[Proj] MGRS/USNG support?

[Oscar van Vlijmen]

>From Eric G. Miller:

> I guess, for this to work the UTM handling would have to change.  Currently,
> you have to specify +zone= and +south (as appropriate), but the library should
> be able to figure that out and report it back in the output.  The user should
> have the option of forcing a zone or allowing the library to determine it.
> Perhaps this is too deep a change?

>From my remarks, April 26:

> Let the projection code calculate what the nearest central meridian is,
> merely based on the input of the longitude of each point.
> But: can (lib)proj do that?

Is it really that obtrusive to change tmerc/utm so, that it calculates the
nearest central meridian itself, if no zone or CM is supplied?
Indeed, processing MGRS strings should be a pre/postprocessor task.