import numpy

from chainer import _backend
# TODO(kmaehashi): `from chainer.backends import cuda` causes circular imports.
# Surprisingly, `import chianer.backends` works as a workaround to avoid, but
# we should fix circular dependencies themselves around `chainer.backends.*`.
import chainer.backends
import chainerx


class CpuDevice(_backend.Device):

    """Device for CPU (NumPy) backend"""

    name = '@numpy'
    xp = numpy
    supported_array_types = (numpy.ndarray,)

    __hash__ = _backend.Device.__hash__

    @staticmethod
    def from_array(array):
        if isinstance(array, numpy.ndarray):
            return CpuDevice()
        return None

    def __eq__(self, other):
        return isinstance(other, CpuDevice)

    def __repr__(self):
        return '<{} (numpy)>'.format(self.__class__.__name__)

    def send_array(self, array):
        return _array_to_cpu(array)

    def is_array_supported(self, array):
        return isinstance(array, numpy.ndarray)


def _to_cpu(array):
    """Converts an array or arrays to NumPy."""
    return _backend._convert_arrays(array, _array_to_cpu)


def _array_to_cpu(array):
    if array is None:
        return None
    if isinstance(array, numpy.ndarray):
        return array
    if isinstance(array, chainer.backends.intel64.mdarray):
        return numpy.asarray(array)
    if isinstance(array, chainerx.ndarray):
        return chainerx.to_numpy(array, copy=False)
    if isinstance(array, chainer.backends.cuda.ndarray):
        with chainer.backends.cuda.get_device_from_array(array):
            return array.get()
    if numpy.isscalar(array):
        return numpy.asarray(array)
    raise TypeError(
        'Array cannot be converted into an numpy.ndarray'
        '\nActual type: {0}.'.format(type(array)))