`svd_full` with empty matrices:

[lkdvos]

Currently svd_full throws when working with arrays that have dimension 0, while LinearAlgebra produces the desired result:

julia> A = rand(4, 0)
4×0 Matrix{Float64}

julia> MatrixAlgebraKit.svd_full(A)
ERROR: BoundsError: attempt to access 4×0 Matrix{Float64} at index [1:0, 1]

julia> LinearAlgebra.svd(A; full=true)
LinearAlgebra.SVD{Float64, Float64, Matrix{Float64}, Vector{Float64}}
U factor:
4×4 Matrix{Float64}:
 1.0  0.0  0.0  0.0
 0.0  1.0  0.0  0.0
 0.0  0.0  1.0  0.0
 0.0  0.0  0.0  1.0
singular values:
Float64[]
Vt factor:
0×0 Matrix{Float64}

[Jutho]

I guess we will need to capture this manually in all the methods, as Lapack does indeed not support empty matrices.