#1138 risch & *integrator-level* unbound

[Barton Willis]

risch(cosh(x)) --> lisp error, but risch(cos(x)) i OK:

(%i4) risch(cosh(x),x);
Maxima encountered a Lisp error:
Error in MACSYMA-TOP-LEVEL [or a callee]: The variable *INTEGRATOR-LEVEL* is unbound.

(%i5) risch(cos(x),x);
(%o5) sin(x)

[Barton Willis]

Logged In: YES
user_id=895922
Originator: YES

A putative fix:

(defmfun $risch (exp var)
(setq *integrator-level* 0)
(with-new-context (context)
(rischint exp var)))

[Raymond Toy]

Logged In: YES
user_id=28849
Originator: NO

*integrator-level* was added fairly recently to break an infinite loop in INTEGRATOR. Instead of setting it in $risch, perhaps it should be bound in hypertrigint1, which looks like the only place risch calls back into sin.lisp (via INTEGRATOR).

[Raymond Toy]

Logged In: YES
user_id=28849
Originator: NO

Fixed as suggested (mostly).

Now risch(cosh(x),x) -> 1/2*(exp(x)-exp(-x))

I suppose that would be better as sinh(x), the answer is right.

risch.lisp, rev 1.10