utilcachecall.py

#!/usr/bin/env python3
# --------------------( LICENSE                            )--------------------
# Copyright (c) 2014-2024 Beartype authors.
# See "LICENSE" for further details.

'''
Project-wide **callable caching utilities** (i.e., low-level callables
performing general-purpose memoization of function and method calls).

This private submodule is *not* intended for importation by downstream callers.
'''

# ....................{ TODO                               }....................
#FIXME: Generalize @callable_cached to revert to the body of the
#@beartype._util.type.decorator.decmemo.func_cached decorator when the passed
#callable accepts *NO* parameters, which can be trivially decided by inspecting
#the code object of this callable. Why do this? Because the @func_cached
#decorator is *INSANELY* fast for this edge case -- substantially faster than
#the current general-purpose @callable_cached approach.

# ....................{ IMPORTS                            }....................
from beartype.roar._roarexc import _BeartypeUtilCallableCachedException
from beartype.typing import Dict
from beartype._data.hint.datahinttyping import CallableT
from beartype._util.func.arg.utilfuncargtest import (
    die_unless_func_args_len_flexible_equal,
    is_func_arg_variadic,
)
from beartype._util.text.utiltextlabel import label_callable
from beartype._util.utilobject import SENTINEL
from functools import wraps

# ....................{ DECORATORS ~ callable              }....................
def callable_cached(func: CallableT) -> CallableT:
    '''
    **Memoize** (i.e., efficiently re-raise all exceptions previously raised by
    the decorated callable when passed the same parameters (i.e., parameters
    that evaluate as equals) as a prior call to that callable if any *or* return
    all values previously returned by that callable otherwise rather than
    inefficiently recalling that callable) the passed callable.

    Specifically, this decorator (in order):

    #. Creates:

       * A local dictionary mapping parameters passed to this callable with the
         values returned by this callable when passed those parameters.
       * A local dictionary mapping parameters passed to this callable with the
         exceptions raised by this callable when passed those parameters.

    #. Creates and returns a closure transparently wrapping this callable with
       memoization. Specifically, this wrapper (in order):

       #. Tests whether this callable has already been called at least once
          with the passed parameters by lookup of those parameters in these
          dictionaries.
       #. If this callable previously raised an exception when passed these
          parameters, this wrapper re-raises the same exception.
       #. Else if this callable returned a value when passed these parameters,
          this wrapper re-returns the same value.
       #. Else, this wrapper:

          #. Calls that callable with those parameters.
          #. If that call raised an exception:

             #. Caches that exception with those parameters in that dictionary.
             #. Raises that exception.

          #. Else:

             #. Caches the value returned by that call with those parameters in
                that dictionary.
             #. Returns that value.

    Caveats
    -------
    **The decorated callable must accept no keyword parameters.** While this
    decorator previously memoized keyword parameters, doing so incurred
    significant performance penalties defeating the purpose of caching. This
    decorator now intentionally memoizes *only* positional parameters.

    **The decorated callable must accept no variadic positional parameters.**
    While memoizing variadic parameters would of course be feasible, this
    decorator has yet to implement support for doing so.

    **The decorated callable should not be a property method** (i.e., either a
    property getter, setter, or deleter subsequently decorated by the
    :class:`property` decorator). Technically, this decorator *can* be used to
    memoize property methods; pragmatically, doing so would be sufficiently
    inefficient as to defeat the intention of memoizing in the first place.

    Efficiency
    ----------
    For efficiency, consider calling the decorated callable with only:

    * **Hashable** (i.e., immutable) arguments. While technically supported,
      every call to the decorated callable passed one or more unhashable
      arguments (e.g., mutable containers like lists and dictionaries) will
      silently *not* be memoized. Equivalently, only calls passed only hashable
      arguments will be memoized. This flexibility enables decorated callables
      to accept unhashable PEP-compliant type hints. Although *all*
      PEP-noncompliant and *most* PEP-compliant type hints are hashable, some
      sadly are not. These include:

      * :pep:`585`-compliant type hints subscripted by one or more unhashable
        objects (e.g., ``collections.abc.Callable[[], str]``, the `PEP
        585`_-compliant type hint annotating piths accepting callables
        accepting no parameters and returning strings).
      * :pep:`586`-compliant type hints subscripted by an unhashable object
        (e.g., ``typing.Literal[[]]``, a literal empty list).
      * :pep:`593`-compliant type hints subscripted by one or more unhashable
        objects (e.g., ``typing.Annotated[typing.Any, []]``, the
        :attr:`typing.Any` singleton annotated by an empty list).

    **This decorator is intentionally not implemented in terms of the stdlib**
    :func:`functools.lru_cache` **decorator,** as that decorator is inefficient
    in the special case of unbounded caching with ``maxsize=None``. Why? Because
    that decorator insists on unconditionally recording irrelevant statistics
    like cache misses and hits. While bounding the number of cached values is
    advisable in the general case (e.g., to avoid exhausting memory merely for
    optional caching), parameters and returns cached by this package are
    sufficiently small in size to render such bounding irrelevant.

    Consider the
    :func:`beartype._util.hint.pep.utilpeptest.is_hint_pep_type_typing`
    function, for example. Each call to that function only accepts a single
    class and returns a boolean. Under conservative assumptions of 4 bytes of
    storage per class reference and 4 byte of storage per boolean reference,
    each call to that function requires caching at most 8 bytes of storage.
    Again, under conservative assumptions of at most 1024 unique type
    annotations for the average downstream consumer, memoizing that function in
    full requires at most 1024 * 8 == 8096 bytes or ~8Kb of storage. Clearly,
    8Kb of overhead is sufficiently negligible to obviate any space concerns
    that would warrant an LRU cache in the first place.

    Parameters
    ----------
    func : CallableT
        Callable to be memoized.

    Returns
    -------
    CallableT
        Closure wrapping this callable with memoization.

    Raises
    ------
    _BeartypeUtilCallableCachedException
        If this callable accepts a variadic positional parameter (e.g.,
        ``*args``).
    '''
    assert callable(func), f'{repr(func)} not callable.'

    #!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
    # CAUTION: Synchronize against the @method_cached_arg_by_id decorator
    # below. For speed, this decorator violates DRY by duplicating logic.
    #!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    # Dictionary mapping a tuple of all flattened parameters passed to each
    # prior call of the decorated callable with the value returned by that call
    # if any (i.e., if that call did *NOT* raise an exception).
    args_flat_to_return_value: Dict[tuple, object] = {}

    # get() method of this dictionary, localized for efficiency.
    args_flat_to_return_value_get = args_flat_to_return_value.get

    # Dictionary mapping a tuple of all flattened parameters passed to each
    # prior call of the decorated callable with the exception raised by that
    # call if any (i.e., if that call raised an exception).
    args_flat_to_exception: Dict[tuple, Exception] = {}

    # get() method of this dictionary, localized for efficiency.
    args_flat_to_exception_get = args_flat_to_exception.get

    @wraps(func)
    def _callable_cached(*args):
        f'''
        Memoized variant of the {func.__name__}() callable.

        See Also
        --------
        :func:`callable_cached`
            Further details.
        '''

        # Object representing all passed positional arguments to be used as the
        # key of various memoized dictionaries, defined as either...
        args_flat = (
            # If passed only one positional argument, minimize space consumption
            # by flattening this tuple of only that argument into that argument.
            # Since tuple items are necessarily hashable, this argument is
            # necessarily hashable and thus permissible as a dictionary key;
            args[0]
            if len(args) == 1 else
            # Else, one or more positional arguments are passed. In this case,
            # reuse this tuple as is.
            args
        )

        # Attempt to...
        try:
            # Exception raised by a prior call to the decorated callable when
            # passed these parameters *OR* the sentinel placeholder otherwise
            # (i.e., if this callable either has yet to be called with these
            # parameters *OR* has but failed to raise an exception).
            #
            # Note that:
            # * This statement raises a "TypeError" exception if any item of
            #   this flattened tuple is unhashable.
            # * A sentinel placeholder (e.g., "SENTINEL") is *NOT* needed here.
            #   The values of the "args_flat_to_exception" dictionary are
            #   guaranteed to *ALL* be exceptions. Since "None" is *NOT* an
            #   exception, disambiguation between "None" and valid dictionary
            #   values is *NOT* needed here. Although a sentinel placeholder
            #   could still be employed, doing so would slightly reduce
            #   efficiency for *NO* real-world gain.
            exception = args_flat_to_exception_get(args_flat)

            # If this callable previously raised an exception when called with
            # these parameters, re-raise the same exception.
            if exception:
                raise exception  # pyright: ignore[reportGeneralTypeIssues]
            # Else, this callable either has yet to be called with these
            # parameters *OR* has but failed to raise an exception.

            # Value returned by a prior call to the decorated callable when
            # passed these parameters *OR* a sentinel placeholder otherwise
            # (i.e., if this callable has yet to be passed these parameters).
            return_value = args_flat_to_return_value_get(args_flat, SENTINEL)

            # If this callable has already been called with these parameters,
            # return the value returned by that prior call.
            if return_value is not SENTINEL:
                return return_value
            # Else, this callable has yet to be called with these parameters.

            # Attempt to...
            try:
                # Call this parameter with these parameters and cache the value
                # returned by this call to these parameters.
                return_value = args_flat_to_return_value[args_flat] = func(
                    *args)
            # If this call raised an exception...
            except Exception as exception:
                # Cache this exception to these parameters.
                args_flat_to_exception[args_flat] = exception

                # Re-raise this exception.
                raise exception
        # If one or more objects either passed to *OR* returned from this call
        # are unhashable, perform this call as is *WITHOUT* memoization. While
        # non-ideal, stability is better than raising a fatal exception.
        except TypeError:
            #FIXME: If testing, emit a non-fatal warning or possibly even raise
            #a fatal exception. In either case, we want our test suite to notify
            #us about this.
            return func(*args)

        # Return this value.
        return return_value

    # Return this wrapper.
    return _callable_cached  # type: ignore[return-value]

# ....................{ DECORATORS ~ method                }....................
def method_cached_arg_by_id(func: CallableT) -> CallableT:
    '''
    **Memoize** (i.e., efficiently re-raise all exceptions previously raised by
    the decorated method when passed the same *exact* parameters (i.e.,
    parameters whose object IDs are equals) as a prior call to that method if
    any *or* return all values previously returned by that method otherwise
    rather than inefficiently recalling that method) the passed method.

    Caveats
    -------
    **This decorator is only intended to decorate bound methods** (i.e., either
    class or instance methods bound to a class or instance). This decorator is
    *not* intended to decorate functions or static methods.

    **This decorator is only intended to decorate a method whose sole argument
    is guaranteed to be a memoized singleton** (e.g.,
    :class:`beartype.door.TypeHint` singleton). In this case, the object
    identifier of that argument uniquely identifies that argument across *all*
    calls to that method -- enabling this decorator to memoize that method.
    Conversely, if that argument is *not* guaranteed to be a memoized singleton,
    this decorator will fail to memoize that method while wasting considerable
    space and time attempting to do so. In short, care is warranted.

    This decorator is a micro-optimized variant of the more general-purpose
    :func:`callable_cached` decorator, which should be preferred in most cases.
    This decorator mostly exists for one specific edge case that the
    :func:`callable_cached` decorator *cannot* by definition support:
    user-defined classes implementing the ``__eq__`` dunder method to internally
    call another method decorated by :func:`callable_cached` accepting an
    instance of the same class. This design pattern appears astonishingly
    frequently, including in our prominent :class:`beartype.door.TypeHint`
    class. This edge case provokes infinite recursion. Consider this
    minimal-length example (MLE) exhibiting the issue:

    .. code-block:: python

       from beartype._util.cache.utilcachecall import callable_cached

       class MuhClass(object):
           def __eq__(self, other: object) -> bool:
               return isinstance(other, MuhClass) and self._is_equal(other)

           @callable_cached
           def _is_equal(self, other: 'MuhClass') -> bool:
               return True

    :func:`callable_cached` internally caches the ``other`` argument passed to
    the ``_is_equal()`` method as keys of various internal dictionaries. When
    passed the same ``other`` argument, subsequent calls to that method lookup
    that ``other`` argument in those dictionaries. Since dictionary lookups
    implicitly call the ``other.__eq__()`` method to resolve key collisions
    *and* since the ``__eq__()`` method has been overridden in terms of the
    ``_is_equal()`` method, infinite recursion results.

    This decorator circumvents this issue by internally looking up the object
    identifier of the passed argument rather than that argument itself, which
    then avoids implicitly calling the ``__eq__()`` method of that argument.

    Parameters
    ----------
    func : CallableT
        Callable to be memoized.

    Returns
    -------
    CallableT
        Closure wrapping this callable with memoization.

    Raises
    ------
    _BeartypeUtilCallableCachedException
        If this callable accepts either:

        * *No* parameters.
        * Two or more parameters.
        * A variadic positional parameter (e.g., ``*args``).

    See Also
    --------
    :func:`callable_cached`
        Further details.
    '''
    assert callable(func), f'{repr(func)} not callable.'

    # Avoid circular import dependencies.
    from beartype._util.func.utilfuncwrap import unwrap_func_all

    # Lowest-level wrappee callable wrapped by this wrapper callable.
    func_wrappee = unwrap_func_all(func)

    # If this wrappee accepts either zero, one, *OR* three or more flexible
    # parameters (i.e., parameters passable as either positional or keyword
    # arguments), raise an exception.
    die_unless_func_args_len_flexible_equal(
        func=func_wrappee,
        func_args_len_flexible=2,
        exception_cls=_BeartypeUtilCallableCachedException,
        # Avoid unnecessary callable unwrapping as a negligible optimization.
        is_unwrap=False,
    )
    # Else, this wrappee accepts exactly one flexible parameter.

    # If this wrappee accepts variadic arguments, raise an exception.
    if is_func_arg_variadic(func_wrappee):
        raise _BeartypeUtilCallableCachedException(
            f'@method_cached_arg_by_id {label_callable(func)} '
            f'variadic arguments uncacheable.'
        )
    # Else, this wrappee accepts *NO* variadic arguments.

    #!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
    # CAUTION: Synchronize against the @callable_cached decorator above. For
    # speed, this decorator violates DRY by duplicating logic.
    #!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    # Dictionary mapping a tuple of all flattened parameters passed to each
    # prior call of the decorated callable with the value returned by that call
    # if any (i.e., if that call did *NOT* raise an exception).
    args_flat_to_return_value: Dict[tuple, object] = {}

    # get() method of this dictionary, localized for efficiency.
    args_flat_to_return_value_get = args_flat_to_return_value.get

    # Dictionary mapping a tuple of all flattened parameters passed to each
    # prior call of the decorated callable with the exception raised by that
    # call if any (i.e., if that call raised an exception).
    args_flat_to_exception: Dict[tuple, Exception] = {}

    # get() method of this dictionary, localized for efficiency.
    args_flat_to_exception_get = args_flat_to_exception.get

    @wraps(func)
    def _method_cached(self_or_cls, arg):
        f'''
        Memoized variant of the {func.__name__}() callable.

        See Also
        --------
        :func:`callable_cached`
            Further details.
        '''

        # Object identifiers of the sole positional parameters passed to the
        # decorated method.
        args_flat = (id(self_or_cls), id(arg))

        # Attempt to...
        try:
            # Exception raised by a prior call to the decorated callable when
            # passed these parameters *OR* the sentinel placeholder otherwise
            # (i.e., if this callable either has yet to be called with these
            # parameters *OR* has but failed to raise an exception).
            #
            # Note that:
            # * This statement raises a "TypeError" exception if any item of
            #   this flattened tuple is unhashable.
            # * A sentinel placeholder (e.g., "SENTINEL") is *NOT* needed here.
            #   The values of the "args_flat_to_exception" dictionary are
            #   guaranteed to *ALL* be exceptions. Since "None" is *NOT* an
            #   exception, disambiguation between "None" and valid dictionary
            #   values is *NOT* needed here. Although a sentinel placeholder
            #   could still be employed, doing so would slightly reduce
            #   efficiency for *NO* real-world gain.
            exception = args_flat_to_exception_get(args_flat)

            # If this callable previously raised an exception when called with
            # these parameters, re-raise the same exception.
            if exception:
                raise exception  # pyright: ignore[reportGeneralTypeIssues]
            # Else, this callable either has yet to be called with these
            # parameters *OR* has but failed to raise an exception.

            # Value returned by a prior call to the decorated callable when
            # passed these parameters *OR* a sentinel placeholder otherwise
            # (i.e., if this callable has yet to be passed these parameters).
            return_value = args_flat_to_return_value_get(
                args_flat, SENTINEL)

            # If this callable has already been called with these parameters,
            # return the value returned by that prior call.
            if return_value is not SENTINEL:
                return return_value
            # Else, this callable has yet to be called with these parameters.

            # Attempt to...
            try:
                # Call this parameter with these parameters and cache the value
                # returned by this call to these parameters.
                return_value = args_flat_to_return_value[args_flat] = func(
                    self_or_cls, arg)
            # If this call raised an exception...
            except Exception as exception:
                # Cache this exception to these parameters.
                args_flat_to_exception[args_flat] = exception

                # Re-raise this exception.
                raise exception
        # If one or more objects either passed to *OR* returned from this call
        # are unhashable, perform this call as is *WITHOUT* memoization. While
        # non-ideal, stability is better than raising a fatal exception.
        except TypeError:
            #FIXME: If testing, emit a non-fatal warning or possibly even raise
            #a fatal exception. In either case, we want our test suite to notify
            #us about this.
            return func(self_or_cls, arg)

        # Return this value.
        return return_value

    # Return this wrapper.
    return _method_cached  # type: ignore[return-value]

# ....................{ DECORATORS ~ property              }....................
def property_cached(func: CallableT) -> CallableT:
    '''
    **Memoize** (i.e., efficiently cache and return all previously returned
    values of the passed property method as well as all previously raised
    exceptions of that method previously rather than inefficiently recalling
    that method) the passed **property method method** (i.e., either a property
    getter, setter, or deleter subsequently decorated by the :class:`property`
    decorator).

    On the first access of a property decorated with this decorator (in order):

    #. The passed method implementing this property is called.
    #. The value returned by this property is internally cached into a private
       attribute of the object to which this method is bound.
    #. This value is returned.

    On each subsequent access of this property, this cached value is returned as
    is *without* calling the decorated method. Hence, the decorated method is
    called at most once for each object exposing this property.

    Caveats
    -------
    **This decorator must be preceded by an explicit usage of the standard**
    :class:`property` **decorator.** Although this decorator could be trivially
    refactored to automatically decorate the returned property method by the
    :class:`property` decorator, doing so would violate static type-checking
    expectations -- introducing far more issues than it would solve.

    **This decorator should always be preferred over the standard**
    :func:`functools.cached_property` **decorator available under Python >=
    3.8.** This decorator is substantially more efficient in both space and time
    than that decorator -- which is, of course, the entire point of caching.

    **This decorator does not destroy bound property methods.** Technically, the
    most efficient means of caching a property value into an instance is to
    replace the property method currently bound to that instance with an
    instance variable initialized to that value (e.g., as documented by this
    `StackOverflow answer`_). Since a property should only ever be treated as an
    instance variable, there superficially exists little harm in dynamically
    changing the type of the former to the latter. Sadly, doing so introduces
    numerous subtle issues with *no* plausible workaround. Notably, replacing
    property methods by instance variables:

    * Permits callers to erroneously set **read-only properties** (i.e.,
      properties lacking setter methods), a profound violation of one of the
      principle use cases for properties.
    * Prevents pickling logic elsewhere from automatically excluding cached
      property values, forcing these values to *always* be pickled to disk.
      This is bad. Cached property values are *always* safely recreatable in
      memory (and hence need *not* be pickled) and typically space-consumptive
      in memory (and hence best *not* pickled). The slight efficiency gain from
      replacing property methods by instance variables is hardly worth the
      significant space loss from pickling these variables.

    .. _StackOverflow answer:
        https://stackoverflow.com/a/36684652/2809027

    Parameters
    ----------
    func : CallableT
        Property method to be memoized.

    Returns
    -------
    CallableT
        Dynamically generated function wrapping this property with memoization.
    '''
    assert callable(func), f'{repr(func)} not callable.'

    # Name of the private instance variable to which this decorator caches the
    # value returned by the decorated property method.
    property_var_name = (
        _PROPERTY_CACHED_VAR_NAME_PREFIX + func.__name__)

    # Raw string of Python statements comprising the body of this wrapper.
    #
    # Note that this implementation intentionally avoids calling our
    # higher-level beartype._util.func.utilfuncmake.make_func() factory function
    # for dynamically generating functions. Although this implementation could
    # certainly be refactored in terms of that factory, doing so would
    # needlessly reduce debuggability and portability for *NO* tangible gain.
    func_body = _PROPERTY_CACHED_CODE.format(
        property_var_name=property_var_name)

    # Dictionary mapping from local attribute names to values. For efficiency,
    # only attributes required by the body of this wrapper are copied from the
    # current namespace. (See below.)
    local_attrs = {'__property_method': func}

    # Dynamically define this wrapper as a closure of this decorator. For
    # obscure and presumably uninteresting reasons, Python fails to locally
    # declare this closure when the locals() dictionary is passed; to capture
    # this closure, a local dictionary must be passed instead.
    exec(func_body, globals(), local_attrs)

    # Return this wrapper method.
    return local_attrs['property_method_cached']

# ....................{ PRIVATE ~ constants : var          }....................
_CALLABLE_CACHED_VAR_NAME_PREFIX = '__beartype_cached__'
'''
Substring prefixing the names of all private instance variables to which all
caching decorators (e.g., :func:`property_cached`) cache values returned by
decorated callables.

This prefix:

* Guarantees uniqueness across *all* instances -- including those instantiated
  from official Python and unofficial third-party classes and those internally
  defined by this application. Doing so permits logic elsewhere (e.g., pickling
  filtering) to uniquely match and act upon these variables.
* Is intentionally prefixed by double rather than single underscores (i.e.,
  ``"__"`` rather than ``"_"``), ensuring that our
  :meth:`beartype._check.forward.reference.fwdrefmeta.BeartypeForwardRefMeta.__getattr__`
  dunder method ignores the private instance variables cached by our cached
  :meth:`beartype._check.forward.reference.fwdrefmeta.BeartypeForwardRefMeta.__type_beartype__`
  property.
'''


_FUNCTION_CACHED_VAR_NAME = (
    f'{_CALLABLE_CACHED_VAR_NAME_PREFIX}function_value')
'''
Name of the private instance variable to which the :func:`func_cached`
decorator statically caches the value returned by the decorated function.
'''


_PROPERTY_CACHED_VAR_NAME_PREFIX = (
    f'{_CALLABLE_CACHED_VAR_NAME_PREFIX}property_')
'''
Substring prefixing the names of all private instance variables to which the
:func:`property_cached` decorator dynamically caches the value returned by the
decorated property method.
'''

# ....................{ PRIVATE ~ constants : code         }....................
_PROPERTY_CACHED_CODE = '''
@wraps(__property_method)
def property_method_cached(self, __property_method=__property_method):
    try:
        return self.{property_var_name}
    except AttributeError:
        self.{property_var_name} = __property_method(self)
        return self.{property_var_name}
'''
'''
Raw string of Python statements comprising the body of the wrapper function
dynamically generated by the :func:`property_cached` decorator.

These statements include (in order):

* A :mod:`functools.wraps` decoration propagating the name, docstring, and other
  identifying metadata of the original function to this wrapper.
* A private ``__property_method`` parameter set to the underlying property
  getter method. In theory, the ``func`` parameter passed to the
  :func:`property_cached` decorator should be accessible as a closure-style
  local in this code. For unknown reasons (presumably, a subtle bug in the
  :func:`exec` builtin), this is not the case. Instead, a closure-style local
  must be simulated by passing the ``func`` parameter to this function at
  function definition time as the default value of an arbitrary parameter.

Design
------
While there exist numerous alternative implementations for caching properties,
the approach implemented below has been profiled to be the most efficient.
Alternatives include (in order of decreasing efficiency):

* Dynamically getting and setting a property-specific key-value pair of the
  internal dictionary for the current object, timed to be approximately 1.5
  times as slow as exception handling: e.g.,

.. code-block:: python
   if not {property_name!r} in self.__dict__:
       self.__dict__[{property_name!r}] = __property_method(self)
   return self.__dict__[{property_name!r}]

* Dynamically getting and setting a property-specific attribute of the current
  object (e.g., the internal dictionary for the current object), timed to be
  approximately 1.5 times as slow as exception handling: e.g.,

.. code-block:: python
   if not hasattr(self, {property_name!r}):
       setattr(self, {property_name!r}, __property_method(self))
   return getattr(self, {property_name!r})
'''


#FIXME: Uncomment to debug memoization-specific issues. *sigh*
# def callable_cached(func: _CallableT) -> _CallableT: return func
# def property_cached(func: _CallableT) -> _CallableT: return func