Conflict with gparts changes

Conflicts:
        src/runner.c

Corrected sod shock ICs

This fixes issue #135. I had forgotten to push the updated version of the Sod shock IC script when updating the smoothing length definitions last week.

I have also added the definition of the unit system used in the run to the ICs. These are currently not used but we are working on converting ICs to internal unit system with Tom. 

See merge request !139

fix genuine problem detected by cppcheck.



See merge request !141

Parallel gpart sort

That's essentially a copy-paste of the part version. 

Indices are still of type `size_t` to be compatible with the rest of the code. We can move to `int` later on.

See merge request !128

More physical definition of SPH kernel functions

This "solves" #97. 

 - I have split the SPH and gravity kernels into two separate files for ease of use. That explains a lot of the file changes below.
 - Corrected a small typo in the hydro.h files. The mistake was not leading to wrong physics results but would slow down convergence towards the correct h in the ghost task.
 - When switching kernel function, the resolution is not modified any more. This was the main point of #97.
 - The documentation of the kernels has increased and I have updated the .tex documents and figures to be completely consistent with the code.

In normal mode (i.e. without editing const.h) the results are identical. 

See merge request !138

Parameter file

First version of parameter file parser. Should be able to:
- Read integers
- Read floats
- Read doubles
- Read strings
- Ignore comments that start with '#'
- Validate that integers are not written as floats in the input file
- Check whether two values are present on the same line
- Print the contents of the structure that holds all parameters read from the file

I had to add `../src/parser.c` to the Makefile.am in the tests directory to get `make check` working. Not sure if this is the correct way of doing it.

See merge request !129

First implementation of gpart motion

The changelog is rather big but the changes are rather straightforward:

 - The acceleration of the `gparts` is now called `a_grav` instead of just `a`. Makes grepping through the files much easier.
 - I have added the command-line options `-g` and `-G` for external- and self-gravity respectively. These build the correct policies into the engine. Separating both (at least for now) allows to work on the two aspect separately.
 - There are tests in critical places to make sure that linked pairs part-gpart are at the same position.
 - The `doinit()`, `dodrift()` and `dokick()` tasks have been updated to also move the gparts.
 
All the tests pass successfully. I am seeking your opinion on the implementation of the tasks. 

See merge request !135