Make it possible to correct for beam a second time

Created by: aroffringa

This makes it e.g. possible to first apply the dipole beam to some direction and then the full beam when phased to another direction. This was requested by Frits.

When invert=true (i.e. normal situation when applying beam to data), and the beam has already been applied to a certain direction, this is first undone and then the requested beam to the requested direction is applied. Some examples:

• Applying an invert=true beam to direction A, then again to direction A, is now accepted, and ends up with the beam in direction A (second run doesn't change anything).
• Applying an invert=true element beam to direction A, then applying an invert=true array beam to direction B, would end up with the array beam applied to direction B (i.e. element beam to direction A is undone).

Fully negating the beam having applied it is also possible (and was already possible). It requires an invert=false beam correction with the same direction and mode setting as was applied. The invert=xxx option is pretty confusing, and I'll file a ticket that we should make the beam setting such that we specify what we want as output, not what operation we want to do.

Fixes #221 (closed).