#3545 wrong definite integral for %e^(-2*%i*%pi*t)*(%e^(2*%i*%pi*t)+1)

[lanride]

maxima 5.42.2 Lisp: SBCL 1.4.10.debian

simple function gives right answer (see u1,u2)
but rewriting u1 into v1 gives wrong result for definite integral, indefinite is right

(%i27)  u1: %e^(-2*%i*%pi*t)+1;
(u1)    %e^(-2*%i*%pi*t)+1
(%i36)  u2: ratsimp(integrate(u1,t,-1/6,1/6));
(u2)    (2*%pi+3^(3/2))/(6*%pi)                         OK
(%i37)  v1: %e^(-2*%i*%pi*t)*(%e^(2*%i*%pi*t)+1);
(v1)    %e^(-2*%i*%pi*t)*(%e^(2*%i*%pi*t)+1)
(%i38)  ratsimp(integrate(v1,t,-1/6,1/6));
(%o38)  sqrt(3)/(2*%pi)                                    BAD

[David Scherfgen]

Correct in current Maxima, closing this ticket.