Integrals with lb > ub

[ArnoStrouwen]

Similar to how Inf bounds are handled, we should handle this case by an automatic transformation.
Particularly because the Inf bound handling implicitly assumes that lb < ub, and integrals from +inf to a might give incorrect results.

[ygol55]

I'm pretty sure that's not supported for a reason, while I couldn't found cases where the lower bound is higher then the higher bound. I guess that's just isn't needed (meaning for infinite positive lower bound).