Here is the simplest example of the problem. If I type

assume(j, integer, j>0); assume(k, integer, k > 0);

followed by

int(cos(j*t)*cos(k*t), t = -Pi..Pi);

Maple gives the answer 0, not, as I would have expected, Pi*KroneckerDelta[j,k].

Is there a way to force Maple to alllow for both the possibilities that j = k and j <> k, and thereby to give the general result using the Kronecker delta (or equivalent)?