[GRASS-user] r.out.gdal bug? problems with createopt

GRASS GRASS user
,
1

Hi List,

I am trying to export a floating point raster into a tiff, BUT without writing
a GeoHeader (no GeoTiff!).

( I want to do this rather unusual thing, because the PCI Geomatics Software
has Problems with the Gauss Kruger Projection. It misinterprets a good
looking GeoTiff and writes only garbage in the GeoHeader of its output. But
if one feeds it just coordinates they line up perfectly.)

r.out.gdal [...] createopt="TFW=YES,PROFILE=BASELINE"
writes always a GeoHeader!

According to the GDAL documentation (http://www.gdal.org/frmt_gtiff.html) it
should not write any Geo-Stuff, if given the option PROFILE=BASELINE.

I then tested the scripted version
r.out.gdal.sh [...] createopt="TFW=YES,PROFILE=BASELINE"
and luckily it behaves and writes a tif without GeoHeader.

cheers,
  robert

2

Hi Robert,

it seems to be the same bug as it was found and fixed in v.out.ogr
some days ago by Martin.

I debugged a bit and found that the second part
"PROFILE=BASELINE" isn't passed. It gets lost in
main.c, line 554.

v.out.ogr's main.c (line 249 etc) has a working solution
but my try to port it to r.out.gdal failed (no idea why, should be
rather easy - but it segfaults for me).

So - we are close to a solution :slight_smile:

Markus

Robert Nuske wrote:

Hi List,

I am trying to export a floating point raster into a tiff, BUT without
writing
a GeoHeader (no GeoTiff!).

( I want to do this rather unusual thing, because the PCI Geomatics
Software
has Problems with the Gauss Kruger Projection. It misinterprets a good
looking GeoTiff and writes only garbage in the GeoHeader of its output.
But
if one feeds it just coordinates they line up perfectly.)

r.out.gdal [...] createopt="TFW=YES,PROFILE=BASELINE"
writes always a GeoHeader!

According to the GDAL documentation (http://www.gdal.org/frmt_gtiff.html)
it
should not write any Geo-Stuff, if given the option PROFILE=BASELINE.

I then tested the scripted version
r.out.gdal.sh [...] createopt="TFW=YES,PROFILE=BASELINE"
and luckily it behaves and writes a tif without GeoHeader.

cheers,
  robert

_______________________________________________
grassuser mailing list
grassuser@grass.itc.it
http://grass.itc.it/mailman/listinfo/grassuser

--
View this message in context: http://www.nabble.com/r.out.gdal-bug--problems-with-createopt-tf4504650.html#a12847602
Sent from the Grass - Users mailing list archive at Nabble.com.

3

Moin Markus,

thanks for looking into my problem.

it seems to be the same bug as it was found and fixed in v.out.ogr
some days ago by Martin.

I debugged a bit and found that the second part
"PROFILE=BASELINE" isn't passed. It gets lost in
main.c, line 554.

I am not sure this is the only problem.

r.out.gdal [...] createopt="PROFILE=BASELINE"
doesn't work either

cheers,
  robert
  

v.out.ogr's main.c (line 249 etc) has a working solution
but my try to port it to r.out.gdal failed (no idea why, should be
rather easy - but it segfaults for me).

So - we are close to a solution :slight_smile:

Markus

Robert Nuske wrote:
> Hi List,
>
> I am trying to export a floating point raster into a tiff, BUT without
> writing
> a GeoHeader (no GeoTiff!).
>
> ( I want to do this rather unusual thing, because the PCI Geomatics
> Software
> has Problems with the Gauss Kruger Projection. It misinterprets a good
> looking GeoTiff and writes only garbage in the GeoHeader of its output.
> But
> if one feeds it just coordinates they line up perfectly.)
>
> r.out.gdal [...] createopt="TFW=YES,PROFILE=BASELINE"
> writes always a GeoHeader!
>
> According to the GDAL documentation (http://www.gdal.org/frmt_gtiff.html)
> it
> should not write any Geo-Stuff, if given the option PROFILE=BASELINE.
>
> I then tested the scripted version
> r.out.gdal.sh [...] createopt="TFW=YES,PROFILE=BASELINE"
> and luckily it behaves and writes a tif without GeoHeader.
>
> cheers,
> robert
>
> _______________________________________________
> grassuser mailing list
> grassuser@grass.itc.it
> http://grass.itc.it/mailman/listinfo/grassuser

4

Markus Neteler wrote:

Hi Robert,

it seems to be the same bug as it was found and fixed in v.out.ogr
some days ago by Martin.

I debugged a bit and found that the second part
"PROFILE=BASELINE" isn't passed. It gets lost in
main.c, line 554.

v.out.ogr's main.c (line 249 etc) has a working solution
but my try to port it to r.out.gdal failed (no idea why, should be
rather easy - but it segfaults for me).

What about moving this multi-option code to
G_store_multi() or something like this?

Modules with multi-option parameters are
- r.out.ogr
- v.out.ogr
- v.transform (new column parameter)
- maybe others

?

Markus
--
View this message in context: http://www.nabble.com/r.out.gdal-bug--problems-with-createopt-tf4504650.html#a12855513
Sent from the Grass - Users mailing list archive at Nabble.com.