Fixes #515

Fixes #513

Fixes issue #245

* updated GPSPropagator, DOPComputer, tests and tutorial DOPComputation
* added SEM and Yuma parsers and tests
* updated messages

Users can now tune max degree, max eccentricity power and max frequency
in true longitude for short period terms.

This setting is independent from the mean elements
settings in DSST tutorial.

Also increased output accuracy.

The tutorial still does not run properly.

The data comes from Telespazio contribution in the Orekit labs.

Work In Progress, the program compiles correctly, but is not tested yet.

The interpolation can either use a fixed number of points or a
maximum time gap between points, for each mean elements
integration step.

This attempt is clearly not successful yet ...

This feature is intended for debugging purposes, as the coefficients
seem very wrong in some cases and seem to depend on propagation
duration.

The coefficients are provided as spacecraft state additional parameters.

This is mainly intended for test and validation purposes.

We may need to use mean orbit as input and create osculating
orbits as output for example.

DSST.

The propagation of mean elements with the DSST propagator have been
roughly validated:
* zonal, drag and SRP contributions look OK,
* tesseral resonant is not far from being good, but needs more tests.