RotationWrapper unnecessarily calculates core of block instead of using DataSetBlock.data_unpadded

[yousefmoazzam]

As the title says, the following logic is unnecessary and likely can be made tidier by using block.data_unpadded:

httomo/httomo/method_wrappers/rotation.py

Lines 154 to 161 in 5c5c57b

	# Extract core of block, in case padding slices are present
	core_angles_start = 0
	core_angles_stop = block.shape_unpadded[0]
	if block.is_padded:
	core_angles_start = block.padding[0]
	core_angles_stop = core_angles_start + block.shape_unpadded[0]
	data = block.data[core_angles_start:core_angles_stop, slice_for_cor, :]

Plus, the logic to extract the core out of a padded block is already tested for DataSetBlock.data_unpadded:

httomo/tests/runner/test_dataset_block.py

Lines 532 to 552 in 5c5c57b

	def test_get_padded_and_unpadded_data():
	padding = (2, 2)
	data = np.ones((10, 10, 10), dtype=np.float32)
	# set the padding areas to something else
	data[: padding[0], :, :] = -1
	data[10 - padding[1] :, :, :] = -2
	angles = np.linspace(0, math.pi, 30, dtype=np.float32)
	block = DataSetBlock(
	data=data,
	aux_data=AuxiliaryData(angles=angles),
	block_start=-2,
	chunk_start=8,
	slicing_dim=0,
	global_shape=(30, 10, 10),
	padding=padding,
	)
	np.testing.assert_array_equal(block.data, data)
	unpadded = block.data_unpadded
	assert unpadded.shape == (6, 10, 10)
	np.testing.assert_array_equal(unpadded, 1.0)

whereas the code extracting the block's core in the rotation wrapper isn't.