Contract epsilon tensors to get kronecker deltas with the sympy.tensor module

[Hersh Singh]

Hi all,

I was wondering if, using the sympy.tensor module, I can contract two epsilon tensors to get kronecker deltas? 

In particular, if I have something like

from sympy.tensor.tensor import TensorIndexType, tensor_indices
Spin = TensorIndexType('Spin', dummy_fmt='S', dim=3)
i, j, k, l, m = tensor_indices('i, j, k, l, m', Spin)

delta = Spin.delta
eps = Spin.epsilon

q = eps(i,j,k)*eps(-i,l,m)

print(q)

then I get 

>>> Eps(S_0, j, k)*Eps(-S_0, l, m)

In 3 dimensions, I would want to contract the $ \epsilon^{ijk} \epsilon^{ilm} $ to give me

 $\delta^{jl}\delta^{km} - \delta^{jm}\delta^{lk}$$.  So, I would like to have something like `q.epsilon_to_delta` to get

q.epsilon_to_delta()
print(q)
>>> KD(j,l)*KD(k,m) - KD(j,m)*KD(l,k)

Is this possible?  

Thanks!