Jon Britton wrote:
> Hi again,
> 
> I'm just looking at the existing  WCS bindings you linked to.  The binding 
> classes I've looked at all seem to be unchanged since they were generated.  
> Did you only implement the ones you were interested in? If so, which ones?
This is the general strategy when working with object models that are 
EMF based. There is a generic binding class (ComplexEMFBinding) which 
uses EMF reflection to do all the parsing and encoding, so a particular 
binding class is not needed. But in some cases the need to specialize is 
there.

So many of those generated binding classes are redundant... should 
probably be removed. If you look at the WCSConfiguration class you 
should see all the "special purpose" bindings which are registered.
> 
> Also, how do you handle the WCS versions when generating the bindings?  I 
> can't see any mention of the version in these bindings, but surely different 
> versions will generate different bindings?  I am assuming that I would need 
> to generate them for each version of WCS I wish to implement, but can see how 
> this could quickly become confusing because many of the bindings will have 
> the same name.
Good question. There are basically two approaches. The first as you 
mention is new bindings + new model for each new version of the schema. 
However, this is quite a bit of work. The second alternative is to try 
to use a single model, re-use bindings, and then specialize for each 
version of the schema.

In the case of WCS, the schema changes radically between versions 1.0 
and 1.1 so it is not practical to use the second approach. However in 
the case of WFS the schema remains quite similar moving from version 1.0 
to version 1.1. So the approach was to generate the latest version of 
the model and have difference bindings for the 1.0 and 1.1 schema.

Hope that helps.

-Justin
> 
> Thanks,
> 
> Jon
> 
> 
> 
> Justin Deoliveira ha scritto:
>> Hi Jon,
>>
>> Yes I believe it does. As far as I know the WCS schema namespace uri is 
>> "http://www.opengis.net/wcs/1.1";.
> 
> Actually OGC used "http://www.opengis.net/wcs/1.1.1"; for the 1.1.1
> corrigendum, and in the second corrigendum they took that back
> and now the schemas for 1.1.2 say "http://www.opengis.net/wcs/1.1";.
> GeoServer implements 1.1.1, the second corrigendum is only a few
> months old and we have no sponsoring on the WCS side.
> 
>> When andrea implemented WCS 1.1 in GeoServer he had to generate 
>> bindings, and I remember it was quite a pain because of these issues 
>> with the namespace URI.
> 
> The pain was due to the fact the schema was split among many
> files, but there were other issues thought. I could not figure
> them out and asked you. Don't remember exactly what we did.
> 
> Anyways, the exiting bindings are here:
> http://svn.geotools.org/branches/2.5.x/modules/extension/xsd/xsd-wcs/
> 
> but we only ever used them to parse request and encode
> the responses, not the opposite (generate requests, parse
> the responses). Afaik they are mostly plain jane EMF
> based, so they should work both ways, but they were
> never tested from a client side.
> 
> You can also have a look at the code in the GeoServer WCS 1.1
> implementation here:
> http://svn.codehaus.org/geoserver/branches/1.7.x/src/wcs1_1/
> 
> Cheers
> Andrea
> 


-- 
Justin Deoliveira
OpenGeo - http://opengeo.org
Enterprise support for open source geospatial.

------------------------------------------------------------------------------
_______________________________________________
Geotools-gt2-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-gt2-users

Reply via email to