[gdal-dev] WCS 1.1 in GDAL

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

[gdal-dev] WCS 1.1 in GDAL

Ari Jolma-2
It seems to me that WCS 1.1 has been dormant and perhaps nobody has
really used it ever(?).

I just found the gdalgmlcoverage.cpp file in gcore, which is only used
by the WCS driver and only in its 1.1 part and it hasn't really been
changed since Frank wrote it ten years ago. I basically rewrote the same
thing since WCS 2.0 uses GML coverages too.

There are some things in the code that make it look such that it's just
been there. Also the things I encountered when trying the 1.1.

Interesting.

Ari


_______________________________________________
gdal-dev mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/gdal-dev
Reply | Threaded
Open this post in threaded view
|

Re: WCS 1.1 in GDAL

Ari Jolma-2
Ari Jolma kirjoitti 01.11.2017 klo 11:04:
> It seems to me that WCS 1.1 has been dormant and perhaps nobody has
> really used it ever(?).
>
> I just found the gdalgmlcoverage.cpp file in gcore, which is only used
> by the WCS driver and only in its 1.1 part

Actually, in 1.0, since 1.1 does not seem to use the gml:RectifiedGrid,
however, 2.0 uses it again.

Anyway, the question is, should gdalgmlcoverage.cpp be in the gcore? The
GDAL sources are sometimes confusing. For example there is gmlutils.cpp
- which is used in WCS because of GML_IsSRSLatLongOrder, but that is
ogr/ogrsf_frmts/gml.

> and it hasn't really been changed since Frank wrote it ten years ago.
> I basically rewrote the same thing since WCS 2.0 uses GML coverages too.
>
> There are some things in the code that make it look such that it's
> just been there. Also the things I encountered when trying the 1.1.
>
> Interesting.
>
> Ari
>
>

_______________________________________________
gdal-dev mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/gdal-dev
Reply | Threaded
Open this post in threaded view
|

Re: WCS 1.1 in GDAL

Even Rouault-2

On mercredi 1 novembre 2017 13:05:09 CET Ari Jolma wrote:

> Ari Jolma kirjoitti 01.11.2017 klo 11:04:

> > It seems to me that WCS 1.1 has been dormant and perhaps nobody has

> > really used it ever(?).

> >

> > I just found the gdalgmlcoverage.cpp file in gcore, which is only used

> > by the WCS driver and only in its 1.1 part

>

> Actually, in 1.0, since 1.1 does not seem to use the gml:RectifiedGrid,

> however, 2.0 uses it again.

>

> Anyway, the question is, should gdalgmlcoverage.cpp be in the gcore?

 

Feel free to move it to frmts/wcs as it only used by it. You'd probably want to remove the declaration of it from gcore/gdal_priv.h as well.

Looking at GDALParseGMLCoverage() code, I see a "GDALJP2Metadata" remaining CPLDebug message, but it is no longer used by the GMLJP2 code which has a specialized version of it now. You may have a look at GDALJP2Metadata::ParseGMLCoverageDesc() in gcore/gdaljp2metadata.cpp by the way, in particular the logic for axis inversion around lines 1055-1089. Note: there are also stuff specific of GMLJP2 issues at the end of the function, so I would *not* recommend blindly porting all this.

 

> The

> GDAL sources are sometimes confusing. For example there is gmlutils.cpp

> - which is used in WCS

 

In your in-progress work I suppose. Can't see that in trunk.

Could probably make sense to move it in ogr/ then (we have already ogr/gml2ogrgeometry.cpp and ogr2gmlgeometry.cpp there)

 

--

Spatialys - Geospatial professional services

http://www.spatialys.com


_______________________________________________
gdal-dev mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/gdal-dev
Reply | Threaded
Open this post in threaded view
|

Re: WCS 1.1 in GDAL

Ari Jolma-2
Even Rouault kirjoitti 01.11.2017 klo 13:20:

On mercredi 1 novembre 2017 13:05:09 CET Ari Jolma wrote:

> Anyway, the question is, should gdalgmlcoverage.cpp be in the gcore?

 

Feel free to move it to frmts/wcs as it only used by it. You'd probably want to remove the declaration of it from gcore/gdal_priv.h as well.


Should be done now ok.

Looking at GDALParseGMLCoverage() code, I see a "GDALJP2Metadata" remaining CPLDebug message, but it is no longer used by the GMLJP2 code which has a specialized version of it now. You may have a look at GDALJP2Metadata::ParseGMLCoverageDesc() in gcore/gdaljp2metadata.cpp by the way, in particular the logic for axis inversion around lines 1055-1089. Note: there are also stuff specific of GMLJP2 issues at the end of the function, so I would *not* recommend blindly porting all this.


Thanks for the reminder.

 

> The

> GDAL sources are sometimes confusing. For example there is gmlutils.cpp

> - which is used in WCS

 

In your in-progress work I suppose. Can't see that in trunk.

Could probably make sense to move it in ogr/ then (we have already ogr/gml2ogrgeometry.cpp and ogr2gmlgeometry.cpp there)


Yes, I looked around for existing code to check for the axis order swap.

I'm beginning to see light at the end of the tunnel and perhaps could really start testing with 2.0.1 servers soon.

There's undocumented "WCS_SDS:" method for accessing WCS servers and putting data from different times into bands. I'll need to study that a bit and maybe implement same kind of system.

Ari

 

--

Spatialys - Geospatial professional services

http://www.spatialys.com



_______________________________________________
gdal-dev mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/gdal-dev