vllm/lib/python3.10/site-packages/cupy/testing/_loops.py

import functools
import inspect
import os
import random
from typing import Tuple, Type
import traceback
import unittest
import warnings

import numpy

import cupy
from cupy.exceptions import AxisError
from cupy.testing import _array
from cupy.testing import _parameterized
import cupyx
import cupyx.scipy.sparse

from cupy.testing._pytest_impl import is_available


if is_available():
    import _pytest.outcomes
    _is_pytest_available = True
    _skip_classes: Tuple[Type, ...] = (
        unittest.SkipTest, _pytest.outcomes.Skipped)
else:
    _is_pytest_available = False
    _skip_classes = unittest.SkipTest,


def _format_exception(exc):
    if exc is None:
        return None
    # TODO(kataoka): Use traceback.format_exception(exc) in Python 3.10
    return ''.join(traceback.TracebackException.from_exception(exc).format())


def _call_func(impl, args, kw):
    # Note that `_pytest.outcomes.Skipped` is derived from BaseException.
    exceptions = Exception,
    if _is_pytest_available:
        exceptions += _pytest.outcomes.Skipped,

    try:
        result = impl(*args, **kw)
        error = None
    except exceptions as e:
        tb = e.__traceback__
        if tb.tb_next is None:
            # failed before impl is called, e.g. invalid kw
            raise e
        result = None
        error = e

    return result, error


def _call_func_cupy(impl, args, kw, name, sp_name, scipy_name):
    assert isinstance(name, str)
    assert sp_name is None or isinstance(sp_name, str)
    assert scipy_name is None or isinstance(scipy_name, str)
    kw = kw.copy()

    # Run cupy
    if sp_name:
        kw[sp_name] = cupyx.scipy.sparse
    if scipy_name:
        kw[scipy_name] = cupyx.scipy
    kw[name] = cupy
    result, error = _call_func(impl, args, kw)
    return result, error


def _call_func_numpy(impl, args, kw, name, sp_name, scipy_name):
    assert isinstance(name, str)
    assert sp_name is None or isinstance(sp_name, str)
    assert scipy_name is None or isinstance(scipy_name, str)
    kw = kw.copy()

    # Run numpy
    kw[name] = numpy
    if sp_name:
        import scipy.sparse
        kw[sp_name] = scipy.sparse
    if scipy_name:
        import scipy
        kw[scipy_name] = scipy
    result, error = _call_func(impl, args, kw)
    return result, error


def _call_func_numpy_cupy(impl, args, kw, name, sp_name, scipy_name):
    # Run cupy
    cupy_result, cupy_error = _call_func_cupy(
        impl, args, kw, name, sp_name, scipy_name)

    # Run numpy
    numpy_result, numpy_error = _call_func_numpy(
        impl, args, kw, name, sp_name, scipy_name)

    return (
        cupy_result, cupy_error,
        numpy_result, numpy_error)


_numpy_errors = [
    AttributeError, Exception, IndexError, TypeError, ValueError,
    NotImplementedError, DeprecationWarning,
    AxisError, numpy.linalg.LinAlgError,
]


def _check_numpy_cupy_error_compatible(cupy_error, numpy_error):
    """Checks if try/except blocks are equivalent up to public error classes
    """

    return all(isinstance(cupy_error, err) == isinstance(numpy_error, err)
               for err in _numpy_errors)


def _fail_test_with_unexpected_errors(
        tb, msg_format, cupy_error, numpy_error):
    # Fails the test due to unexpected errors raised from the test.
    # msg_format may include format placeholders:
    # '{cupy_error}' '{numpy_error}'

    msg = msg_format.format(
        cupy_error=_format_exception(cupy_error),
        numpy_error=_format_exception(numpy_error))

    # Fail the test with the traceback of the error (for pytest --pdb)
    raise AssertionError(msg).with_traceback(tb)


def _check_cupy_numpy_error(cupy_error, numpy_error,
                            accept_error=False):
    # Skip the test if both raised SkipTest.
    if (isinstance(cupy_error, _skip_classes)
            and isinstance(numpy_error, _skip_classes)):
        if cupy_error.__class__ is not numpy_error.__class__:
            raise AssertionError(
                'Both numpy and cupy were skipped but with different '
                'exceptions.')
        if cupy_error.args != numpy_error.args:
            raise AssertionError(
                'Both numpy and cupy were skipped but with different causes.')
        raise numpy_error  # reraise SkipTest

    # Check if the error was not raised from test code.
    if os.environ.get('CUPY_CI', '') != '' and cupy_error is not None:
        frame = traceback.extract_tb(cupy_error.__traceback__)[-1]
        filename = os.path.basename(frame.filename)
        if filename == 'test_helper.py':
            # Allows errors from the test code for testing helpers.
            pass
        elif filename.startswith('test_'):
            _fail_test_with_unexpected_errors(
                cupy_error.__traceback__,
                'Error was raised from test code.\n\n{cupy_error}',
                cupy_error, None)

    # For backward compatibility
    if accept_error is True:
        accept_error = Exception
    elif not accept_error:
        accept_error = ()
    # TODO(oktua): expected_regexp like numpy.testing.assert_raises_regex
    if cupy_error is None and numpy_error is None:
        raise AssertionError(
            'Both cupy and numpy are expected to raise errors, but not')
    elif cupy_error is None:
        _fail_test_with_unexpected_errors(
            numpy_error.__traceback__,
            'Only numpy raises error\n\n{numpy_error}',
            None, numpy_error)
    elif numpy_error is None:
        _fail_test_with_unexpected_errors(
            cupy_error.__traceback__,
            'Only cupy raises error\n\n{cupy_error}',
            cupy_error, None)

    elif not _check_numpy_cupy_error_compatible(cupy_error, numpy_error):
        _fail_test_with_unexpected_errors(
            cupy_error.__traceback__,
            '''Different types of errors occurred

cupy
{cupy_error}

numpy
{numpy_error}
''',
            cupy_error, numpy_error)

    elif not (isinstance(cupy_error, accept_error)
              and isinstance(numpy_error, accept_error)):
        _fail_test_with_unexpected_errors(
            cupy_error.__traceback__,
            '''Both cupy and numpy raise exceptions

cupy
{cupy_error}

numpy
{numpy_error}
''',
            cupy_error, numpy_error)


def _signed_counterpart(dtype):
    return numpy.dtype(numpy.dtype(dtype).char.lower()).type


def _make_positive_masks(impl, args, kw, name, sp_name, scipy_name):
    # Returns a mask of output arrays that indicates valid elements for
    # comparison. See the comment at the caller.
    ks = [k for k, v in kw.items() if v in _unsigned_dtypes]
    for k in ks:
        kw[k] = _signed_counterpart(kw[k])
    result, error = _call_func_cupy(
        impl, args, kw, name, sp_name, scipy_name)
    assert error is None
    if not isinstance(result, (tuple, list)):
        result = result,
    return [cupy.asnumpy(r) >= 0 for r in result]


def _contains_signed_and_unsigned(kw):
    def isdtype(v):
        if isinstance(v, numpy.dtype):
            return True
        elif isinstance(v, str):  # expecting dtype character codes
            return True
        elif isinstance(v, type) and issubclass(v, numpy.number):
            return True
        else:
            return False
    vs = set(v for v in kw.values() if isdtype(v))
    return any(d in vs for d in _unsigned_dtypes) and \
        any(d in vs for d in _float_dtypes + _signed_dtypes)


def _wraps_partial(wrapped, *names):
    # Only `wrapped` function have args of `names`.
    def decorator(impl):
        impl = functools.wraps(wrapped)(impl)
        impl.__signature__ = inspect.signature(
            functools.partial(wrapped, **{name: None for name in names}))
        return impl
    return decorator


def _wraps_partial_xp(wrapped, name, sp_name, scipy_name):
    names = [name, sp_name, scipy_name]
    names = [n for n in names if n is not None]
    return _wraps_partial(wrapped, *names)


def _make_decorator(check_func, name, type_check, contiguous_check,
                    accept_error, sp_name=None, scipy_name=None,
                    check_sparse_format=True):
    assert isinstance(name, str)
    assert sp_name is None or isinstance(sp_name, str)
    assert scipy_name is None or isinstance(scipy_name, str)

    def decorator(impl):
        @_wraps_partial_xp(impl, name, sp_name, scipy_name)
        def test_func(*args, **kw):
            # Run cupy and numpy
            (
                cupy_result, cupy_error,
                numpy_result, numpy_error) = (
                    _call_func_numpy_cupy(
                        impl, args, kw, name, sp_name, scipy_name))
            assert cupy_result is not None or cupy_error is not None
            assert numpy_result is not None or numpy_error is not None

            # Check errors raised
            if cupy_error or numpy_error:
                _check_cupy_numpy_error(cupy_error,
                                        numpy_error,
                                        accept_error=accept_error)
                return

            # Check returned arrays

            if not isinstance(cupy_result, (tuple, list)):
                cupy_result = cupy_result,
            if not isinstance(numpy_result, (tuple, list)):
                numpy_result = numpy_result,

            assert len(cupy_result) == len(numpy_result)

            # Check types
            cupy_numpy_result_ndarrays = [
                _convert_output_to_ndarray(
                    cupy_r, numpy_r, sp_name, check_sparse_format)
                for cupy_r, numpy_r in zip(cupy_result, numpy_result)]

            # Check dtypes
            numpy_version = numpy.lib.NumpyVersion(numpy.__version__)
            if type_check and numpy_version < '2.0.0':
                for cupy_r, numpy_r in cupy_numpy_result_ndarrays:
                    if cupy_r.dtype != numpy_r.dtype:
                        raise AssertionError(
                            '''ndarrays of different dtypes are returned.
cupy: {}
numpy: {}'''.format(cupy_r.dtype, numpy_r.dtype))

            # Check contiguities
            if contiguous_check:
                for cupy_r, numpy_r in zip(cupy_result, numpy_result):
                    if isinstance(numpy_r, numpy.ndarray):
                        if (numpy_r.flags.c_contiguous
                                and not cupy_r.flags.c_contiguous):
                            raise AssertionError(
                                'The state of c_contiguous flag is false. '
                                '(cupy_result:{} numpy_result:{})'.format(
                                    cupy_r.flags.c_contiguous,
                                    numpy_r.flags.c_contiguous))
                        if (numpy_r.flags.f_contiguous
                                and not cupy_r.flags.f_contiguous):
                            raise AssertionError(
                                'The state of f_contiguous flag is false. '
                                '(cupy_result:{} numpy_result:{})'.format(
                                    cupy_r.flags.f_contiguous,
                                    numpy_r.flags.f_contiguous))

            # Check shapes
            for cupy_r, numpy_r in cupy_numpy_result_ndarrays:
                assert cupy_r.shape == numpy_r.shape

            masks = [None] * len(cupy_result)
            if _contains_signed_and_unsigned(kw):
                needs_mask = [
                    cupy_r.dtype in _unsigned_dtypes
                    for cupy_r in cupy_result]
                if any(needs_mask):
                    masks = _make_positive_masks(
                        impl, args, kw, name, sp_name, scipy_name)
                    for i, flag in enumerate(needs_mask):
                        if not flag:
                            masks[i] = None

            # Check item values
            for (cupy_r, numpy_r), mask in zip(
                    cupy_numpy_result_ndarrays, masks):
                # Behavior of assigning a negative value to an unsigned integer
                # variable is undefined.
                # nVidia GPUs and Intel CPUs behave differently.
                # To avoid this difference, we need to ignore dimensions whose
                # values are negative.

                skip = False
                if mask is not None:
                    if cupy_r.shape == ():
                        skip = (mask == 0).all()
                    else:
                        cupy_r = cupy_r[mask].get()
                        numpy_r = numpy_r[mask]

                if not skip:
                    check_func(cupy_r, numpy_r)
        return test_func
    return decorator


def _convert_output_to_ndarray(c_out, n_out, sp_name, check_sparse_format):
    """Checks type of cupy/numpy results and returns cupy/numpy ndarrays.

    Args:
        c_out (cupy.ndarray, cupyx.scipy.sparse matrix, cupy.poly1d or scalar):
            cupy result
        n_out (numpy.ndarray, scipy.sparse matrix, numpy.poly1d or scalar):
            numpy result
        sp_name(str or None): Argument name whose value is either
            ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
            argument is given for the modules.
        check_sparse_format (bool): If ``True``, consistency of format of
            sparse matrix is also checked. Default is ``True``.

    Returns:
        The tuple of cupy.ndarray and numpy.ndarray.
    """
    if sp_name is not None and cupyx.scipy.sparse.issparse(c_out):
        # Sparse output case.
        import scipy.sparse
        assert scipy.sparse.issparse(n_out)
        if check_sparse_format:
            assert c_out.format == n_out.format
        return c_out.toarray(), n_out.toarray()
    if (isinstance(c_out, cupy.ndarray)
            and isinstance(n_out, (numpy.ndarray, numpy.generic))):
        # ndarray output case.
        return c_out, n_out
    if isinstance(c_out, cupy.poly1d) and isinstance(n_out, numpy.poly1d):
        # poly1d output case.
        assert c_out.variable == n_out.variable
        return c_out.coeffs, n_out.coeffs
    if isinstance(c_out, numpy.generic) and isinstance(n_out, numpy.generic):
        # numpy scalar output case.
        return c_out, n_out
    if numpy.isscalar(c_out) and numpy.isscalar(n_out):
        # python scalar output case.
        return cupy.array(c_out), numpy.array(n_out)
    raise AssertionError(
        'numpy and cupy returns different type of return value:\n'
        'cupy: {}\nnumpy: {}'.format(
            type(c_out), type(n_out)))


def _check_tolerance_keys(rtol, atol):
    def _check(tol):
        if isinstance(tol, dict):
            for k in tol.keys():
                if type(k) is type:
                    continue
                if type(k) is str and k == 'default':
                    continue
                msg = ('Keys of the tolerance dictionary need to be type '
                       'objects as `numpy.float32` and `cupy.float32` or '
                       '`\'default\'` string.')
                raise TypeError(msg)
    _check(rtol)
    _check(atol)


def _resolve_tolerance(type_check, result, rtol, atol):
    def _resolve(dtype, tol):
        if isinstance(tol, dict):
            tol1 = tol.get(dtype.type)
            if tol1 is None:
                tol1 = tol.get('default')
                if tol1 is None:
                    raise TypeError(
                        'Can not find tolerance for {}'.format(dtype.type))
            return tol1
        else:
            return tol

    dtype = result.dtype
    rtol1 = _resolve(dtype, rtol)
    atol1 = _resolve(dtype, atol)
    return rtol1, atol1


def numpy_cupy_allclose(rtol=1e-7, atol=0, err_msg='', verbose=True,
                        name='xp', type_check=True, accept_error=False,
                        sp_name=None, scipy_name=None, contiguous_check=True,
                        *, _check_sparse_format=True):
    """Decorator that checks NumPy results and CuPy ones are close.

    Args:
         rtol(float or dict): Relative tolerance. Besides a float value, a
             dictionary that maps a dtypes to a float value can be supplied to
             adjust tolerance per dtype. If the dictionary has ``'default'``
             string as its key, its value is used as the default tolerance in
             case any dtype keys do not match.
         atol(float or dict): Absolute tolerance. Besides a float value, a
             dictionary can be supplied as ``rtol``.
         err_msg(str): The error message to be printed in case of failure.
         verbose(bool): If ``True``, the conflicting values are
             appended to the error message.
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         type_check(bool): If ``True``, consistency of dtype is also checked.
         accept_error(bool, Exception or tuple of Exception): Specify
             acceptable errors. When both NumPy test and CuPy test raises the
             same type of errors, and the type of the errors is specified with
             this argument, the errors are ignored and not raised.
             If it is ``True`` all error types are acceptable.
             If it is ``False`` no error is acceptable.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.
         contiguous_check(bool): If ``True``, consistency of contiguity is
             also checked.

    Decorated test fixture is required to return the arrays whose values are
    close between ``numpy`` case and ``cupy`` case.
    For example, this test case checks ``numpy.zeros`` and ``cupy.zeros``
    should return same value.

    >>> import unittest
    >>> from cupy import testing
    >>> class TestFoo(unittest.TestCase):
    ...
    ...     @testing.numpy_cupy_allclose()
    ...     def test_foo(self, xp):
    ...         # ...
    ...         # Prepare data with xp
    ...         # ...
    ...
    ...         xp_result = xp.zeros(10)
    ...         return xp_result

    .. seealso:: :func:`cupy.testing.assert_allclose`
    """
    _check_tolerance_keys(rtol, atol)

    # When `type_check` is `False`, cupy result and numpy result may have
    # different dtypes so we can not determine the dtype to use from the
    # tolerance associations.
    if not type_check:
        if isinstance(rtol, dict) or isinstance(atol, dict):
            raise TypeError('When `type_check` is `False`, `rtol` and `atol` '
                            'must be supplied as float.')

    def check_func(c, n):
        rtol1, atol1 = _resolve_tolerance(type_check, c, rtol, atol)
        _array.assert_allclose(c, n, rtol1, atol1, err_msg, verbose)
    return _make_decorator(check_func, name, type_check, contiguous_check,
                           accept_error, sp_name, scipy_name,
                           _check_sparse_format)


def numpy_cupy_array_almost_equal(decimal=6, err_msg='', verbose=True,
                                  name='xp', type_check=True,
                                  accept_error=False, sp_name=None,
                                  scipy_name=None):
    """Decorator that checks NumPy results and CuPy ones are almost equal.

    Args:
         decimal(int): Desired precision.
         err_msg(str): The error message to be printed in case of failure.
         verbose(bool): If ``True``, the conflicting values
             are appended to the error message.
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         type_check(bool): If ``True``, consistency of dtype is also checked.
         accept_error(bool, Exception or tuple of Exception): Specify
             acceptable errors. When both NumPy test and CuPy test raises the
             same type of errors, and the type of the errors is specified with
             this argument, the errors are ignored and not raised.
             If it is ``True`` all error types are acceptable.
             If it is ``False`` no error is acceptable.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.

    Decorated test fixture is required to return the same arrays
    in the sense of :func:`cupy.testing.assert_array_almost_equal`
    (except the type of array module) even if ``xp`` is ``numpy`` or ``cupy``.

    .. seealso:: :func:`cupy.testing.assert_array_almost_equal`
    """
    def check_func(x, y):
        _array.assert_array_almost_equal(x, y, decimal, err_msg, verbose)
    return _make_decorator(check_func, name, type_check, False,
                           accept_error, sp_name, scipy_name)


def numpy_cupy_array_almost_equal_nulp(nulp=1, name='xp', type_check=True,
                                       accept_error=False, sp_name=None,
                                       scipy_name=None):
    """Decorator that checks results of NumPy and CuPy are equal w.r.t. spacing.

    Args:
         nulp(int): The maximum number of unit in the last place for tolerance.
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         type_check(bool): If ``True``, consistency of dtype is also checked.
         accept_error(bool, Exception or tuple of Exception): Specify
             acceptable errors. When both NumPy test and CuPy test raises the
             same type of errors, and the type of the errors is specified with
             this argument, the errors are ignored and not raised.
             If it is ``True``, all error types are acceptable.
             If it is ``False``, no error is acceptable.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.

    Decorated test fixture is required to return the same arrays
    in the sense of :func:`cupy.testing.assert_array_almost_equal_nulp`
    (except the type of array module) even if ``xp`` is ``numpy`` or ``cupy``.

    .. seealso:: :func:`cupy.testing.assert_array_almost_equal_nulp`
    """  # NOQA
    def check_func(x, y):
        _array.assert_array_almost_equal_nulp(x, y, nulp)
    return _make_decorator(check_func, name, type_check, False,
                           accept_error, sp_name, scipy_name=None)


def numpy_cupy_array_max_ulp(maxulp=1, dtype=None, name='xp', type_check=True,
                             accept_error=False, sp_name=None,
                             scipy_name=None):
    """Decorator that checks results of NumPy and CuPy ones are equal w.r.t. ulp.

    Args:
         maxulp(int): The maximum number of units in the last place
             that elements of resulting two arrays can differ.
         dtype(numpy.dtype): Data-type to convert the resulting
             two array to if given.
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         type_check(bool): If ``True``, consistency of dtype is also checked.
         accept_error(bool, Exception or tuple of Exception): Specify
             acceptable errors. When both NumPy test and CuPy test raises the
             same type of errors, and the type of the errors is specified with
             this argument, the errors are ignored and not raised.
             If it is ``True`` all error types are acceptable.
             If it is ``False`` no error is acceptable.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.

    Decorated test fixture is required to return the same arrays
    in the sense of :func:`assert_array_max_ulp`
    (except the type of array module) even if ``xp`` is ``numpy`` or ``cupy``.

    .. seealso:: :func:`cupy.testing.assert_array_max_ulp`

    """  # NOQA
    def check_func(x, y):
        _array.assert_array_max_ulp(x, y, maxulp, dtype)
    return _make_decorator(check_func, name, type_check, False,
                           accept_error, sp_name, scipy_name)


def numpy_cupy_array_equal(err_msg='', verbose=True, name='xp',
                           type_check=True, accept_error=False, sp_name=None,
                           scipy_name=None, strides_check=False):
    """Decorator that checks NumPy results and CuPy ones are equal.

    Args:
         err_msg(str): The error message to be printed in case of failure.
         verbose(bool): If ``True``, the conflicting values are
             appended to the error message.
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         type_check(bool): If ``True``, consistency of dtype is also checked.
         accept_error(bool, Exception or tuple of Exception): Specify
             acceptable errors. When both NumPy test and CuPy test raises the
             same type of errors, and the type of the errors is specified with
             this argument, the errors are ignored and not raised.
             If it is ``True`` all error types are acceptable.
             If it is ``False`` no error is acceptable.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.
         strides_check(bool): If ``True``, consistency of strides is also
             checked.

    Decorated test fixture is required to return the same arrays
    in the sense of :func:`numpy_cupy_array_equal`
    (except the type of array module) even if ``xp`` is ``numpy`` or ``cupy``.

    .. seealso:: :func:`cupy.testing.assert_array_equal`
    """
    def check_func(x, y):
        _array.assert_array_equal(x, y, err_msg, verbose, strides_check)
    return _make_decorator(check_func, name, type_check, False,
                           accept_error, sp_name, scipy_name)


def numpy_cupy_array_list_equal(
        err_msg='', verbose=True, name='xp', sp_name=None, scipy_name=None):
    """Decorator that checks the resulting lists of NumPy and CuPy's one are equal.

    Args:
         err_msg(str): The error message to be printed in case of failure.
         verbose(bool): If ``True``, the conflicting values are appended
             to the error message.
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.

    Decorated test fixture is required to return the same list of arrays
    (except the type of array module) even if ``xp`` is ``numpy`` or ``cupy``.

    .. seealso:: :func:`cupy.testing.assert_array_list_equal`
    """  # NOQA
    warnings.warn(
        'numpy_cupy_array_list_equal is deprecated.'
        ' Use numpy_cupy_array_equal instead.',
        DeprecationWarning)

    def check_func(x, y):
        _array.assert_array_equal(x, y, err_msg, verbose)
    return _make_decorator(check_func, name, False, False,
                           False, sp_name, scipy_name)


def numpy_cupy_array_less(err_msg='', verbose=True, name='xp',
                          type_check=True, accept_error=False, sp_name=None,
                          scipy_name=None):
    """Decorator that checks the CuPy result is less than NumPy result.

    Args:
         err_msg(str): The error message to be printed in case of failure.
         verbose(bool): If ``True``, the conflicting values are
             appended to the error message.
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         type_check(bool): If ``True``, consistency of dtype is also checked.
         accept_error(bool, Exception or tuple of Exception): Specify
             acceptable errors. When both NumPy test and CuPy test raises the
             same type of errors, and the type of the errors is specified with
             this argument, the errors are ignored and not raised.
             If it is ``True`` all error types are acceptable.
             If it is ``False`` no error is acceptable.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.

    Decorated test fixture is required to return the smaller array
    when ``xp`` is ``cupy`` than the one when ``xp`` is ``numpy``.

    .. seealso:: :func:`cupy.testing.assert_array_less`
    """
    def check_func(x, y):
        _array.assert_array_less(x, y, err_msg, verbose)
    return _make_decorator(check_func, name, type_check, False,
                           accept_error, sp_name, scipy_name)


def numpy_cupy_equal(name='xp', sp_name=None, scipy_name=None):
    """Decorator that checks NumPy results are equal to CuPy ones.

    Args:
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.

    Decorated test fixture is required to return the same results
    even if ``xp`` is ``numpy`` or ``cupy``.
    """
    def decorator(impl):
        @_wraps_partial_xp(impl, name, sp_name, scipy_name)
        def test_func(*args, **kw):
            # Run cupy and numpy
            (
                cupy_result, cupy_error,
                numpy_result, numpy_error) = (
                    _call_func_numpy_cupy(
                        impl, args, kw, name, sp_name, scipy_name))

            if cupy_error or numpy_error:
                _check_cupy_numpy_error(
                    cupy_error, numpy_error,
                    accept_error=False)
                return

            if cupy_result != numpy_result:
                message = '''Results are not equal:
cupy: %s
numpy: %s''' % (str(cupy_result), str(numpy_result))
                raise AssertionError(message)
        return test_func
    return decorator


def numpy_cupy_raises(name='xp', sp_name=None, scipy_name=None,
                      accept_error=Exception):
    """Decorator that checks the NumPy and CuPy throw same errors.

    Args:
         name(str): Argument name whose value is either
             ``numpy`` or ``cupy`` module.
         sp_name(str or None): Argument name whose value is either
             ``scipy.sparse`` or ``cupyx.scipy.sparse`` module. If ``None``, no
             argument is given for the modules.
         scipy_name(str or None): Argument name whose value is either ``scipy``
             or ``cupyx.scipy`` module. If ``None``, no argument is given for
             the modules.
         accept_error(bool, Exception or tuple of Exception): Specify
             acceptable errors. When both NumPy test and CuPy test raises the
             same type of errors, and the type of the errors is specified with
             this argument, the errors are ignored and not raised.
             If it is ``True`` all error types are acceptable.
             If it is ``False`` no error is acceptable.

    Decorated test fixture is required throw same errors
    even if ``xp`` is ``numpy`` or ``cupy``.
    """
    warnings.warn(
        'cupy.testing.numpy_cupy_raises is deprecated.',
        DeprecationWarning)

    def decorator(impl):
        @_wraps_partial_xp(impl, name, sp_name, scipy_name)
        def test_func(*args, **kw):
            # Run cupy and numpy
            (
                cupy_result, cupy_error,
                numpy_result, numpy_error) = (
                    _call_func_numpy_cupy(
                        impl, args, kw, name, sp_name, scipy_name))

            _check_cupy_numpy_error(cupy_error,
                                    numpy_error,
                                    accept_error=accept_error)
        return test_func
    return decorator


def for_dtypes(dtypes, name='dtype'):
    """Decorator for parameterized dtype test.

    Args:
         dtypes(list of dtypes): dtypes to be tested.
         name(str): Argument name to which specified dtypes are passed.

    This decorator adds a keyword argument specified by ``name``
    to the test fixture. Then, it runs the fixtures in parallel
    by passing the each element of ``dtypes`` to the named
    argument.
    """
    def decorator(impl):
        @_wraps_partial(impl, name)
        def test_func(*args, **kw):
            for dtype in dtypes:
                try:
                    kw[name] = numpy.dtype(dtype).type
                    impl(*args, **kw)
                except _skip_classes as e:
                    print('skipped: {} = {} ({})'.format(name, dtype, e))
                except Exception:
                    print(name, 'is', dtype)
                    raise

        return test_func
    return decorator


_complex_dtypes = (numpy.complex64, numpy.complex128)
_regular_float_dtypes = (numpy.float64, numpy.float32)
_float_dtypes = _regular_float_dtypes + (numpy.float16,)
_signed_dtypes = tuple(numpy.dtype(i).type for i in 'bhilq')
_unsigned_dtypes = tuple(numpy.dtype(i).type for i in 'BHILQ')
_int_dtypes = _signed_dtypes + _unsigned_dtypes
_int_bool_dtypes = _int_dtypes + (numpy.bool_,)
_regular_dtypes = _regular_float_dtypes + _int_bool_dtypes
_dtypes = _float_dtypes + _int_bool_dtypes


def _make_all_dtypes(no_float16, no_bool, no_complex):
    if no_float16:
        dtypes = _regular_float_dtypes
    else:
        dtypes = _float_dtypes

    if no_bool:
        dtypes += _int_dtypes
    else:
        dtypes += _int_bool_dtypes

    if not no_complex:
        dtypes += _complex_dtypes

    return dtypes


def for_all_dtypes(name='dtype', no_float16=False, no_bool=False,
                   no_complex=False):
    """Decorator that checks the fixture with all dtypes.

    Args:
         name(str): Argument name to which specified dtypes are passed.
         no_float16(bool): If ``True``, ``numpy.float16`` is
             omitted from candidate dtypes.
         no_bool(bool): If ``True``, ``numpy.bool_`` is
             omitted from candidate dtypes.
         no_complex(bool): If ``True``, ``numpy.complex64`` and
             ``numpy.complex128`` are omitted from candidate dtypes.

    dtypes to be tested: ``numpy.complex64`` (optional),
    ``numpy.complex128`` (optional),
    ``numpy.float16`` (optional), ``numpy.float32``,
    ``numpy.float64``, ``numpy.dtype('b')``, ``numpy.dtype('h')``,
    ``numpy.dtype('i')``, ``numpy.dtype('l')``, ``numpy.dtype('q')``,
    ``numpy.dtype('B')``, ``numpy.dtype('H')``, ``numpy.dtype('I')``,
    ``numpy.dtype('L')``, ``numpy.dtype('Q')``, and ``numpy.bool_`` (optional).

    The usage is as follows.
    This test fixture checks if ``cPickle`` successfully reconstructs
    :class:`cupy.ndarray` for various dtypes.
    ``dtype`` is an argument inserted by the decorator.

    >>> import unittest
    >>> from cupy import testing
    >>> class TestNpz(unittest.TestCase):
    ...
    ...     @testing.for_all_dtypes()
    ...     def test_pickle(self, dtype):
    ...         a = testing.shaped_arange((2, 3, 4), dtype=dtype)
    ...         s = pickle.dumps(a)
    ...         b = pickle.loads(s)
    ...         testing.assert_array_equal(a, b)

    Typically, we use this decorator in combination with
    decorators that check consistency between NumPy and CuPy like
    :func:`cupy.testing.numpy_cupy_allclose`.
    The following is such an example.

    >>> import unittest
    >>> from cupy import testing
    >>> class TestMean(unittest.TestCase):
    ...
    ...     @testing.for_all_dtypes()
    ...     @testing.numpy_cupy_allclose()
    ...     def test_mean_all(self, xp, dtype):
    ...         a = testing.shaped_arange((2, 3), xp, dtype)
    ...         return a.mean()

    .. seealso:: :func:`cupy.testing.for_dtypes`
    """
    return for_dtypes(_make_all_dtypes(no_float16, no_bool, no_complex),
                      name=name)


def for_float_dtypes(name='dtype', no_float16=False):
    """Decorator that checks the fixture with float dtypes.

    Args:
         name(str): Argument name to which specified dtypes are passed.
         no_float16(bool): If ``True``, ``numpy.float16`` is
             omitted from candidate dtypes.

    dtypes to be tested are ``numpy.float16`` (optional), ``numpy.float32``,
    and ``numpy.float64``.

    .. seealso:: :func:`cupy.testing.for_dtypes`,
        :func:`cupy.testing.for_all_dtypes`
    """
    if no_float16:
        return for_dtypes(_regular_float_dtypes, name=name)
    else:
        return for_dtypes(_float_dtypes, name=name)


def for_signed_dtypes(name='dtype'):
    """Decorator that checks the fixture with signed dtypes.

    Args:
         name(str): Argument name to which specified dtypes are passed.

    dtypes to be tested are ``numpy.dtype('b')``, ``numpy.dtype('h')``,
    ``numpy.dtype('i')``, ``numpy.dtype('l')``, and ``numpy.dtype('q')``.

    .. seealso:: :func:`cupy.testing.for_dtypes`,
        :func:`cupy.testing.for_all_dtypes`
    """
    return for_dtypes(_signed_dtypes, name=name)


def for_unsigned_dtypes(name='dtype'):
    """Decorator that checks the fixture with unsigned dtypes.

    Args:
         name(str): Argument name to which specified dtypes are passed.

    dtypes to be tested are ``numpy.dtype('B')``, ``numpy.dtype('H')``,

     ``numpy.dtype('I')``, ``numpy.dtype('L')``, and ``numpy.dtype('Q')``.

    .. seealso:: :func:`cupy.testing.for_dtypes`,
        :func:`cupy.testing.for_all_dtypes`
    """
    return for_dtypes(_unsigned_dtypes, name=name)


def for_int_dtypes(name='dtype', no_bool=False):
    """Decorator that checks the fixture with integer and optionally bool dtypes.

    Args:
         name(str): Argument name to which specified dtypes are passed.
         no_bool(bool): If ``True``, ``numpy.bool_`` is
             omitted from candidate dtypes.

    dtypes to be tested are ``numpy.dtype('b')``, ``numpy.dtype('h')``,
    ``numpy.dtype('i')``, ``numpy.dtype('l')``, ``numpy.dtype('q')``,
    ``numpy.dtype('B')``, ``numpy.dtype('H')``, ``numpy.dtype('I')``,
    ``numpy.dtype('L')``, ``numpy.dtype('Q')``, and ``numpy.bool_`` (optional).

    .. seealso:: :func:`cupy.testing.for_dtypes`,
        :func:`cupy.testing.for_all_dtypes`
    """  # NOQA
    if no_bool:
        return for_dtypes(_int_dtypes, name=name)
    else:
        return for_dtypes(_int_bool_dtypes, name=name)


def for_complex_dtypes(name='dtype'):
    """Decorator that checks the fixture with complex dtypes.

    Args:
         name(str): Argument name to which specified dtypes are passed.

    dtypes to be tested are ``numpy.complex64`` and ``numpy.complex128``.

    .. seealso:: :func:`cupy.testing.for_dtypes`,
        :func:`cupy.testing.for_all_dtypes`
    """
    return for_dtypes(_complex_dtypes, name=name)


def for_dtypes_combination(types, names=('dtype',), full=None):
    """Decorator that checks the fixture with a product set of dtypes.

    Args:
         types(list of dtypes): dtypes to be tested.
         names(list of str): Argument names to which dtypes are passed.
         full(bool): If ``True``, then all combinations
             of dtypes will be tested.
             Otherwise, the subset of combinations will be tested
             (see the description below).

    Decorator adds the keyword arguments specified by ``names``
    to the test fixture. Then, it runs the fixtures in parallel
    with passing (possibly a subset of) the product set of dtypes.
    The range of dtypes is specified by ``types``.

    The combination of dtypes to be tested changes depending
    on the option ``full``. If ``full`` is ``True``,
    all combinations of ``types`` are tested.
    Sometimes, such an exhaustive test can be costly.
    So, if ``full`` is ``False``, only a subset of possible combinations
    is randomly sampled. If ``full`` is ``None``, the behavior is
    determined by an environment variable ``CUPY_TEST_FULL_COMBINATION``.
    If the value is set to ``'1'``, it behaves as if ``full=True``, and
    otherwise ``full=False``.
    """
    types = list(types)

    if len(types) == 1:
        name, = names
        return for_dtypes(types, name)

    if full is None:
        full = int(os.environ.get('CUPY_TEST_FULL_COMBINATION', '0')) != 0

    if full:
        combination = _parameterized.product({name: types for name in names})
    else:
        ts = []
        for _ in range(len(names)):
            # Make shuffled list of types for each name
            shuffled_types = types[:]
            random.shuffle(shuffled_types)
            ts.append(types + shuffled_types)

        combination = [tuple(zip(names, typs)) for typs in zip(*ts)]
        # Remove duplicate entries
        combination = [dict(assoc_list) for assoc_list in set(combination)]

    def decorator(impl):
        @_wraps_partial(impl, *names)
        def test_func(*args, **kw):
            for dtypes in combination:
                kw_copy = kw.copy()
                kw_copy.update(dtypes)

                try:
                    impl(*args, **kw_copy)
                except _skip_classes as e:
                    msg = ', '.join(
                        '{} = {}'.format(name, dtype)
                        for name, dtype in dtypes.items())
                    print('skipped: {} ({})'.format(msg, e))
                except Exception:
                    print(dtypes)
                    raise

        return test_func
    return decorator


def for_all_dtypes_combination(names=('dtyes',),
                               no_float16=False, no_bool=False, full=None,
                               no_complex=False):
    """Decorator that checks the fixture with a product set of all dtypes.

    Args:
         names(list of str): Argument names to which dtypes are passed.
         no_float16(bool): If ``True``, ``numpy.float16`` is
             omitted from candidate dtypes.
         no_bool(bool): If ``True``, ``numpy.bool_`` is
             omitted from candidate dtypes.
         full(bool): If ``True``, then all combinations of dtypes
             will be tested.
             Otherwise, the subset of combinations will be tested
             (see description in :func:`cupy.testing.for_dtypes_combination`).
         no_complex(bool): If, True, ``numpy.complex64`` and
             ``numpy.complex128`` are omitted from candidate dtypes.

    .. seealso:: :func:`cupy.testing.for_dtypes_combination`
    """
    types = _make_all_dtypes(no_float16, no_bool, no_complex)
    return for_dtypes_combination(types, names, full)


def for_signed_dtypes_combination(names=('dtype',), full=None):
    """Decorator for parameterized test w.r.t. the product set of signed dtypes.

    Args:
         names(list of str): Argument names to which dtypes are passed.
         full(bool): If ``True``, then all combinations of dtypes
             will be tested.
             Otherwise, the subset of combinations will be tested
             (see description in :func:`cupy.testing.for_dtypes_combination`).

    .. seealso:: :func:`cupy.testing.for_dtypes_combination`
    """  # NOQA
    return for_dtypes_combination(_signed_dtypes, names=names, full=full)


def for_unsigned_dtypes_combination(names=('dtype',), full=None):
    """Decorator for parameterized test w.r.t. the product set of unsigned dtypes.

    Args:
         names(list of str): Argument names to which dtypes are passed.
         full(bool): If ``True``, then all combinations of dtypes
             will be tested.
             Otherwise, the subset of combinations will be tested
             (see description in :func:`cupy.testing.for_dtypes_combination`).

    .. seealso:: :func:`cupy.testing.for_dtypes_combination`
    """  # NOQA
    return for_dtypes_combination(_unsigned_dtypes, names=names, full=full)


def for_int_dtypes_combination(names=('dtype',), no_bool=False, full=None):
    """Decorator for parameterized test w.r.t. the product set of int and boolean.

    Args:
         names(list of str): Argument names to which dtypes are passed.
         no_bool(bool): If ``True``, ``numpy.bool_`` is
             omitted from candidate dtypes.
         full(bool): If ``True``, then all combinations of dtypes
             will be tested.
             Otherwise, the subset of combinations will be tested
             (see description in :func:`cupy.testing.for_dtypes_combination`).

    .. seealso:: :func:`cupy.testing.for_dtypes_combination`
    """  # NOQA
    if no_bool:
        types = _int_dtypes
    else:
        types = _int_bool_dtypes
    return for_dtypes_combination(types, names, full)


def for_orders(orders, name='order'):
    """Decorator to parameterize tests with order.

    Args:
         orders(list of order): orders to be tested.
         name(str): Argument name to which the specified order is passed.

    This decorator adds a keyword argument specified by ``name``
    to the test fixtures. Then, the fixtures run by passing each element of
    ``orders`` to the named argument.

    """
    def decorator(impl):
        @_wraps_partial(impl, name)
        def test_func(*args, **kw):
            for order in orders:
                try:
                    kw[name] = order
                    impl(*args, **kw)
                except Exception:
                    print(name, 'is', order)
                    raise

        return test_func
    return decorator


def for_CF_orders(name='order'):
    """Decorator that checks the fixture with orders 'C' and 'F'.

    Args:
         name(str): Argument name to which the specified order is passed.

    .. seealso:: :func:`cupy.testing.for_all_dtypes`

    """
    return for_orders([None, 'C', 'F', 'c', 'f'], name)


def for_contiguous_axes(name='axis'):
    '''Decorator for parametrizing tests with possible contiguous axes.

    Args:
        name(str): Argument name to which specified axis are passed.

    .. note::
        1. Adapted from tests/cupy_tests/fft_tests/test_fft.py.
        2. Example: for ``shape = (1, 2, 3)``, the tested axes are
            ``[(2,), (1, 2), (0, 1, 2)]`` for the C order, and
            ``[(0,), (0, 1), (0, 1, 2)]`` for the F order.
    '''
    def decorator(impl):
        @_wraps_partial(impl, name)
        def test_func(self, *args, **kw):
            ndim = len(self.shape)
            order = self.order
            for i in range(ndim):
                a = ()
                if order in ('c', 'C'):
                    for j in range(ndim-1, i-1, -1):
                        a = (j,) + a
                elif order in ('f', 'F'):
                    for j in range(0, i+1):
                        a = a + (j,)
                else:
                    raise ValueError('Please specify the array order.')
                try:
                    kw[name] = a
                    impl(self, *args, **kw)
                except Exception:
                    print(name, 'is', a, ', ndim is', ndim, ', shape is',
                          self.shape, ', order is', order)
                    raise
        return test_func
    return decorator