[GRASS-user] Re: r.in.wms bug (rchirgwin@ozemail.com.au)

[Hamish]

Richard Chirgwin wrote:
> Some progress on this, but I suspect there's something "under the 
> covers" that somebody needs to know about.
> 
> The r.in.wms documentation says:
> *r.in.wms* [-*ldockp*] [*output*=/string/] *mapserver*=/string/ 
> [*layers*=/string/[,/string/,...]] [*styles*=/string/[,/string/,...]]
> [*srs*=/string/] *format*=/string/ [*wmsquery*=/string/] 
> *maxcols*=/integer/ *maxrows*=/integer/ [*tileoptions*=/string/] 
> [*region*=/string/] [*folder*=/string/] [*wgetoptions*=/string/] 
> [*curloptions*=/string/] *method*=/string/ [*v*=/integer/] 
> [--*overwrite*] [--*verbose*] [--*quiet*]
> 
> Note that "mapserver" appears before layers. If I use this syntax, I
> get:
> ERROR: Required parameter <layers> not set:
>     (Layers to request from map server).
> 
> However, if I reverse the order of the mapserver and the layers:
> r.in.wms output=Background layers=psma:all mapserver=<URL>

The parser doesn't really care about the order*; this lends credence to
the theory that you need to quote your strings from the terminal and
the command gets side-tracked with a &|;= before getting to the end.
In the case where layer= comes first it will truncate the mapserver=
option and it will still fail, but later on in the script.

[*] If your mapserver URL contains an unquoted "=" the parser might
treat that part as an option.

> - then things work okay. So is it an error in the documentation, or
> has a more recent release changed the r.in.wms script?

Did this once work for you in an earlier version of GRASS?


Hamish



      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs