Fixed bug in HLLC Riemann solver flux calculation.

This fixes #216 (again, but now all cases).

I basically overlooked a tiny detail in the real GIZMO and changed the order of some operations that cannot be interchanged. Square test is now verified to work with all 3 possible Riemann solvers.

See merge request !257

Updated the ICs for the EAGLE_25 test case

Here are the new ICs if you want to try them out.

See merge request !256

Added conversion of box dimensions

Fixes #217. We need to convert the box size to the internal system of units, not just the particle properties.

See merge request !254

Implementation Pressure entropy SPH

Here is a new flavour of SPH that works and gives better result than the default Gadget-2 scheme.

The only change affecting the rest of the code is the addition of the condition in `engine_init_particles()` to create a second density loop over the particles for the special conversions that are required in this scheme. Other than that, it just fits in the rest of the framework.

See merge request !252

Made sure primitive variable velocity is correctly initialized.

This fixes #216. The initial velocities are now initialized (again) in `hydro_convert_quantities` by using `p->v`, before they are used to set the momentum. I also removed the `if` condition from `hydro_end_density`, as we do not longer care if the primitive variables are overwritten after the initialization step.

See merge request !253

(cherry picked from commit e22730e917d6bf96028982d5842f84c517f0835f)

(cherry picked from commit 2c9c86429a9cf195ef671163fd261f9cafb0c910)

(cherry picked from commit ce0055f95a0eb5052081858c883e477bafef2fd2)

Add some sanity checking when splitting cells.

Stops after 52 splits. This depth is way beyond our precision.


See merge request !250

This depth is way beyond our precision.

Re-factored the whole external potential mechanism to match what is done for cooling

Also, renamed 'disk' to 'disc' and improved the documentation and description of the parameters. 

@pdraper this is just code rearrangement, no changes to the actual physics has been done.

See merge request !249