1
Fork 0
mirror of https://github.com/RYGhub/royalnet.git synced 2024-11-22 19:14:20 +00:00

Remove the refactored royalnet.engineer.teleporter module

This commit is contained in:
Steffo 2022-05-02 18:44:44 +02:00
parent 8fed7ba510
commit d40351a6f7
Signed by: steffo
GPG key ID: 6965406171929D01

View file

@ -1,276 +0,0 @@
"""
This module contains the :class:`.Teleporter` class and its exceptions.
"""
from __future__ import annotations
import inspect
import logging
import pydantic
import royalnet.royaltyping as t
from . import exc
Value = t.TypeVar("Value")
log = logging.getLogger(__name__)
class TeleporterError(exc.EngineerException, pydantic.ValidationError):
"""
The base class for errors in :mod:`royalnet.engineer.teleporter`.
"""
class InTeleporterError(TeleporterError):
"""
The input parameters validation failed.
"""
class OutTeleporterError(TeleporterError):
"""
The return value validation failed.
"""
class Teleporter:
"""
A :class:`.Teleporter` is a function wrapper which uses :mod:`pydantic` to perform type checking
"""
def __init__(self,
f: t.Callable[..., t.Any],
validate_input: bool = True,
validate_output: bool = True):
self.f: t.Callable = f
"""
The function which is having its parameters and return value validated.
"""
self.InputModel: t.Type[pydantic.BaseModel] = self._create_input_model() if validate_input else None
"""
The :mod:`pydantic` model used to validate input parameters.
"""
self.OutputModel: t.Type[pydantic.BaseModel] = self._create_output_model() if validate_output else None
"""
The :mod:`pydantic` model used to validate the return value.
"""
def __repr__(self):
if self.InputModel and self.OutputModel:
validation = "validating input and output"
elif self.InputModel:
validation = "validating only input"
elif self.OutputModel:
validation = "validating only output"
else:
validation = "not validating anything"
return f"<{self.__class__.__qualname__} {validation}>"
@staticmethod
def _parameter_to_field(param: inspect.Parameter, **kwargs) -> t.Tuple[type, pydantic.fields.FieldInfo]:
"""
Convert a :class:`inspect.Parameter` to a type-field :class:`tuple`, which can be easily passed to
:func:`pydantic.create_model`.
If the parameter is already a :class:`pydantic.FieldInfo` (created by :func:`pydantic.Field`), it will be
returned as the value, without creating a new model.
:param param: The :class:`inspect.Parameter` to convert.
:param kwargs: Additional kwargs to pass to the field.
:return: A :class:`tuple`, where the first element is a :class:`type` and the second is a
:class:`pydantic.Field`.
"""
if isinstance(param.default, pydantic.fields.FieldInfo):
log.debug(f"Parameter {param} is a pydantic.Field, leaving it untouched...")
return (
param.annotation,
param.default
)
else:
log.debug(f"Parameter {param} is not a pydantic.Field, converting it to one...")
return (
param.annotation,
pydantic.Field(
default=param.default if param.default is not inspect.Parameter.empty else ...,
title=param.name,
**kwargs,
),
)
class TeleporterDefaultConfig(pydantic.BaseConfig):
"""
A :mod:`pydantic` model config which allows for arbitrary types.
"""
arbitrary_types_allowed = True
def get_model_config(self):
"""
Get the :mod:`pydantic` config to use in both input and output, if :meth:`.get_input_model_config` and
:meth:`.get_output_model_config` are not overridden.
:return: A :mod:`pydantic` config.
"""
log.debug("Getting default model config...")
return self.TeleporterDefaultConfig
def get_input_model_config(self):
"""
Get the :mod:`pydantic` config to use while creating input models.
:return: A :mod:`pydantic` config.
"""
log.debug("Getting common model config...")
return self.get_model_config()
def get_output_model_config(self):
"""
Get the :mod:`pydantic` config to use while creating output models.
:return: A :mod:`pydantic` config.
"""
log.debug("Getting common model config...")
return self.get_model_config()
def _create_input_model(self,
**extra_fields) -> t.Type[pydantic.BaseModel]:
"""
Create a pydantic model based on the arguments of the :attr:`f` function.
Arguments starting with ``_`` are ignored.
The model is created using the config obtained through :meth:`.get_input_model_config` .
:param extra_fields: Extra fields to be added to the model.
:return: The created pydantic model.
"""
log.debug(f"Getting function signature of: {self.f!r}")
signature: inspect.Signature = inspect.signature(self.f)
log.debug(f"Converting parameter annotations of {self.f!r} to fields...")
fields = {
key: self._parameter_to_field(value)
for key, value in signature.parameters.items()
if not key.startswith("_")
}
log.debug(f"Creating input model with parsed fields {fields!r} and extra fields {extra_fields!r}...")
return pydantic.create_model(
f"{self.__class__.__name__}InputModel",
__config__=self.get_input_model_config(),
**fields,
**extra_fields
)
def _create_output_model(self) -> t.Type[pydantic.BaseModel]:
"""
Create a pydantic model based on the return value of the :attr:`f` function.
The model is created using the config obtained through :meth:`.get_output_model_config` .
:return: The created pydantic model.
"""
log.debug(f"Getting function signature of: {self.f!r}")
signature: inspect.Signature = inspect.signature(self.f)
log.debug(f"Creating output model...")
return pydantic.create_model(
f"{self.__class__.__name__}OutputModel",
__config__=self.get_output_model_config(),
__root__=(signature.return_annotation, pydantic.Field(..., title="Returns"))
)
def teleport_in(self, **kwargs) -> pydantic.BaseModel:
"""
Instantiate the :attr:`.InputModel` with the passed kwargs.
:param kwargs: The keyword arguments that should be passed to the model.
:return: The created model.
:raises .InTeleporterError: If the kwargs fail the validation.
"""
log.debug(f"Teleporting in: {kwargs!r}")
try:
return self.InputModel(**kwargs)
except pydantic.ValidationError as e:
log.error(f"Teleport in failed: {e!r}")
raise InTeleporterError(errors=e.raw_errors, model=e.model)
def teleport_out(self, value: Value) -> pydantic.BaseModel:
"""
Instantiate the :attr:`.OutputModel` with the passed value.
:param value: The value that should be validated.
:return: The created model.
:raises .OutTeleporterError: If the value fails the validation.
"""
log.debug(f"Teleporting out: {value!r}")
try:
return self.OutputModel(__root__=value)
except pydantic.ValidationError as e:
log.error(f"Teleport out failed: {e!r}")
raise OutTeleporterError(errors=e.raw_errors, model=e.model)
@staticmethod
def _split_kwargs(**kwargs) -> t.Tuple[t.Dict[str, t.Any], t.Dict[str, t.Any]]:
"""
Split the passed kwargs in two different :class:`dict`:
- One containing the arguments that **do not start with ``_``**;
- Another containing the ones which do.
:return: A tuple of :class:`dict`, where the second contains the ones starting with ``_``, and the first
contains the rest.
"""
model_params = {}
extra_params = {}
for key, value in kwargs.items():
if key.startswith("_"):
log.debug(f"Found extra keyword argument: {key}")
extra_params[key] = value
else:
log.debug(f"Found model keyword argument: {key}")
model_params[key] = value
return model_params, extra_params
def _run(self, **kwargs) -> t.Any:
"""
Run the :class:`.Teleporter` synchronously.
"""
if self.InputModel:
log.debug("Validating input...")
model_kwargs, extra_kwargs = self._split_kwargs(**kwargs)
model_kwargs = self.teleport_in(**kwargs).dict()
kwargs = {**model_kwargs, **extra_kwargs}
result = self.f(**kwargs)
if self.OutputModel:
result = self.teleport_out(result).__root__
return result
async def _run_async(self, **kwargs) -> t.Awaitable[t.Any]:
"""
Run the :class:`.Teleporter` asynchronously.
"""
if self.InputModel:
log.debug("Validating input...")
model_kwargs, extra_kwargs = self._split_kwargs(**kwargs)
model_kwargs = self.teleport_in(**kwargs).dict()
kwargs = {**model_kwargs, **extra_kwargs}
result = await self.f(**kwargs)
if self.OutputModel:
result = self.teleport_out(result).__root__
return result
def __call__(self, **kwargs) -> t.Any:
if inspect.iscoroutinefunction(self.f):
return self._run_async(**kwargs)
else:
return self._run(**kwargs)
__all__ = (
"InTeleporterError",
"OutTeleporterError",
"Teleporter",
"TeleporterError",
)