dev(narugo): add other transforms

from unittest import skipUnless

import pytest

from PIL import Image

from hbutils.testing import tmatrix

from imgutils.preprocess.torchvision import _get_interpolation_mode

from imgutils.preprocess.torchvision import _get_interpolation_mode, create_torchvision_transforms

from test.testings import get_testfile

try:

    import torchvision

            _ = _get_interpolation_mode(None)

        with pytest.raises(TypeError):

            _ = _get_interpolation_mode([])

    @skipUnless(_TORCHVISION_AVAILABLE, 'Torchvision required.')

    @pytest.mark.parametrize(*tmatrix({

        'src_image': [

            'png_640.png',

            'png_640_m90.png',

        ],

        ('mode', 'channels'): [

            ('I', 1), ('I;16', 1), ('F', 1), ('1', 1), ('L', 1), ('P', 1),

            ('LA', 2),

            ('RGB', 3), ('YCbCr', 3),

            ('RGBA', 4), ('CMYK', 4),

        ]

    }))

    def test_maybe_to_tensor(self, src_image, mode, channels):

        image = Image.open(get_testfile(src_image))

        image = image.convert(mode)

        assert image.mode == mode

        ttrans = create_torchvision_transforms({'type': 'maybe_to_tensor'})

        result = ttrans(image)

        assert tuple(result.shape) == (channels, image.height, image.width)

    @skipUnless(_TORCHVISION_AVAILABLE, 'Torchvision required.')

    def test_maybe_to_tensor_repr(self):

        ttrans = create_torchvision_transforms({'type': 'maybe_to_tensor'})

        assert repr(ttrans) == 'MaybeToTensor()'

    @skipUnless(_TORCHVISION_AVAILABLE, 'Torchvision required.')

    def test_maybe_to_tensor_np(self):

        import torch

        input_ = torch.randn(3, 384, 384)

        ttrans = create_torchvision_transforms({'type': 'maybe_to_tensor'})

        assert torch.allclose(ttrans(input_), input_)

    @skipUnless(_TORCHVISION_AVAILABLE, 'Torchvision required.')

    @pytest.mark.parametrize(*tmatrix({

        'src_image': [

            'png_640.png',

            'png_640_m90.png',

        ],

        ('mode', 'channels'): [

            ('I', 1), ('I;16', 1), ('F', 1), ('1', 1), ('L', 1), ('P', 1),

            ('LA', 2),

            ('RGB', 3), ('YCbCr', 3),

            ('RGBA', 4), ('CMYK', 4),

        ]

    }))

    def test_to_tensor(self, src_image, mode, channels):

        image = Image.open(get_testfile(src_image))

        image = image.convert(mode)

        assert image.mode == mode

        ttrans = create_torchvision_transforms({'type': 'to_tensor'})

        result = ttrans(image)

        assert tuple(result.shape) == (channels, image.height, image.width)