Constructing matrices over finite fields silently fails if multiplicative generator not used

Bonsoir, not sure if this is related to any of the other recent `GF` matrix bugs, but there seems to be a silent failure to construct a matrix over a finite field if the multiplicative element is not used in the second argument. To see what I mean, consider the following example: ```gap gap> A := Matrix(GF(2), [[1, 1, 1], [0, 1, 1], [0, 0, 1]]); [ [ 1, 1, 1 ], [ 0, 1, 1 ], [ 0, 0, 1 ] ] gap> A^-1; [ [ 1, -1, 0 ], [ 0, 1, -1 ], [ 0, 0, 1 ] ] gap> A[1][1] in GF(2); false gap> B := Matrix(GF(2), Z(2)^0*[[1, 1, 1], [0, 1, 1], [0, 0, 1]]); <a 3x3 matrix over GF2> gap> B^-1; <a 3x3 matrix over GF2> gap> Display(B^-1); 1 1 . . 1 1 . . 1 gap> B[1][1] in GF(2); true ``` Despite the underlying semiring `GF(2)` being specified in the first component, since `Z(2)` is not used to define the matrix `A`, the resulting matrix is not a matrix over semiring object at all. This is a silent failure and quite an unintuitive footgun. Not sure if this is a bug or just a case of me holding it wrong, but the result is very unintuitive and can lead to hard to debug errors further down the line. I think the fix might be to either: 1. Raise an error if the matrix in the second component of `Matrix(GF(p^d), A)` does not have all its elements in `GF(p^d)` or 2. Automatically multiply the matrix in the second component by `Z(p^d)^0`. I think 2. might be more user friendly, but there might be some hidden cost or subtle mathematical nuance I'm missing. Initially filed as https://github.com/semigroups/Semigroups/issues/1186