitions(schema)
+            s = apply_discriminator(s, discriminator, definitions)
+        return s
+
+    return simplify_schema_references(_core_utils.walk_core_schema(schema, inner))
+
+
+def apply_discriminator(
+    schema: core_schema.CoreSchema,
+    discriminator: str | Discriminator,
+    definitions: dict[str, core_schema.CoreSchema] | None = None,
+) -> core_schema.CoreSchema:
+    """Applies the discriminator and returns a new core schema.
+
+    Args:
+        schema: The input schema.
+        discriminator: The name of the field which will serve as the discriminator.
+        definitions: A mapping of schema ref to schema.
+
+    Returns:
+        The new core schema.
+
+    Raises:
+        TypeError:
+            - If `discriminator` is used with invalid union variant.
+            - If `discriminator` is used with `Union` type with one variant.
+            - If `discriminator` value mapped to multiple choices.
+        MissingDefinitionForUnionRef:
+            If the definition for ref is missing.
+        PydanticUserError:
+            - If a model in union doesn't have a discriminator field.
+            - If discriminator field has a non-string alias.
+            - If discriminator fields have different aliases.
+            - If discriminator field not of type `Literal`.
+    """
+    from ..types import Discriminator
+
+    if isinstance(discriminator, Discriminator):
+        if isinstance(discriminator.discriminator, str):
+            discriminator = discriminator.discriminator
+        else:
+            return discriminator._convert_schema(schema)
+
+    return _ApplyInferredDiscriminator(discriminator, definitions or {}).apply(schema)
+
+
+class _ApplyInferredDiscriminator:
+    """This class is used to convert an input schema containing a union schema into one where that union is
+    replaced with a tagged-union, with all the associated debugging and performance benefits.
+
+    This is done by:
+    * Validating that the input schema is compatible with the provided discriminator
+    * Introspecting the schema to determine which discriminator values should map to which union choices
+    * Handling various edge cases such as 'definitions', 'default', 'nullable' schemas, and more
+
+    I have chosen to implement the conversion algorithm in this class, rather than a function,
+    to make it easier to maintain state while recursively walking the provided CoreSchema.
+    """
+
+    def __init__(self, discriminator: str, definitions: dict[str, core_schema.CoreSchema]):
+        # `discriminator` should be the name of the field which will serve as the discriminator.
+        # It must be the python name of the field, and *not* the field's alias. Note that as of now,
+        # all members of a discriminated union _must_ use a field with the same name as the discriminator.
+        # This may change if/when we expose a way to manually specify the TaggedUnionSchema's choices.
+        self.discriminator = discriminator
+
+        # `definitions` should contain a mapping of schema ref to schema for all schemas which might
+        # be referenced by some choice
+        self.definitions = definitions
+
+        # `_discriminator_alias` will hold the value, if present, of the alias for the discriminator
+        #
+        # Note: following the v1 implementation, we currently disallow the use of different aliases
+        # for different choices. This is not a limitation of pydantic_core, but if we try to handle
+        # this, the inference logic gets complicated very quickly, and could result in confusing
+        # debugging challenges for users making subtle mistakes.
+        #
+        # Rather than trying to do the most powerful inference possible, I think we should eventually
+        # expose a way to more-manually control the way the TaggedUnionSchema is constructed through
+        # the use of a new type which would be placed as an Annotation on the Union type. This would
+        # provide the full flexibility/power of pydantic_core's TaggedUnionSchema where necessary for
+        # more complex cases, without over-complicating the inference logic for the common cases.
+        self._discriminator_alias: str | None = None
+
+        # `_should_be_nullable` indicates whether the converted union has `None` as an allowed value.
+        # If `None` is an acceptable value of the (possibly-wrapped) union, we ignore it while
+        # constructing the TaggedUnionSchema, but set the `_should_be_nullable` attribute to True.
+        # Once we have constructed the TaggedUnionSchema, if `_should_be_nullable` is True, we ensure
+        # that the final schema gets wrapped as a NullableSchema. This has the same semantics on the
+        # python side, but resolves the issue that `None` cannot correspond to any discriminator values.
+        self._should_be_nullable = False
+
+        # `_is_nullable` is used to track if the final produced schema will definitely be nullable;
+        # we set it to True if the input schema is wrapped in a nullable schema that we know will be preserved
+        # as an indication that, even if None is discovered as one of the union choices, we will not need to wrap
+        # the final value in another nullable schema.
+        #
+        # This is more complicated than just checking for the final outermost schema having type 'nullable' thanks
+        # to the possible presence of other wrapper schemas such as DefinitionsSchema, WithDefaultSchema, etc.
+        self._is_nullable = False
+
+        # `_choices_to_handle` serves as a stack of choices to add to the tagged union. Initially, choices
+        # from the union in the wrapped schema will be appended to this list, and the recursive choice-handling
+        # algorithm may add more choices to this stack as (nested) unions are encountered.
+        self._choices_to_handle: list[core_schema.CoreSchema] = []
+
+        # `_tagged_union_choices` is built during the call to `apply`, and will hold the choices to be included
+        # in the output TaggedUnionSchema that will replace the union from the input schema
+        self._tagged_union_choices: dict[Hashable, core_schema.CoreSchema] = {}
+
+        # `_used` is changed to True after applying the discriminator to prevent accidental re-use
+        self._used = False
+
+    def apply(self, schema: core_schema.CoreSchema) -> core_schema.CoreSchema:
+        """Return a new CoreSchema based on `schema` that uses a tagged-union with the discriminator provided
+        to this class.
+
+        Args:
+            schema: The input schema.
+
+        Returns:
+            The new core schema.
+
+        Raises:
+            TypeError:
+                - If `discriminator` is used with invalid union variant.
+                - If `discriminator` is used with `Union` type with one variant.
+                - If `discriminator` value mapped to multiple choices.
+            ValueError:
+                If the definition for ref is missing.
+            PydanticUserError:
+                - If a model in union doesn't have a discriminator field.
+                - If discriminator field has a non-string alias.
+                - If discriminator fields have different aliases.
+                - If discriminator field not of type `Literal`.
+        """
+        self.definitions.update(collect_definitions(schema))
+        assert not self._used
+        schema = self._apply_to_root(schema)
+        if self._should_be_nullable and not self._is_nullable:
+            schema = core_schema.nullable_schema(schema)
+        self._used = True
+        new_defs = collect_definitions(schema)
+        missing_defs = self.definitions.keys() - new_defs.keys()
+        if missing_defs:
+            schema = core_schema.definitions_schema(schema, [self.definitions[ref] for ref in missing_defs])
+        return schema
+
+    def _apply_to_root(self, schema: core_schema.CoreSchema) -> core_schema.CoreSchema:
+        """This method handles the outer-most stage of recursion over the input schema:
+        unwrapping nullable or definitions schemas, and calling the `_handle_choice`
+        method iteratively on the choices extracted (recursively) from the possibly-wrapped union.
+        """
+        if schema['type'] == 'nullable':
+            self._is_nullable = True
+            wrapped = self._apply_to_root(schema['schema'])
+            nullable_wrapper = schema.copy()
+            nullable_wrapper['schema'] = wrapped
+            return nullable_wrapper
+
+        if schema['type'] == 'definitions':
+            wrapped = self._apply_to_root(schema['schema'])
+            definitions_wrapper = schema.copy()
+            definitions_wrapper['schema'] = wrapped
+            return definitions_wrapper
+
+        if schema['type'] != 'union':
+            # If the schema is not a union, it probably means it just had a single member and
+            # was flattened by pydantic_core.
+            # However, it still may make sense to apply the discriminator to this schema,
+            # as a way to get discriminated-union-style error messages, so we allow this here.
+            schema = core_schema.union_schema([schema])
+
+        # Reverse the choices list before extending the stack so that they get handled in the order they occur
+        choices_schemas = [v[0] if isinstance(v, tuple) else v for v in schema['choices'][::-1]]
+        self._choices_to_handle.extend(choices_schemas)
+        while self._choices_to_handle:
+            choice = self._choices_to_handle.pop()
+            self._handle_choice(choice)
+
+        if self._discriminator_alias is not None and self._discriminator_alias != self.discriminator:
+            # * We need to annotate `discriminator` as a union here to handle both branches of this conditional
+            # * We need to annotate `discriminator` as list[list[str | int]] and not list[list[str]] due to the
+            #   invariance of list, and because list[list[str | int]] is the type of the discriminator argument
+            #   to tagged_union_schema below
+            # * See the docstring of pydantic_core.core_schema.tagged_union_schema for more details about how to
+            #   interpret the value of the discriminator argument to tagged_union_schema. (The list[list[str]] here
+            #   is the appropriate way to provide a list of fallback attributes to check for a discriminator value.)
+            discriminator: str | list[list[str | int]] = [[self.discriminator], [self._discriminator_alias]]
+        else:
+            discriminator = self.discriminator
+        return core_schema.tagged_union_schema(
+            choices=self._tagged_union_choices,
+            discriminator=discriminator,
+            custom_error_type=schema.get('custom_error_type'),
+            custom_error_message=schema.get('custom_error_message'),
+            custom_error_context=schema.get('custom_error_context'),
+            strict=False,
+            from_attributes=True,
+            ref=schema.get('ref'),
+            metadata=schema.get('metadata'),
+            serialization=schema.get('serialization'),
+        )
+
+    def _handle_choice(self, choice: core_schema.CoreSchema) -> None:
+        """This method handles the "middle" stage of recursion over the input schema.
+        Specifically, it is responsible for handling each choice of the outermost union
+        (and any "coalesced" choices obtained from inner unions).
+
+        Here, "handling" entails:
+        * Coalescing nested unions and compatible tagged-unions
+        * Tracking the presence of 'none' and 'nullable' schemas occurring as choices
+        * Validating that each allowed discriminator value maps to a unique choice
+        * Updating the _tagged_union_choices mapping that will ultimately be used to build the TaggedUnionSchema.
+        """
+        if choice['type'] == 'definition-ref':
+            if choice['schema_ref'] not in self.definitions:
+                raise MissingDefinitionForUnionRef(choice['schema_ref'])
+
+        if choice['type'] == 'none':
+            self._should_be_nullable = True
+        elif choice['type'] == 'definitions':
+            self._handle_choice(choice['schema'])
+        elif choice['type'] == 'nullable':
+            self._should_be_nullable = True
+            self._handle_choice(choice['schema'])  # unwrap the nullable schema
+        elif choice['type'] == 'union':
+            # Reverse the choices list before extending the stack so that they get handled in the order they occur
+            choices_schemas = [v[0] if isinstance(v, tuple) else v for v in choice['choices'][::-1]]
+            self._choices_to_handle.extend(choices_schemas)
+        elif choice['type'] not in {
+            'model',
+            'typed-dict',
+            'tagged-union',
+            'lax-or-strict',
+            'dataclass',
+            'dataclass-args',
+            'definition-ref',
+        } and not _core_utils.is_function_with_inner_schema(choice):
+            # We should eventually handle 'definition-ref' as well
+            raise TypeError(
+                f'{choice["type"]!r} is not a valid discriminated union variant;'
+                ' should be a `BaseModel` or `dataclass`'
+            )
+        else:
+            if choice['type'] == 'tagged-union' and self._is_discriminator_shared(choice):
+                # In this case, this inner tagged-union is compatible with the outer tagged-union,
+                # and its choices can be coalesced into the outer TaggedUnionSchema.
+                subchoices = [x for x in choice['choices'].values() if not isinstance(x, (str, int))]
+                # Reverse the choices list before extending the stack so that they get handled in the order they occur
+                self._choices_to_handle.extend(subchoices[::-1])
+                return
+
+            inferred_discriminator_values = self._infer_discriminator_values_for_choice(choice, source_name=None)
+            self._set_unique_choice_for_values(choice, inferred_discriminator_values)
+
+    def _is_discriminator_shared(self, choice: core_schema.TaggedUnionSchema) -> bool:
+        """This method returns a boolean indicating whether the discriminator for the `choice`
+        is the same as that being used for the outermost tagged union. This is used to
+        determine whether this TaggedUnionSchema choice should be "coalesced" into the top level,
+        or whether it should be treated as a separate (nested) choice.
+        """
+        inner_discriminator = choice['discriminator']
+        return inner_discriminator == self.discriminator or (
+            isinstance(inner_discriminator, list)
+            and (self.discriminator in inner_discriminator or [self.discriminator] in inner_discriminator)
+        )
+
+    def _infer_discriminator_values_for_choice(  # noqa C901
+        self, choice: core_schema.CoreSchema, source_name: str | None
+    ) -> list[str | int]:
+        """This function recurses over `choice`, extracting all discriminator values that should map to this choice.
+
+        `model_name` is accepted for the purpose of producing useful error messages.
+        """
+        if choice['type'] == 'definitions':
+            return self._infer_discriminator_values_for_choice(choice['schema'], source_name=source_name)
+        elif choice['type'] == 'function-plain':
+            raise TypeError(
+                f'{choice["type"]!r} is not a valid discriminated union variant;'
+                ' should be a `BaseModel` or `dataclass`'
+            )
+        elif _core_utils.is_function_with_inner_schema(choice):
+            return self._infer_discriminator_values_for_choice(choice['schema'], source_name=source_name)
+        elif choice['type'] == 'lax-or-strict':
+            return sorted(
+                set(
+                    self._infer_discriminator_values_for_choice(choice['lax_schema'], source_name=None)
+                    + self._infer_discriminator_values_for_choice(choice['strict_schema'], source_name=None)
+                )
+            )
+
+        elif choice['type'] == 'tagged-union':
+            values: list[str | int] = []
+            # Ignore str/int "choices" since these are just references to other choices
+            subchoices = [x for x in choice['choices'].values() if not isinstance(x, (str, int))]
+            for subchoice in subchoices:
+                subchoice_values = self._infer_discriminator_values_for_choice(subchoice, source_name=None)
+                values.extend(subchoice_values)
+            return values
+
+        elif choice['type'] == 'union':
+            values = []
+            for subchoice in choice['choices']:
+                subchoice_schema = subchoice[0] if isinstance(subchoice, tuple) else subchoice
+                subchoice_values = self._infer_discriminator_values_for_choice(subchoice_schema, source_name=None)
+                values.extend(subchoice_values)
+            return values
+
+        elif choice['type'] == 'nullable':
+            self._should_be_nullable = True
+            return self._infer_discriminator_values_for_choice(choice['schema'], source_name=None)
+
+        elif choice['type'] == 'model':
+            return self._infer_discriminator_values_for_choice(choice['schema'], source_name=choice['cls'].__name__)
+
+        elif choice['type'] == 'dataclass':
+            return self._infer_discriminator_values_for_choice(choice['schema'], source_name=choice['cls'].__name__)
+
+        elif choice['type'] == 'model-fields':
+            return self._infer_discriminator_values_for_model_choice(choice, source_name=source_name)
+
+        elif choice['type'] == 'dataclass-args':
+            return self._infer_discriminator_values_for_dataclass_choice(choice, source_name=source_name)
+
+        elif choice['type'] == 'typed-dict':
+            return self._infer_discriminator_values_for_typed_dict_choice(choice, source_name=source_name)
+
+        elif choice['type'] == 'definition-ref':
+            schema_ref = choice['schema_ref']
+            if schema_ref not in self.definitions:
+                raise MissingDefinitionForUnionRef(schema_ref)
+            return self._infer_discriminator_values_for_choice(self.definitions[schema_ref], source_name=source_name)
+        else:
+            raise TypeError(
+                f'{choice["type"]!r} is not a valid discriminated union variant;'
+                ' should be a `BaseModel` or `dataclass`'
+            )
+
+    def _infer_discriminator_values_for_typed_dict_choice(
+        self, choice: core_schema.TypedDictSchema, source_name: str | None = None
+    ) -> list[str | int]:
+        """This method just extracts the _infer_discriminator_values_for_choice logic specific to TypedDictSchema
+        for the sake of readability.
+        """
+        source = 'TypedDict' if source_name is None else f'TypedDict {source_name!r}'
+        field = choice['fields'].get(self.discriminator)
+        if field is None:
+            raise PydanticUserError(
+                f'{source} needs a discriminator field for key {self.discriminator!r}', code='discriminator-no-field'
+            )
+        return self._infer_discriminator_values_for_field(field, source)
+
+    def _infer_discriminator_values_for_model_choice(
+        self, choice: core_schema.ModelFieldsSchema, source_name: str | None = None
+    ) -> list[str | int]:
+        source = 'ModelFields' if source_name is None else f'Model {source_name!r}'
+        field = choice['fields'].get(self.discriminator)
+        if field is None:
+            raise PydanticUserError(
+                f'{source} needs a discriminator field for key {self.discriminator!r}', code='discriminator-no-field'
+            )
+        return self._infer_discriminator_values_for_field(field, source)
+
+    def _infer_discriminator_values_for_dataclass_choice(
+        self, choice: core_schema.DataclassArgsSchema, source_name: str | None = None
+    ) -> list[str | int]:
+        source = 'DataclassArgs' if source_name is None else f'Dataclass {source_name!r}'
+        for field in choice['fields']:
+            if field['name'] == self.discriminator:
+                break
+        else:
+            raise PydanticUserError(
+                f'{source} needs a discriminator field for key {self.discriminator!r}', code='discriminator-no-field'
+            )
+        return self._infer_discriminator_values_for_field(field, source)
+
+    def _infer_discriminator_values_for_field(self, field: CoreSchemaField, source: str) -> list[str | int]:
+        if field['type'] == 'computed-field':
+            # This should never occur as a discriminator, as it is only relevant to serialization
+            return []
+        alias = field.get('validation_alias', self.discriminator)
+        if not isinstance(alias, str):
+            raise PydanticUserError(
+                f'Alias {alias!r} is not supported in a discriminated union', code='discriminator-alias-type'
+            )
+        if self._discriminator_alias is None:
+            self._discriminator_alias = alias
+        elif self._discriminator_alias != alias:
+            raise PydanticUserError(
+                f'Aliases for discriminator {self.discriminator!r} must be the same '
+                f'(got {alias}, {self._discriminator_alias})',
+                code='discriminator-alias',
+            )
+        return self._infer_discriminator_values_for_inner_schema(field['schema'], source)
+
+    def _infer_discriminator_values_for_inner_schema(
+        self, schema: core_schema.CoreSchema, source: str
+    ) -> list[str | int]:
+        """When inferring discriminator values for a field, we typically extract the expected values from a literal
+        schema. This function does that, but also handles nested unions and defaults.
+        """
+        if schema['type'] == 'literal':
+            return schema['expected']
+
+        elif schema['type'] == 'union':
+            # Generally when multiple values are allowed they should be placed in a single `Literal`, but
+            # we add this case to handle the situation where a field is annotated as a `Union` of `Literal`s.
+            # For example, this lets us handle `Union[Literal['key'], Union[Literal['Key'], Literal['KEY']]]`
+            values: list[Any] = []
+            for choice in schema['choices']:
+                choice_schema = choice[0] if isinstance(choice, tuple) else choice
+                choice_values = self._infer_discriminator_values_for_inner_schema(choice_schema, source)
+                values.extend(choice_values)
+            return values
+
+        elif schema['type'] == 'default':
+            # This will happen if the field has a default value; we ignore it while extracting the discriminator values
+            return self._infer_discriminator_values_for_inner_schema(schema['schema'], source)
+
+        elif schema['type'] == 'function-after':
+            # After validators don't affect the discriminator values
+            return self._infer_discriminator_values_for_inner_schema(schema['schema'], source)
+
+        elif schema['type'] in {'function-before', 'function-wrap', 'function-plain'}:
+            validator_type = repr(schema['type'].split('-')[1])
+            raise PydanticUserError(
+                f'Cannot use a mode={validator_type} validator in the'
+                f' discriminator field {self.discriminator!r} of {source}',
+                code='discriminator-validator',
+            )
+
+        else:
+            raise PydanticUserError(
+                f'{source} needs field {self.discriminator!r} to be of type `Literal`',
+                code='discriminator-needs-literal',
+            )
+
+    def _set_unique_choice_for_values(self, choice: core_schema.CoreSchema, values: Sequence[str | int]) -> None:
+        """This method updates `self.tagged_union_choices` so that all provided (discriminator) `values` map to the
+        provided `choice`, validating that none of these values already map to another (different) choice.
+        """
+        for discriminator_value in values:
+            if discriminator_value in self._tagged_union_choices:
+                # It is okay if `value` is already in tagged_union_choices as long as it maps to the same value.
+                # Because tagged_union_choices may map values to other values, we need to walk the choices dict
+                # until we get to a "real" choice, and confirm that is equal to the one assigned.
+                existing_choice = self._tagged_union_choices[discriminator_value]
+                if existing_choice != choice:
+                    raise TypeError(
+                        f'Value {discriminator_value!r} for discriminator '
+                        f'{self.discriminator!r} mapped to multiple choices'
+                    )
+            else:
+                self._tagged_union_choices[discriminator_value] = choice

lib/pydantic/_internal/_fields.py

@@ -0,0 +1,319 @@
+"""Private logic related to fields (the `Field()` function and `FieldInfo` class), and arguments to `Annotated`."""
+from __future__ import annotations as _annotations
+
+import dataclasses
+import sys
+import warnings
+from copy import copy
+from functools import lru_cache
+from typing import TYPE_CHECKING, Any
+
+from pydantic_core import PydanticUndefined
+
+from pydantic.errors import PydanticUserError
+
+from . import _typing_extra
+from ._config import ConfigWrapper
+from ._repr import Representation
+from ._typing_extra import get_cls_type_hints_lenient, get_type_hints, is_classvar, is_finalvar
+
+if TYPE_CHECKING:
+    from annotated_types import BaseMetadata
+
+    from ..fields import FieldInfo
+    from ..main import BaseModel
+    from ._dataclasses import StandardDataclass
+    from ._decorators import DecoratorInfos
+
+
+def get_type_hints_infer_globalns(
+    obj: Any,
+    localns: dict[str, Any] | None = None,
+    include_extras: bool = False,
+) -> dict[str, Any]:
+    """Gets type hints for an object by inferring the global namespace.
+
+    It uses the `typing.get_type_hints`, The only thing that we do here is fetching
+    global namespace from `obj.__module__` if it is not `None`.
+
+    Args:
+        obj: The object to get its type hints.
+        localns: The local namespaces.
+        include_extras: Whether to recursively include annotation metadata.
+
+    Returns:
+        The object type hints.
+    """
+    module_name = getattr(obj, '__module__', None)
+    globalns: dict[str, Any] | None = None
+    if module_name:
+        try:
+            globalns = sys.modules[module_name].__dict__
+        except KeyError:
+            # happens occasionally, see https://github.com/pydantic/pydantic/issues/2363
+            pass
+    return get_type_hints(obj, globalns=globalns, localns=localns, include_extras=include_extras)
+
+
+class PydanticMetadata(Representation):
+    """Base class for annotation markers like `Strict`."""
+
+    __slots__ = ()
+
+
+def pydantic_general_metadata(**metadata: Any) -> BaseMetadata:
+    """Create a new `_PydanticGeneralMetadata` class with the given metadata.
+
+    Args:
+        **metadata: The metadata to add.
+
+    Returns:
+        The new `_PydanticGeneralMetadata` class.
+    """
+    return _general_metadata_cls()(metadata)  # type: ignore
+
+
+@lru_cache(maxsize=None)
+def _general_metadata_cls() -> type[BaseMetadata]:
+    """Do it this way to avoid importing `annotated_types` at import time."""
+    from annotated_types import BaseMetadata
+
+    class _PydanticGeneralMetadata(PydanticMetadata, BaseMetadata):
+        """Pydantic general metadata like `max_digits`."""
+
+        def __init__(self, metadata: Any):
+            self.__dict__ = metadata
+
+    return _PydanticGeneralMetadata  # type: ignore
+
+
+def collect_model_fields(  # noqa: C901
+    cls: type[BaseModel],
+    bases: tuple[type[Any], ...],
+    config_wrapper: ConfigWrapper,
+    types_namespace: dict[str, Any] | None,
+    *,
+    typevars_map: dict[Any, Any] | None = None,
+) -> tuple[dict[str, FieldInfo], set[str]]:
+    """Collect the fields of a nascent pydantic model.
+
+    Also collect the names of any ClassVars present in the type hints.
+
+    The returned value is a tuple of two items: the fields dict, and the set of ClassVar names.
+
+    Args:
+        cls: BaseModel or dataclass.
+        bases: Parents of the class, generally `cls.__bases__`.
+        config_wrapper: The config wrapper instance.
+        types_namespace: Optional extra namespace to look for types in.
+        typevars_map: A dictionary mapping type variables to their concrete types.
+
+    Returns:
+        A tuple contains fields and class variables.
+
+    Raises:
+        NameError:
+            - If there is a conflict between a field name and protected namespaces.
+            - If there is a field other than `root` in `RootModel`.
+            - If a field shadows an attribute in the parent model.
+    """
+    from ..fields import FieldInfo
+
+    type_hints = get_cls_type_hints_lenient(cls, types_namespace)
+
+    # https://docs.python.org/3/howto/annotations.html#accessing-the-annotations-dict-of-an-object-in-python-3-9-and-older
+    # annotations is only used for finding fields in parent classes
+    annotations = cls.__dict__.get('__annotations__', {})
+    fields: dict[str, FieldInfo] = {}
+
+    class_vars: set[str] = set()
+    for ann_name, ann_type in type_hints.items():
+        if ann_name == 'model_config':
+            # We never want to treat `model_config` as a field
+            # Note: we may need to change this logic if/when we introduce a `BareModel` class with no
+            # protected namespaces (where `model_config` might be allowed as a field name)
+            continue
+        for protected_namespace in config_wrapper.protected_namespaces:
+            if ann_name.startswith(protected_namespace):
+                for b in bases:
+                    if hasattr(b, ann_name):
+                        from ..main import BaseModel
+
+                        if not (issubclass(b, BaseModel) and ann_name in b.model_fields):
+                            raise NameError(
+                                f'Field "{ann_name}" conflicts with member {getattr(b, ann_name)}'
+                                f' of protected namespace "{protected_namespace}".'
+                            )
+                else:
+                    valid_namespaces = tuple(
+                        x for x in config_wrapper.protected_namespaces if not ann_name.startswith(x)
+                    )
+                    warnings.warn(
+                        f'Field "{ann_name}" has conflict with protected namespace "{protected_namespace}".'
+                        '\n\nYou may be able to resolve this warning by setting'
+                        f" `model_config['protected_namespaces'] = {valid_namespaces}`.",
+                        UserWarning,
+                    )
+        if is_classvar(ann_type):
+            class_vars.add(ann_name)
+            continue
+        if _is_finalvar_with_default_val(ann_type, getattr(cls, ann_name, PydanticUndefined)):
+            class_vars.add(ann_name)
+            continue
+        if not is_valid_field_name(ann_name):
+            continue
+        if cls.__pydantic_root_model__ and ann_name != 'root':
+            raise NameError(
+                f"Unexpected field with name {ann_name!r}; only 'root' is allowed as a field of a `RootModel`"
+            )
+
+        # when building a generic model with `MyModel[int]`, the generic_origin check makes sure we don't get
+        # "... shadows an attribute" errors
+        generic_origin = getattr(cls, '__pydantic_generic_metadata__', {}).get('origin')
+        for base in bases:
+            dataclass_fields = {
+                field.name for field in (dataclasses.fields(base) if dataclasses.is_dataclass(base) else ())
+            }
+            if hasattr(base, ann_name):
+                if base is generic_origin:
+                    # Don't error when "shadowing" of attributes in parametrized generics
+                    continue
+
+                if ann_name in dataclass_fields:
+                    # Don't error when inheriting stdlib dataclasses whose fields are "shadowed" by defaults being set
+                    # on the class instance.
+                    continue
+                warnings.warn(
+                    f'Field name "{ann_name}" shadows an attribute in parent "{base.__qualname__}"; ',
+                    UserWarning,
+                )
+
+        try:
+            default = getattr(cls, ann_name, PydanticUndefined)
+            if default is PydanticUndefined:
+                raise AttributeError
+        except AttributeError:
+            if ann_name in annotations:
+                field_info = FieldInfo.from_annotation(ann_type)
+            else:
+                # if field has no default value and is not in __annotations__ this means that it is
+                # defined in a base class and we can take it from there
+                model_fields_lookup: dict[str, FieldInfo] = {}
+                for x in cls.__bases__[::-1]:
+                    model_fields_lookup.update(getattr(x, 'model_fields', {}))
+                if ann_name in model_fields_lookup:
+                    # The field was present on one of the (possibly multiple) base classes
+                    # copy the field to make sure typevar substitutions don't cause issues with the base classes
+                    field_info = copy(model_fields_lookup[ann_name])
+                else:
+                    # The field was not found on any base classes; this seems to be caused by fields not getting
+                    # generated thanks to models not being fully defined while initializing recursive models.
+                    # Nothing stops us from just creating a new FieldInfo for this type hint, so we do this.
+                    field_info = FieldInfo.from_annotation(ann_type)
+        else:
+            field_info = FieldInfo.from_annotated_attribute(ann_type, default)
+            # attributes which are fields are removed from the class namespace:
+            # 1. To match the behaviour of annotation-only fields
+            # 2. To avoid false positives in the NameError check above
+            try:
+                delattr(cls, ann_name)
+            except AttributeError:
+                pass  # indicates the attribute was on a parent class
+
+        # Use cls.__dict__['__pydantic_decorators__'] instead of cls.__pydantic_decorators__
+        # to make sure the decorators have already been built for this exact class
+        decorators: DecoratorInfos = cls.__dict__['__pydantic_decorators__']
+        if ann_name in decorators.computed_fields:
+            raise ValueError("you can't override a field with a computed field")
+        fields[ann_name] = field_info
+
+    if typevars_map:
+        for field in fields.values():
+            field.apply_typevars_map(typevars_map, types_namespace)
+
+    return fields, class_vars
+
+
+def _is_finalvar_with_default_val(type_: type[Any], val: Any) -> bool:
+    from ..fields import FieldInfo
+
+    if not is_finalvar(type_):
+        return False
+    elif val is PydanticUndefined:
+        return False
+    elif isinstance(val, FieldInfo) and (val.default is PydanticUndefined and val.default_factory is None):
+        return False
+    else:
+        return True
+
+
+def collect_dataclass_fields(
+    cls: type[StandardDataclass], types_namespace: dict[str, Any] | None, *, typevars_map: dict[Any, Any] | None = None
+) -> dict[str, FieldInfo]:
+    """Collect the fields of a dataclass.
+
+    Args:
+        cls: dataclass.
+        types_namespace: Optional extra namespace to look for types in.
+        typevars_map: A dictionary mapping type variables to their concrete types.
+
+    Returns:
+        The dataclass fields.
+    """
+    from ..fields import FieldInfo
+
+    fields: dict[str, FieldInfo] = {}
+    dataclass_fields: dict[str, dataclasses.Field] = cls.__dataclass_fields__
+    cls_localns = dict(vars(cls))  # this matches get_cls_type_hints_lenient, but all tests pass with `= None` instead
+
+    source_module = sys.modules.get(cls.__module__)
+    if source_module is not None:
+        types_namespace = {**source_module.__dict__, **(types_namespace or {})}
+
+    for ann_name, dataclass_field in dataclass_fields.items():
+        ann_type = _typing_extra.eval_type_lenient(dataclass_field.type, types_namespace, cls_localns)
+        if is_classvar(ann_type):
+            continue
+
+        if (
+            not dataclass_field.init
+            and dataclass_field.default == dataclasses.MISSING
+            and dataclass_field.default_factory == dataclasses.MISSING
+        ):
+            # TODO: We should probably do something with this so that validate_assignment behaves properly
+            #   Issue: https://github.com/pydantic/pydantic/issues/5470
+            continue
+
+        if isinstance(dataclass_field.default, FieldInfo):
+            if dataclass_field.default.init_var:
+                if dataclass_field.default.init is False:
+                    raise PydanticUserError(
+                        f'Dataclass field {ann_name} has init=False and init_var=True, but these are mutually exclusive.',
+                        code='clashing-init-and-init-var',
+                    )
+
+                # TODO: same note as above re validate_assignment
+                continue
+            field_info = FieldInfo.from_annotated_attribute(ann_type, dataclass_field.default)
+        else:
+            field_info = FieldInfo.from_annotated_attribute(ann_type, dataclass_field)
+
+        fields[ann_name] = field_info
+
+        if field_info.default is not PydanticUndefined and isinstance(getattr(cls, ann_name, field_info), FieldInfo):
+            # We need this to fix the default when the "default" from __dataclass_fields__ is a pydantic.FieldInfo
+            setattr(cls, ann_name, field_info.default)
+
+    if typevars_map:
+        for field in fields.values():
+            field.apply_typevars_map(typevars_map, types_namespace)
+
+    return fields
+
+
+def is_valid_field_name(name: str) -> bool:
+    return not name.startswith('_')
+
+
+def is_valid_privateattr_name(name: str) -> bool:
+    return name.startswith('_') and not name.startswith('__')

lib/pydantic/_internal/_forward_ref.py

@@ -0,0 +1,23 @@
+from __future__ import annotations as _annotations
+
+from dataclasses import dataclass
+from typing import Union
+
+
+@dataclass
+class PydanticRecursiveRef:
+    type_ref: str
+
+    __name__ = 'PydanticRecursiveRef'
+    __hash__ = object.__hash__
+
+    def __call__(self) -> None:
+        """Defining __call__ is necessary for the `typing` module to let you use an instance of
+        this class as the result of resolving a standard ForwardRef.
+        """
+
+    def __or__(self, other):
+        return Union[self, other]  # type: ignore
+
+    def __ror__(self, other):
+        return Union[other, self]  # type: ignore

lib/pydantic/_internal/_generics.py

@@ -0,0 +1,517 @@
+from __future__ import annotations
+
+import sys
+import types
+import typing
+from collections import ChainMap
+from contextlib import contextmanager
+from contextvars import ContextVar
+from types import prepare_class
+from typing import TYPE_CHECKING, Any, Iterator, List, Mapping, MutableMapping, Tuple, TypeVar
+from weakref import WeakValueDictionary
+
+import typing_extensions
+
+from ._core_utils import get_type_ref
+from ._forward_ref import PydanticRecursiveRef
+from ._typing_extra import TypeVarType, typing_base
+from ._utils import all_identical, is_model_class
+
+if sys.version_info >= (3, 10):
+    from typing import _UnionGenericAlias  # type: ignore[attr-defined]
+
+if TYPE_CHECKING:
+    from ..main import BaseModel
+
+GenericTypesCacheKey = Tuple[Any, Any, Tuple[Any, ...]]
+
+# Note: We want to remove LimitedDict, but to do this, we'd need to improve the handling of generics caching.
+#   Right now, to handle recursive generics, we some types must remain cached for brief periods without references.
+#   By chaining the WeakValuesDict with a LimitedDict, we have a way to retain caching for all types with references,
+#   while also retaining a limited number of types even without references. This is generally enough to build
+#   specific recursive generic models without losing required items out of the cache.
+
+KT = TypeVar('KT')
+VT = TypeVar('VT')
+_LIMITED_DICT_SIZE = 100
+if TYPE_CHECKING:
+
+    class LimitedDict(dict, MutableMapping[KT, VT]):
+        def __init__(self, size_limit: int = _LIMITED_DICT_SIZE):
+            ...
+
+else:
+
+    class LimitedDict(dict):
+        """Limit the size/length of a dict used for caching to avoid unlimited increase in memory usage.
+
+        Since the dict is ordered, and we always remove elements from the beginning, this is effectively a FIFO cache.
+        """
+
+        def __init__(self, size_limit: int = _LIMITED_DICT_SIZE):
+            self.size_limit = size_limit
+            super().__init__()
+
+        def __setitem__(self, __key: Any, __value: Any) -> None:
+            super().__setitem__(__key, __value)
+            if len(self) > self.size_limit:
+                excess = len(self) - self.size_limit + self.size_limit // 10
+                to_remove = list(self.keys())[:excess]
+                for key in to_remove:
+                    del self[key]
+
+
+# weak dictionaries allow the dynamically created parametrized versions of generic models to get collected
+# once they are no longer referenced by the caller.
+if sys.version_info >= (3, 9):  # Typing for weak dictionaries available at 3.9
+    GenericTypesCache = WeakValueDictionary[GenericTypesCacheKey, 'type[BaseModel]']
+else:
+    GenericTypesCache = WeakValueDictionary
+
+if TYPE_CHECKING:
+
+    class DeepChainMap(ChainMap[KT, VT]):  # type: ignore
+        ...
+
+else:
+
+    class DeepChainMap(ChainMap):
+        """Variant of ChainMap that allows direct updates to inner scopes.
+
+        Taken from https://docs.python.org/3/library/collections.html#collections.ChainMap,
+        with some light modifications for this use case.
+        """
+
+        def clear(self) -> None:
+            for mapping in self.maps:
+                mapping.clear()
+
+        def __setitem__(self, key: KT, value: VT) -> None:
+            for mapping in self.maps:
+                mapping[key] = value
+
+        def __delitem__(self, key: KT) -> None:
+            hit = False
+            for mapping in self.maps:
+                if key in mapping:
+                    del mapping[key]
+                    hit = True
+            if not hit:
+                raise KeyError(key)
+
+
+# Despite the fact that LimitedDict _seems_ no longer necessary, I'm very nervous to actually remove it
+# and discover later on that we need to re-add all this infrastructure...
+# _GENERIC_TYPES_CACHE = DeepChainMap(GenericTypesCache(), LimitedDict())
+
+_GENERIC_TYPES_CACHE = GenericTypesCache()
+
+
+class PydanticGenericMetadata(typing_extensions.TypedDict):
+    origin: type[BaseModel] | None  # analogous to typing._GenericAlias.__origin__
+    args: tuple[Any, ...]  # analogous to typing._GenericAlias.__args__
+    parameters: tuple[type[Any], ...]  # analogous to typing.Generic.__parameters__
+
+
+def create_generic_submodel(
+    model_name: str, origin: type[BaseModel], args: tuple[Any, ...], params: tuple[Any, ...]
+) -> type[BaseModel]:
+    """Dynamically create a submodel of a provided (generic) BaseModel.
+
+    This is used when producing concrete parametrizations of generic models. This function
+    only *creates* the new subclass; the schema/validators/serialization must be updated to
+    reflect a concrete parametrization elsewhere.
+
+    Args:
+        model_name: The name of the newly created model.
+        origin: The base class for the new model to inherit from.
+        args: A tuple of generic metadata arguments.
+        params: A tuple of generic metadata parameters.
+
+    Returns:
+        The created submodel.
+    """
+    namespace: dict[str, Any] = {'__module__': origin.__module__}
+    bases = (origin,)
+    meta, ns, kwds = prepare_class(model_name, bases)
+    namespace.update(ns)
+    created_model = meta(
+        model_name,
+        bases,
+        namespace,
+        __pydantic_generic_metadata__={
+            'origin': origin,
+            'args': args,
+            'parameters': params,
+        },
+        __pydantic_reset_parent_namespace__=False,
+        **kwds,
+    )
+
+    model_module, called_globally = _get_caller_frame_info(depth=3)
+    if called_globally:  # create global reference and therefore allow pickling
+        object_by_reference = None
+        reference_name = model_name
+        reference_module_globals = sys.modules[created_model.__module__].__dict__
+        while object_by_reference is not created_model:
+            object_by_reference = reference_module_globals.setdefault(reference_name, created_model)
+            reference_name += '_'
+
+    return created_model
+
+
+def _get_caller_frame_info(depth: int = 2) -> tuple[str | None, bool]:
+    """Used inside a function to check whether it was called globally.
+
+    Args:
+        depth: The depth to get the frame.
+
+    Returns:
+        A tuple contains `module_name` and `called_globally`.
+
+    Raises:
+        RuntimeError: If the function is not called inside a function.
+    """
+    try:
+        previous_caller_frame = sys._getframe(depth)
+    except ValueError as e:
+        raise RuntimeError('This function must be used inside another function') from e
+    except AttributeError:  # sys module does not have _getframe function, so there's nothing we can do about it
+        return None, False
+    frame_globals = previous_caller_frame.f_globals
+    return frame_globals.get('__name__'), previous_caller_frame.f_locals is frame_globals
+
+
+DictValues: type[Any] = {}.values().__class__
+
+
+def iter_contained_typevars(v: Any) -> Iterator[TypeVarType]:
+    """Recursively iterate through all subtypes and type args of `v` and yield any typevars that are found.
+
+    This is inspired as an alternative to directly accessing the `__parameters__` attribute of a GenericAlias,
+    since __parameters__ of (nested) generic BaseModel subclasses won't show up in that list.
+    """
+    if isinstance(v, TypeVar):
+        yield v
+    elif is_model_class(v):
+        yield from v.__pydantic_generic_metadata__['parameters']
+    elif isinstance(v, (DictValues, list)):
+        for var in v:
+            yield from iter_contained_typevars(var)
+    else:
+        args = get_args(v)
+        for arg in args:
+            yield from iter_contained_typevars(arg)
+
+
+def get_args(v: Any) -> Any:
+    pydantic_generic_metadata: PydanticGenericMetadata | None = getattr(v, '__pydantic_generic_metadata__', None)
+    if pydantic_generic_metadata:
+        return pydantic_generic_metadata.get('args')
+    return typing_extensions.get_args(v)
+
+
+def get_origin(v: Any) -> Any:
+    pydantic_generic_metadata: PydanticGenericMetadata | None = getattr(v, '__pydantic_generic_metadata__', None)
+    if pydantic_generic_metadata:
+        return pydantic_generic_metadata.get('origin')
+    return typing_extensions.get_origin(v)
+
+
+def get_standard_typevars_map(cls: type[Any]) -> dict[TypeVarType, Any] | None:
+    """Package a generic type's typevars and parametrization (if present) into a dictionary compatible with the
+    `replace_types` function. Specifically, this works with standard typing generics and typing._GenericAlias.
+    """
+    origin = get_origin(cls)
+    if origin is None:
+        return None
+    if not hasattr(origin, '__parameters__'):
+        return None
+
+    # In this case, we know that cls is a _GenericAlias, and origin is the generic type
+    # So it is safe to access cls.__args__ and origin.__parameters__
+    args: tuple[Any, ...] = cls.__args__  # type: ignore
+    parameters: tuple[TypeVarType, ...] = origin.__parameters__
+    return dict(zip(parameters, args))
+
+
+def get_model_typevars_map(cls: type[BaseModel]) -> dict[TypeVarType, Any] | None:
+    """Package a generic BaseModel's typevars and concrete parametrization (if present) into a dictionary compatible
+    with the `replace_types` function.
+
+    Since BaseModel.__class_getitem__ does not produce a typing._GenericAlias, and the BaseModel generic info is
+    stored in the __pydantic_generic_metadata__ attribute, we need special handling here.
+    """
+    # TODO: This could be unified with `get_standard_typevars_map` if we stored the generic metadata
+    #   in the __origin__, __args__, and __parameters__ attributes of the model.
+    generic_metadata = cls.__pydantic_generic_metadata__
+    origin = generic_metadata['origin']
+    args = generic_metadata['args']
+    return dict(zip(iter_contained_typevars(origin), args))
+
+
+def replace_types(type_: Any, type_map: Mapping[Any, Any] | None) -> Any:
+    """Return type with all occurrences of `type_map` keys recursively replaced with their values.
+
+    Args:
+        type_: The class or generic alias.
+        type_map: Mapping from `TypeVar` instance to concrete types.
+
+    Returns:
+        A new type representing the basic structure of `type_` with all
+        `typevar_map` keys recursively replaced.
+
+    Example:
+        ```py
+        from typing import List, Tuple, Union
+
+        from pydantic._internal._generics import replace_types
+
+        replace_types(Tuple[str, Union[List[str], float]], {str: int})
+        #> Tuple[int, Union[List[int], float]]
+        ```
+    """
+    if not type_map:
+        return type_
+
+    type_args = get_args(type_)
+    origin_type = get_origin(type_)
+
+    if origin_type is typing_extensions.Annotated:
+        annotated_type, *annotations = type_args
+        annotated = replace_types(annotated_type, type_map)
+        for annotation in annotations:
+            annotated = typing_extensions.Annotated[annotated, annotation]
+        return annotated
+
+    # Having type args is a good indicator that this is a typing module
+    # class instantiation or a generic alias of some sort.
+    if type_args:
+        resolved_type_args = tuple(replace_types(arg, type_map) for arg in type_args)
+        if all_identical(type_args, resolved_type_args):
+            # If all arguments are the same, there is no need to modify the
+            # type or create a new object at all
+            return type_
+        if (
+            origin_type is not None
+            and isinstance(type_, typing_base)
+            and not isinstance(origin_type, typing_base)
+            and getattr(type_, '_name', None) is not None
+        ):
+            # In python < 3.9 generic aliases don't exist so any of these like `list`,
+            # `type` or `collections.abc.Callable` need to be translated.
+            # See: https://www.python.org/dev/peps/pep-0585
+            origin_type = getattr(typing, type_._name)
+        assert origin_type is not None
+        # PEP-604 syntax (Ex.: list | str) is represented with a types.UnionType object that does not have __getitem__.
+        # We also cannot use isinstance() since we have to compare types.
+        if sys.version_info >= (3, 10) and origin_type is types.UnionType:
+            return _UnionGenericAlias(origin_type, resolved_type_args)
+        # NotRequired[T] and Required[T] don't support tuple type resolved_type_args, hence the condition below
+        return origin_type[resolved_type_args[0] if len(resolved_type_args) == 1 else resolved_type_args]
+
+    # We handle pydantic generic models separately as they don't have the same
+    # semantics as "typing" classes or generic aliases
+
+    if not origin_type and is_model_class(type_):
+        parameters = type_.__pydantic_generic_metadata__['parameters']
+        if not parameters:
+            return type_
+        resolved_type_args = tuple(replace_types(t, type_map) for t in parameters)
+        if all_identical(parameters, resolved_type_args):
+            return type_
+        return type_[resolved_type_args]
+
+    # Handle special case for typehints that can have lists as arguments.
+    # `typing.Callable[[int, str], int]` is an example for this.
+    if isinstance(type_, (List, list)):
+        resolved_list = list(replace_types(element, type_map) for element in type_)
+        if all_identical(type_, resolved_list):
+            return type_
+        return resolved_list
+
+    # If all else fails, we try to resolve the type directly and otherwise just
+    # return the input with no modifications.
+    return type_map.get(type_, type_)
+
+
+def has_instance_in_type(type_: Any, isinstance_target: Any) -> bool:
+    """Checks if the type, or any of its arbitrary nested args, satisfy
+    `isinstance(<type>, isinstance_target)`.
+    """
+    if isinstance(type_, isinstance_target):
+        return True
+
+    type_args = get_args(type_)
+    origin_type = get_origin(type_)
+
+    if origin_type is typing_extensions.Annotated:
+        annotated_type, *annotations = type_args
+        return has_instance_in_type(annotated_type, isinstance_target)
+
+    # Having type args is a good indicator that this is a typing module
+    # class instantiation or a generic alias of some sort.
+    if any(has_instance_in_type(a, isinstance_target) for a in type_args):
+        return True
+
+    # Handle special case for typehints that can have lists as arguments.
+    # `typing.Callable[[int, str], int]` is an example for this.
+    if isinstance(type_, (List, list)) and not isinstance(type_, typing_extensions.ParamSpec):
+        if any(has_instance_in_type(element, isinstance_target) for element in type_):
+            return True
+
+    return False
+
+
+def check_parameters_count(cls: type[BaseModel], parameters: tuple[Any, ...]) -> None:
+    """Check the generic model parameters count is equal.
+
+    Args:
+        cls: The generic model.
+        parameters: A tuple of passed parameters to the generic model.
+
+    Raises:
+        TypeError: If the passed parameters count is not equal to generic model parameters count.
+    """
+    actual = len(parameters)
+    expected = len(cls.__pydantic_generic_metadata__['parameters'])
+    if actual != expected:
+        description = 'many' if actual > expected else 'few'
+        raise TypeError(f'Too {description} parameters for {cls}; actual {actual}, expected {expected}')
+
+
+_generic_recursion_cache: ContextVar[set[str] | None] = ContextVar('_generic_recursion_cache', default=None)
+
+
+@contextmanager
+def generic_recursion_self_type(
+    origin: type[BaseModel], args: tuple[Any, ...]
+) -> Iterator[PydanticRecursiveRef | None]:
+    """This contextmanager should be placed around the recursive calls used to build a generic type,
+    and accept as arguments the generic origin type and the type arguments being passed to it.
+
+    If the same origin and arguments are observed twice, it implies that a self-reference placeholder
+    can be used while building the core schema, and will produce a schema_ref that will be valid in the
+    final parent schema.
+    """
+    previously_seen_type_refs = _generic_recursion_cache.get()
+    if previously_seen_type_refs is None:
+        previously_seen_type_refs = set()
+        token = _generic_recursion_cache.set(previously_seen_type_refs)
+    else:
+        token = None
+
+    try:
+        type_ref = get_type_ref(origin, args_override=args)
+        if type_ref in previously_seen_type_refs:
+            self_type = PydanticRecursiveRef(type_ref=type_ref)
+            yield self_type
+        else:
+            previously_seen_type_refs.add(type_ref)
+            yield None
+    finally:
+        if token:
+            _generic_recursion_cache.reset(token)
+
+
+def recursively_defined_type_refs() -> set[str]:
+    visited = _generic_recursion_cache.get()
+    if not visited:
+        return set()  # not in a generic recursion, so there are no types
+
+    return visited.copy()  # don't allow modifications
+
+
+def get_cached_generic_type_early(parent: type[BaseModel], typevar_values: Any) -> type[BaseModel] | None:
+    """The use of a two-stage cache lookup approach was necessary to have the highest performance possible for
+    repeated calls to `__class_getitem__` on generic types (which may happen in tighter loops during runtime),
+    while still ensuring that certain alternative parametrizations ultimately resolve to the same type.
+
+    As a concrete example, this approach was necessary to make Model[List[T]][int] equal to Model[List[int]].
+    The approach could be modified to not use two different cache keys at different points, but the
+    _early_cache_key is optimized to be as quick to compute as possible (for repeated-access speed), and the
+    _late_cache_key is optimized to be as "correct" as possible, so that two types that will ultimately be the
+    same after resolving the type arguments will always produce cache hits.
+
+    If we wanted to move to only using a single cache key per type, we would either need to always use the
+    slower/more computationally intensive logic associated with _late_cache_key, or would need to accept
+    that Model[List[T]][int] is a different type than Model[List[T]][int]. Because we rely on subclass relationships
+    during validation, I think it is worthwhile to ensure that types that are functionally equivalent are actually
+    equal.
+    """
+    return _GENERIC_TYPES_CACHE.get(_early_cache_key(parent, typevar_values))
+
+
+def get_cached_generic_type_late(
+    parent: type[BaseModel], typevar_values: Any, origin: type[BaseModel], args: tuple[Any, ...]
+) -> type[BaseModel] | None:
+    """See the docstring of `get_cached_generic_type_early` for more information about the two-stage cache lookup."""
+    cached = _GENERIC_TYPES_CACHE.get(_late_cache_key(origin, args, typevar_values))
+    if cached is not None:
+        set_cached_generic_type(parent, typevar_values, cached, origin, args)
+    return cached
+
+
+def set_cached_generic_type(
+    parent: type[BaseModel],
+    typevar_values: tuple[Any, ...],
+    type_: type[BaseModel],
+    origin: type[BaseModel] | None = None,
+    args: tuple[Any, ...] | None = None,
+) -> None:
+    """See the docstring of `get_cached_generic_type_early` for more information about why items are cached with
+    two different keys.
+    """
+    _GENERIC_TYPES_CACHE[_early_cache_key(parent, typevar_values)] = type_
+    if len(typevar_values) == 1:
+        _GENERIC_TYPES_CACHE[_early_cache_key(parent, typevar_values[0])] = type_
+    if origin and args:
+        _GENERIC_TYPES_CACHE[_late_cache_key(origin, args, typevar_values)] = type_
+
+
+def _union_orderings_key(typevar_values: Any) -> Any:
+    """This is intended to help differentiate between Union types with the same arguments in different order.
+
+    Thanks to caching internal to the `typing` module, it is not possible to distinguish between
+    List[Union[int, float]] and List[Union[float, int]] (and similarly for other "parent" origins besides List)
+    because `typing` considers Union[int, float] to be equal to Union[float, int].
+
+    However, you _can_ distinguish between (top-level) Union[int, float] vs. Union[float, int].
+    Because we parse items as the first Union type that is successful, we get slightly more consistent behavior
+    if we make an effort to distinguish the ordering of items in a union. It would be best if we could _always_
+    get the exact-correct order of items in the union, but that would require a change to the `typing` module itself.
+    (See https://github.com/python/cpython/issues/86483 for reference.)
+    """
+    if isinstance(typevar_values, tuple):
+        args_data = []
+        for value in typevar_values:
+            args_data.append(_union_orderings_key(value))
+        return tuple(args_data)
+    elif typing_extensions.get_origin(typevar_values) is typing.Union:
+        return get_args(typevar_values)
+    else:
+        return ()
+
+
+def _early_cache_key(cls: type[BaseModel], typevar_values: Any) -> GenericTypesCacheKey:
+    """This is intended for minimal computational overhead during lookups of cached types.
+
+    Note that this is overly simplistic, and it's possible that two different cls/typevar_values
+    inputs would ultimately result in the same type being created in BaseModel.__class_getitem__.
+    To handle this, we have a fallback _late_cache_key that is checked later if the _early_cache_key
+    lookup fails, and should result in a cache hit _precisely_ when the inputs to __class_getitem__
+    would result in the same type.
+    """
+    return cls, typevar_values, _union_orderings_key(typevar_values)
+
+
+def _late_cache_key(origin: type[BaseModel], args: tuple[Any, ...], typevar_values: Any) -> GenericTypesCacheKey:
+    """This is intended for use later in the process of creating a new type, when we have more information
+    about the exact args that will be passed. If it turns out that a different set of inputs to
+    __class_getitem__ resulted in the same inputs to the generic type creation process, we can still
+    return the cached type, and update the cache with the _early_cache_key as well.
+    """
+    # The _union_orderings_key is placed at the start here to ensure there cannot be a collision with an
+    # _early_cache_key, as that function will always produce a BaseModel subclass as the first item in the key,
+    # whereas this function will always produce a tuple as the first item in the key.
+    return _union_orderings_key(typevar_values), origin, args

lib/pydantic/_internal/_git.py

@@ -0,0 +1,26 @@
+"""Git utilities, adopted from mypy's git utilities (https://github.com/python/mypy/blob/master/mypy/git.py)."""
+from __future__ import annotations
+
+import os
+import subprocess
+
+
+def is_git_repo(dir: str) -> bool:
+    """Is the given directory version-controlled with git?"""
+    return os.path.exists(os.path.join(dir, '.git'))
+
+
+def have_git() -> bool:
+    """Can we run the git executable?"""
+    try:
+        subprocess.check_output(['git', '--help'])
+        return True
+    except subprocess.CalledProcessError:
+        return False
+    except OSError:
+        return False
+
+
+def git_revision(dir: str) -> str:
+    """Get the SHA-1 of the HEAD of a git repository."""
+    return subprocess.check_output(['git', 'rev-parse', '--short', 'HEAD'], cwd=dir).decode('utf-8').strip()

lib/pydantic/_internal/_internal_dataclass.py

@@ -0,0 +1,10 @@
+import sys
+from typing import Any, Dict
+
+dataclass_kwargs: Dict[str, Any]
+
+# `slots` is available on Python >= 3.10
+if sys.version_info >= (3, 10):
+    slots_true = {'slots': True}
+else:
+    slots_true = {}

lib/pydantic/_internal/_known_annotated_metadata.py

@@ -0,0 +1,410 @@
+from __future__ import annotations
+
+from collections import defaultdict
+from copy import copy
+from functools import partial
+from typing import TYPE_CHECKING, Any, Callable, Iterable
+
+from pydantic_core import CoreSchema, PydanticCustomError, to_jsonable_python
+from pydantic_core import core_schema as cs
+
+from ._fields import PydanticMetadata
+
+if TYPE_CHECKING:
+    from ..annotated_handlers import GetJsonSchemaHandler
+
+
+STRICT = {'strict'}
+SEQUENCE_CONSTRAINTS = {'min_length', 'max_length'}
+INEQUALITY = {'le', 'ge', 'lt', 'gt'}
+NUMERIC_CONSTRAINTS = {'multiple_of', 'allow_inf_nan', *INEQUALITY}
+
+STR_CONSTRAINTS = {*SEQUENCE_CONSTRAINTS, *STRICT, 'strip_whitespace', 'to_lower', 'to_upper', 'pattern'}
+BYTES_CONSTRAINTS = {*SEQUENCE_CONSTRAINTS, *STRICT}
+
+LIST_CONSTRAINTS = {*SEQUENCE_CONSTRAINTS, *STRICT}
+TUPLE_CONSTRAINTS = {*SEQUENCE_CONSTRAINTS, *STRICT}
+SET_CONSTRAINTS = {*SEQUENCE_CONSTRAINTS, *STRICT}
+DICT_CONSTRAINTS = {*SEQUENCE_CONSTRAINTS, *STRICT}
+GENERATOR_CONSTRAINTS = {*SEQUENCE_CONSTRAINTS, *STRICT}
+
+FLOAT_CONSTRAINTS = {*NUMERIC_CONSTRAINTS, *STRICT}
+INT_CONSTRAINTS = {*NUMERIC_CONSTRAINTS, *STRICT}
+BOOL_CONSTRAINTS = STRICT
+UUID_CONSTRAINTS = STRICT
+
+DATE_TIME_CONSTRAINTS = {*NUMERIC_CONSTRAINTS, *STRICT}
+TIMEDELTA_CONSTRAINTS = {*NUMERIC_CONSTRAINTS, *STRICT}
+TIME_CONSTRAINTS = {*NUMERIC_CONSTRAINTS, *STRICT}
+LAX_OR_STRICT_CONSTRAINTS = STRICT
+
+UNION_CONSTRAINTS = {'union_mode'}
+URL_CONSTRAINTS = {
+    'max_length',
+    'allowed_schemes',
+    'host_required',
+    'default_host',
+    'default_port',
+    'default_path',
+}
+
+TEXT_SCHEMA_TYPES = ('str', 'bytes', 'url', 'multi-host-url')
+SEQUENCE_SCHEMA_TYPES = ('list', 'tuple', 'set', 'frozenset', 'generator', *TEXT_SCHEMA_TYPES)
+NUMERIC_SCHEMA_TYPES = ('float', 'int', 'date', 'time', 'timedelta', 'datetime')
+
+CONSTRAINTS_TO_ALLOWED_SCHEMAS: dict[str, set[str]] = defaultdict(set)
+for constraint in STR_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(TEXT_SCHEMA_TYPES)
+for constraint in BYTES_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('bytes',))
+for constraint in LIST_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('list',))
+for constraint in TUPLE_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('tuple',))
+for constraint in SET_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('set', 'frozenset'))
+for constraint in DICT_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('dict',))
+for constraint in GENERATOR_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('generator',))
+for constraint in FLOAT_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('float',))
+for constraint in INT_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('int',))
+for constraint in DATE_TIME_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('date', 'time', 'datetime'))
+for constraint in TIMEDELTA_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('timedelta',))
+for constraint in TIME_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('time',))
+for schema_type in (*TEXT_SCHEMA_TYPES, *SEQUENCE_SCHEMA_TYPES, *NUMERIC_SCHEMA_TYPES, 'typed-dict', 'model'):
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS['strict'].add(schema_type)
+for constraint in UNION_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('union',))
+for constraint in URL_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('url', 'multi-host-url'))
+for constraint in BOOL_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('bool',))
+for constraint in UUID_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('uuid',))
+for constraint in LAX_OR_STRICT_CONSTRAINTS:
+    CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint].update(('lax-or-strict',))
+
+
+def add_js_update_schema(s: cs.CoreSchema, f: Callable[[], dict[str, Any]]) -> None:
+    def update_js_schema(s: cs.CoreSchema, handler: GetJsonSchemaHandler) -> dict[str, Any]:
+        js_schema = handler(s)
+        js_schema.update(f())
+        return js_schema
+
+    if 'metadata' in s:
+        metadata = s['metadata']
+        if 'pydantic_js_functions' in s:
+            metadata['pydantic_js_functions'].append(update_js_schema)
+        else:
+            metadata['pydantic_js_functions'] = [update_js_schema]
+    else:
+        s['metadata'] = {'pydantic_js_functions': [update_js_schema]}
+
+
+def as_jsonable_value(v: Any) -> Any:
+    if type(v) not in (int, str, float, bytes, bool, type(None)):
+        return to_jsonable_python(v)
+    return v
+
+
+def expand_grouped_metadata(annotations: Iterable[Any]) -> Iterable[Any]:
+    """Expand the annotations.
+
+    Args:
+        annotations: An iterable of annotations.
+
+    Returns:
+        An iterable of expanded annotations.
+
+    Example:
+        ```py
+        from annotated_types import Ge, Len
+
+        from pydantic._internal._known_annotated_metadata import expand_grouped_metadata
+
+        print(list(expand_grouped_metadata([Ge(4), Len(5)])))
+        #> [Ge(ge=4), MinLen(min_length=5)]
+        ```
+    """
+    import annotated_types as at
+
+    from pydantic.fields import FieldInfo  # circular import
+
+    for annotation in annotations:
+        if isinstance(annotation, at.GroupedMetadata):
+            yield from annotation
+        elif isinstance(annotation, FieldInfo):
+            yield from annotation.metadata
+            # this is a bit problematic in that it results in duplicate metadata
+            # all of our "consumers" can handle it, but it is not ideal
+            # we probably should split up FieldInfo into:
+            # - annotated types metadata
+            # - individual metadata known only to Pydantic
+            annotation = copy(annotation)
+            annotation.metadata = []
+            yield annotation
+        else:
+            yield annotation
+
+
+def apply_known_metadata(annotation: Any, schema: CoreSchema) -> CoreSchema | None:  # noqa: C901
+    """Apply `annotation` to `schema` if it is an annotation we know about (Gt, Le, etc.).
+    Otherwise return `None`.
+
+    This does not handle all known annotations. If / when it does, it can always
+    return a CoreSchema and return the unmodified schema if the annotation should be ignored.
+
+    Assumes that GroupedMetadata has already been expanded via `expand_grouped_metadata`.
+
+    Args:
+        annotation: The annotation.
+        schema: The schema.
+
+    Returns:
+        An updated schema with annotation if it is an annotation we know about, `None` otherwise.
+
+    Raises:
+        PydanticCustomError: If `Predicate` fails.
+    """
+    import annotated_types as at
+
+    from . import _validators
+
+    schema = schema.copy()
+    schema_update, other_metadata = collect_known_metadata([annotation])
+    schema_type = schema['type']
+    for constraint, value in schema_update.items():
+        if constraint not in CONSTRAINTS_TO_ALLOWED_SCHEMAS:
+            raise ValueError(f'Unknown constraint {constraint}')
+        allowed_schemas = CONSTRAINTS_TO_ALLOWED_SCHEMAS[constraint]
+
+        if schema_type in allowed_schemas:
+            if constraint == 'union_mode' and schema_type == 'union':
+                schema['mode'] = value  # type: ignore  # schema is UnionSchema
+            else:
+                schema[constraint] = value
+            continue
+
+        if constraint == 'allow_inf_nan' and value is False:
+            return cs.no_info_after_validator_function(
+                _validators.forbid_inf_nan_check,
+                schema,
+            )
+        elif constraint == 'pattern':
+            # insert a str schema to make sure the regex engine matches
+            return cs.chain_schema(
+                [
+                    schema,
+                    cs.str_schema(pattern=value),
+                ]
+            )
+        elif constraint == 'gt':
+            s = cs.no_info_after_validator_function(
+                partial(_validators.greater_than_validator, gt=value),
+                schema,
+            )
+            add_js_update_schema(s, lambda: {'gt': as_jsonable_value(value)})
+            return s
+        elif constraint == 'ge':
+            return cs.no_info_after_validator_function(
+                partial(_validators.greater_than_or_equal_validator, ge=value),
+                schema,
+            )
+        elif constraint == 'lt':
+            return cs.no_info_after_validator_function(
+                partial(_validators.less_than_validator, lt=value),
+                schema,
+            )
+        elif constraint == 'le':
+            return cs.no_info_after_validator_function(
+                partial(_validators.less_than_or_equal_validator, le=value),
+                schema,
+            )
+        elif constraint == 'multiple_of':
+            return cs.no_info_after_validator_function(
+                partial(_validators.multiple_of_validator, multiple_of=value),
+                schema,
+            )
+        elif constraint == 'min_length':
+            s = cs.no_info_after_validator_function(
+                partial(_validators.min_length_validator, min_length=value),
+                schema,
+            )
+            add_js_update_schema(s, lambda: {'minLength': (as_jsonable_value(value))})
+            return s
+        elif constraint == 'max_length':
+            s = cs.no_info_after_validator_function(
+                partial(_validators.max_length_validator, max_length=value),
+                schema,
+            )
+            add_js_update_schema(s, lambda: {'maxLength': (as_jsonable_value(value))})
+            return s
+        elif constraint == 'strip_whitespace':
+            return cs.chain_schema(
+                [
+                    schema,
+                    cs.str_schema(strip_whitespace=True),
+                ]
+            )
+        elif constraint == 'to_lower':
+            return cs.chain_schema(
+                [
+                    schema,
+                    cs.str_schema(to_lower=True),
+                ]
+            )
+        elif constraint == 'to_upper':
+            return cs.chain_schema(
+                [
+                    schema,
+                    cs.str_schema(to_upper=True),
+                ]
+            )
+        elif constraint == 'min_length':
+            return cs.no_info_after_validator_function(
+                partial(_validators.min_length_validator, min_length=annotation.min_length),
+                schema,
+            )
+        elif constraint == 'max_length':
+            return cs.no_info_after_validator_function(
+                partial(_validators.max_length_validator, max_length=annotation.max_length),
+                schema,
+            )
+        else:
+            raise RuntimeError(f'Unable to apply constraint {constraint} to schema {schema_type}')
+
+    for annotation in other_metadata:
+        if isinstance(annotation, at.Gt):
+            return cs.no_info_after_validator_function(
+                partial(_validators.greater_than_validator, gt=annotation.gt),
+                schema,
+            )
+        elif isinstance(annotation, at.Ge):
+            return cs.no_info_after_validator_function(
+                partial(_validators.greater_than_or_equal_validator, ge=annotation.ge),
+                schema,
+            )
+        elif isinstance(annotation, at.Lt):
+            return cs.no_info_after_validator_function(
+                partial(_validators.less_than_validator, lt=annotation.lt),
+                schema,
+            )
+        elif isinstance(annotation, at.Le):
+            return cs.no_info_after_validator_function(
+                partial(_validators.less_than_or_equal_validator, le=annotation.le),
+                schema,
+            )
+        elif isinstance(annotation, at.MultipleOf):
+            return cs.no_info_after_validator_function(
+                partial(_validators.multiple_of_validator, multiple_of=annotation.multiple_of),
+                schema,
+            )
+        elif isinstance(annotation, at.MinLen):
+            return cs.no_info_after_validator_function(
+                partial(_validators.min_length_validator, min_length=annotation.min_length),
+                schema,
+            )
+        elif isinstance(annotation, at.MaxLen):
+            return cs.no_info_after_validator_function(
+                partial(_validators.max_length_validator, max_length=annotation.max_length),
+                schema,
+            )
+        elif isinstance(annotation, at.Predicate):
+            predicate_name = f'{annotation.func.__qualname__} ' if hasattr(annotation.func, '__qualname__') else ''
+
+            def val_func(v: Any) -> Any:
+                # annotation.func may also raise an exception, let it pass through
+                if not annotation.func(v):
+                    raise PydanticCustomError(
+                        'predicate_failed',
+                        f'Predicate {predicate_name}failed',  # type: ignore
+                    )
+                return v
+
+            return cs.no_info_after_validator_function(val_func, schema)
+        # ignore any other unknown metadata
+        return None
+
+    return schema
+
+
+def collect_known_metadata(annotations: Iterable[Any]) -> tuple[dict[str, Any], list[Any]]:
+    """Split `annotations` into known metadata and unknown annotations.
+
+    Args:
+        annotations: An iterable of annotations.
+
+    Returns:
+        A tuple contains a dict of known metadata and a list of unknown annotations.
+
+    Example:
+        ```py
+        from annotated_types import Gt, Len
+
+        from pydantic._internal._known_annotated_metadata import collect_known_metadata
+
+        print(collect_known_metadata([Gt(1), Len(42), ...]))
+        #> ({'gt': 1, 'min_length': 42}, [Ellipsis])
+        ```
+    """
+    import annotated_types as at
+
+    annotations = expand_grouped_metadata(annotations)
+
+    res: dict[str, Any] = {}
+    remaining: list[Any] = []
+    for annotation in annotations:
+        # isinstance(annotation, PydanticMetadata) also covers ._fields:_PydanticGeneralMetadata
+        if isinstance(annotation, PydanticMetadata):
+            res.update(annotation.__dict__)
+        # we don't use dataclasses.asdict because that recursively calls asdict on the field values
+        elif isinstance(annotation, at.MinLen):
+            res.update({'min_length': annotation.min_length})
+        elif isinstance(annotation, at.MaxLen):
+            res.update({'max_length': annotation.max_length})
+        elif isinstance(annotation, at.Gt):
+            res.update({'gt': annotation.gt})
+        elif isinstance(annotation, at.Ge):
+            res.update({'ge': annotation.ge})
+        elif isinstance(annotation, at.Lt):
+            res.update({'lt': annotation.lt})
+        elif isinstance(annotation, at.Le):
+            res.update({'le': annotation.le})
+        elif isinstance(annotation, at.MultipleOf):
+            res.update({'multiple_of': annotation.multiple_of})
+        elif isinstance(annotation, type) and issubclass(annotation, PydanticMetadata):
+            # also support PydanticMetadata classes being used without initialisation,
+            # e.g. `Annotated[int, Strict]` as well as `Annotated[int, Strict()]`
+            res.update({k: v for k, v in vars(annotation).items() if not k.startswith('_')})
+        else:
+            remaining.append(annotation)
+    # Nones can sneak in but pydantic-core will reject them
+    # it'd be nice to clean things up so we don't put in None (we probably don't _need_ to, it was just easier)
+    # but this is simple enough to kick that can down the road
+    res = {k: v for k, v in res.items() if v is not None}
+    return res, remaining
+
+
+def check_metadata(metadata: dict[str, Any], allowed: Iterable[str], source_type: Any) -> None:
+    """A small utility function to validate that the given metadata can be applied to the target.
+    More than saving lines of code, this gives us a consistent error message for all of our internal implementations.
+
+    Args:
+        metadata: A dict of metadata.
+        allowed: An iterable of allowed metadata.
+        source_type: The source type.
+
+    Raises:
+        TypeError: If there is metadatas that can't be applied on source type.
+    """
+    unknown = metadata.keys() - set(allowed)
+    if unknown:
+        raise TypeError(
+            f'The following constraints cannot be applied to {source_type!r}: {", ".join([f"{k!r}" for k in unknown])}'
+        )

lib/pydantic/_internal/_mock_val_ser.py

@@ -0,0 +1,140 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Callable, Generic, TypeVar
+
+from pydantic_core import SchemaSerializer, SchemaValidator
+from typing_extensions import Literal
+
+from ..errors import PydanticErrorCodes, PydanticUserError
+
+if TYPE_CHECKING:
+    from ..dataclasses import PydanticDataclass
+    from ..main import BaseModel
+
+
+ValSer = TypeVar('ValSer', SchemaValidator, SchemaSerializer)
+
+
+class MockValSer(Generic[ValSer]):
+    """Mocker for `pydantic_core.SchemaValidator` or `pydantic_core.SchemaSerializer` which optionally attempts to
+    rebuild the thing it's mocking when one of its methods is accessed and raises an error if that fails.
+    """
+
+    __slots__ = '_error_message', '_code', '_val_or_ser', '_attempt_rebuild'
+
+    def __init__(
+        self,
+        error_message: str,
+        *,
+        code: PydanticErrorCodes,
+        val_or_ser: Literal['validator', 'serializer'],
+        attempt_rebuild: Callable[[], ValSer | None] | None = None,
+    ) -> None:
+        self._error_message = error_message
+        self._val_or_ser = SchemaValidator if val_or_ser == 'validator' else SchemaSerializer
+        self._code: PydanticErrorCodes = code
+        self._attempt_rebuild = attempt_rebuild
+
+    def __getattr__(self, item: str) -> None:
+        __tracebackhide__ = True
+        if self._attempt_rebuild:
+            val_ser = self._attempt_rebuild()
+            if val_ser is not None:
+                return getattr(val_ser, item)
+
+        # raise an AttributeError if `item` doesn't exist
+        getattr(self._val_or_ser, item)
+        raise PydanticUserError(self._error_message, code=self._code)
+
+    def rebuild(self) -> ValSer | None:
+        if self._attempt_rebuild:
+            val_ser = self._attempt_rebuild()
+            if val_ser is not None:
+                return val_ser
+            else:
+                raise PydanticUserError(self._error_message, code=self._code)
+        return None
+
+
+def set_model_mocks(cls: type[BaseModel], cls_name: str, undefined_name: str = 'all referenced types') -> None:
+    """Set `__pydantic_validator__` and `__pydantic_serializer__` to `MockValSer`s on a model.
+
+    Args:
+        cls: The model class to set the mocks on
+        cls_name: Name of the model class, used in error messages
+        undefined_name: Name of the undefined thing, used in error messages
+    """
+    undefined_type_error_message = (
+        f'`{cls_name}` is not fully defined; you should define {undefined_name},'
+        f' then call `{cls_name}.model_rebuild()`.'
+    )
+
+    def attempt_rebuild_validator() -> SchemaValidator | None:
+        if cls.model_rebuild(raise_errors=False, _parent_namespace_depth=5) is not False:
+            return cls.__pydantic_validator__
+        else:
+            return None
+
+    cls.__pydantic_validator__ = MockValSer(  # type: ignore[assignment]
+        undefined_type_error_message,
+        code='class-not-fully-defined',
+        val_or_ser='validator',
+        attempt_rebuild=attempt_rebuild_validator,
+    )
+
+    def attempt_rebuild_serializer() -> SchemaSerializer | None:
+        if cls.model_rebuild(raise_errors=False, _parent_namespace_depth=5) is not False:
+            return cls.__pydantic_serializer__
+        else:
+            return None
+
+    cls.__pydantic_serializer__ = MockValSer(  # type: ignore[assignment]
+        undefined_type_error_message,
+        code='class-not-fully-defined',
+        val_or_ser='serializer',
+        attempt_rebuild=attempt_rebuild_serializer,
+    )
+
+
+def set_dataclass_mocks(
+    cls: type[PydanticDataclass], cls_name: str, undefined_name: str = 'all referenced types'
+) -> None:
+    """Set `__pydantic_validator__` and `__pydantic_serializer__` to `MockValSer`s on a dataclass.
+
+    Args:
+        cls: The model class to set the mocks on
+        cls_name: Name of the model class, used in error messages
+        undefined_name: Name of the undefined thing, used in error messages
+    """
+    from ..dataclasses import rebuild_dataclass
+
+    undefined_type_error_message = (
+        f'`{cls_name}` is not fully defined; you should define {undefined_name},'
+        f' then call `pydantic.dataclasses.rebuild_dataclass({cls_name})`.'
+    )
+
+    def attempt_rebuild_validator() -> SchemaValidator | None:
+        if rebuild_dataclass(cls, raise_errors=False, _parent_namespace_depth=5) is not False:
+            return cls.__pydantic_validator__
+        else:
+            return None
+
+    cls.__pydantic_validator__ = MockValSer(  # type: ignore[assignment]
+        undefined_type_error_message,
+        code='class-not-fully-defined',
+        val_or_ser='validator',
+        attempt_rebuild=attempt_rebuild_validator,
+    )
+
+    def attempt_rebuild_serializer() -> SchemaSerializer | None:
+        if rebuild_dataclass(cls, raise_errors=False, _parent_namespace_depth=5) is not False:
+            return cls.__pydantic_serializer__
+        else:
+            return None
+
+    cls.__pydantic_serializer__ = MockValSer(  # type: ignore[assignment]
+        undefined_type_error_message,
+        code='class-not-fully-defined',
+        val_or_ser='validator',
+        attempt_rebuild=attempt_rebuild_serializer,
+    )

lib/pydantic/_internal/_model_construction.py

@@ -0,0 +1,637 @@
+"""Private logic for creating models."""
+from __future__ import annotations as _annotations
+
+import operator
+import typing
+import warnings
+import weakref
+from abc import ABCMeta
+from functools import partial
+from types import FunctionType
+from typing import Any, Callable, Generic
+
+import typing_extensions
+from pydantic_core import PydanticUndefined, SchemaSerializer
+from typing_extensions import dataclass_transform, deprecated
+
+from ..errors import PydanticUndefinedAnnotation, PydanticUserError
+from ..plugin._schema_validator import create_schema_validator
+from ..warnings import GenericBeforeBaseModelWarning, PydanticDeprecatedSince20
+from ._config import ConfigWrapper
+from ._decorators import DecoratorInfos, PydanticDescriptorProxy, get_attribute_from_bases
+from ._fields import collect_model_fields, is_valid_field_name, is_valid_privateattr_name
+from ._generate_schema import GenerateSchema
+from ._generics import PydanticGenericMetadata, get_model_typevars_map
+from ._mock_val_ser import MockValSer, set_model_mocks
+from ._schema_generation_shared import CallbackGetCoreSchemaHandler
+from ._signature import generate_pydantic_signature
+from ._typing_extra import get_cls_types_namespace, is_annotated, is_classvar, parent_frame_namespace
+from ._utils import ClassAttribute, SafeGetItemProxy
+from ._validate_call import ValidateCallWrapper
+
+if typing.TYPE_CHECKING:
+    from ..fields import Field as PydanticModelField
+    from ..fields import FieldInfo, ModelPrivateAttr
+    from ..main import BaseModel
+else:
+    # See PyCharm issues https://youtrack.jetbrains.com/issue/PY-21915
+    # and https://youtrack.jetbrains.com/issue/PY-51428
+    DeprecationWarning = PydanticDeprecatedSince20
+    PydanticModelField = object()
+
+object_setattr = object.__setattr__
+
+
+class _ModelNamespaceDict(dict):
+    """A dictionary subclass that intercepts attribute setting on model classes and
+    warns about overriding of decorators.
+    """
+
+    def __setitem__(self, k: str, v: object) -> None:
+        existing: Any = self.get(k, None)
+        if existing and v is not existing and isinstance(existing, PydanticDescriptorProxy):
+            warnings.warn(f'`{k}` overrides an existing Pydantic `{existing.decorator_info.decorator_repr}` decorator')
+
+        return super().__setitem__(k, v)
+
+
+@dataclass_transform(kw_only_default=True, field_specifiers=(PydanticModelField,))
+class ModelMetaclass(ABCMeta):
+    def __new__(
+        mcs,
+        cls_name: str,
+        bases: tuple[type[Any], ...],
+        namespace: dict[str, Any],
+        __pydantic_generic_metadata__: PydanticGenericMetadata | None = None,
+        __pydantic_reset_parent_namespace__: bool = True,
+        _create_model_module: str | None = None,
+        **kwargs: Any,
+    ) -> type:
+        """Metaclass for creating Pydantic models.
+
+        Args:
+            cls_name: The name of the class to be created.
+            bases: The base classes of the class to be created.
+            namespace: The attribute dictionary of the class to be created.
+            __pydantic_generic_metadata__: Metadata for generic models.
+            __pydantic_reset_parent_namespace__: Reset parent namespace.
+            _create_model_module: The module of the class to be created, if created by `create_model`.
+            **kwargs: Catch-all for any other keyword arguments.
+
+        Returns:
+            The new class created by the metaclass.
+        """
+        # Note `ModelMetaclass` refers to `BaseModel`, but is also used to *create* `BaseModel`, so we rely on the fact
+        # that `BaseModel` itself won't have any bases, but any subclass of it will, to determine whether the `__new__`
+        # call we're in the middle of is for the `BaseModel` class.
+        if bases:
+            base_field_names, class_vars, base_private_attributes = mcs._collect_bases_data(bases)
+
+            config_wrapper = ConfigWrapper.for_model(bases, namespace, kwargs)
+            namespace['model_config'] = config_wrapper.config_dict
+            private_attributes = inspect_namespace(
+                namespace, config_wrapper.ignored_types, class_vars, base_field_names
+            )
+            if private_attributes:
+                original_model_post_init = get_model_post_init(namespace, bases)
+                if original_model_post_init is not None:
+                    # if there are private_attributes and a model_post_init function, we handle both
+
+                    def wrapped_model_post_init(self: BaseModel, __context: Any) -> None:
+                        """We need to both initialize private attributes and call the user-defined model_post_init
+                        method.
+                        """
+                        init_private_attributes(self, __context)
+                        original_model_post_init(self, __context)
+
+                    namespace['model_post_init'] = wrapped_model_post_init
+                else:
+                    namespace['model_post_init'] = init_private_attributes
+
+            namespace['__class_vars__'] = class_vars
+            namespace['__private_attributes__'] = {**base_private_attributes, **private_attributes}
+
+            cls: type[BaseModel] = super().__new__(mcs, cls_name, bases, namespace, **kwargs)  # type: ignore
+
+            from ..main import BaseModel
+
+            mro = cls.__mro__
+            if Generic in mro and mro.index(Generic) < mro.index(BaseModel):
+                warnings.warn(
+                    GenericBeforeBaseModelWarning(
+                        'Classes should inherit from `BaseModel` before generic classes (e.g. `typing.Generic[T]`) '
+                        'for pydantic generics to work properly.'
+                    ),
+                    stacklevel=2,
+                )
+
+            cls.__pydantic_custom_init__ = not getattr(cls.__init__, '__pydantic_base_init__', False)
+            cls.__pydantic_post_init__ = None if cls.model_post_init is BaseModel.model_post_init else 'model_post_init'
+
+            cls.__pydantic_decorators__ = DecoratorInfos.build(cls)
+
+            # Use the getattr below to grab the __parameters__ from the `typing.Generic` parent class
+            if __pydantic_generic_metadata__:
+                cls.__pydantic_generic_metadata__ = __pydantic_generic_metadata__
+            else:
+                parent_parameters = getattr(cls, '__pydantic_generic_metadata__', {}).get('parameters', ())
+                parameters = getattr(cls, '__parameters__', None) or parent_parameters
+                if parameters and parent_parameters and not all(x in parameters for x in parent_parameters):
+                    combined_parameters = parent_parameters + tuple(x for x in parameters if x not in parent_parameters)
+                    parameters_str = ', '.join([str(x) for x in combined_parameters])
+                    generic_type_label = f'typing.Generic[{parameters_str}]'
+                    error_message = (
+                        f'All parameters must be present on typing.Generic;'
+                        f' you should inherit from {generic_type_label}.'
+                    )
+                    if Generic not in bases:  # pragma: no cover
+                        # We raise an error here not because it is desirable, but because some cases are mishandled.
+                        # It would be nice to remove this error and still have things behave as expected, it's just
+                        # challenging because we are using a custom `__class_getitem__` to parametrize generic models,
+                        # and not returning a typing._GenericAlias from it.
+                        bases_str = ', '.join([x.__name__ for x in bases] + [generic_type_label])
+                        error_message += (
+                            f' Note: `typing.Generic` must go last: `class {cls.__name__}({bases_str}): ...`)'
+                        )
+                    raise TypeError(error_message)
+
+                cls.__pydantic_generic_metadata__ = {
+                    'origin': None,
+                    'args': (),
+                    'parameters': parameters,
+                }
+
+            cls.__pydantic_complete__ = False  # Ensure this specific class gets completed
+
+            # preserve `__set_name__` protocol defined in https://peps.python.org/pep-0487
+            # for attributes not in `new_namespace` (e.g. private attributes)
+            for name, obj in private_attributes.items():
+                obj.__set_name__(cls, name)
+
+            if __pydantic_reset_parent_namespace__:
+                cls.__pydantic_parent_namespace__ = build_lenient_weakvaluedict(parent_frame_namespace())
+            parent_namespace = getattr(cls, '__pydantic_parent_namespace__', None)
+            if isinstance(parent_namespace, dict):
+                parent_namespace = unpack_lenient_weakvaluedict(parent_namespace)
+
+            types_namespace = get_cls_types_namespace(cls, parent_namespace)
+            set_model_fields(cls, bases, config_wrapper, types_namespace)
+
+            if config_wrapper.frozen and '__hash__' not in namespace:
+                set_default_hash_func(cls, bases)
+
+            complete_model_class(
+                cls,
+                cls_name,
+                config_wrapper,
+                raise_errors=False,
+                types_namespace=types_namespace,
+                create_model_module=_create_model_module,
+            )
+
+            # If this is placed before the complete_model_class call above,
+            # the generic computed fields return type is set to PydanticUndefined
+            cls.model_computed_fields = {k: v.info for k, v in cls.__pydantic_decorators__.computed_fields.items()}
+
+            # using super(cls, cls) on the next line ensures we only call the parent class's __pydantic_init_subclass__
+            # I believe the `type: ignore` is only necessary because mypy doesn't realize that this code branch is
+            # only hit for _proper_ subclasses of BaseModel
+            super(cls, cls).__pydantic_init_subclass__(**kwargs)  # type: ignore[misc]
+            return cls
+        else:
+            # this is the BaseModel class itself being created, no logic required
+            return super().__new__(mcs, cls_name, bases, namespace, **kwargs)
+
+    if not typing.TYPE_CHECKING:  # pragma: no branch
+        # We put `__getattr__` in a non-TYPE_CHECKING block because otherwise, mypy allows arbitrary attribute access
+
+        def __getattr__(self, item: str) -> Any:
+            """This is necessary to keep attribute access working for class attribute access."""
+            private_attributes = self.__dict__.get('__private_attributes__')
+            if private_attributes and item in private_attributes:
+                return private_attributes[item]
+            if item == '__pydantic_core_schema__':
+                # This means the class didn't get a schema generated for it, likely because there was an undefined reference
+                maybe_mock_validator = getattr(self, '__pydantic_validator__', None)
+                if isinstance(maybe_mock_validator, MockValSer):
+                    rebuilt_validator = maybe_mock_validator.rebuild()
+                    if rebuilt_validator is not None:
+                        # In this case, a validator was built, and so `__pydantic_core_schema__` should now be set
+                        return getattr(self, '__pydantic_core_schema__')
+            raise AttributeError(item)
+
+    @classmethod
+    def __prepare__(cls, *args: Any, **kwargs: Any) -> dict[str, object]:
+        return _ModelNamespaceDict()
+
+    def __instancecheck__(self, instance: Any) -> bool:
+        """Avoid calling ABC _abc_subclasscheck unless we're pretty sure.
+
+        See #3829 and python/cpython#92810
+        """
+        return hasattr(instance, '__pydantic_validator__') and super().__instancecheck__(instance)
+
+    @staticmethod
+    def _collect_bases_data(bases: tuple[type[Any], ...]) -> tuple[set[str], set[str], dict[str, ModelPrivateAttr]]:
+        from ..main import BaseModel
+
+        field_names: set[str] = set()
+        class_vars: set[str] = set()
+        private_attributes: dict[str, ModelPrivateAttr] = {}
+        for base in bases:
+            if issubclass(base, BaseModel) and base is not BaseModel:
+                # model_fields might not be defined yet in the case of generics, so we use getattr here:
+                field_names.update(getattr(base, 'model_fields', {}).keys())
+                class_vars.update(base.__class_vars__)
+                private_attributes.update(base.__private_attributes__)
+        return field_names, class_vars, private_attributes
+
+    @property
+    @deprecated('The `__fields__` attribute is deprecated, use `model_fields` instead.', category=None)
+    def __fields__(self) -> dict[str, FieldInfo]:
+        warnings.warn(
+            'The `__fields__` attribute is deprecated, use `model_fields` instead.', PydanticDeprecatedSince20
+        )
+        return self.model_fields  # type: ignore
+
+    def __dir__(self) -> list[str]:
+        attributes = list(super().__dir__())
+        if '__fields__' in attributes:
+            attributes.remove('__fields__')
+        return attributes
+
+
+def init_private_attributes(self: BaseModel, __context: Any) -> None:
+    """This function is meant to behave like a BaseModel method to initialise private attributes.
+
+    It takes context as an argument since that's what pydantic-core passes when calling it.
+
+    Args:
+        self: The BaseModel instance.
+        __context: The context.
+    """
+    if getattr(self, '__pydantic_private__', None) is None:
+        pydantic_private = {}
+        for name, private_attr in self.__private_attributes__.items():
+            default = private_attr.get_default()
+            if default is not PydanticUndefined:
+                pydantic_private[name] = default
+        object_setattr(self, '__pydantic_private__', pydantic_private)
+
+
+def get_model_post_init(namespace: dict[str, Any], bases: tuple[type[Any], ...]) -> Callable[..., Any] | None:
+    """Get the `model_post_init` method from the namespace or the class bases, or `None` if not defined."""
+    if 'model_post_init' in namespace:
+        return namespace['model_post_init']
+
+    from ..main import BaseModel
+
+    model_post_init = get_attribute_from_bases(bases, 'model_post_init')
+    if model_post_init is not BaseModel.model_post_init:
+        return model_post_init
+
+
+def inspect_namespace(  # noqa C901
+    namespace: dict[str, Any],
+    ignored_types: tuple[type[Any], ...],
+    base_class_vars: set[str],
+    base_class_fields: set[str],
+) -> dict[str, ModelPrivateAttr]:
+    """Iterate over the namespace and:
+    * gather private attributes
+    * check for items which look like fields but are not (e.g. have no annotation) and warn.
+
+    Args:
+        namespace: The attribute dictionary of the class to be created.
+        ignored_types: A tuple of ignore types.
+        base_class_vars: A set of base class class variables.
+        base_class_fields: A set of base class fields.
+
+    Returns:
+        A dict contains private attributes info.
+
+    Raises:
+        TypeError: If there is a `__root__` field in model.
+        NameError: If private attribute name is invalid.
+        PydanticUserError:
+            - If a field does not have a type annotation.
+            - If a field on base class was overridden by a non-annotated attribute.
+    """
+    from ..fields import FieldInfo, ModelPrivateAttr, PrivateAttr
+
+    all_ignored_types = ignored_types + default_ignored_types()
+
+    private_attributes: dict[str, ModelPrivateAttr] = {}
+    raw_annotations = namespace.get('__annotations__', {})
+
+    if '__root__' in raw_annotations or '__root__' in namespace:
+        raise TypeError("To define root models, use `pydantic.RootModel` rather than a field called '__root__'")
+
+    ignored_names: set[str] = set()
+    for var_name, value in list(namespace.items()):
+        if var_name == 'model_config':
+            continue
+        elif (
+            isinstance(value, type)
+            and value.__module__ == namespace['__module__']
+            and value.__qualname__.startswith(namespace['__qualname__'])
+        ):
+            # `value` is a nested type defined in this namespace; don't error
+            continue
+        elif isinstance(value, all_ignored_types) or value.__class__.__module__ == 'functools':
+            ignored_names.add(var_name)
+            continue
+        elif isinstance(value, ModelPrivateAttr):
+            if var_name.startswith('__'):
+                raise NameError(
+                    'Private attributes must not use dunder names;'
+                    f' use a single underscore prefix instead of {var_name!r}.'
+                )
+            elif is_valid_field_name(var_name):
+                raise NameError(
+                    'Private attributes must not use valid field names;'
+                    f' use sunder names, e.g. {"_" + var_name!r} instead of {var_name!r}.'
+                )
+            private_attributes[var_name] = value
+            del namespace[var_name]
+        elif isinstance(value, FieldInfo) and not is_valid_field_name(var_name):
+            suggested_name = var_name.lstrip('_') or 'my_field'  # don't suggest '' for all-underscore name
+            raise NameError(
+                f'Fields must not use names with leading underscores;'
+                f' e.g., use {suggested_name!r} instead of {var_name!r}.'
+            )
+
+        elif var_name.startswith('__'):
+            continue
+        elif is_valid_privateattr_name(var_name):
+            if var_name not in raw_annotations or not is_classvar(raw_annotations[var_name]):
+                private_attributes[var_name] = PrivateAttr(default=value)
+                del namespace[var_name]
+        elif var_name in base_class_vars:
+            continue
+        elif var_name not in raw_annotations:
+            if var_name in base_class_fields:
+                raise PydanticUserError(
+                    f'Field {var_name!r} defined on a base class was overridden by a non-annotated attribute. '
+                    f'All field definitions, including overrides, require a type annotation.',
+                    code='model-field-overridden',
+                )
+            elif isinstance(value, FieldInfo):
+                raise PydanticUserError(
+                    f'Field {var_name!r} requires a type annotation', code='model-field-missing-annotation'
+                )
+            else:
+                raise PydanticUserError(
+                    f'A non-annotated attribute was detected: `{var_name} = {value!r}`. All model fields require a '
+                    f'type annotation; if `{var_name}` is not meant to be a field, you may be able to resolve this '
+                    f"error by annotating it as a `ClassVar` or updating `model_config['ignored_types']`.",
+                    code='model-field-missing-annotation',
+                )
+
+    for ann_name, ann_type in raw_annotations.items():
+        if (
+            is_valid_privateattr_name(ann_name)
+            and ann_name not in private_attributes
+            and ann_name not in ignored_names
+            and not is_classvar(ann_type)
+            and ann_type not in all_ignored_types
+            and getattr(ann_type, '__module__', None) != 'functools'
+        ):
+            if is_annotated(ann_type):
+                _, *metadata = typing_extensions.get_args(ann_type)
+                private_attr = next((v for v in metadata if isinstance(v, ModelPrivateAttr)), None)
+                if private_attr is not None:
+                    private_attributes[ann_name] = private_attr
+                    continue
+            private_attributes[ann_name] = PrivateAttr()
+
+    return private_attributes
+
+
+def set_default_hash_func(cls: type[BaseModel], bases: tuple[type[Any], ...]) -> None:
+    base_hash_func = get_attribute_from_bases(bases, '__hash__')
+    new_hash_func = make_hash_func(cls)
+    if base_hash_func in {None, object.__hash__} or getattr(base_hash_func, '__code__', None) == new_hash_func.__code__:
+        # If `__hash__` is some default, we generate a hash function.
+        # It will be `None` if not overridden from BaseModel.
+        # It may be `object.__hash__` if there is another
+        # parent class earlier in the bases which doesn't override `__hash__` (e.g. `typing.Generic`).
+        # It may be a value set by `set_default_hash_func` if `cls` is a subclass of another frozen model.
+        # In the last case we still need a new hash function to account for new `model_fields`.
+        cls.__hash__ = new_hash_func
+
+
+def make_hash_func(cls: type[BaseModel]) -> Any:
+    getter = operator.itemgetter(*cls.model_fields.keys()) if cls.model_fields else lambda _: 0
+
+    def hash_func(self: Any) -> int:
+        try:
+            return hash(getter(self.__dict__))
+        except KeyError:
+            # In rare cases (such as when using the deprecated copy method), the __dict__ may not contain
+            # all model fields, which is how we can get here.
+            # getter(self.__dict__) is much faster than any 'safe' method that accounts for missing keys,
+            # and wrapping it in a `try` doesn't slow things down much in the common case.
+            return hash(getter(SafeGetItemProxy(self.__dict__)))
+
+    return hash_func
+
+
+def set_model_fields(
+    cls: type[BaseModel], bases: tuple[type[Any], ...], config_wrapper: ConfigWrapper, types_namespace: dict[str, Any]
+) -> None:
+    """Collect and set `cls.model_fields` and `cls.__class_vars__`.
+
+    Args:
+        cls: BaseModel or dataclass.
+        bases: Parents of the class, generally `cls.__bases__`.
+        config_wrapper: The config wrapper instance.
+        types_namespace: Optional extra namespace to look for types in.
+    """
+    typevars_map = get_model_typevars_map(cls)
+    fields, class_vars = collect_model_fields(cls, bases, config_wrapper, types_namespace, typevars_map=typevars_map)
+
+    cls.model_fields = fields
+    cls.__class_vars__.update(class_vars)
+
+    for k in class_vars:
+        # Class vars should not be private attributes
+        #     We remove them _here_ and not earlier because we rely on inspecting the class to determine its classvars,
+        #     but private attributes are determined by inspecting the namespace _prior_ to class creation.
+        #     In the case that a classvar with a leading-'_' is defined via a ForwardRef (e.g., when using
+        #     `__future__.annotations`), we want to remove the private attribute which was detected _before_ we knew it
+        #     evaluated to a classvar
+
+        value = cls.__private_attributes__.pop(k, None)
+        if value is not None and value.default is not PydanticUndefined:
+            setattr(cls, k, value.default)
+
+
+def complete_model_class(
+    cls: type[BaseModel],
+    cls_name: str,
+    config_wrapper: ConfigWrapper,
+    *,
+    raise_errors: bool = True,
+    types_namespace: dict[str, Any] | None,
+    create_model_module: str | None = None,
+) -> bool:
+    """Finish building a model class.
+
+    This logic must be called after class has been created since validation functions must be bound
+    and `get_type_hints` requires a class object.
+
+    Args:
+        cls: BaseModel or dataclass.
+        cls_name: The model or dataclass name.
+        config_wrapper: The config wrapper instance.
+        raise_errors: Whether to raise errors.
+        types_namespace: Optional extra namespace to look for types in.
+        create_model_module: The module of the class to be created, if created by `create_model`.
+
+    Returns:
+        `True` if the model is successfully completed, else `False`.
+
+    Raises:
+        PydanticUndefinedAnnotation: If `PydanticUndefinedAnnotation` occurs in`__get_pydantic_core_schema__`
+            and `raise_errors=True`.
+    """
+    typevars_map = get_model_typevars_map(cls)
+    gen_schema = GenerateSchema(
+        config_wrapper,
+        types_namespace,
+        typevars_map,
+    )
+
+    handler = CallbackGetCoreSchemaHandler(
+        partial(gen_schema.generate_schema, from_dunder_get_core_schema=False),
+        gen_schema,
+        ref_mode='unpack',
+    )
+
+    if config_wrapper.defer_build:
+        set_model_mocks(cls, cls_name)
+        return False
+
+    try:
+        schema = cls.__get_pydantic_core_schema__(cls, handler)
+    except PydanticUndefinedAnnotation as e:
+        if raise_errors:
+            raise
+        set_model_mocks(cls, cls_name, f'`{e.name}`')
+        return False
+
+    core_config = config_wrapper.core_config(cls)
+
+    try:
+        schema = gen_schema.clean_schema(schema)
+    except gen_schema.CollectedInvalid:
+        set_model_mocks(cls, cls_name)
+        return False
+
+    # debug(schema)
+    cls.__pydantic_core_schema__ = schema
+
+    cls.__pydantic_validator__ = create_schema_validator(
+        schema,
+        cls,
+        create_model_module or cls.__module__,
+        cls.__qualname__,
+        'create_model' if create_model_module else 'BaseModel',
+        core_config,
+        config_wrapper.plugin_settings,
+    )
+    cls.__pydantic_serializer__ = SchemaSerializer(schema, core_config)
+    cls.__pydantic_complete__ = True
+
+    # set __signature__ attr only for model class, but not for its instances
+    cls.__signature__ = ClassAttribute(
+        '__signature__',
+        generate_pydantic_signature(init=cls.__init__, fields=cls.model_fields, config_wrapper=config_wrapper),
+    )
+    return True
+
+
+class _PydanticWeakRef:
+    """Wrapper for `weakref.ref` that enables `pickle` serialization.
+
+    Cloudpickle fails to serialize `weakref.ref` objects due to an arcane error related
+    to abstract base classes (`abc.ABC`). This class works around the issue by wrapping
+    `weakref.ref` instead of subclassing it.
+
+    See https://github.com/pydantic/pydantic/issues/6763 for context.
+
+    Semantics:
+        - If not pickled, behaves the same as a `weakref.ref`.
+        - If pickled along with the referenced object, the same `weakref.ref` behavior
+          will be maintained between them after unpickling.
+        - If pickled without the referenced object, after unpickling the underlying
+          reference will be cleared (`__call__` will always return `None`).
+    """
+
+    def __init__(self, obj: Any):
+        if obj is None:
+            # The object will be `None` upon deserialization if the serialized weakref
+            # had lost its underlying object.
+            self._wr = None
+        else:
+            self._wr = weakref.ref(obj)
+
+    def __call__(self) -> Any:
+        if self._wr is None:
+            return None
+        else:
+            return self._wr()
+
+    def __reduce__(self) -> tuple[Callable, tuple[weakref.ReferenceType | None]]:
+        return _PydanticWeakRef, (self(),)
+
+
+def build_lenient_weakvaluedict(d: dict[str, Any] | None) -> dict[str, Any] | None:
+    """Takes an input dictionary, and produces a new value that (invertibly) replaces the values with weakrefs.
+
+    We can't just use a WeakValueDictionary because many types (including int, str, etc.) can't be stored as values
+    in a WeakValueDictionary.
+
+    The `unpack_lenient_weakvaluedict` function can be used to reverse this operation.
+    """
+    if d is None:
+        return None
+    result = {}
+    for k, v in d.items():
+        try:
+            proxy = _PydanticWeakRef(v)
+        except TypeError:
+            proxy = v
+        result[k] = proxy
+    return result
+
+
+def unpack_lenient_weakvaluedict(d: dict[str, Any] | None) -> dict[str, Any] | None:
+    """Inverts the transform performed by `build_lenient_weakvaluedict`."""
+    if d is None:
+        return None
+
+    result = {}
+    for k, v in d.items():
+        if isinstance(v, _PydanticWeakRef):
+            v = v()
+            if v is not None:
+                result[k] = v
+        else:
+            result[k] = v
+    return result
+
+
+def default_ignored_types() -> tuple[type[Any], ...]:
+    from ..fields import ComputedFieldInfo
+
+    return (
+        FunctionType,
+        property,
+        classmethod,
+        staticmethod,
+        PydanticDescriptorProxy,
+        ComputedFieldInfo,
+        ValidateCallWrapper,
+    )

lib/pydantic/_internal/_repr.py

@@ -0,0 +1,117 @@
+"""Tools to provide pretty/human-readable display of objects."""
+from __future__ import annotations as _annotations
+
+import types
+import typing
+from typing import Any
+
+import typing_extensions
+
+from . import _typing_extra
+
+if typing.TYPE_CHECKING:
+    ReprArgs: typing_extensions.TypeAlias = 'typing.Iterable[tuple[str | None, Any]]'
+    RichReprResult: typing_extensions.TypeAlias = (
+        'typing.Iterable[Any | tuple[Any] | tuple[str, Any] | tuple[str, Any, Any]]'
+    )
+
+
+class PlainRepr(str):
+    """String class where repr doesn't include quotes. Useful with Representation when you want to return a string
+    representation of something that is valid (or pseudo-valid) python.
+    """
+
+    def __repr__(self) -> str:
+        return str(self)
+
+
+class Representation:
+    # Mixin to provide `__str__`, `__repr__`, and `__pretty__` and `__rich_repr__` methods.
+    # `__pretty__` is used by [devtools](https://python-devtools.helpmanual.io/).
+    # `__rich_repr__` is used by [rich](https://rich.readthedocs.io/en/stable/pretty.html).
+    # (this is not a docstring to avoid adding a docstring to classes which inherit from Representation)
+
+    # we don't want to use a type annotation here as it can break get_type_hints
+    __slots__ = tuple()  # type: typing.Collection[str]
+
+    def __repr_args__(self) -> ReprArgs:
+        """Returns the attributes to show in __str__, __repr__, and __pretty__ this is generally overridden.
+
+        Can either return:
+        * name - value pairs, e.g.: `[('foo_name', 'foo'), ('bar_name', ['b', 'a', 'r'])]`
+        * or, just values, e.g.: `[(None, 'foo'), (None, ['b', 'a', 'r'])]`
+        """
+        attrs_names = self.__slots__
+        if not attrs_names and hasattr(self, '__dict__'):
+            attrs_names = self.__dict__.keys()
+        attrs = ((s, getattr(self, s)) for s in attrs_names)
+        return [(a, v) for a, v in attrs if v is not None]
+
+    def __repr_name__(self) -> str:
+        """Name of the instance's class, used in __repr__."""
+        return self.__class__.__name__
+
+    def __repr_str__(self, join_str: str) -> str:
+        return join_str.join(repr(v) if a is None else f'{a}={v!r}' for a, v in self.__repr_args__())
+
+    def __pretty__(self, fmt: typing.Callable[[Any], Any], **kwargs: Any) -> typing.Generator[Any, None, None]:
+        """Used by devtools (https://python-devtools.helpmanual.io/) to pretty print objects."""
+        yield self.__repr_name__() + '('
+        yield 1
+        for name, value in self.__repr_args__():
+            if name is not None:
+                yield name + '='
+            yield fmt(value)
+            yield ','
+            yield 0
+        yield -1
+        yield ')'
+
+    def __rich_repr__(self) -> RichReprResult:
+        """Used by Rich (https://rich.readthedocs.io/en/stable/pretty.html) to pretty print objects."""
+        for name, field_repr in self.__repr_args__():
+            if name is None:
+                yield field_repr
+            else:
+                yield name, field_repr
+
+    def __str__(self) -> str:
+        return self.__repr_str__(' ')
+
+    def __repr__(self) -> str:
+        return f'{self.__repr_name__()}({self.__repr_str__(", ")})'
+
+
+def display_as_type(obj: Any) -> str:
+    """Pretty representation of a type, should be as close as possible to the original type definition string.
+
+    Takes some logic from `typing._type_repr`.
+    """
+    if isinstance(obj, types.FunctionType):
+        return obj.__name__
+    elif obj is ...:
+        return '...'
+    elif isinstance(obj, Representation):
+        return repr(obj)
+    elif isinstance(obj, typing_extensions.TypeAliasType):
+        return str(obj)
+
+    if not isinstance(obj, (_typing_extra.typing_base, _typing_extra.WithArgsTypes, type)):
+        obj = obj.__class__
+
+    if _typing_extra.origin_is_union(typing_extensions.get_origin(obj)):
+        args = ', '.join(map(display_as_type, typing_extensions.get_args(obj)))
+        return f'Union[{args}]'
+    elif isinstance(obj, _typing_extra.WithArgsTypes):
+        if typing_extensions.get_origin(obj) == typing_extensions.Literal:
+            args = ', '.join(map(repr, typing_extensions.get_args(obj)))
+        else:
+            args = ', '.join(map(display_as_type, typing_extensions.get_args(obj)))
+        try:
+            return f'{obj.__qualname__}[{args}]'
+        except AttributeError:
+            return str(obj)  # handles TypeAliasType in 3.12
+    elif isinstance(obj, type):
+        return obj.__qualname__
+    else:
+        return repr(obj).replace('typing.', '').replace('typing_extensions.', '')

lib/pydantic/_internal/_schema_generation_shared.py

@@ -0,0 +1,124 @@
+"""Types and utility functions used by various other internal tools."""
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Any, Callable
+
+from pydantic_core import core_schema
+from typing_extensions import Literal
+
+from ..annotated_handlers import GetCoreSchemaHandler, GetJsonSchemaHandler
+
+if TYPE_CHECKING:
+    from ..json_schema import GenerateJsonSchema, JsonSchemaValue
+    from ._core_utils import CoreSchemaOrField
+    from ._generate_schema import GenerateSchema
+
+    GetJsonSchemaFunction = Callable[[CoreSchemaOrField, GetJsonSchemaHandler], JsonSchemaValue]
+    HandlerOverride = Callable[[CoreSchemaOrField], JsonSchemaValue]
+
+
+class GenerateJsonSchemaHandler(GetJsonSchemaHandler):
+    """JsonSchemaHandler implementation that doesn't do ref unwrapping by default.
+
+    This is used for any Annotated metadata so that we don't end up with conflicting
+    modifications to the definition schema.
+
+    Used internally by Pydantic, please do not rely on this implementation.
+    See `GetJsonSchemaHandler` for the handler API.
+    """
+
+    def __init__(self, generate_json_schema: GenerateJsonSchema, handler_override: HandlerOverride | None) -> None:
+        self.generate_json_schema = generate_json_schema
+        self.handler = handler_override or generate_json_schema.generate_inner
+        self.mode = generate_json_schema.mode
+
+    def __call__(self, __core_schema: CoreSchemaOrField) -> JsonSchemaValue:
+        return self.handler(__core_schema)
+
+    def resolve_ref_schema(self, maybe_ref_json_schema: JsonSchemaValue) -> JsonSchemaValue:
+        """Resolves `$ref` in the json schema.
+
+        This returns the input json schema if there is no `$ref` in json schema.
+
+        Args:
+            maybe_ref_json_schema: The input json schema that may contains `$ref`.
+
+        Returns:
+            Resolved json schema.
+
+        Raises:
+            LookupError: If it can't find the definition for `$ref`.
+        """
+        if '$ref' not in maybe_ref_json_schema:
+            return maybe_ref_json_schema
+        ref = maybe_ref_json_schema['$ref']
+        json_schema = self.generate_json_schema.get_schema_from_definitions(ref)
+        if json_schema is None:
+            raise LookupError(
+                f'Could not find a ref for {ref}.'
+                ' Maybe you tried to call resolve_ref_schema from within a recursive model?'
+            )
+        return json_schema
+
+
+class CallbackGetCoreSchemaHandler(GetCoreSchemaHandler):
+    """Wrapper to use an arbitrary function as a `GetCoreSchemaHandler`.
+
+    Used internally by Pydantic, please do not rely on this implementation.
+    See `GetCoreSchemaHandler` for the handler API.
+    """
+
+    def __init__(
+        self,
+        handler: Callable[[Any], core_schema.CoreSchema],
+        generate_schema: GenerateSchema,
+        ref_mode: Literal['to-def', 'unpack'] = 'to-def',
+    ) -> None:
+        self._handler = handler
+        self._generate_schema = generate_schema
+        self._ref_mode = ref_mode
+
+    def __call__(self, __source_type: Any) -> core_schema.CoreSchema:
+        schema = self._handler(__source_type)
+        ref = schema.get('ref')
+        if self._ref_mode == 'to-def':
+            if ref is not None:
+                self._generate_schema.defs.definitions[ref] = schema
+                return core_schema.definition_reference_schema(ref)
+            return schema
+        else:  # ref_mode = 'unpack
+            return self.resolve_ref_schema(schema)
+
+    def _get_types_namespace(self) -> dict[str, Any] | None:
+        return self._generate_schema._types_namespace
+
+    def generate_schema(self, __source_type: Any) -> core_schema.CoreSchema:
+        return self._generate_schema.generate_schema(__source_type)
+
+    @property
+    def field_name(self) -> str | None:
+        return self._generate_schema.field_name_stack.get()
+
+    def resolve_ref_schema(self, maybe_ref_schema: core_schema.CoreSchema) -> core_schema.CoreSchema:
+        """Resolves reference in the core schema.
+
+        Args:
+            maybe_ref_schema: The input core schema that may contains reference.
+
+        Returns:
+            Resolved core schema.
+
+        Raises:
+            LookupError: If it can't find the definition for reference.
+        """
+        if maybe_ref_schema['type'] == 'definition-ref':
+            ref = maybe_ref_schema['schema_ref']
+            if ref not in self._generate_schema.defs.definitions:
+                raise LookupError(
+                    f'Could not find a ref for {ref}.'
+                    ' Maybe you tried to call resolve_ref_schema from within a recursive model?'
+                )
+            return self._generate_schema.defs.definitions[ref]
+        elif maybe_ref_schema['type'] == 'definitions':
+            return self.resolve_ref_schema(maybe_ref_schema['schema'])
+        return maybe_ref_schema

lib/pydantic/_internal/_signature.py

@@ -0,0 +1,164 @@
+from __future__ import annotations
+
+import dataclasses
+from inspect import Parameter, Signature, signature
+from typing import TYPE_CHECKING, Any, Callable
+
+from pydantic_core import PydanticUndefined
+
+from ._config import ConfigWrapper
+from ._utils import is_valid_identifier
+
+if TYPE_CHECKING:
+    from ..fields import FieldInfo
+
+
+def _field_name_for_signature(field_name: str, field_info: FieldInfo) -> str:
+    """Extract the correct name to use for the field when generating a signature.
+
+    Assuming the field has a valid alias, this will return the alias. Otherwise, it will return the field name.
+    First priority is given to the validation_alias, then the alias, then the field name.
+
+    Args:
+        field_name: The name of the field
+        field_info: The corresponding FieldInfo object.
+
+    Returns:
+        The correct name to use when generating a signature.
+    """
+
+    def _alias_if_valid(x: Any) -> str | None:
+        """Return the alias if it is a valid alias and identifier, else None."""
+        return x if isinstance(x, str) and is_valid_identifier(x) else None
+
+    return _alias_if_valid(field_info.alias) or _alias_if_valid(field_info.validation_alias) or field_name
+
+
+def _process_param_defaults(param: Parameter) -> Parameter:
+    """Modify the signature for a parameter in a dataclass where the default value is a FieldInfo instance.
+
+    Args:
+        param (Parameter): The parameter
+
+    Returns:
+        Parameter: The custom processed parameter
+    """
+    from ..fields import FieldInfo
+
+    param_default = param.default
+    if isinstance(param_default, FieldInfo):
+        annotation = param.annotation
+        # Replace the annotation if appropriate
+        # inspect does "clever" things to show annotations as strings because we have
+        # `from __future__ import annotations` in main, we don't want that
+        if annotation == 'Any':
+            annotation = Any
+
+        # Replace the field default
+        default = param_default.default
+        if default is PydanticUndefined:
+            if param_default.default_factory is PydanticUndefined:
+                default = Signature.empty
+            else:
+                # this is used by dataclasses to indicate a factory exists:
+                default = dataclasses._HAS_DEFAULT_FACTORY  # type: ignore
+        return param.replace(
+            annotation=annotation, name=_field_name_for_signature(param.name, param_default), default=default
+        )
+    return param
+
+
+def _generate_signature_parameters(  # noqa: C901 (ignore complexity, could use a refactor)
+    init: Callable[..., None],
+    fields: dict[str, FieldInfo],
+    config_wrapper: ConfigWrapper,
+) -> dict[str, Parameter]:
+    """Generate a mapping of parameter names to Parameter objects for a pydantic BaseModel or dataclass."""
+    from itertools import islice
+
+    present_params = signature(init).parameters.values()
+    merged_params: dict[str, Parameter] = {}
+    var_kw = None
+    use_var_kw = False
+
+    for param in islice(present_params, 1, None):  # skip self arg
+        # inspect does "clever" things to show annotations as strings because we have
+        # `from __future__ import annotations` in main, we don't want that
+        if fields.get(param.name):
+            # exclude params with init=False
+            if getattr(fields[param.name], 'init', True) is False:
+                continue
+            param = param.replace(name=_field_name_for_signature(param.name, fields[param.name]))
+        if param.annotation == 'Any':
+            param = param.replace(annotation=Any)
+        if param.kind is param.VAR_KEYWORD:
+            var_kw = param
+            continue
+        merged_params[param.name] = param
+
+    if var_kw:  # if custom init has no var_kw, fields which are not declared in it cannot be passed through
+        allow_names = config_wrapper.populate_by_name
+        for field_name, field in fields.items():
+            # when alias is a str it should be used for signature generation
+            param_name = _field_name_for_signature(field_name, field)
+
+            if field_name in merged_params or param_name in merged_params:
+                continue
+
+            if not is_valid_identifier(param_name):
+                if allow_names:
+                    param_name = field_name
+                else:
+                    use_var_kw = True
+                    continue
+
+            kwargs = {} if field.is_required() else {'default': field.get_default(call_default_factory=False)}
+            merged_params[param_name] = Parameter(
+                param_name, Parameter.KEYWORD_ONLY, annotation=field.rebuild_annotation(), **kwargs
+            )
+
+    if config_wrapper.extra == 'allow':
+        use_var_kw = True
+
+    if var_kw and use_var_kw:
+        # Make sure the parameter for extra kwargs
+        # does not have the same name as a field
+        default_model_signature = [
+            ('self', Parameter.POSITIONAL_ONLY),
+            ('data', Parameter.VAR_KEYWORD),
+        ]
+        if [(p.name, p.kind) for p in present_params] == default_model_signature:
+            # if this is the standard model signature, use extra_data as the extra args name
+            var_kw_name = 'extra_data'
+        else:
+            # else start from var_kw
+            var_kw_name = var_kw.name
+
+        # generate a name that's definitely unique
+        while var_kw_name in fields:
+            var_kw_name += '_'
+        merged_params[var_kw_name] = var_kw.replace(name=var_kw_name)
+
+    return merged_params
+
+
+def generate_pydantic_signature(
+    init: Callable[..., None], fields: dict[str, FieldInfo], config_wrapper: ConfigWrapper, is_dataclass: bool = False
+) -> Signature:
+    """Generate signature for a pydantic BaseModel or dataclass.
+
+    Args:
+        init: The class init.
+        fields: The model fields.
+        config_wrapper: The config wrapper instance.
+        is_dataclass: Whether the model is a dataclass.
+
+    Returns:
+        The dataclass/BaseModel subclass signature.
+    """
+    merged_params = _generate_signature_parameters(init, fields, config_wrapper)
+
+    if is_dataclass:
+        merged_params = {k: _process_param_defaults(v) for k, v in merged_params.items()}
+
+    return Signature(parameters=list(merged_params.values()), return_annotation=None)

lib/pydantic/_internal/_std_types_schema.py

@@ -0,0 +1,714 @@
+"""Logic for generating pydantic-core schemas for standard library types.
+
+Import of this module is deferred since it contains imports of many standard library modules.
+"""
+from __future__ import annotations as _annotations
+
+import collections
+import collections.abc
+import dataclasses
+import decimal
+import inspect
+import os
+import typing
+from enum import Enum
+from functools import partial
+from ipaddress import IPv4Address, IPv4Interface, IPv4Network, IPv6Address, IPv6Interface, IPv6Network
+from typing import Any, Callable, Iterable, TypeVar
+
+import typing_extensions
+from pydantic_core import (
+    CoreSchema,
+    MultiHostUrl,
+    PydanticCustomError,
+    PydanticOmit,
+    Url,
+    core_schema,
+)
+from typing_extensions import get_args, get_origin
+
+from pydantic.errors import PydanticSchemaGenerationError
+from pydantic.fields import FieldInfo
+from pydantic.types import Strict
+
+from ..config import ConfigDict
+from ..json_schema import JsonSchemaValue, update_json_schema
+from . import _known_annotated_metadata, _typing_extra, _validators
+from ._core_utils import get_type_ref
+from ._internal_dataclass import slots_true
+from ._schema_generation_shared import GetCoreSchemaHandler, GetJsonSchemaHandler
+
+if typing.TYPE_CHECKING:
+    from ._generate_schema import GenerateSchema
+
+    StdSchemaFunction = Callable[[GenerateSchema, type[Any]], core_schema.CoreSchema]
+
+
+@dataclasses.dataclass(**slots_true)
+class SchemaTransformer:
+    get_core_schema: Callable[[Any, GetCoreSchemaHandler], CoreSchema]
+    get_json_schema: Callable[[CoreSchema, GetJsonSchemaHandler], JsonSchemaValue]
+
+    def __get_pydantic_core_schema__(self, source_type: Any, handler: GetCoreSchemaHandler) -> CoreSchema:
+        return self.get_core_schema(source_type, handler)
+
+    def __get_pydantic_json_schema__(self, schema: CoreSchema, handler: GetJsonSchemaHandler) -> JsonSchemaValue:
+        return self.get_json_schema(schema, handler)
+
+
+def get_enum_core_schema(enum_type: type[Enum], config: ConfigDict) -> CoreSchema:
+    cases: list[Any] = list(enum_type.__members__.values())
+
+    enum_ref = get_type_ref(enum_type)
+    description = None if not enum_type.__doc__ else inspect.cleandoc(enum_type.__doc__)
+    if description == 'An enumeration.':  # This is the default value provided by enum.EnumMeta.__new__; don't use it
+        description = None
+    updates = {'title': enum_type.__name__, 'description': description}
+    updates = {k: v for k, v in updates.items() if v is not None}
+
+    def get_json_schema(_, handler: GetJsonSchemaHandler) -> JsonSchemaValue:
+        json_schema = handler(core_schema.literal_schema([x.value for x in cases], ref=enum_ref))
+        original_schema = handler.resolve_ref_schema(json_schema)
+        update_json_schema(original_schema, updates)
+        return json_schema
+
+    if not cases:
+        # Use an isinstance check for enums with no cases.
+        # The most important use case for this is creating TypeVar bounds for generics that should
+        # be restricted to enums. This is more consistent than it might seem at first, since you can only
+        # subclass enum.Enum (or subclasses of enum.Enum) if all parent classes have no cases.
+        # We use the get_json_schema function when an Enum subclass has been declared with no cases
+        # so that we can still generate a valid json schema.
+        return core_schema.is_instance_schema(enum_type, metadata={'pydantic_js_functions': [get_json_schema]})
+
+    use_enum_values = config.get('use_enum_values', False)
+
+    if len(cases) == 1:
+        expected = repr(cases[0].value)
+    else:
+        expected = ', '.join([repr(case.value) for case in cases[:-1]]) + f' or {cases[-1].value!r}'
+
+    def to_enum(__input_value: Any) -> Enum:
+        try:
+            enum_field = enum_type(__input_value)
+            if use_enum_values:
+                return enum_field.value
+            return enum_field
+        except ValueError:
+            # The type: ignore on the next line is to ignore the requirement of LiteralString
+            raise PydanticCustomError('enum', f'Input should be {expected}', {'expected': expected})  # type: ignore
+
+    strict_python_schema = core_schema.is_instance_schema(enum_type)
+    if use_enum_values:
+        strict_python_schema = core_schema.chain_schema(
+            [strict_python_schema, core_schema.no_info_plain_validator_function(lambda x: x.value)]
+        )
+
+    to_enum_validator = core_schema.no_info_plain_validator_function(to_enum)
+    if issubclass(enum_type, int):
+        # this handles `IntEnum`, and also `Foobar(int, Enum)`
+        updates['type'] = 'integer'
+        lax = core_schema.chain_schema([core_schema.int_schema(), to_enum_validator])
+        # Disallow float from JSON due to strict mode
+        strict = core_schema.json_or_python_schema(
+            json_schema=core_schema.no_info_after_validator_function(to_enum, core_schema.int_schema()),
+            python_schema=strict_python_schema,
+        )
+    elif issubclass(enum_type, str):
+        # this handles `StrEnum` (3.11 only), and also `Foobar(str, Enum)`
+        updates['type'] = 'string'
+        lax = core_schema.chain_schema([core_schema.str_schema(), to_enum_validator])
+        strict = core_schema.json_or_python_schema(
+            json_schema=core_schema.no_info_after_validator_function(to_enum, core_schema.str_schema()),
+            python_schema=strict_python_schema,
+        )
+    elif issubclass(enum_type, float):
+        updates['type'] = 'numeric'
+        lax = core_schema.chain_schema([core_schema.float_schema(), to_enum_validator])
+        strict = core_schema.json_or_python_schema(
+            json_schema=core_schema.no_info_after_validator_function(to_enum, core_schema.float_schema()),
+            python_schema=strict_python_schema,
+        )
+    else:
+        lax = to_enum_validator
+        strict = core_schema.json_or_python_schema(json_schema=to_enum_validator, python_schema=strict_python_schema)
+    return core_schema.lax_or_strict_schema(
+        lax_schema=lax, strict_schema=strict, ref=enum_ref, metadata={'pydantic_js_functions': [get_json_schema]}
+    )
+
+
+@dataclasses.dataclass(**slots_true)
+class InnerSchemaValidator:
+    """Use a fixed CoreSchema, avoiding interference from outward annotations."""
+
+    core_schema: CoreSchema
+    js_schema: JsonSchemaValue | None = None
+    js_core_schema: CoreSchema | None = None
+    js_schema_update: JsonSchemaValue | None = None
+
+    def __get_pydantic_json_schema__(self, _schema: CoreSchema, handler: GetJsonSchemaHandler) -> JsonSchemaValue:
+        if self.js_schema is not None:
+            return self.js_schema
+        js_schema = handler(self.js_core_schema or self.core_schema)
+        if self.js_schema_update is not None:
+            js_schema.update(self.js_schema_update)
+        return js_schema
+
+    def __get_pydantic_core_schema__(self, _source_type: Any, _handler: GetCoreSchemaHandler) -> CoreSchema:
+        return self.core_schema
+
+
+def decimal_prepare_pydantic_annotations(
+    source: Any, annotations: Iterable[Any], config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    if source is not decimal.Decimal:
+        return None
+
+    metadata, remaining_annotations = _known_annotated_metadata.collect_known_metadata(annotations)
+
+    config_allow_inf_nan = config.get('allow_inf_nan')
+    if config_allow_inf_nan is not None:
+        metadata.setdefault('allow_inf_nan', config_allow_inf_nan)
+
+    _known_annotated_metadata.check_metadata(
+        metadata, {*_known_annotated_metadata.FLOAT_CONSTRAINTS, 'max_digits', 'decimal_places'}, decimal.Decimal
+    )
+    return source, [InnerSchemaValidator(core_schema.decimal_schema(**metadata)), *remaining_annotations]
+
+
+def datetime_prepare_pydantic_annotations(
+    source_type: Any, annotations: Iterable[Any], _config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    import datetime
+
+    metadata, remaining_annotations = _known_annotated_metadata.collect_known_metadata(annotations)
+    if source_type is datetime.date:
+        sv = InnerSchemaValidator(core_schema.date_schema(**metadata))
+    elif source_type is datetime.datetime:
+        sv = InnerSchemaValidator(core_schema.datetime_schema(**metadata))
+    elif source_type is datetime.time:
+        sv = InnerSchemaValidator(core_schema.time_schema(**metadata))
+    elif source_type is datetime.timedelta:
+        sv = InnerSchemaValidator(core_schema.timedelta_schema(**metadata))
+    else:
+        return None
+    # check now that we know the source type is correct
+    _known_annotated_metadata.check_metadata(metadata, _known_annotated_metadata.DATE_TIME_CONSTRAINTS, source_type)
+    return (source_type, [sv, *remaining_annotations])
+
+
+def uuid_prepare_pydantic_annotations(
+    source_type: Any, annotations: Iterable[Any], _config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    # UUIDs have no constraints - they are fixed length, constructing a UUID instance checks the length
+
+    from uuid import UUID
+
+    if source_type is not UUID:
+        return None
+
+    return (source_type, [InnerSchemaValidator(core_schema.uuid_schema()), *annotations])
+
+
+def path_schema_prepare_pydantic_annotations(
+    source_type: Any, annotations: Iterable[Any], _config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    import pathlib
+
+    if source_type not in {
+        os.PathLike,
+        pathlib.Path,
+        pathlib.PurePath,
+        pathlib.PosixPath,
+        pathlib.PurePosixPath,
+        pathlib.PureWindowsPath,
+    }:
+        return None
+
+    metadata, remaining_annotations = _known_annotated_metadata.collect_known_metadata(annotations)
+    _known_annotated_metadata.check_metadata(metadata, _known_annotated_metadata.STR_CONSTRAINTS, source_type)
+
+    construct_path = pathlib.PurePath if source_type is os.PathLike else source_type
+
+    def path_validator(input_value: str) -> os.PathLike[Any]:
+        try:
+            return construct_path(input_value)
+        except TypeError as e:
+            raise PydanticCustomError('path_type', 'Input is not a valid path') from e
+
+    constrained_str_schema = core_schema.str_schema(**metadata)
+
+    instance_schema = core_schema.json_or_python_schema(
+        json_schema=core_schema.no_info_after_validator_function(path_validator, constrained_str_schema),
+        python_schema=core_schema.is_instance_schema(source_type),
+    )
+
+    strict: bool | None = None
+    for annotation in annotations:
+        if isinstance(annotation, Strict):
+            strict = annotation.strict
+
+    schema = core_schema.lax_or_strict_schema(
+        lax_schema=core_schema.union_schema(
+            [
+                instance_schema,
+                core_schema.no_info_after_validator_function(path_validator, constrained_str_schema),
+            ],
+            custom_error_type='path_type',
+            custom_error_message='Input is not a valid path',
+            strict=True,
+        ),
+        strict_schema=instance_schema,
+        serialization=core_schema.to_string_ser_schema(),
+        strict=strict,
+    )
+
+    return (
+        source_type,
+        [
+            InnerSchemaValidator(schema, js_core_schema=constrained_str_schema, js_schema_update={'format': 'path'}),
+            *remaining_annotations,
+        ],
+    )
+
+
+def dequeue_validator(
+    input_value: Any, handler: core_schema.ValidatorFunctionWrapHandler, maxlen: None | int
+) -> collections.deque[Any]:
+    if isinstance(input_value, collections.deque):
+        maxlens = [v for v in (input_value.maxlen, maxlen) if v is not None]
+        if maxlens:
+            maxlen = min(maxlens)
+        return collections.deque(handler(input_value), maxlen=maxlen)
+    else:
+        return collections.deque(handler(input_value), maxlen=maxlen)
+
+
+@dataclasses.dataclass(**slots_true)
+class SequenceValidator:
+    mapped_origin: type[Any]
+    item_source_type: type[Any]
+    min_length: int | None = None
+    max_length: int | None = None
+    strict: bool = False
+
+    def serialize_sequence_via_list(
+        self, v: Any, handler: core_schema.SerializerFunctionWrapHandler, info: core_schema.SerializationInfo
+    ) -> Any:
+        items: list[Any] = []
+        for index, item in enumerate(v):
+            try:
+                v = handler(item, index)
+            except PydanticOmit:
+                pass
+            else:
+                items.append(v)
+
+        if info.mode_is_json():
+            return items
+        else:
+            return self.mapped_origin(items)
+
+    def __get_pydantic_core_schema__(self, source_type: Any, handler: GetCoreSchemaHandler) -> CoreSchema:
+        if self.item_source_type is Any:
+            items_schema = None
+        else:
+            items_schema = handler.generate_schema(self.item_source_type)
+
+        metadata = {'min_length': self.min_length, 'max_length': self.max_length, 'strict': self.strict}
+
+        if self.mapped_origin in (list, set, frozenset):
+            if self.mapped_origin is list:
+                constrained_schema = core_schema.list_schema(items_schema, **metadata)
+            elif self.mapped_origin is set:
+                constrained_schema = core_schema.set_schema(items_schema, **metadata)
+            else:
+                assert self.mapped_origin is frozenset  # safety check in case we forget to add a case
+                constrained_schema = core_schema.frozenset_schema(items_schema, **metadata)
+
+            schema = constrained_schema
+        else:
+            # safety check in case we forget to add a case
+            assert self.mapped_origin in (collections.deque, collections.Counter)
+
+            if self.mapped_origin is collections.deque:
+                # if we have a MaxLen annotation might as well set that as the default maxlen on the deque
+                # this lets us re-use existing metadata annotations to let users set the maxlen on a dequeue
+                # that e.g. comes from JSON
+                coerce_instance_wrap = partial(
+                    core_schema.no_info_wrap_validator_function,
+                    partial(dequeue_validator, maxlen=metadata.get('max_length', None)),
+                )
+            else:
+                coerce_instance_wrap = partial(core_schema.no_info_after_validator_function, self.mapped_origin)
+
+            constrained_schema = core_schema.list_schema(items_schema, **metadata)
+
+            check_instance = core_schema.json_or_python_schema(
+                json_schema=core_schema.list_schema(),
+                python_schema=core_schema.is_instance_schema(self.mapped_origin),
+            )
+
+            serialization = core_schema.wrap_serializer_function_ser_schema(
+                self.serialize_sequence_via_list, schema=items_schema or core_schema.any_schema(), info_arg=True
+            )
+
+            strict = core_schema.chain_schema([check_instance, coerce_instance_wrap(constrained_schema)])
+
+            if metadata.get('strict', False):
+                schema = strict
+            else:
+                lax = coerce_instance_wrap(constrained_schema)
+                schema = core_schema.lax_or_strict_schema(lax_schema=lax, strict_schema=strict)
+            schema['serialization'] = serialization
+
+        return schema
+
+
+SEQUENCE_ORIGIN_MAP: dict[Any, Any] = {
+    typing.Deque: collections.deque,
+    collections.deque: collections.deque,
+    list: list,
+    typing.List: list,
+    set: set,
+    typing.AbstractSet: set,
+    typing.Set: set,
+    frozenset: frozenset,
+    typing.FrozenSet: frozenset,
+    typing.Sequence: list,
+    typing.MutableSequence: list,
+    typing.MutableSet: set,
+    # this doesn't handle subclasses of these
+    # parametrized typing.Set creates one of these
+    collections.abc.MutableSet: set,
+    collections.abc.Set: frozenset,
+}
+
+
+def identity(s: CoreSchema) -> CoreSchema:
+    return s
+
+
+def sequence_like_prepare_pydantic_annotations(
+    source_type: Any, annotations: Iterable[Any], _config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    origin: Any = get_origin(source_type)
+
+    mapped_origin = SEQUENCE_ORIGIN_MAP.get(origin, None) if origin else SEQUENCE_ORIGIN_MAP.get(source_type, None)
+    if mapped_origin is None:
+        return None
+
+    args = get_args(source_type)
+
+    if not args:
+        args = (Any,)
+    elif len(args) != 1:
+        raise ValueError('Expected sequence to have exactly 1 generic parameter')
+
+    item_source_type = args[0]
+
+    metadata, remaining_annotations = _known_annotated_metadata.collect_known_metadata(annotations)
+    _known_annotated_metadata.check_metadata(metadata, _known_annotated_metadata.SEQUENCE_CONSTRAINTS, source_type)
+
+    return (source_type, [SequenceValidator(mapped_origin, item_source_type, **metadata), *remaining_annotations])
+
+
+MAPPING_ORIGIN_MAP: dict[Any, Any] = {
+    typing.DefaultDict: collections.defaultdict,
+    collections.defaultdict: collections.defaultdict,
+    collections.OrderedDict: collections.OrderedDict,
+    typing_extensions.OrderedDict: collections.OrderedDict,
+    dict: dict,
+    typing.Dict: dict,
+    collections.Counter: collections.Counter,
+    typing.Counter: collections.Counter,
+    # this doesn't handle subclasses of these
+    typing.Mapping: dict,
+    typing.MutableMapping: dict,
+    # parametrized typing.{Mutable}Mapping creates one of these
+    collections.abc.MutableMapping: dict,
+    collections.abc.Mapping: dict,
+}
+
+
+def defaultdict_validator(
+    input_value: Any, handler: core_schema.ValidatorFunctionWrapHandler, default_default_factory: Callable[[], Any]
+) -> collections.defaultdict[Any, Any]:
+    if isinstance(input_value, collections.defaultdict):
+        default_factory = input_value.default_factory
+        return collections.defaultdict(default_factory, handler(input_value))
+    else:
+        return collections.defaultdict(default_default_factory, handler(input_value))
+
+
+def get_defaultdict_default_default_factory(values_source_type: Any) -> Callable[[], Any]:
+    def infer_default() -> Callable[[], Any]:
+        allowed_default_types: dict[Any, Any] = {
+            typing.Tuple: tuple,
+            tuple: tuple,
+            collections.abc.Sequence: tuple,
+            collections.abc.MutableSequence: list,
+            typing.List: list,
+            list: list,
+            typing.Sequence: list,
+            typing.Set: set,
+            set: set,
+            typing.MutableSet: set,
+            collections.abc.MutableSet: set,
+            collections.abc.Set: frozenset,
+            typing.MutableMapping: dict,
+            typing.Mapping: dict,
+            collections.abc.Mapping: dict,
+            collections.abc.MutableMapping: dict,
+            float: float,
+            int: int,
+            str: str,
+            bool: bool,
+        }
+        values_type_origin = get_origin(values_source_type) or values_source_type
+        instructions = 'set using `DefaultDict[..., Annotated[..., Field(default_factory=...)]]`'
+        if isinstance(values_type_origin, TypeVar):
+
+            def type_var_default_factory() -> None:
+                raise RuntimeError(
+                    'Generic defaultdict cannot be used without a concrete value type or an'
+                    ' explicit default factory, ' + instructions
+                )
+
+            return type_var_default_factory
+        elif values_type_origin not in allowed_default_types:
+            # a somewhat subjective set of types that have reasonable default values
+            allowed_msg = ', '.join([t.__name__ for t in set(allowed_default_types.values())])
+            raise PydanticSchemaGenerationError(
+                f'Unable to infer a default factory for keys of type {values_source_type}.'
+                f' Only {allowed_msg} are supported, other types require an explicit default factory'
+                ' ' + instructions
+            )
+        return allowed_default_types[values_type_origin]
+
+    # Assume Annotated[..., Field(...)]
+    if _typing_extra.is_annotated(values_source_type):
+        field_info = next((v for v in get_args(values_source_type) if isinstance(v, FieldInfo)), None)
+    else:
+        field_info = None
+    if field_info and field_info.default_factory:
+        default_default_factory = field_info.default_factory
+    else:
+        default_default_factory = infer_default()
+    return default_default_factory
+
+
+@dataclasses.dataclass(**slots_true)
+class MappingValidator:
+    mapped_origin: type[Any]
+    keys_source_type: type[Any]
+    values_source_type: type[Any]
+    min_length: int | None = None
+    max_length: int | None = None
+    strict: bool = False
+
+    def serialize_mapping_via_dict(self, v: Any, handler: core_schema.SerializerFunctionWrapHandler) -> Any:
+        return handler(v)
+
+    def __get_pydantic_core_schema__(self, source_type: Any, handler: GetCoreSchemaHandler) -> CoreSchema:
+        if self.keys_source_type is Any:
+            keys_schema = None
+        else:
+            keys_schema = handler.generate_schema(self.keys_source_type)
+        if self.values_source_type is Any:
+            values_schema = None
+        else:
+            values_schema = handler.generate_schema(self.values_source_type)
+
+        metadata = {'min_length': self.min_length, 'max_length': self.max_length, 'strict': self.strict}
+
+        if self.mapped_origin is dict:
+            schema = core_schema.dict_schema(keys_schema, values_schema, **metadata)
+        else:
+            constrained_schema = core_schema.dict_schema(keys_schema, values_schema, **metadata)
+            check_instance = core_schema.json_or_python_schema(
+                json_schema=core_schema.dict_schema(),
+                python_schema=core_schema.is_instance_schema(self.mapped_origin),
+            )
+
+            if self.mapped_origin is collections.defaultdict:
+                default_default_factory = get_defaultdict_default_default_factory(self.values_source_type)
+                coerce_instance_wrap = partial(
+                    core_schema.no_info_wrap_validator_function,
+                    partial(defaultdict_validator, default_default_factory=default_default_factory),
+                )
+            else:
+                coerce_instance_wrap = partial(core_schema.no_info_after_validator_function, self.mapped_origin)
+
+            serialization = core_schema.wrap_serializer_function_ser_schema(
+                self.serialize_mapping_via_dict,
+                schema=core_schema.dict_schema(
+                    keys_schema or core_schema.any_schema(), values_schema or core_schema.any_schema()
+                ),
+                info_arg=False,
+            )
+
+            strict = core_schema.chain_schema([check_instance, coerce_instance_wrap(constrained_schema)])
+
+            if metadata.get('strict', False):
+                schema = strict
+            else:
+                lax = coerce_instance_wrap(constrained_schema)
+                schema = core_schema.lax_or_strict_schema(lax_schema=lax, strict_schema=strict)
+                schema['serialization'] = serialization
+
+        return schema
+
+
+def mapping_like_prepare_pydantic_annotations(
+    source_type: Any, annotations: Iterable[Any], _config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    origin: Any = get_origin(source_type)
+
+    mapped_origin = MAPPING_ORIGIN_MAP.get(origin, None) if origin else MAPPING_ORIGIN_MAP.get(source_type, None)
+    if mapped_origin is None:
+        return None
+
+    args = get_args(source_type)
+
+    if not args:
+        args = (Any, Any)
+    elif mapped_origin is collections.Counter:
+        # a single generic
+        if len(args) != 1:
+            raise ValueError('Expected Counter to have exactly 1 generic parameter')
+        args = (args[0], int)  # keys are always an int
+    elif len(args) != 2:
+        raise ValueError('Expected mapping to have exactly 2 generic parameters')
+
+    keys_source_type, values_source_type = args
+
+    metadata, remaining_annotations = _known_annotated_metadata.collect_known_metadata(annotations)
+    _known_annotated_metadata.check_metadata(metadata, _known_annotated_metadata.SEQUENCE_CONSTRAINTS, source_type)
+
+    return (
+        source_type,
+        [
+            MappingValidator(mapped_origin, keys_source_type, values_source_type, **metadata),
+            *remaining_annotations,
+        ],
+    )
+
+
+def ip_prepare_pydantic_annotations(
+    source_type: Any, annotations: Iterable[Any], _config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    def make_strict_ip_schema(tp: type[Any]) -> CoreSchema:
+        return core_schema.json_or_python_schema(
+            json_schema=core_schema.no_info_after_validator_function(tp, core_schema.str_schema()),
+            python_schema=core_schema.is_instance_schema(tp),
+        )
+
+    if source_type is IPv4Address:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.lax_or_strict_schema(
+                    lax_schema=core_schema.no_info_plain_validator_function(_validators.ip_v4_address_validator),
+                    strict_schema=make_strict_ip_schema(IPv4Address),
+                    serialization=core_schema.to_string_ser_schema(),
+                ),
+                lambda _1, _2: {'type': 'string', 'format': 'ipv4'},
+            ),
+            *annotations,
+        ]
+    if source_type is IPv4Network:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.lax_or_strict_schema(
+                    lax_schema=core_schema.no_info_plain_validator_function(_validators.ip_v4_network_validator),
+                    strict_schema=make_strict_ip_schema(IPv4Network),
+                    serialization=core_schema.to_string_ser_schema(),
+                ),
+                lambda _1, _2: {'type': 'string', 'format': 'ipv4network'},
+            ),
+            *annotations,
+        ]
+    if source_type is IPv4Interface:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.lax_or_strict_schema(
+                    lax_schema=core_schema.no_info_plain_validator_function(_validators.ip_v4_interface_validator),
+                    strict_schema=make_strict_ip_schema(IPv4Interface),
+                    serialization=core_schema.to_string_ser_schema(),
+                ),
+                lambda _1, _2: {'type': 'string', 'format': 'ipv4interface'},
+            ),
+            *annotations,
+        ]
+
+    if source_type is IPv6Address:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.lax_or_strict_schema(
+                    lax_schema=core_schema.no_info_plain_validator_function(_validators.ip_v6_address_validator),
+                    strict_schema=make_strict_ip_schema(IPv6Address),
+                    serialization=core_schema.to_string_ser_schema(),
+                ),
+                lambda _1, _2: {'type': 'string', 'format': 'ipv6'},
+            ),
+            *annotations,
+        ]
+    if source_type is IPv6Network:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.lax_or_strict_schema(
+                    lax_schema=core_schema.no_info_plain_validator_function(_validators.ip_v6_network_validator),
+                    strict_schema=make_strict_ip_schema(IPv6Network),
+                    serialization=core_schema.to_string_ser_schema(),
+                ),
+                lambda _1, _2: {'type': 'string', 'format': 'ipv6network'},
+            ),
+            *annotations,
+        ]
+    if source_type is IPv6Interface:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.lax_or_strict_schema(
+                    lax_schema=core_schema.no_info_plain_validator_function(_validators.ip_v6_interface_validator),
+                    strict_schema=make_strict_ip_schema(IPv6Interface),
+                    serialization=core_schema.to_string_ser_schema(),
+                ),
+                lambda _1, _2: {'type': 'string', 'format': 'ipv6interface'},
+            ),
+            *annotations,
+        ]
+
+    return None
+
+
+def url_prepare_pydantic_annotations(
+    source_type: Any, annotations: Iterable[Any], _config: ConfigDict
+) -> tuple[Any, list[Any]] | None:
+    if source_type is Url:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.url_schema(),
+                lambda cs, handler: handler(cs),
+            ),
+            *annotations,
+        ]
+    if source_type is MultiHostUrl:
+        return source_type, [
+            SchemaTransformer(
+                lambda _1, _2: core_schema.multi_host_url_schema(),
+                lambda cs, handler: handler(cs),
+            ),
+            *annotations,
+        ]
+
+
+PREPARE_METHODS: tuple[Callable[[Any, Iterable[Any], ConfigDict], tuple[Any, list[Any]] | None], ...] = (
+    decimal_prepare_pydantic_annotations,
+    sequence_like_prepare_pydantic_annotations,
+    datetime_prepare_pydantic_annotations,
+    uuid_prepare_pydantic_annotations,
+    path_schema_prepare_pydantic_annotations,
+    mapping_like_prepare_pydantic_annotations,
+    ip_prepare_pydantic_annotations,
+    url_prepare_pydantic_annotations,
+)

lib/pydantic/_internal/_typing_extra.py

@@ -0,0 +1,469 @@
+"""Logic for interacting with type annotations, mostly extensions, shims and hacks to wrap python's typing module."""
+from __future__ import annotations as _annotations
+
+import dataclasses
+import sys
+import types
+import typing
+from collections.abc import Callable
+from functools import partial
+from types import GetSetDescriptorType
+from typing import TYPE_CHECKING, Any, Final
+
+from typing_extensions import Annotated, Literal, TypeAliasType, TypeGuard, get_args, get_origin
+
+if TYPE_CHECKING:
+    from ._dataclasses import StandardDataclass
+
+try:
+    from typing import _TypingBase  # type: ignore[attr-defined]
+except ImportError:
+    from typing import _Final as _TypingBase  # type: ignore[attr-defined]
+
+typing_base = _TypingBase
+
+
+if sys.version_info < (3, 9):
+    # python < 3.9 does not have GenericAlias (list[int], tuple[str, ...] and so on)
+    TypingGenericAlias = ()
+else:
+    from typing import GenericAlias as TypingGenericAlias  # type: ignore
+
+
+if sys.version_info < (3, 11):
+    from typing_extensions import NotRequired, Required
+else:
+    from typing import NotRequired, Required  # noqa: F401
+
+
+if sys.version_info < (3, 10):
+
+    def origin_is_union(tp: type[Any] | None) -> bool:
+        return tp is typing.Union
+
+    WithArgsTypes = (TypingGenericAlias,)
+
+else:
+
+    def origin_is_union(tp: type[Any] | None) -> bool:
+        return tp is typing.Union or tp is types.UnionType
+
+    WithArgsTypes = typing._GenericAlias, types.GenericAlias, types.UnionType  # type: ignore[attr-defined]
+
+
+if sys.version_info < (3, 10):
+    NoneType = type(None)
+    EllipsisType = type(Ellipsis)
+else:
+    from types import NoneType as NoneType
+
+
+LITERAL_TYPES: set[Any] = {Literal}
+if hasattr(typing, 'Literal'):
+    LITERAL_TYPES.add(typing.Literal)  # type: ignore
+
+NONE_TYPES: tuple[Any, ...] = (None, NoneType, *(tp[None] for tp in LITERAL_TYPES))
+
+
+TypeVarType = Any  # since mypy doesn't allow the use of TypeVar as a type
+
+
+def is_none_type(type_: Any) -> bool:
+    return type_ in NONE_TYPES
+
+
+def is_callable_type(type_: type[Any]) -> bool:
+    return type_ is Callable or get_origin(type_) is Callable
+
+
+def is_literal_type(type_: type[Any]) -> bool:
+    return Literal is not None and get_origin(type_) in LITERAL_TYPES
+
+
+def literal_values(type_: type[Any]) -> tuple[Any, ...]:
+    return get_args(type_)
+
+
+def all_literal_values(type_: type[Any]) -> list[Any]:
+    """This method is used to retrieve all Literal values as
+    Literal can be used recursively (see https://www.python.org/dev/peps/pep-0586)
+    e.g. `Literal[Literal[Literal[1, 2, 3], "foo"], 5, None]`.
+    """
+    if not is_literal_type(type_):
+        return [type_]
+
+    values = literal_values(type_)
+    return list(x for value in values for x in all_literal_values(value))
+
+
+def is_annotated(ann_type: Any) -> bool:
+    from ._utils import lenient_issubclass
+
+    origin = get_origin(ann_type)
+    return origin is not None and lenient_issubclass(origin, Annotated)
+
+
+def is_namedtuple(type_: type[Any]) -> bool:
+    """Check if a given class is a named tuple.
+    It can be either a `typing.NamedTuple` or `collections.namedtuple`.
+    """
+    from ._utils import lenient_issubclass
+
+    return lenient_issubclass(type_, tuple) and hasattr(type_, '_fields')
+
+
+test_new_type = typing.NewType('test_new_type', str)
+
+
+def is_new_type(type_: type[Any]) -> bool:
+    """Check whether type_ was created using typing.NewType.
+
+    Can't use isinstance because it fails <3.10.
+    """
+    return isinstance(type_, test_new_type.__class__) and hasattr(type_, '__supertype__')  # type: ignore[arg-type]
+
+
+def _check_classvar(v: type[Any] | None) -> bool:
+    if v is None:
+        return False
+
+    return v.__class__ == typing.ClassVar.__class__ and getattr(v, '_name', None) == 'ClassVar'
+
+
+def is_classvar(ann_type: type[Any]) -> bool:
+    if _check_classvar(ann_type) or _check_classvar(get_origin(ann_type)):
+        return True
+
+    # this is an ugly workaround for class vars that contain forward references and are therefore themselves
+    # forward references, see #3679
+    if ann_type.__class__ == typing.ForwardRef and ann_type.__forward_arg__.startswith('ClassVar['):  # type: ignore
+        return True
+
+    return False
+
+
+def _check_finalvar(v: type[Any] | None) -> bool:
+    """Check if a given type is a `typing.Final` type."""
+    if v is None:
+        return False
+
+    return v.__class__ == Final.__class__ and (sys.version_info < (3, 8) or getattr(v, '_name', None) == 'Final')
+
+
+def is_finalvar(ann_type: Any) -> bool:
+    return _check_finalvar(ann_type) or _check_finalvar(get_origin(ann_type))
+
+
+def parent_frame_namespace(*, parent_depth: int = 2) -> dict[str, Any] | None:
+    """We allow use of items in parent namespace to get around the issue with `get_type_hints` only looking in the
+    global module namespace. See https://github.com/pydantic/pydantic/issues/2678#issuecomment-1008139014 -> Scope
+    and suggestion at the end of the next comment by @gvanrossum.
+
+    WARNING 1: it matters exactly where this is called. By default, this function will build a namespace from the
+    parent of where it is called.
+
+    WARNING 2: this only looks in the parent namespace, not other parents since (AFAIK) there's no way to collect a
+    dict of exactly what's in scope. Using `f_back` would work sometimes but would be very wrong and confusing in many
+    other cases. See https://discuss.python.org/t/is-there-a-way-to-access-parent-nested-namespaces/20659.
+    """
+    frame = sys._getframe(parent_depth)
+    # if f_back is None, it's the global module namespace and we don't need to include it here
+    if frame.f_back is None:
+        return None
+    else:
+        return frame.f_locals
+
+
+def add_module_globals(obj: Any, globalns: dict[str, Any] | None = None) -> dict[str, Any]:
+    module_name = getattr(obj, '__module__', None)
+    if module_name:
+        try:
+            module_globalns = sys.modules[module_name].__dict__
+        except KeyError:
+            # happens occasionally, see https://github.com/pydantic/pydantic/issues/2363
+            pass
+        else:
+            if globalns:
+                return {**module_globalns, **globalns}
+            else:
+                # copy module globals to make sure it can't be updated later
+                return module_globalns.copy()
+
+    return globalns or {}
+
+
+def get_cls_types_namespace(cls: type[Any], parent_namespace: dict[str, Any] | None = None) -> dict[str, Any]:
+    ns = add_module_globals(cls, parent_namespace)
+    ns[cls.__name__] = cls
+    return ns
+
+
+def get_cls_type_hints_lenient(obj: Any, globalns: dict[str, Any] | None = None) -> dict[str, Any]:
+    """Collect annotations from a class, including those from parent classes.
+
+    Unlike `typing.get_type_hints`, this function will not error if a forward reference is not resolvable.
+    """
+    hints = {}
+    for base in reversed(obj.__mro__):
+        ann = base.__dict__.get('__annotations__')
+        localns = dict(vars(base))
+        if ann is not None and ann is not GetSetDescriptorType:
+            for name, value in ann.items():
+                hints[name] = eval_type_lenient(value, globalns, localns)
+    return hints
+
+
+def eval_type_lenient(value: Any, globalns: dict[str, Any] | None = None, localns: dict[str, Any] | None = None) -> Any:
+    """Behaves like typing._eval_type, except it won't raise an error if a forward reference can't be resolved."""
+    if value is None:
+        value = NoneType
+    elif isinstance(value, str):
+        value = _make_forward_ref(value, is_argument=False, is_class=True)
+
+    try:
+        return eval_type_backport(value, globalns, localns)
+    except NameError:
+        # the point of this function is to be tolerant to this case
+        return value
+
+
+def eval_type_backport(
+    value: Any, globalns: dict[str, Any] | None = None, localns: dict[str, Any] | None = None
+) -> Any:
+    """Like `typing._eval_type`, but falls back to the `eval_type_backport` package if it's
+    installed to let older Python versions use newer typing features.
+    Specifically, this transforms `X | Y` into `typing.Union[X, Y]`
+    and `list[X]` into `typing.List[X]` etc. (for all the types made generic in PEP 585)
+    if the original syntax is not supported in the current Python version.
+    """
+    try:
+        return typing._eval_type(  # type: ignore
+            value, globalns, localns
+        )
+    except TypeError as e:
+        if not (isinstance(value, typing.ForwardRef) and is_backport_fixable_error(e)):
+            raise
+        try:
+            from eval_type_backport import eval_type_backport
+        except ImportError:
+            raise TypeError(
+                f'You have a type annotation {value.__forward_arg__!r} '
+                f'which makes use of newer typing features than are supported in your version of Python. '
+                f'To handle this error, you should either remove the use of new syntax '
+                f'or install the `eval_type_backport` package.'
+            ) from e
+
+        return eval_type_backport(value, globalns, localns, try_default=False)
+
+
+def is_backport_fixable_error(e: TypeError) -> bool:
+    msg = str(e)
+    return msg.startswith('unsupported operand type(s) for |: ') or "' object is not subscriptable" in msg
+
+
+def get_function_type_hints(
+    function: Callable[..., Any], *, include_keys: set[str] | None = None, types_namespace: dict[str, Any] | None = None
+) -> dict[str, Any]:
+    """Like `typing.get_type_hints`, but doesn't convert `X` to `Optional[X]` if the default value is `None`, also
+    copes with `partial`.
+    """
+    if isinstance(function, partial):
+        annotations = function.func.__annotations__
+    else:
+        annotations = function.__annotations__
+
+    globalns = add_module_globals(function)
+    type_hints = {}
+    for name, value in annotations.items():
+        if include_keys is not None and name not in include_keys:
+            continue
+        if value is None:
+            value = NoneType
+        elif isinstance(value, str):
+            value = _make_forward_ref(value)
+
+        type_hints[name] = eval_type_backport(value, globalns, types_namespace)
+
+    return type_hints
+
+
+if sys.version_info < (3, 9, 8) or (3, 10) <= sys.version_info < (3, 10, 1):
+
+    def _make_forward_ref(
+        arg: Any,
+        is_argument: bool = True,
+        *,
+        is_class: bool = False,
+    ) -> typing.ForwardRef:
+        """Wrapper for ForwardRef that accounts for the `is_class` argument missing in older versions.
+        The `module` argument is omitted as it breaks <3.9.8, =3.10.0 and isn't used in the calls below.
+
+        See https://github.com/python/cpython/pull/28560 for some background.
+        The backport happened on 3.9.8, see:
+        https://github.com/pydantic/pydantic/discussions/6244#discussioncomment-6275458,
+        and on 3.10.1 for the 3.10 branch, see:
+        https://github.com/pydantic/pydantic/issues/6912
+
+        Implemented as EAFP with memory.
+        """
+        return typing.ForwardRef(arg, is_argument)
+
+else:
+    _make_forward_ref = typing.ForwardRef
+
+
+if sys.version_info >= (3, 10):
+    get_type_hints = typing.get_type_hints
+
+else:
+    """
+    For older versions of python, we have a custom implementation of `get_type_hints` which is a close as possible to
+    the implementation in CPython 3.10.8.
+    """
+
+    @typing.no_type_check
+    def get_type_hints(  # noqa: C901
+        obj: Any,
+        globalns: dict[str, Any] | None = None,
+        localns: dict[str, Any] | None = None,
+        include_extras: bool = False,
+    ) -> dict[str, Any]:  # pragma: no cover
+        """Taken verbatim from python 3.10.8 unchanged, except:
+        * type annotations of the function definition above.
+        * prefixing `typing.` where appropriate
+        * Use `_make_forward_ref` instead of `typing.ForwardRef` to handle the `is_class` argument.
+
+        https://github.com/python/cpython/blob/aaaf5174241496afca7ce4d4584570190ff972fe/Lib/typing.py#L1773-L1875
+
+        DO NOT CHANGE THIS METHOD UNLESS ABSOLUTELY NECESSARY.
+        ======================================================
+
+        Return type hints for an object.
+
+        This is often the same as obj.__annotations__, but it handles
+        forward references encoded as string literals, adds Optional[t] if a
+        default value equal to None is set and recursively replaces all
+        'Annotated[T, ...]' with 'T' (unless 'include_extras=True').
+
+        The argument may be a module, class, method, or function. The annotations
+        are returned as a dictionary. For classes, annotations include also
+        inherited members.
+
+        TypeError is raised if the argument is not of a type that can contain
+        annotations, and an empty dictionary is returned if no annotations are
+        present.
+
+        BEWARE -- the behavior of globalns and localns is counterintuitive
+        (unless you are familiar with how eval() and exec() work).  The
+        search order is locals first, then globals.
+
+        - If no dict arguments are passed, an attempt is made to use the
+          globals from obj (or the respective module's globals for classes),
+          and these are also used as the locals.  If the object does not appear
+          to have globals, an empty dictionary is used.  For classes, the search
+          order is globals first then locals.
+
+        - If one dict argument is passed, it is used for both globals and
+          locals.
+
+        - If two dict arguments are passed, they specify globals and
+          locals, respectively.
+        """
+        if getattr(obj, '__no_type_check__', None):
+            return {}
+        # Classes require a special treatment.
+        if isinstance(obj, type):
+            hints = {}
+            for base in reversed(obj.__mro__):
+                if globalns is None:
+                    base_globals = getattr(sys.modules.get(base.__module__, None), '__dict__', {})
+                else:
+                    base_globals = globalns
+                ann = base.__dict__.get('__annotations__', {})
+                if isinstance(ann, types.GetSetDescriptorType):
+                    ann = {}
+                base_locals = dict(vars(base)) if localns is None else localns
+                if localns is None and globalns is None:
+                    # This is surprising, but required.  Before Python 3.10,
+                    # get_type_hints only evaluated the globalns of
+                    # a class.  To maintain backwards compatibility, we reverse
+                    # the globalns and localns order so that eval() looks into
+                    # *base_globals* first rather than *base_locals*.
+                    # This only affects ForwardRefs.
+                    base_globals, base_locals = base_locals, base_globals
+                for name, value in ann.items():
+                    if value is None:
+                        value = type(None)
+                    if isinstance(value, str):
+                        value = _make_forward_ref(value, is_argument=False, is_class=True)
+
+                    value = eval_type_backport(value, base_globals, base_locals)
+                    hints[name] = value
+            if not include_extras and hasattr(typing, '_strip_annotations'):
+                return {
+                    k: typing._strip_annotations(t)  # type: ignore
+                    for k, t in hints.items()
+                }
+            else:
+                return hints
+
+        if globalns is None:
+            if isinstance(obj, types.ModuleType):
+                globalns = obj.__dict__
+            else:
+                nsobj = obj
+                # Find globalns for the unwrapped object.
+                while hasattr(nsobj, '__wrapped__'):
+                    nsobj = nsobj.__wrapped__
+                globalns = getattr(nsobj, '__globals__', {})
+            if localns is None:
+                localns = globalns
+        elif localns is None:
+            localns = globalns
+        hints = getattr(obj, '__annotations__', None)
+        if hints is None:
+            # Return empty annotations for something that _could_ have them.
+            if isinstance(obj, typing._allowed_types):  # type: ignore
+                return {}
+            else:
+                raise TypeError(f'{obj!r} is not a module, class, method, ' 'or function.')
+        defaults = typing._get_defaults(obj)  # type: ignore
+        hints = dict(hints)
+        for name, value in hints.items():
+            if value is None:
+                value = type(None)
+            if isinstance(value, str):
+                # class-level forward refs were handled above, this must be either
+                # a module-level annotation or a function argument annotation
+
+                value = _make_forward_ref(
+                    value,
+                    is_argument=not isinstance(obj, types.ModuleType),
+                    is_class=False,
+                )
+            value = eval_type_backport(value, globalns, localns)
+            if name in defaults and defaults[name] is None:
+                value = typing.Optional[value]
+            hints[name] = value
+        return hints if include_extras else {k: typing._strip_annotations(t) for k, t in hints.items()}  # type: ignore
+
+
+def is_dataclass(_cls: type[Any]) -> TypeGuard[type[StandardDataclass]]:
+    # The dataclasses.is_dataclass function doesn't seem to provide TypeGuard functionality,
+    # so I created this convenience function
+    return dataclasses.is_dataclass(_cls)
+
+
+def origin_is_type_alias_type(origin: Any) -> TypeGuard[TypeAliasType]:
+    return isinstance(origin, TypeAliasType)
+
+
+if sys.version_info >= (3, 10):
+
+    def is_generic_alias(type_: type[Any]) -> bool:
+        return isinstance(type_, (types.GenericAlias, typing._GenericAlias))  # type: ignore[attr-defined]
+
+else:
+
+    def is_generic_alias(type_: type[Any]) -> bool:
+        return isinstance(type_, typing._GenericAlias)  # type: ignore

lib/pydantic/_internal/_utils.py

@@ -0,0 +1,362 @@
+"""Bucket of reusable internal utilities.
+
+This should be reduced as much as possible with functions only used in one place, moved to that place.
+"""
+from __future__ import annotations as _annotations
+
+import dataclasses
+import keyword
+import typing
+import weakref
+from collections import OrderedDict, defaultdict, deque
+from copy import deepcopy
+from itertools import zip_longest
+from types import BuiltinFunctionType, CodeType, FunctionType, GeneratorType, LambdaType, ModuleType
+from typing import Any, Mapping, TypeVar
+
+from typing_extensions import TypeAlias, TypeGuard
+
+from . import _repr, _typing_extra
+
+if typing.TYPE_CHECKING:
+    MappingIntStrAny: TypeAlias = 'typing.Mapping[int, Any] | typing.Mapping[str, Any]'
+    AbstractSetIntStr: TypeAlias = 'typing.AbstractSet[int] | typing.AbstractSet[str]'
+    from ..main import BaseModel
+
+
+# these are types that are returned unchanged by deepcopy
+IMMUTABLE_NON_COLLECTIONS_TYPES: set[type[Any]] = {
+    int,
+    float,
+    complex,
+    str,
+    bool,
+    bytes,
+    type,
+    _typing_extra.NoneType,
+    FunctionType,
+    BuiltinFunctionType,
+    LambdaType,
+    weakref.ref,
+    CodeType,
+    # note: including ModuleType will differ from behaviour of deepcopy by not producing error.
+    # It might be not a good idea in general, but considering that this function used only internally
+    # against default values of fields, this will allow to actually have a field with module as default value
+    ModuleType,
+    NotImplemented.__class__,
+    Ellipsis.__class__,
+}
+
+# these are types that if empty, might be copied with simple copy() instead of deepcopy()
+BUILTIN_COLLECTIONS: set[type[Any]] = {
+    list,
+    set,
+    tuple,
+    frozenset,
+    dict,
+    OrderedDict,
+    defaultdict,
+    deque,
+}
+
+
+def sequence_like(v: Any) -> bool:
+    return isinstance(v, (list, tuple, set, frozenset, GeneratorType, deque))
+
+
+def lenient_isinstance(o: Any, class_or_tuple: type[Any] | tuple[type[Any], ...] | None) -> bool:  # pragma: no cover
+    try:
+        return isinstance(o, class_or_tuple)  # type: ignore[arg-type]
+    except TypeError:
+        return False
+
+
+def lenient_issubclass(cls: Any, class_or_tuple: Any) -> bool:  # pragma: no cover
+    try:
+        return isinstance(cls, type) and issubclass(cls, class_or_tuple)
+    except TypeError:
+        if isinstance(cls, _typing_extra.WithArgsTypes):
+            return False
+        raise  # pragma: no cover
+
+
+def is_model_class(cls: Any) -> TypeGuard[type[BaseModel]]:
+    """Returns true if cls is a _proper_ subclass of BaseModel, and provides proper type-checking,
+    unlike raw calls to lenient_issubclass.
+    """
+    from ..main import BaseModel
+
+    return lenient_issubclass(cls, BaseModel) and cls is not BaseModel
+
+
+def is_valid_identifier(identifier: str) -> bool:
+    """Checks that a string is a valid identifier and not a Python keyword.
+    :param identifier: The identifier to test.
+    :return: True if the identifier is valid.
+    """
+    return identifier.isidentifier() and not keyword.iskeyword(identifier)
+
+
+KeyType = TypeVar('KeyType')
+
+
+def deep_update(mapping: dict[KeyType, Any], *updating_mappings: dict[KeyType, Any]) -> dict[KeyType, Any]:
+    updated_mapping = mapping.copy()
+    for updating_mapping in updating_mappings:
+        for k, v in updating_mapping.items():
+            if k in updated_mapping and isinstance(updated_mapping[k], dict) and isinstance(v, dict):
+                updated_mapping[k] = deep_update(updated_mapping[k], v)
+            else:
+                updated_mapping[k] = v
+    return updated_mapping
+
+
+def update_not_none(mapping: dict[Any, Any], **update: Any) -> None:
+    mapping.update({k: v for k, v in update.items() if v is not None})
+
+
+T = TypeVar('T')
+
+
+def unique_list(
+    input_list: list[T] | tuple[T, ...],
+    *,
+    name_factory: typing.Callable[[T], str] = str,
+) -> list[T]:
+    """Make a list unique while maintaining order.
+    We update the list if another one with the same name is set
+    (e.g. model validator overridden in subclass).
+    """
+    result: list[T] = []
+    result_names: list[str] = []
+    for v in input_list:
+        v_name = name_factory(v)
+        if v_name not in result_names:
+            result_names.append(v_name)
+            result.append(v)
+        else:
+            result[result_names.index(v_name)] = v
+
+    return result
+
+
+class ValueItems(_repr.Representation):
+    """Class for more convenient calculation of excluded or included fields on values."""
+
+    __slots__ = ('_items', '_type')
+
+    def __init__(self, value: Any, items: AbstractSetIntStr | MappingIntStrAny) -> None:
+        items = self._coerce_items(items)
+
+        if isinstance(value, (list, tuple)):
+            items = self._normalize_indexes(items, len(value))  # type: ignore
+
+        self._items: MappingIntStrAny = items  # type: ignore
+
+    def is_excluded(self, item: Any) -> bool:
+        """Check if item is fully excluded.
+
+        :param item: key or index of a value
+        """
+        return self.is_true(self._items.get(item))
+
+    def is_included(self, item: Any) -> bool:
+        """Check if value is contained in self._items.
+
+        :param item: key or index of value
+        """
+        return item in self._items
+
+    def for_element(self, e: int | str) -> AbstractSetIntStr | MappingIntStrAny | None:
+        """:param e: key or index of element on value
+        :return: raw values for element if self._items is dict and contain needed element
+        """
+        item = self._items.get(e)  # type: ignore
+        return item if not self.is_true(item) else None
+
+    def _normalize_indexes(self, items: MappingIntStrAny, v_length: int) -> dict[int | str, Any]:
+        """:param items: dict or set of indexes which will be normalized
+        :param v_length: length of sequence indexes of which will be
+
+        >>> self._normalize_indexes({0: True, -2: True, -1: True}, 4)
+        {0: True, 2: True, 3: True}
+        >>> self._normalize_indexes({'__all__': True}, 4)
+        {0: True, 1: True, 2: True, 3: True}
+        """
+        normalized_items: dict[int | str, Any] = {}
+        all_items = None
+        for i, v in items.items():
+            if not (isinstance(v, typing.Mapping) or isinstance(v, typing.AbstractSet) or self.is_true(v)):
+                raise TypeError(f'Unexpected type of exclude value for index "{i}" {v.__class__}')
+            if i == '__all__':
+                all_items = self._coerce_value(v)
+                continue
+            if not isinstance(i, int):
+                raise TypeError(
+                    'Excluding fields from a sequence of sub-models or dicts must be performed index-wise: '
+                    'expected integer keys or keyword "__all__"'
+                )
+            normalized_i = v_length + i if i < 0 else i
+            normalized_items[normalized_i] = self.merge(v, normalized_items.get(normalized_i))
+
+        if not all_items:
+            return normalized_items
+        if self.is_true(all_items):
+            for i in range(v_length):
+                normalized_items.setdefault(i, ...)
+            return normalized_items
+        for i in range(v_length):
+            normalized_item = normalized_items.setdefault(i, {})
+            if not self.is_true(normalized_item):
+                normalized_items[i] = self.merge(all_items, normalized_item)
+        return normalized_items
+
+    @classmethod
+    def merge(cls, base: Any, override: Any, intersect: bool = False) -> Any:
+        """Merge a `base` item with an `override` item.
+
+        Both `base` and `override` are converted to dictionaries if possible.
+        Sets are converted to dictionaries with the sets entries as keys and
+        Ellipsis as values.
+
+        Each key-value pair existing in `base` is merged with `override`,
+        while the rest of the key-value pairs are updated recursively with this function.
+
+        Merging takes place based on the "union" of keys if `intersect` is
+        set to `False` (default) and on the intersection of keys if
+        `intersect` is set to `True`.
+        """
+        override = cls._coerce_value(override)
+        base = cls._coerce_value(base)
+        if override is None:
+            return base
+        if cls.is_true(base) or base is None:
+            return override
+        if cls.is_true(override):
+            return base if intersect else override
+
+        # intersection or union of keys while preserving ordering:
+        if intersect:
+            merge_keys = [k for k in base if k in override] + [k for k in override if k in base]
+        else:
+            merge_keys = list(base) + [k for k in override if k not in base]
+
+        merged: dict[int | str, Any] = {}
+        for k in merge_keys:
+            merged_item = cls.merge(base.get(k), override.get(k), intersect=intersect)
+            if merged_item is not None:
+                merged[k] = merged_item
+
+        return merged
+
+    @staticmethod
+    def _coerce_items(items: AbstractSetIntStr | MappingIntStrAny) -> MappingIntStrAny:
+        if isinstance(items, typing.Mapping):
+            pass
+        elif isinstance(items, typing.AbstractSet):
+            items = dict.fromkeys(items, ...)  # type: ignore
+        else:
+            class_name = getattr(items, '__class__', '???')
+            raise TypeError(f'Unexpected type of exclude value {class_name}')
+        return items  # type: ignore
+
+    @classmethod
+    def _coerce_value(cls, value: Any) -> Any:
+        if value is None or cls.is_true(value):
+            return value
+        return cls._coerce_items(value)
+
+    @staticmethod
+    def is_true(v: Any) -> bool:
+        return v is True or v is ...
+
+    def __repr_args__(self) -> _repr.ReprArgs:
+        return [(None, self._items)]
+
+
+if typing.TYPE_CHECKING:
+
+    def ClassAttribute(name: str, value: T) -> T:
+        ...
+
+else:
+
+    class ClassAttribute:
+        """Hide class attribute from its instances."""
+
+        __slots__ = 'name', 'value'
+
+        def __init__(self, name: str, value: Any) -> None:
+            self.name = name
+            self.value = value
+
+        def __get__(self, instance: Any, owner: type[Any]) -> None:
+            if instance is None:
+                return self.value
+            raise AttributeError(f'{self.name!r} attribute of {owner.__name__!r} is class-only')
+
+
+Obj = TypeVar('Obj')
+
+
+def smart_deepcopy(obj: Obj) -> Obj:
+    """Return type as is for immutable built-in types
+    Use obj.copy() for built-in empty collections
+    Use copy.deepcopy() for non-empty collections and unknown objects.
+    """
+    obj_type = obj.__class__
+    if obj_type in IMMUTABLE_NON_COLLECTIONS_TYPES:
+        return obj  # fastest case: obj is immutable and not collection therefore will not be copied anyway
+    try:
+        if not obj and obj_type in BUILTIN_COLLECTIONS:
+            # faster way for empty collections, no need to copy its members
+            return obj if obj_type is tuple else obj.copy()  # tuple doesn't have copy method  # type: ignore
+    except (TypeError, ValueError, RuntimeError):
+        # do we really dare to catch ALL errors? Seems a bit risky
+        pass
+
+    return deepcopy(obj)  # slowest way when we actually might need a deepcopy
+
+
+_SENTINEL = object()
+
+
+def all_identical(left: typing.Iterable[Any], right: typing.Iterable[Any]) -> bool:
+    """Check that the items of `left` are the same objects as those in `right`.
+
+    >>> a, b = object(), object()
+    >>> all_identical([a, b, a], [a, b, a])
+    True
+    >>> all_identical([a, b, [a]], [a, b, [a]])  # new list object, while "equal" is not "identical"
+    False
+    """
+    for left_item, right_item in zip_longest(left, right, fillvalue=_SENTINEL):
+        if left_item is not right_item:
+            return False
+    return True
+
+
+@dataclasses.dataclass(frozen=True)
+class SafeGetItemProxy:
+    """Wrapper redirecting `__getitem__` to `get` with a sentinel value as default
+
+    This makes is safe to use in `operator.itemgetter` when some keys may be missing
+    """
+
+    # Define __slots__manually for performances
+    # @dataclasses.dataclass() only support slots=True in python>=3.10
+    __slots__ = ('wrapped',)
+
+    wrapped: Mapping[str, Any]
+
+    def __getitem__(self, __key: str) -> Any:
+        return self.wrapped.get(__key, _SENTINEL)
+
+    # required to pass the object to operator.itemgetter() instances due to a quirk of typeshed
+    # https://github.com/python/mypy/issues/13713
+    # https://github.com/python/typeshed/pull/8785
+    # Since this is typing-only, hide it in a typing.TYPE_CHECKING block
+    if typing.TYPE_CHECKING:
+
+        def __contains__(self, __key: str) -> bool:
+            return self.wrapped.__contains__(__key)

lib/pydantic/_internal/_validate_call.py

@@ -0,0 +1,84 @@
+from __future__ import annotations as _annotations
+
+import inspect
+from functools import partial
+from typing import Any, Awaitable, Callable
+
+import pydantic_core
+
+from ..config import ConfigDict
+from ..plugin._schema_validator import create_schema_validator
+from . import _generate_schema, _typing_extra
+from ._config import ConfigWrapper
+
+
+class ValidateCallWrapper:
+    """This is a wrapper around a function that validates the arguments passed to it, and optionally the return value."""
+
+    __slots__ = (
+        '__pydantic_validator__',
+        '__name__',
+        '__qualname__',
+        '__annotations__',
+        '__dict__',  # required for __module__
+    )
+
+    def __init__(self, function: Callable[..., Any], config: ConfigDict | None, validate_return: bool):
+        if isinstance(function, partial):
+            func = function.func
+            schema_type = func
+            self.__name__ = f'partial({func.__name__})'
+            self.__qualname__ = f'partial({func.__qualname__})'
+            self.__module__ = func.__module__
+        else:
+            schema_type = function
+            self.__name__ = function.__name__
+            self.__qualname__ = function.__qualname__
+            self.__module__ = function.__module__
+
+        namespace = _typing_extra.add_module_globals(function, None)
+        config_wrapper = ConfigWrapper(config)
+        gen_schema = _generate_schema.GenerateSchema(config_wrapper, namespace)
+        schema = gen_schema.clean_schema(gen_schema.generate_schema(function))
+        core_config = config_wrapper.core_config(self)
+
+        self.__pydantic_validator__ = create_schema_validator(
+            schema,
+            schema_type,
+            self.__module__,
+            self.__qualname__,
+            'validate_call',
+            core_config,
+            config_wrapper.plugin_settings,
+        )
+
+        if validate_return:
+            signature = inspect.signature(function)
+            return_type = signature.return_annotation if signature.return_annotation is not signature.empty else Any
+            gen_schema = _generate_schema.GenerateSchema(config_wrapper, namespace)
+            schema = gen_schema.clean_schema(gen_schema.generate_schema(return_type))
+            validator = create_schema_validator(
+                schema,
+                schema_type,
+                self.__module__,
+                self.__qualname__,
+                'validate_call',
+                core_config,
+                config_wrapper.plugin_settings,
+            )
+            if inspect.iscoroutinefunction(function):
+
+                async def return_val_wrapper(aw: Awaitable[Any]) -> None:
+                    return validator.validate_python(await aw)
+
+                self.__return_pydantic_validator__ = return_val_wrapper
+            else:
+                self.__return_pydantic_validator__ = validator.validate_python
+        else:
+            self.__return_pydantic_validator__ = None
+
+    def __call__(self, *args: Any, **kwargs: Any) -> Any:
+        res = self.__pydantic_validator__.validate_python(pydantic_core.ArgsKwargs(args, kwargs))
+        if self.__return_pydantic_validator__:
+            return self.__return_pydantic_validator__(res)
+        return res

lib/pydantic/_internal/_validators.py

@@ -0,0 +1,278 @@
+"""Validator functions for standard library types.
+
+Import of this module is deferred since it contains imports of many standard library modules.
+"""
+
+from __future__ import annotations as _annotations
+
+import math
+import re
+import typing
+from ipaddress import IPv4Address, IPv4Interface, IPv4Network, IPv6Address, IPv6Interface, IPv6Network
+from typing import Any
+
+from pydantic_core import PydanticCustomError, core_schema
+from pydantic_core._pydantic_core import PydanticKnownError
+
+
+def sequence_validator(
+    __input_value: typing.Sequence[Any],
+    validator: core_schema.ValidatorFunctionWrapHandler,
+) -> typing.Sequence[Any]:
+    """Validator for `Sequence` types, isinstance(v, Sequence) has already been called."""
+    value_type = type(__input_value)
+
+    # We don't accept any plain string as a sequence
+    # Relevant issue: https://github.com/pydantic/pydantic/issues/5595
+    if issubclass(value_type, (str, bytes)):
+        raise PydanticCustomError(
+            'sequence_str',
+            "'{type_name}' instances are not allowed as a Sequence value",
+            {'type_name': value_type.__name__},
+        )
+
+    v_list = validator(__input_value)
+
+    # the rest of the logic is just re-creating the original type from `v_list`
+    if value_type == list:
+        return v_list
+    elif issubclass(value_type, range):
+        # return the list as we probably can't re-create the range
+        return v_list
+    else:
+        # best guess at how to re-create the original type, more custom construction logic might be required
+        return value_type(v_list)  # type: ignore[call-arg]
+
+
+def import_string(value: Any) -> Any:
+    if isinstance(value, str):
+        try:
+            return _import_string_logic(value)
+        except ImportError as e:
+            raise PydanticCustomError('import_error', 'Invalid python path: {error}', {'error': str(e)}) from e
+    else:
+        # otherwise we just return the value and let the next validator do the rest of the work
+        return value
+
+
+def _import_string_logic(dotted_path: str) -> Any:
+    """Inspired by uvicorn — dotted paths should include a colon before the final item if that item is not a module.
+    (This is necessary to distinguish between a submodule and an attribute when there is a conflict.).
+
+    If the dotted path does not include a colon and the final item is not a valid module, importing as an attribute
+    rather than a submodule will be attempted automatically.
+
+    So, for example, the following values of `dotted_path` result in the following returned values:
+    * 'collections': <module 'collections'>
+    * 'collections.abc': <module 'collections.abc'>
+    * 'collections.abc:Mapping': <class 'collections.abc.Mapping'>
+    * `collections.abc.Mapping`: <class 'collections.abc.Mapping'> (though this is a bit slower than the previous line)
+
+    An error will be raised under any of the following scenarios:
+    * `dotted_path` contains more than one colon (e.g., 'collections:abc:Mapping')
+    * the substring of `dotted_path` before the colon is not a valid module in the environment (e.g., '123:Mapping')
+    * the substring of `dotted_path` after the colon is not an attribute of the module (e.g., 'collections:abc123')
+    """
+    from importlib import import_module
+
+    components = dotted_path.strip().split(':')
+    if len(components) > 2:
+        raise ImportError(f"Import strings should have at most one ':'; received {dotted_path!r}")
+
+    module_path = components[0]
+    if not module_path:
+        raise ImportError(f'Import strings should have a nonempty module name; received {dotted_path!r}')
+
+    try:
+        module = import_module(module_path)
+    except ModuleNotFoundError as e:
+        if '.' in module_path:
+            # Check if it would be valid if the final item was separated from its module with a `:`
+            maybe_module_path, maybe_attribute = dotted_path.strip().rsplit('.', 1)
+            try:
+                return _import_string_logic(f'{maybe_module_path}:{maybe_attribute}')
+            except ImportError:
+                pass
+            raise ImportError(f'No module named {module_path!r}') from e
+        raise e
+
+    if len(components) > 1:
+        attribute = components[1]
+        try:
+            return getattr(module, attribute)
+        except AttributeError as e:
+            raise ImportError(f'cannot import name {attribute!r} from {module_path!r}') from e
+    else:
+        return module
+
+
+def pattern_either_validator(__input_value: Any) -> typing.Pattern[Any]:
+    if isinstance(__input_value, typing.Pattern):
+        return __input_value
+    elif isinstance(__input_value, (str, bytes)):
+        # todo strict mode
+        return compile_pattern(__input_value)  # type: ignore
+    else:
+        raise PydanticCustomError('pattern_type', 'Input should be a valid pattern')
+
+
+def pattern_str_validator(__input_value: Any) -> typing.Pattern[str]:
+    if isinstance(__input_value, typing.Pattern):
+        if isinstance(__input_value.pattern, str):
+            return __input_value
+        else:
+            raise PydanticCustomError('pattern_str_type', 'Input should be a string pattern')
+    elif isinstance(__input_value, str):
+        return compile_pattern(__input_value)
+    elif isinstance(__input_value, bytes):
+        raise PydanticCustomError('pattern_str_type', 'Input should be a string pattern')
+    else:
+        raise PydanticCustomError('pattern_type', 'Input should be a valid pattern')
+
+
+def pattern_bytes_validator(__input_value: Any) -> typing.Pattern[bytes]:
+    if isinstance(__input_value, typing.Pattern):
+        if isinstance(__input_value.pattern, bytes):
+            return __input_value
+        else:
+            raise PydanticCustomError('pattern_bytes_type', 'Input should be a bytes pattern')
+    elif isinstance(__input_value, bytes):
+        return compile_pattern(__input_value)
+    elif isinstance(__input_value, str):
+        raise PydanticCustomError('pattern_bytes_type', 'Input should be a bytes pattern')
+    else:
+        raise PydanticCustomError('pattern_type', 'Input should be a valid pattern')
+
+
+PatternType = typing.TypeVar('PatternType', str, bytes)
+
+
+def compile_pattern(pattern: PatternType) -> typing.Pattern[PatternType]:
+    try:
+        return re.compile(pattern)
+    except re.error:
+        raise PydanticCustomError('pattern_regex', 'Input should be a valid regular expression')
+
+
+def ip_v4_address_validator(__input_value: Any) -> IPv4Address:
+    if isinstance(__input_value, IPv4Address):
+        return __input_value
+
+    try:
+        return IPv4Address(__input_value)
+    except ValueError:
+        raise PydanticCustomError('ip_v4_address', 'Input is not a valid IPv4 address')
+
+
+def ip_v6_address_validator(__input_value: Any) -> IPv6Address:
+    if isinstance(__input_value, IPv6Address):
+        return __input_value
+
+    try:
+        return IPv6Address(__input_value)
+    except ValueError:
+        raise PydanticCustomError('ip_v6_address', 'Input is not a valid IPv6 address')
+
+
+def ip_v4_network_validator(__input_value: Any) -> IPv4Network:
+    """Assume IPv4Network initialised with a default `strict` argument.
+
+    See more:
+    https://docs.python.org/library/ipaddress.html#ipaddress.IPv4Network
+    """
+    if isinstance(__input_value, IPv4Network):
+        return __input_value
+
+    try:
+        return IPv4Network(__input_value)
+    except ValueError:
+        raise PydanticCustomError('ip_v4_network', 'Input is not a valid IPv4 network')
+
+
+def ip_v6_network_validator(__input_value: Any) -> IPv6Network:
+    """Assume IPv6Network initialised with a default `strict` argument.
+
+    See more:
+    https://docs.python.org/library/ipaddress.html#ipaddress.IPv6Network
+    """
+    if isinstance(__input_value, IPv6Network):
+        return __input_value
+
+    try:
+        return IPv6Network(__input_value)
+    except ValueError:
+        raise PydanticCustomError('ip_v6_network', 'Input is not a valid IPv6 network')
+
+
+def ip_v4_interface_validator(__input_value: Any) -> IPv4Interface:
+    if isinstance(__input_value, IPv4Interface):
+        return __input_value
+
+    try:
+        return IPv4Interface(__input_value)
+    except ValueError:
+        raise PydanticCustomError('ip_v4_interface', 'Input is not a valid IPv4 interface')
+
+
+def ip_v6_interface_validator(__input_value: Any) -> IPv6Interface:
+    if isinstance(__input_value, IPv6Interface):
+        return __input_value
+
+    try:
+        return IPv6Interface(__input_value)
+    except ValueError:
+        raise PydanticCustomError('ip_v6_interface', 'Input is not a valid IPv6 interface')
+
+
+def greater_than_validator(x: Any, gt: Any) -> Any:
+    if not (x > gt):
+        raise PydanticKnownError('greater_than', {'gt': gt})
+    return x
+
+
+def greater_than_or_equal_validator(x: Any, ge: Any) -> Any:
+    if not (x >= ge):
+        raise PydanticKnownError('greater_than_equal', {'ge': ge})
+    return x
+
+
+def less_than_validator(x: Any, lt: Any) -> Any:
+    if not (x < lt):
+        raise PydanticKnownError('less_than', {'lt': lt})
+    return x
+
+
+def less_than_or_equal_validator(x: Any, le: Any) -> Any:
+    if not (x <= le):
+        raise PydanticKnownError('less_than_equal', {'le': le})
+    return x
+
+
+def multiple_of_validator(x: Any, multiple_of: Any) -> Any:
+    if not (x % multiple_of == 0):
+        raise PydanticKnownError('multiple_of', {'multiple_of': multiple_of})
+    return x
+
+
+def min_length_validator(x: Any, min_length: Any) -> Any:
+    if not (len(x) >= min_length):
+        raise PydanticKnownError(
+            'too_short',
+            {'field_type': 'Value', 'min_length': min_length, 'actual_length': len(x)},
+        )
+    return x
+
+
+def max_length_validator(x: Any, max_length: Any) -> Any:
+    if len(x) > max_length:
+        raise PydanticKnownError(
+            'too_long',
+            {'field_type': 'Value', 'max_length': max_length, 'actual_length': len(x)},
+        )
+    return x
+
+
+def forbid_inf_nan_check(x: Any) -> Any:
+    if not math.isfinite(x):
+        raise PydanticKnownError('finite_number')
+    return x