[Proj] Question on colon DMS format

[Hamish]

> Hamish wrote:
> > GRASS still uses this format, the following are allowed:
> >
> > DDDh
> >  (DDD.DDDDDDh is not allowed, as E has problems with
> >   mistranslation into e.g. e+06; so +/- DDD.DDDDD must be
> >   used there; could probably be coded around but currently isn't)

Gerald wrote:
> I do not understand the "h".

(I just meant hemisphere)
 
> Problems with a NEWS suffix sign sounds like a library problem.
....
> A proper lexical analyser should sense existance of an e|E
> followed by a signed number.

or just my bad memory, or assumption about why something so simple wasn't
implemented (I'm too trusting there was a good reason). all I'm really
sure of is that at present you can't mix degree-only with decimal places
and NSEW in GRASS. Maybe it is a simple libgrass coding fix to enable
it after all. shrug, not relevant for this thread.


> The "D" of FORTRAN infamy is not recognized.

but "U" and "f" maybe?

> In any case proj_dms2rad and preceeding versions have not put a
> restriction on ddd.dddd.

ok, good.

> > GRASS enforces %02d for MM and SS input, but I don't think
> > that's necessary and have considered removing it.
> 
> Don't use %2d as that will cause space for numbers under
> 10.

%02d pads the space with a zero. my point was that GRASS expects to
see 123:45E or 123:05E, but 123:5E is considered an error.

> "%d" will work OK with proj_dms2rad.

ok, good.

> Basic, old time, standard ASCII does *not* have a degree
> mark and that is why I have not used it.

you mean in the days before they discovered the number 129? ;-)



Hamish