This is a problem from the book "Introduction to Lie-Algebra"-Erdmann & Wildon, Chapter 5-"Subalgebras of $gl(V)$"
Let $L=gl(n, \mathbb{C})$. Let $A\subset L$ be a subalgebra of $L$, such that $A$ contains all the diagonal matrices. Prove that $N_{L}(A)=A$, where $$N_{L}(A)=\{\ x\in L : [x,a]\in A \forall a\in A \}\ $$, called the normalizer of $A$ in $L$.
Now whenever I take a particular example for $A$, that is , say $A=d(n,\mathbb{C})$, or $A=t(n,\mathbb{C})$, it becomes easy to compute and we can see that they are self-normalizing. But whenever we have to take an arbitrary subalgebra containing $d(n,\mathbb{C})$, I have no clue about how to compute the elements of $N_{L}(A)$.
The book suggests, two ways- one a direct approach , which I tried without succeeding, another a little sophisticated one- by using the "invariance Lemma", which is there in the book in the same chapter.
I want to have a clue for both the suggested ways. I will really appreciate some help. Thanks in advance!