Functionality to extract submodels

[daavid00]

Using the flag -v, then a submodel can be extracted either by conditions on the init values (e.g., 'fipnum 2,3') or by giving the xypolygon vertices in meters (e.g., 'xypolygon [0,0] [30,0] [30,30] [0,0]'). The cool thing is that the pore volume outside the extracted domain can be added to the submodel boundaries with different options to project it using the -p flag: '0' no porv correction, '1' adds the porv from outside on the boundary of the submodel, '2' adds the corner regions (e.g., below the mini and minj from the input model) to the corners in the submodel, '3' distributes the porv uniformly along the boundary, and '4' distributes it on the whole submodel.

Extracting a submodel is useful to reduce memory and increase efficiency (all ijk numbers are mapped accordingly to the new deck, including segmented wells).